SEARCH

SEARCH BY CITATION

Keywords:

  • Cancer stromal fibroblast;
  • Lung cancer;
  • Fibroblast progenitor cell

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure of Potential Conflicts of Interest
  8. Acknowledgements
  9. References

Recent animal data have suggested that cancer-induced stroma consists of blood-borne fibroblasts as well as tissue-derived fibroblasts. In this study, mononuclear cells isolated from the pulmonary vein blood of lungs resected from lung cancer patients were cultured to confirm the presence of blood-borne fibroblast. In 34% (16 of 47) of the cases, spindle cells with fibroblast morphology proliferated in a disarrayed fashion and were positive for vimentin and collagen type I but negative for both specific myogenic and endothelial markers. The cDNA profiles of blood-borne fibroblasts, tissue-derived (lung) fibroblasts, human vascular smooth muscle cells (HSMCs), and umbilical vein endothelial cells (HUVECs) were clustered with a hierarchical classification algorithm. The profiles of the blood-borne fibroblasts were clearly isolated from those of the tissue-derived fibroblasts, HSMCs, and HUVECs. When carboxyfluorescein succinyl ester (CFSE)-labeled human mononuclear cells from the blood of lung cancer patients were transferred into nonobese diabetic/severe combined immunodeficient (NOD/SCID) mice engrafted with a human lung cancer xenograft, CFSE-labeled fibroblasts were found around the cancer nests. We investigated the several clinicopathological factors of blood-borne fibroblast-positive patients. The blood-borne fibroblast-positive cases had a significantly larger central fibrotic area in primary lung cancer than in the negative cases (123 ± 29 vs. 59 ± 13 mm2; p = .02). Our results indicated that the blood in the vicinity of human lung cancer contains fibroblast progenitor cells that have the capacity to migrate into the cancer stroma and differentiate into fibroblasts having biological characteristics different from those of tissue-derived fibroblasts.

Disclosure of potential conflicts of interest is found at the end of this article.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure of Potential Conflicts of Interest
  8. Acknowledgements
  9. References

Cancer cells live under complex microenvironments constituted by the extracellular matrices and a variety of nonepithelial cell types: immune cells, endothelial cells of blood and lymph vessels, pericytes, and fibroblasts. One of these cell types, tissue fibroblasts, is thought to actively migrate into cancer tissue from the surrounding tissue and produce several growth factors and extracellular matrix proteins that affect the growth of the cancer cells. Fibroblasts are known to be heterogeneous with respect to a number of phenotypic and functional features [1, [2], [3], [4]5], and their heterogeneity may arise not only from activation or differentiation processes but differences in their cells of origin [5]. Investigators, including ourselves, have recently reported finding that bone marrow (BM)-derived cells contain progenitor cells of tissue fibroblasts that are recruited through the circulation to populate peripheral organs [6, [7], [8], [9], [10]11]. In mouse bone marrow transplantation models, BM-derived fibroblasts are engrafted into multiple organs and are efficiently recruited into fibrotic lesions in response to injurious stimuli [7, 11]. The cancer-induced stroma in a mouse model of pancreatic insulinoma and in severe combined immunodeficient (SCID) mice transplanted with a human xenograft consists of various proportions of BM-derived fibroblasts [9, 10, 11], and we have recently shown that the frequency of BM-derived fibroblasts recruited into cancer stroma is significantly correlated with the proportion of the stroma in cancer tissue [12].

Early studies described the presence of fibrocytes called “circulating fibrocytes” in normal peripheral blood [13, 14]. Circulating fibrocytes comprise 0.1%–0.5% of the human nonerythrocytic cell population in peripheral blood and migrate efficiently to injured tissues, including skin wounds and pulmonary fibrosis [8, 13, 14, 15]. Although there is controversy as to whether circulating fibrocytes and BM-derived fibroblasts are the same population, the above findings provide convincing evidence that cancer-induced stroma contains both tissue-derived fibroblasts (residual fibroblasts) and fibroblasts that have migrated in from peripheral blood (blood-borne fibroblasts). However, there have been few reports on blood-borne fibroblasts in human cancer patients.

