Integrin αvβ6 plays a bi-directional regulation role between colon cancer cells and cancer-associated fibroblasts

Almost 130 years ago, Paget [1] put forth the famous ‘seed and soil’ theory to explain the metastatic preference of cancer cells for specific organs. The cancer cells were considered the ‘seeds’, and the specific organ microenvironment was the ‘soil’; interactions between the ‘seeds’ and ‘soil’ were proposed to determine the formation of a secondary tumor. With subsequent research, the concept of the tumor microenvironment (TME) gradually gained empirical support [2,3]. The TME is composed of many types of cells and proteins, such as fibroblasts, inflammatory cells, immune cells, vascular endothelial cells, mesenchymal stem cells (MSCs), and extracellular macromolecules including collagen, laminin, and protein polysaccharides. The TME does not only function as a scaffold to support tissue architecture and integrity but also secretes cytokines, growth factors, and protease.

Fibroblasts are the main cellular components of the TME; they synthesize collagen and other extracellular matrix (ECM) proteins, which provide the structural framework (stroma) for animal tissues. Under normal conditions, fibroblasts are inactive (fibrocytes) in the resting state and are involved in maintenance and tissue metabolism [4]. They are activated when tissues are injured, and play a role in processes such as wound healing and inflammation. Usually, fibroblasts not only synthesize components of the ECM but also secrete proteases. When they are recruited to the TME, cells within the tumor transform the surrounding stromal fibroblasts into cancer-associated fibroblasts (CAFs) [5]. CAFs differ from normal fibroblasts (NFs) in appearance and function [6]. They can survive in the TME for a fairly long period of time, and do not undergo normal apoptosis [7]. CAFs are derived from several cell types. When activated by cells in the surrounding stroma, CAFs can secrete many types of growth factors, cytokines, and proteases, which promote tumor cell proliferation, invasion, and angiogenesis [8]. Activated CAFs express specific proteins, such as α-smooth muscle actin (α-SMA) and fibroblast-activating protein (FAP) [9].

Our previous study showed that αvβ6 is a special integrin subtype that is only expressed in epithelial cells; its major ligand is fibronectin. In normal epithelial cells, αvβ6 can hardly be detected [10], but its expression significantly increases in response to injury and/or inflammation as well as in epithelial tumors, such as those occurring in gastric and colon cancers (colorectal cancers, CRCs) [11–13]. As a cell adhesion molecule, integrin αvβ6 mediates cell–cell and cell–ECM events through binding ECM proteins [14]. Integrin αvβ6 also participates in the epithelial–mesenchymal transition (EMT) of CRC cells, and the αvβ6-ERK-Ets1 signaling pathway plays an important role in this process [15,16].

In the present study, the human CRC cell lines HT-29 and RKO were used to investigate whether integrin αvβ6 mediates the relationship between cancer cells and CAFs, and to determine the underlying mechanisms.

The human normal colonic fibroblast cell line CCD-18Co and human CRC cell lines HT-29, RKO, Caco-2, SW480, LoVo, and SW620 were obtained from the American Type Culture Collection (Gaithersburg, MD, U.S.A.). Stable transfectants of RKO cells expressing pcDNA1neo constructs containing wild-type β6 or HT-29 cells expressing β6 siRNA were prepared as previously reported [16]. Normal human colonic fibroblast cell line CCD-18Co were cultured in fibroblast medium supplemented with 2% FBS (Sigma, St. Louis, MO, U.S.A.), 1% fibroblast growth supplement, and 1% penicillin/streptomycin solution. The CRC cell lines were maintained as monolayers in standard Dulbecco’s modified Eagle’s medium (4.5 g/l glucose; Sigma) with 10% heat-inactivated FBS, 20 mM HEPES, 100 IU/ml penicillin, and 100 μg/ml streptomycin (Merck, Darmstadt, Germany). The cells were incubated at 37°C with 5% CO₂ and saturated humidity. Cells were subcultured until they were more than 90% confluent. Cells at passages 3–6 were used for experiments.

Indirect co-culture of fibroblasts with cancer cells were used in the study to activate the fibroblast as previously described and modified based on the preliminary experiments [17,18]. CCD-18Co cells (3.0 × 10⁵) were seeded in six-well culture plate. Colon cancer cells (3.0 × 10⁵) were seeded in a 0.4-μm polyester membrane of a 24-mm transwell insert and placed in a separate culture plate. The next day, the transwell insert was transferred to the culture plate containing fibroblast cells, and co-cultured for 96 h. Co-cultured cells were cultured in DMEM/F12 medium (10% FBS), change medium every 48 h, and were incubated at 37°C with 5% CO₂ and saturated humidity. In some situation, fibroblast cells were seeded in the insert while colon cancer cells were cultured in the well according to the purpose of the experiments.