In this study, we attempted to identify the circulating progenitor cells of cancer stromal fibroblasts in the blood of lung cancer patients.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure of Potential Conflicts of Interest
  8. Acknowledgements
  9. References

Subjects and Collection of Blood Mononuclear Cells from Pulmonary Vein

The dissected and ligated pulmonary veins of surgically resected lungs contain more than 10 ml of blood. In this study, an 18-gauge needle was inserted into the pulmonary vein, and 10 ml of blood was collected from the pulmonary vein from 47 lungs surgically resected to treat primary non-small cell lung cancer at our hospital. All blood samples were collected after the subjects gave their written informed consent, approved by the Institutional Review Boards at the National Cancer Center.

Peripheral Blood Mononuclear Cell Isolation and Cell Culture

Mononuclear cells from the pulmonary vein were isolated from the buffy coat by Ficoll-paque (Amersham Biosciences, Piscataway, NJ, http://www.amersham.com) density-gradient centrifugation (1.077 g/cm3). Isolated cells were plated at a density of 5 × 106 cells per well in six-well tissue culture dishes (Becton, Dickinson and Company, Mountain View, CA, http://www.bd.com) in Dulbecco's modified Eagle's medium (DMEM) supplemented with 20% heat-inactivated fetal bovine serum (FBS) (Sigma-Aldrich, St. Louis, http://www.sigmaaldrich.com), 50 U/ml penicillin, and 50 μg/ml streptomycin. Nonadherent cells were removed after 72 hours of incubation, and the adherent cells were cultured. If fibroblast colonies were found in the culture dish, the cells were grown until confluence. Cells were detached with trypsin and washed, and culture was continued by dividing 1-second dilution. Every 3 days, the medium containing floating cells was removed, and fresh medium was added. Cells were cultured for up to 8 weeks. Human vascular smooth muscle cells (HSMCs) and human umbilical vein endothelial cells (HUVECs) were obtained from Cambrex (Walkersville, MD, http://www.cambrex.com), and human fibroblasts (WI-38 and MRC-5) were obtained from American Type Culture Collection, Rockville, MD, http://www.atcc.org. All cultures were maintained at 37°C in a humidified atmosphere containing 95% air and 5% CO2.

Microarray Analysis

We used GeneChip Human Genome U133 Plus 2.0 arrays (Affymetrix, Santa Clara, CA, http://www.affymetrix.com) that including 54,675 probe sets for analysis of mRNA expression levels of approximately 47,000 transcripts and variants from 38,500 well-characterized human genes. Target cRNA was generated from 2.5 μg of total RNA from each sample with One-Cycle Target Labeling and Control Reagents (Affymetrix). The procedures for target hybridization, washing, and staining with signal amplification were conducted according to the supplier's protocols (http://www.affymetrix.com/support/technical/manual/expression_manual.affx). The arrays were scanned with a GeneChip Scanner 3000 (Affymetrix), and the intensity of each feature of the array was calculated with GeneChip Operating Software v1.1.1 (Affymetrix). Average intensity was standardized to the target intensity, which was set equal to 1,000, to reliably compare variable multiple arrays. The dendrogram was generated with the Cluster and Treeview programs [16]. The values were log-transformed and median-centered. GeneSpring software (Agilent Technologies, Santa Clara, CA, http://www.agilent.com) and Excel software (Microsoft, Redmond, WA, http://www.microsoft.com) were used to perform the numerical analysis to permit gene selection.

Pathological Studies

All surgical specimens were fixed with 10% formalin and embedded in paraffin. The tumors were cut at approximately 5-mm intervals, and serial 4-μm sections were stained with hematoxylin and eosin, the Alcian blue-periodic acid Schiff method to visualize cytoplasmic mucin production, or the Verhoeff-van Gieson method to visualize collagen fiber and elastic fibers. Lymphatic permeation and pulmonary metastases were evaluated on the sections stained with hematoxylin and eosin. Vascular invasion, pleural invasion, and collapse fibrosis size were evaluated in the sections stained with the Verhoeff-van Gieson method. The size of the central fibrotic region was measured histologically at low power view as previously reported [17]. Two observers (G.I. and A.O.) who were unaware of the clinical data independently reviewed all the pathological slides. The histological diagnoses were based on the revised WHO histological classification [18]. Tumor size was measured as the maximal diameter on the cut section of the lung. The pathological stage was determined according to the classification of the Union Internationale Contre le Cancer. The clinical and pathological characteristics of the patients are summarized in Table 1.