The mouse-anti-human monoclonal antibody R6G9 (IgG_2a), which is directed against the extracellular domain of human integrin subunit β6, was obtained from Chemicon International (Temecula, CA, U.S.A.). The following monoclonal antibodies were obtained from Abcam (Cambridge, MA, U.S.A.): ab5694 and ab28244, which target the CAF markers α-SMA and FAP, respectively: ab27969, which targets transforming growth factor β (TGF-β); and ab9797, ab9695, ab16828, and ab6672, which target the four cytokines stromal cell-derived factor-1 (SDF-1), epidermal growth factor (EGF), basic fibroblast growth factor (bFGF), and interleukin 6 (IL-6). EP1186Y and EP1254 monoclonal antibodies, which target matrix metalloproteinase (MMP) 3 (MMP-3) and MMP-9, respectively, were also purchased from Abcam. Reagents for SDS/PAGE and molecular weight markers were obtained from Bio-Rad Laboratories (Hercules, CA, U.S.A.). The C–X–C chemokine receptor type 4 axis (CXCR4) antagonist AMD3100 was purchased from Santa Cruz Biotechnology (Santa Cruz, CA, U.S.A.).

Total cellular RNA was extracted using TRIzol reagent (Sigma), and cDNA was synthesized according to the manufacturer’s instructions (Promega, Madison, WI, U.S.A.). Equal amounts of cDNA were subjected to PCR analysis. The primer sequences were as follows: β6 integrin forward: 5′-AGGATAGTTCTGTTTCCTGC-3′ and β6 integrin reverse: 5′-ATCATAGGAATATTTGGAGG-3′, which generated a 141-bp amplicon; α-SMA forward: 5′-CGTGGCTACTCCTTCGTG-3′ and α-SMA reverse: 5′-TGATGACCTGCCCGTCT-3′,which generated a 160-bp amplicon; FAP forward: 5′-TCAACTGTGATGGCAAGAGCA-3′ and FAP reverse: 5′-TAGGAAGTGGGTCATGTGGGT-3′, which generated a 219-bp amplicon; GAPDH forward: 5′-GTCAGTGGTGGACCTGACCT-3′ and GAPDH reverse: 5′-TGAGGAGGG GAGATTCAGTG-3′, which generated a 400-bp amplicon and was used as the internal control. The amplification conditions were 33 cycles at 94°C for 30 s, 51°C for 30 s, and 72°C for 30 s. The PCR products were electrophoresed on a 1.5% agarose gel and detected by Ethidium Bromide staining. Data were analyzed with AlphaImager software (Alpha Innotech Co., San Leandro, CA, U.S.A.).

CCD-18Co cells were co-cultured with various CRC cells for 96 h. Both adherent and floating cells were collected and frozen at −80°C. To detect the levels of potentially mechanistic proteins, these proteins were extracted from the cells for the determination of concentration. The protein concentration was measured using the BCA Protein Assay Reagent, and equal amounts of protein (10 μg) were loaded on to a 12.5% SDS/PAGE gel and electrophoresed under non-reducing conditions. After electrophoresis, the proteins were transferred to nitrocellulose membranes. Equivalent protein loading in each lane was confirmed by staining the nitrocellulose membranes with Ponceau-S Stain using the 42-kDa β-actin band as a reference marker. Then, the membranes were probed with primary monoclonal antibodies against key signaling pathway factors followed by peroxidase–conjugated secondary antibodies. Proteins on the membranes were visualized using the Enhanced Chemiluminescence Detection System (Amersham, Aylesbury, U.K.) according to the manufacturer’s instructions. Optical density was analyzed with ImageJ software.

MMP-3 and MMP-9 levels in the conditioned medium obtained from co-cultured cells were assayed using the commercially available kit Biotrak MMP Activity Assay System (Amersham, Aylesbury, U.K.). This assay measured total MMP levels (inactive pro-enzyme artificially activated and endogenous active enzyme), and MMP secretion was calculated on a per-cell basis.

The levels of TGF-β1 and SDF-1 were detected using the appropriate enzyme-linked immunoassay (ELISA) kit (R&D systems, Minneapolis, MN, U.S.A.) according to the manufacturer’s instructions.

The invasive ability of the CRC cells was analyzed in 24-well Boyden chambers with polycarbonate membranes (8-μm pore size; CoStar, Boston, MA, U.S.A.). The membranes were pre-coated with 50 μl Matrigel (BD Biosciences, San Diego, CA, U.S.A.) to form matrix barriers. Colon cancer cells were re-suspended in 100 μl serum-free medium at a concentration of 1 × 10⁶ cells/ml, and added to the upper chamber; 1 × 10⁵ fibroblast cells were seeded in the lower compartments with 600 μl medium containing 10% FBS. After incubation, the cells remaining on the upper surface of the membrane were removed. Cells on the lower surface of the membrane were fixed and stained with Crystal Violet, and counted under a light microscope at 200× magnification.

The results were expressed as the mean ± S.D. Comparisons between the two groups were performed with the Student’s ttest. Comparisons of multiple samples and rates were performed using the single-factor ANOVA. All statistical analyses were performed using SPSS for Windows version 16.0. P-values less than 0.05 were considered statistically significant.

To investigate the expression of β6 in CRC cell lines, we selected six types of CRC cell lines and performed RT-PCR to detect β6 mRNA levels. The results showed that β6 mRNA expression was high in HT-29, Caco-2, Lovo, and SW620 CRC cell lines, with the highest expression observed in HT-29 cells and the lowest expression found in RKO cells (Figure 1A). To investigate the effects of these CRC cells on CCD-18Co fibroblasts, we co-cultured them with CCD-18Co fibroblasts for 96 h. Next, we performed RT-PCR to detect the mRNA levels of α-SMA and FAP. The results of these assays showed that α-SMA and FAP mRNA levels in CCD-18Co fibroblasts varied according to the type of CRC cell line. The mRNA level of α-SMA was tightly correlated with β6 expression and exhibited the same expression pattern, as shown in Figure 1B. Similar results were observed with FAP mRNA expression (Figure 1C).

To avoid the individual difference between NFs and CAFs used in the study effecting the result of transwell experiments, invasion experiment was done with CAFs activated by cancer cells and those without cancer cellls pretreatment. There was no difference observed between NFs and CAFs (Figure 1D).

To investigate the relationship between β6 expression in CRC cells with the fibroblast markers α-SMA and FAP, we selected HT-29 and RKO cells, which had the highest and lowest expression levels of β6, respectively. We established β6 knockdown HT-29 cells (siβ6) via siRNA technology and β6 overexpressing RKO cells (β6 overexpression) via plasmid transfection. Meanwhile, we also established β6 siRNA negative control HT-29 cells (siNC) and mock plasmid transfection RKO cells (Mock). Then CCD-18Co fibroblasts were co-cultured with these CRC cells for 96 h, followed by RT-PCR and Western blotting to detect the mRNA and protein expression of α-SMA and FAP, respectively, in the fibroblasts. In β6 knockdown cells, the decreased expression of β6 was accompanied by a significant decrease in α-SMA and FAP mRNA expression in CCD-18Co fibroblasts (*P<0.05 and **P<0.01, respectively). Accordingly, the overexpression of β6 in RKO CRC cells resulted in a significant increase in α-SMA and FAP mRNA expression in CCD-18Co fibroblasts (**P<0.01 and ***P<0.001, respectively), as shown in Figure 2A,B. The same expression patterns of α-SMA and FAP were observed at the protein level (Figure 2C,D).

TGF-β is secreted by many cell types in a latent form in which it is complexed with two other polypeptides: latent TGF-β-binding protein and latency-associated peptide (LAP), a protein derived from the N-terminal region of the TGF-β gene product. To determine if TGF-β can be activated by integrin αvβ6, we incubated 10 μg TGF-β-LAP with the CRC cell lines for 24 h. Then, we collected the cell culture media for use in ELISA. We found that active TGF-β levels significantly decreased in β6 HT-29 knockdown cells (*P<0.05) but increased in RKO cells overexpressing β6 plasmid (*P<0.05), as shown in Figure 3A,B, respectively.

We treated CCD-18Co fibroblasts with 10 μg/ml TGF-β1 for 24 h, and subsequently performed RT-PCR and Western blotting to detect the mRNA and protein levels, of α-SMA and FAP, respectively. The dose of TGF-β was based on previous studies and our preliminary experiments [19–21]. Treatment with TGF-β1 significantly increased the mRNA and protein levels of α-SMA and FAP (**P<0.01), as shown in Figure 3C,D, respectively.