Table Table 1.. Patient characteristics and clinicopathological factors
Thumbnail image of

Immunocytochemistry

To evaluate blood-borne fibroblast phenotype, cells were grown on eight-chamber slides (Lab-Tec; Poly Labo, Strasbourg, France) and fixed with formalin at room temperature. Endogenous peroxidases were inactivated with 3% H2O2 in methanol. Nonspecific antibody binding was blocked by incubation with 3% bovine serum albumin. Primary antibodies against human CD14 (TUK4), CD31 (JC70A), CD34 (TUK3), CD117 (104D2), glial fibrillary acidic protein (GFAP) (6F-2), S-100 (all from DAKO, Glostrup, Denmark, http://www.dako.com), CD44 (Novocastra Ltd., Newcastle upon Tyne, U.K., http://www.novocastra.co.uk), CD45 (eBiosciences, San Diego, CA, http://www.ebioscience.com), collagen type I (Calbiochem, San Diego, http://www.emdbiosciences.com), and smooth muscle actin (1A4; Sigma-Aldrich) were used.

Animal Care and Lung Cancer Xenograft

Female nonobese diabetic/severe combined immunodeficient (NOD/SCID) mice were obtained from CLEA Japan, Inc. (Tokyo, http://www.clea-japan.com). In accordance with the Institutional Guidelines, all animals were maintained under specific-pathogen-free, temperature-controlled-air conditions throughout this study. Written approval of all animal experiments (K03-011) was obtained from the local Animal Experiment Committee of the National Cancer Center Research Institute. The NOD/SCID mice used in all of the experiments were 6–8 weeks of age. We subcutaneously inoculated NOD/SCID recipients with a transplantable human large cell neuroendocrine carcinoma of the lung (613LCNEC) as previously described [11]. Two weeks later, 1 × 107 of carboxyfluorescein succinyl ester (CFSE)-labeled mononuclear cells from a pulmonary vein were inoculated by cardiac puncture. Two days later, tumors were resected and snap-frozen until they were used in this study.

CFSE Labeling

Isolated mononuclear cells were labeled with CFSE (Dojindo, Kumamoto, Japan, http://www.dojindo.com). In brief, after incubation with CFSE (final concentration, 20M) at 37°C for 15 minutes, cells were washed twice in 10 ml of phosphate-buffered saline and then incubated in 10 ml of DMEM for 2–3 hours, with a medium change every 30 minutes.

Immunohistochemical and Immunofluorescence Analysis

Five-micrometer sections of frozen tissue from the cancer xenografts were fixed in 10% formaldehyde for 5 minutes, and endogenous peroxidases were inactivated with 3% H2O2 in methanol. Sections were incubated for 1 hour with primary antibody against HLA-class 1 (W6/32; Sigma-Aldrich) at a 1:10 dilution and then with the DAKO EnVision+System-HRP (DAKO). The reaction products were stained with diaminobenzidine.