To investigate the role of integrin αvβ6 in regulating active fibroblasts, we co-cultured CCD-18Co fibroblasts with the above-mentioned four types of CRC cells for 96 h, and followed by RT-PCR and Western blotting to detect the mRNA and protein expression, respectively, of α-SMA and FAP in fibroblasts. The results showed a positive correlation between β6 expression in CRC cells, and the mRNA levels of α-SMA and FAP in CCD-18Co fibroblasts, as shown in Figure 3E,F. This correlation was also observed at the protein level (Figure 3G).

To investigate the role of integrin αvβ6 and active CAFs in regulating cancer invasion, we incubated the co-culture of active fibroblasts CCD-18Co (CAFs) or NFs with the above-mentioned four types of CRC cells in upper and lower chamber of the transwell migration assay system for 24 h. Then, we observed the cell morphology and counted the number of cells in the lower chamber. The results showed that in HT-29 colon cancer cells and the siβ6 knockdown group, integrin αvβ6 promoted CRC cell invasion in the absence or presence of co-cultured CAFs, and the number of the invasive cancer cells significantly increased (**P<0.01 and ***P<0.001, respectively). Likewise, CAFs induced cell invasion when co-cultured with CRC cells with high or low β6 expression (*P<0.05), as shown in Figure 4A. These results were confirmed in RKO CRC cells and the β6 overexpression group (*P<0.05, **P<0.01, ***P<0.001, respectively), as shown in Figure 4B.

To further investigate the role of integrin αvβ6 and active CAFs in the regulation of cancer cell invasion, we co-cultured active fibroblasts CCD-18Co (CAFs) or NFs with the above-mentioned four types of CRC cells for 24 h. Then, we collected the cell culture media and performed MMP ELISA. Integrin αvβ6 induced secretion of MMP-3 in the absence or presence of co-cultured CAFs (*P<0.05); meanwhile, the CAFs also significantly induced MMP-3 secretion from CRC cells in the presence (**P<0.01) or absence (*P<0.05) of β6 expression, as shown in Figure 4C. Integrin αvβ6 also induced secretion of MMP-9 in the absence or presence of co-cultured with CAFs (*P<0.05); the CAFs also significantly induced MMP-9 secretion from HT-29 CRC cells and RKO β6 overexpression CRC cells (**P<0.01 and ***P<0.001, respectively). This type of induction effect was even found in siβ6 CRC cells (*P<0.05), as shown in Figure 4D.

To investigate whether the cytokines secreted by CAFs are responsible for the progression of CRC cells, we co-cultured the CCD-18Co fibroblasts with CRC cells for 96 h. Then, we collected the culture medium and performed ELISA. We measured the levels of SDF-1, EGF, bFGF, and IL-6 and found that only SDF-1 in HT-29 cells and CCD-18Co co-culture medium was significantly increased (**P<0.01); there were little changes in expression of the other three cytokines (Figure 5A,B). To determine if integrin αvβ6 influences the expression of SDF-1, we collected cell culture media from the above-mentioned CRC cells after they were co-cultured with CCD-18Co for 96 h. The results showed that integrin αvβ6 increased the secretion of SDF-1 by CAFs, which was decreased upon β6 knockdown by siRNA (**P<0.01; Figure 5C,D).

To investigate if the SDF-1/CXCR4 axis was involved in CAF-mediated induction of CRC cell progression, we co-cultured HT-29 or RKO CRC cells with NFs or active CAFs for 24 h, followed by the transwell invasion assay. The number of invasive CRC cells significantly increased when co-cultured with CAFs (**P<0.01). However, when the co-cultured cells were pretreated with AMD3100 (a CXCR4 antagonist), these effects almost disappeared (Figure 5E,F).

Malignant tumors are regarded as the result of an imbalance between tumor cells and normal cells [22], and the TME plays a very important role in this process. The TME is a complex environment for the survival of tumor cells and is composed of multiple proteins and cell types. The TME not only plays a supporting role for various cells, including tumor cells, immune cells, vascular endothelial cells, and fibroblasts but also forms the ECM (including collagen, laminin, and proteoglycan complexes) and secretes multiple cytokines, such as SDF-1 [23], IL-6 [24], bFGF [25], and EGF [26]. The cancer cell can re-construe the TME; moreover, the re-construed TME can also further influence cancer cell behaviors and conditions, which illustrates the cross-talk between cancer cells and the TME [27].