For the immunofluorescence analysis, frozen sections were cut at 5 μm and air-dried for 5 minutes. Mouse monoclonal anti-α-smooth muscle actin (SMA)-Cy3 (1A4; Sigma-Aldrich) and anti-human HLA class 1 were used at dilutions of 1:400 and 1:10, respectively. Rabbit polyclonal anti-vimentin (H-84; reactive with human and mouse vimentin; Santa Cruz Biotechnology Inc., Santa Cruz, CA, http://www.scbt.com) was used at a 1:25 dilution. After the sections were washed, Alexa Fluor 488 goat anti-mouse IgG and Alexa Fluor 546 goat anti-rabbit IgG were used as the secondary antibodies. The sections were examined with an inverted microscope at an excitation wavelength of 488 nm for CFSE and Alexa Fluor 488, 543 nm for Alexa Fluor 546 and α-SMA-Cy3, and 633 nm for DRAQ5 (for nuclear staining). After nuclear staining with DRAQ-5 (Alexis Biochemical, Lorrach, Germany, http://www.alexis-corp.com), the sections were examined with an LSM5 PASCAL confocal imaging system (Carl Zeiss, Jena, Germany, http://www.zeiss.com). Confocal images were stored as digital files and viewed with Photoshop software (Adobe, Mountain View, CA, http://www.adobe.com).

Western Blotting

Cells were solubilized in lysis buffer (300 mM NaCl, 0.5% Nonidet P40, 1 mM EDTA, 20 mM HEPES-NaOH, pH 7.9, 50 μM leupepsin, 50 μM pepstain, 50 μM aprotinin, 10 mM sodium fluoride, 1 mM phenylmethylsulfonyl fluoride, 1 mM dithiothreitol, and 250 mM sodium orthovanadate). Protein lysates were separated by electrophoresis in 7.5% acrylamide/bisacrylamide (29:1) gels containing SDS and were transferred to nitrocellulose membranes. The membranes were incubated overnight at 4°C with antibody against vascular endothelial (VE)-cadherin (Santa Cruz Biotechnology), CD31 (DAKO), and Calponin (Novocastra) in Tris-buffered saline (TBS). The membranes were then washed with horseradish peroxidase-conjugated anti-mouse IgG in TBS. Antibody binding was detected with the Amersham enhanced chemiluminescence system.

Statistical Analysis

Data are expressed as means ± SE. Differences in measured variables between the experimental and control groups were assessed by Student's t test. Statistical calculations were performed on a Windows personal computer with GraphPad Prism software (GraphPad Software, Inc., San Diego, http://www.graphpad.com). p values <0.05 were considered statistically significant.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure of Potential Conflicts of Interest
  8. Acknowledgements
  9. References

Identification of Fibroblast Progenitor Cells in the Blood of Human Lung Cancer Patients

In 16 of the 47 cases, 1–2 fibroblastic colonies per 5 × 106 mononuclear cells were identified after 4–51 days of cul-ture (average, 23 ± 4 days) (Fig. 1A, 1B; Table 1). The adherent cells grew in a disarrayed fashion and were distinguished by their long projections, indicating fibroblast morphology (Fig. 1C, 1D). In some cases, stellate-shaped cells with thinner cytoplasm appeared web-like (Fig. 1E, 1F). Colonies became bigger and subsequently merged, resulting in the formation of a monolayer of typical spindle-shaped fibroblasts. After these cultures became confluent, the cells were split into two plates. However, these ex vivo outgrowth fibroblasts grew for up to three passages before senescence in all cases, and the passage number was unrelated to the age of the patient (Table 1).

thumbnail image

Figure Figure 1.. Morphological character and immunophenotype of blood-borne fibroblasts. (A): Colony formation after mononuclear cell seeding. At 1 week of culture, a single outgrowth colony was present at the center of the field. Many mononuclear cells were attached to the colony. (B): High-power view of (A). (C): The colonies grew larger, and the adherent spindle cells grew in a disarrayed fashion and were distinguished by their long projections, indicating fibroblasts. (D): High-power of view of (C). (E): The stellate-shaped cells with thinner cytoplasm had a web-like appearance. (F): High-power of view of (E). (G): Outgrowth fibroblasts from peripheral blood were universally positive for collagen type I and vimentin. Most of the outgrowth fibroblasts were negative for CD45 and CD31, but a small number were positive. (H): Western blotting of blood-borne fibroblasts. Blood-borne fibroblasts were negative for the specific endothelial markers VE-cadherin and CD31 and the muscle-specific marker Calponin. Human smooth muscle cells and HUVECs were used as positive controls for the specific smooth muscle marker Calponin and the endothelial markers VE-cadherin and CD31, respectively. WI-38 lung fibroblasts were negative for the endothelial markers but weakly positive for Calponin, implying myofibroblast differentiation. Abbreviations: Ex-Fb, ex vivo outgrowth fibroblast; HUVEC, human umbilical vein endothelial cell; SMC, smooth muscle cell; VE, vascular endothelial.