Fibroblasts comprise a large proportion of cells within the TME. When they are recruited to the malignant tumor zone, they become transformed into CAFs, which are sometimes called reactive stromal fibroblasts, myofibroblasts, or activated fibroblasts [28]. CAFs are mainly derived from three kinds of cells. The first type is the stromal fibroblast, which is thought to be the primary source of CAFs via the process of transdifferentiation of fibroblasts to myofibroblasts [29]. Multiple growth factors secreted by cancer cells induce the transformation of fibroblasts during this process [30,31]. Second, some cancer cells can be transformed into MSCs during the EMT, which can further differentiate into CAFs [32]. Third, some CAFS are derived from endothelial cells via the endothelial–mesenchymal transition [33]. A small proportion of CAFs are also derived from other cell types.

FAP is one of the important marker proteins of activated CAFs. In normal NFs, FAP expression can hardly be detected, but its expression increases when the fibroblasts are activated [34]. α-SMA is another marker of activated fibroblasts [35], and α-SMA-positive myofibroblasts are the main constituents of the cell matrix. CAFs play a very important role in regulating the progression of cancer. When they make contact with different tumors, they secrete multiple cytokines that promote tumor cell proliferation. Matsumoto and Nakamura [36] found that activated CAFs can release the hepatocyte growth factor, which mediates tumor–stromal interactions. Wang et al. [37] found that stromal fibroblasts can produce EGF in lung cancer and combine with the EGFR in lung cancer cells to induce drug resistance. Yu et al. [38] found that stromal fibroblasts can be transformed into CAFs by paracrine TGF-β signaling, leading to the induction of breast cancer cell proliferation; our data are in accordance with these results. SDF-1 also called CXCL-12, is the ligand of CXCR4, both of which belong to the CXC chemokine family [39]. Studies from our group and others found that the SDF-1/CXCR4 axis plays a significant role in the progression and metastasis of CRC cells [40,41]. Orimo et al. [42] confirmed that after CAFs are activated in human breast cancer, they secrete SDF-1 to promote cancer cell progression via the SDF-1/CXCR4 axis.

In our study, we found that CRC cells induce a morphology change in fibroblasts isolated in the TME, similar to myofibroblasts (data not shown). Compared with negative integrin αvβ6 expression CRC cells, positive integrin αvβ6 expression in CRC cells accelerate this process. Activated fibroblasts expressed proteins such as α-SMA and FAP, which are typical molecular marker of CAFs. Further investigations are needed to determine the underlying mechanisms. All types of CRC cells in the TME can secrete TGF-β-LAP. Although TGF-β with LAP is an inactive form, integrin αvβ6 can activate TGF-β by degrading LAP. Activated TGF-β can further induce fibroblast activation; subsequently, activated CAFs can induce CRC cell invasion. However, this induction effect was strengthened when CRC cells expressed integrin αvβ6, and it could be inhibited by the CXCR4 antagonist AMD3100. Hence, these data suggest that CAFs induce colon cancer progression via the SDF-1/CXCR4 axis.

Overall, the results of the present study showed that integrin αvβ6 plays a role in the bi-directional regulation of CRC cells and CAFs. Thus, drugs targetting integrin αvβ6 may be potential therapeutic targets for epithelial cell cancers expressing integrin αvβ6.

This work was supported by the National Natural Sciences Foundation of China [grant number 81572414]; and the Natural Sciences Foundation of Shandong Province [grant number ZR2015HM042].

The authors declare that there are no competing interests associated with the manuscript.

J.N. was responsible for designing the study and critical review of manuscript. C.P. and X.Z. were responsible for designing and performing the study, literature research, and manuscript drafting. W.X., H.G., and Z.L. participated in the cell culturing and statistical analysis. N.L., Z.X., and C.G. participated in the design of the study and sequence alignment. Z.H., W.N., and R.F. participated in sequence alignment and immunoassays. M.A. participated in the designing of the study. S.B. and X.Z. participated in the language editing. All authors approved the final version of the manuscript.

α-SMA

α-smooth muscle actin

bFGF

basic fibroblast growth factor

CAF

cancer-associated fibroblast

CRC

colorectal cancer

CXCR4

C–X–C chemokine receptor type 4

ECM

extracellular matrix

EGF

epidermal growth factor

EMT

epithelial–mesenchymal transition

FAP

fibroblast-activating protein

IL-6

interleukin 6

LAP

latency-associated peptide

MMP

matrix metalloproteinase

MSC

mesenchymal stem cell

NF

normal fibroblast

SDF-1

stromal cell-derived factor-1

TGF-β

transforming growth factor β

TME

tumor microenvironment