Download figure to PowerPoint

Table Table 2.. Summary of immunophenotype of ex vivo outgrowth fibroblasts
Thumbnail image of

Immunophenotype of Blood-Borne Fibroblasts

We performed immunocytochemistry analysis to evaluate the immunophenotype of the blood-borne fibroblasts. The fibroblasts from peripheral blood were universally positive for CD44, collagen type I, smooth muscle actin, and vimentin but negative for the monocyte marker CD14, stem cell markers CD34 and CD117, and neural markers GFAP and S-100 (Fig. 1G). A small number of cells were positive for CD45 and CD31, and they were intermingled with marker-negative cells (Fig. 1G). Table 2 summarizes the immunophenotypes of the blood-borne fibroblasts and two human lung fibroblast cells, MRC-5 and WI-38. Blood-borne fibroblasts exhibited an immunophenotype similar to that of two lung fibroblasts. To further confirm that the blood-borne fibroblasts differentiate into neither smooth muscle cells nor endothelial cells, cell lysates from each cell type were run on SDS-polyacrylamide gel electrophoresis and immunoblotted with the antibodies to the specific endothelial marker VE-cadherin and the muscle-specific marker Calponin. Blood-borne fibroblasts did not react with either the endothelial marker or smooth muscle marker (Fig. 1H).

Table Table 3.. cDNA microarray search for genes that are differentially expressed between blood-borne fibroblasts and tissue fibroblasts
Thumbnail image of

Comparison Between the Expression Profiles of Blood-Borne Fibroblast and Tissue- Derived Fibroblast

Genome-wide screening for genes with different expression patterns in blood-borne fibroblasts and tissue-derived (lung) fibroblasts was performed by using a microarray containing 54,675 probes. First, we conducted an unsupervised clustering analysis using various sets of genes. Representative results using 550 genes are shown in Figure 2. A gene was selected if more than five of nine nonimmortalized human mesenchymal cell samples, including HUVECs, HSMCs, human lung fibroblasts (WI-38 and MRC-5), and blood-borne fibroblasts (cases 5, 43, 45, 48, and 49) expressed, and if more than two of the nine samples showed more than threefold changes compared with an average expression level of each gene. The clustering algorithm separated the profiles into two distinct clusters. The first cluster included HUVECs, HSMCs, WI-38, and MRC-5. The second cluster contained all blood-borne fibroblasts. The profiles of blood-borne fibroblasts were the most isolated profiles from those of the WI-38 and MRC-5, suggesting that the blood-borne fibroblasts have expression profiles independent of the tissue-derived fibroblasts.

thumbnail image

Figure Figure 2.. Metaprofile comparison with hierarchical clustering algorithm. Trees represent the proportional distance between each metaprofile. MRC-5 and WI-38 are human lung fibroblasts; No.5, No.43, No.45, No.48, and No.49 are blood-borne fibroblasts. Abbreviations: HUVEC, human umbilical vein endothelial cell; SMC, smooth muscle cell.

Download figure to PowerPoint

Next, we selected 612 genes whose expression in all of the five blood-borne fibroblasts (cases 5, 43, 45, 48, and 49) was more than threefold up- or downregulated compared with the two lung fibroblasts (WI-38 and MRC-5) (247 upregulated and 365 downregulated genes). Of these 612 genes, those that changed more than 25-fold in all five blood-borne fibroblasts were shown in Table 3. They included genes that encode signal transduction (upregulated genes: sorbin and SH3 domain-containing 2 [SORBS2], filaggrin [FLG], etc.), transcription factors (upregulated genes: zinc finger protein 423 [ZNF423], etc.; downregulated gene: forkhead box F [FOXBF]), extracellular matrix components and their receptors (upregulated genes: cartilage oligomeric matrix protein 1 [COMP], hyaluronan and proteoglycan link protein 1 [HAPLN1], microfibrillar associated protein 5 [MFAP5], elastin [ELN], etc.), and enzyme (upregulated genes: metallophosphoesterase domain-containing 2 [MPPED2], hyaluronan synthase 1 [HAS1], etc.; downregulated gene: hydroxysteroid [17-β] dehydrogenase 2 [HSD17B2]). Since all of the genes are expected to express with a large difference between the two types of the fibroblasts, our extensive gene list provides important candidates for characterizing the blood-borne fibroblasts.

Fibroblast Progenitors in the Blood in the Vicinity of the Lung Cancer Are Recruited into the Cancer Stroma and Differentiate into Cancer- Stromal Fibroblasts

CFSE-labeled human mononuclear cells from pulmonary vein blood were injected into NOD/SCID mice that had been subcutaneously implanted with a transplantable human lung cancer cell line, and on day 2 after the injection, the mice were sacrificed. As shown in Figure 3A, many CFSE-labeled spindle cells were scattered around the cancer nests. When mononuclear cells unlabeled with CFSE were injected, on the other hand, no CFSE-positive cells were observed (Fig. 3B). Furthermore, anti- HLA class 1 antibody was used to identify the human fibroblasts (Fig. 3C–3E). When mice transplanted with a lung cancer xenograft were injected with medium alone, no HLA class 1-positive cells were found around the cancer nests (Fig. 3C). When human mononuclear cells from pulmonary vein blood were injected, however, a large number of HLA class 1-positive cells, including morphologically identified fibroblasts (Fig. 3E, arrows) and inflammatory cells (Fig. 3D, arrowhead), infiltrated into the cancer stroma. This was also confirmed by the immunofluorescence study. Figure 3F showed a result without primary antibodies (negative control study). When medium alone was injected, HLA class 1-positive spindle cells did not exist (Fig. 3G). After the injection of human pulmonary vein mononuclear cells, HLA class 1/vimentin or HLA class 1/α-SMA double-positive spindle cells in the cancer-induced stroma were found, indicating that these cells were fibroblasts or myofibroblasts (Fig. 3H, 3I, respectively). Thus, peripheral blood in the vicinity of lung cancers contains fibroblast progenitor cells that have the capacity to migrate into cancer stroma and differentiate into fibroblasts.

thumbnail image

Figure Figure 3.. Fibroblast progenitors in the blood in the vicinity of lung Ca were recruited into Ca stroma and differentiated into fibroblasts. (A): CFSE-labeled human mononuclear cells from pulmonary vein blood were injected into NOD/SCID mice that had been subcutaneously implanted with a human transplantable lung Ca cell line (613LCNEC). The tumor was removed 2 days after injection and analyzed. CFSE-labeled spindle cells were scattered around the Ca nests. The upper left panel shows CFSE fluorescence. The lower left panel shows cells stained with DRAQ5 to identify nucleated cells. The lower right panel shows a composite of both fluorophores. Inset, high-power view of merged image. (B): No CFSE-positive cells were observed when mononuclear cells without CFSE labeling were injected. (C): When mice transplanted with a lung Ca xenograft were injected with medium alone (negative control), no HLA class 1-positive cells could be found around the Ca nests. Note that the human lung Ca xenografts showed a strong positive reaction for HLA class 1. (D): After injection with human mononuclear cells, a large number of HLA class 1 positive cells infiltrated the Ca stroma. The boxed area contains HLA class 1 positive fibroblasts. The arrowhead points to human inflammatory cells. (E): High-power view of the boxed area in (D). Arrows point to fibroblasts identified morphologically. (F): Negative control study without primary antibodies. The upper left and upper right panels show the results of Alexa Fluor 488 and Alexa Fluor 546 fluorescence, respectively. The lower left panel shows cells stained with DRAQ5 to identify nucleated cells. The lower right panel shows a composite of both fluorophores. (G): When mice that had been transplanted with a lung Ca xenograft were injected with medium alone (negative control), no HLA class 1 positive cells could be found around the Ca nests. Note that the human lung Ca xenografts showed positive reaction for HLA class 1 and vimentin. The upper left panel shows HLA class 1 fluorescence. The upper right panel shows cells immunostained with vimentin antibody in the same area. The lower left panel shows cells stained with DRAQ5 to identify nucleated cells. The lower right panel shows a composite of both fluorophores. (H): After injection with human mononuclear cells, HLA class 1/vimentin double-positive cells were found around the Ca nests. The upper left panel shows HLA class 1 fluorescence. The upper right panel shows cells immunostained with vimentin antibody in the same area. The lower left panel shows cells stained with DRAQ5 to identify nucleated cells. The lower right panel shows a composite of both fluorophores. (I): Colocalization of HLA class 1/α-SMA on fibroblasts. The upper left panel shows HLA class 1 fluorescence. The upper right panel shows cells immunostained with α-SMA antibody in the same area. The lower left panel shows cells stained with DRAQ5 to identify nucleated cells. The lower right panel shows a composite of both fluorophores. Abbreviations: Ca, cancer; CFSE, carboxyfluorescein succinyl ester.

Download figure to PowerPoint

In Blood-Borne Fibroblast-Positive Cases, a Significantly Larger Area of Central Collapse Fibrosis Within the Primary Lung Cancer Was Observed

Table 4 shows the correlations between clinicopathological factors and blood-borne fibroblast-positive cases. The central collapse fibrosis size in the lung cancer in the blood-borne fibroblast-positive cases measured 123 ± 29 mm2, as opposed to 59 ± 13 mm2 in the negative cases, and the difference was significant (p < .05). However, there were no significant correlations with any other clinicopathological factors, including age, tumor volume, or vascular invasion.

Table Table 4.. Correlation between ex vivo outgrowth fibroblast-positive cases and clinicopathological factors
Thumbnail image of

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure of Potential Conflicts of Interest
  8. Acknowledgements
  9. References

In this study, we showed clearly that fibroblast progenitor cells are present in the blood in the vicinity of human lung cancers. Recent reports have revealed that peripheral blood contains several mesenchymal cell progenitors [19, [20], [21], [22], [23], [24], [25], [26], [27]28], for example, progenitors for smooth muscle cells [19, 20], vascular endothelial cells [21, [22], [23]24], and fibrocytes [13, [14]15, 25]. These progenitor cells have the ability to migrate from bloodstream and differentiate into functional mesenchymal cells in the appropriate microenvironments. Blood-borne fibroblasts would be different from the circulating fibrocytes, monocyte-derived mesenchymal progenitor (MOMP) [29], smooth muscle cells, and endothelial cells. Blood-borne fibroblasts are CD14− and CD34−, whereas circulating fibrocytes and MOMP have been reported to be positive for both markers. The blood-borne fibroblasts grew in a disarrayed fashion and were distinguished by their long projections, whereas smooth muscle progenitor cells and endothelial progenitor cells are known to exhibit “hill and valley” and “cobblestone” morphology, respectively. Furthermore, blood-borne fibroblasts do not react with specific antibodies against myogenic and endothelial markers.

The blood-borne fibroblast-positive cases had a significantly larger central collapse fibrosis size in their primary lung cancer than the negative cases. This finding may indicate that more fibroblast progenitor cells were present in the blood of the positive cases than in the blood of the negative cases. Alternatively, there may be one or more leukocyte subpopulations that affect the differentiation of fibroblast progenitor cells via direct contact or by releasing secretary factors, and the proportions of the leukocyte subpopulations may differ in positive and negative cases. We collected peripheral mononuclear cells from 10 healthy volunteers and cultured them under the same conditions; however, we were unable to find blood-borne fibroblasts in any of them (data not shown). These results also suggest that the blood microenvironment in the vicinity of cancer, which is also influenced by the pathophysiological characteristics of the primary cancer, is quite different in each case.

Normal human cells undergo a limited number of cell divisions in culture before entering a nondividing state called senescence, and the average number of population doublings has been found to be approximately 40 [30, [31]32]. Blood-borne fibroblasts were passaged only 1–3 times (corresponding to ∼20 population doublings) before reaching premature senescence. Although exactly the same may not occur in vivo, it is speculated that blood-borne fibroblasts in a cancer-induced stroma have a short life span. Since our previous studies showed that only 2.2% of bone marrow-derived fibroblasts displayed proliferative activity in the cancer-induced stroma [9], blood-borne fibroblasts may constantly be recruited from peripheral blood; however, most of these cells are quiescent rather than rapidly growing.

Surprisingly, the cDNA profiles of the blood-borne fibroblasts were the most separate profiles from those of the human lung fibroblasts WI-38 and MRC-5. This means that blood-borne fibroblasts are a distinct subpopulation within the cancer-induced stroma and play a role in cancer progression. As shown in Table 3, many genes involved in extracellular matrix (ECM) components were found to be upregulated in blood-borne fibroblasts as compared with tissue-derived fibroblasts. Elastin is the main component of elastic fibers, and after interactions with the S-Gal elastin receptor, elastin-derived peptides, which are degradation products of elastin, promote melanoma cell invasion through a three-dimensional type I collagen matrix by upregulating MMP-2 activation [33, [34]35]. Hyaluronan (HA) performs a variety of functions, including space filling, joint lubrication, and provision of a matrix through which cells can migrate. HASs are plasma membrane enzymes synthesizing HA, and overexpression of HASs has been correlated with tumor aggressiveness [36, [37]38]. Higher expression of Elastin and HAS genes in blood-borne fibroblasts may mean that dynamic migration of these cells into the tumor stroma is capable of exerting a dramatic impact on the growth and progression of malignant cells in vivo.

Some investigators may suspect that blood-borne fibroblasts are derived from vascular wall cells [39, 40] contaminated during the needle insertion procedure. However, we think that ex vivo outgrowth fibroblasts in the current study were derived from fibroblast precursor cells in the blood, for the following reasons. First, even when we opened a ligated pulmonary vein with scissors and directly collected the blood without inserting the syringe, we observed outgrowth fibroblasts on the culture dish. Second, when we cultured vascular wall cells, they showed a capacity for extended growth in culture (>10 passages) (data not shown), in contrast to blood-borne fibroblasts, which become senescent after >1–3 passages. Taken together, these findings make it extremely unlikely that the blood-borne fibroblasts in this study were derived from contaminating adult vascular wall cells.

In conclusion, we here report finding that human peripheral blood in the vicinity of lung cancer contains fibroblast progenitor cells that have the capacity to migrate into cancer stroma and differentiate into stromal fibroblasts. Recruited blood-borne fibroblasts produce several growth factors and ECMs, which may influence cancer cell proliferation. It is important to investigate the role of blood-borne fibroblasts in cancer progression, and the results should improve our understanding of the complex constituents of cancer tissue and open new possibilities for the diagnosis and treatment of cancer.

Disclosure of Potential Conflicts of Interest

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure of Potential Conflicts of Interest
  8. Acknowledgements
  9. References

The authors indicate no potential conflicts of interest.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure of Potential Conflicts of Interest
  8. Acknowledgements
  9. References

We are grateful to Yoko Okuhara, Chie Okumura, Hiroko Hashimoto, and Mai Okumoto for technical support and to Motoko Suzaki for preparing the manuscript. H.F. is a recipient of a Research Resident Fellowship from Foundation for Promotion of Cancer Research. This work was supported in part by a Grant-in-Aid for Cancer Research from the Ministry of Health, Labour and Welfare; a Grant for Scientific Research Expenses for Health Labour and Welfare Programs; the Foundation for the Promotion of Cancer Research, 3rd-Term Comprehensive 10-Year Strategy for Cancer Control; the program for promotion of Fundamental Studies in Health Sciences of the National Institute of Biomedical Innovation; and Special Coordination Funds for Promoting Science and Technology from the Ministry of Education, Culture, Sports, Science and Technology of Japan.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure of Potential Conflicts of Interest
  8. Acknowledgements
  9. References