PI3K–Akt signaling is critical for the development, progression, and metastasis of malignant tumors, but its role in the tumor microenvironment has been relatively little studied. Here, we report that the Akt substrate Girdin, an actin-binding protein that regulates cell migration, is expressed and activated by Akt phosphorylation in cancer-associated fibroblasts (CAF) and blood vessels within the tumor microenvironment. Lewis lung tumors grafted into mice defective in Akt-mediated Girdin phosphorylation (SA transgenic mice) exhibited a decrease in both CAF infiltration and tumor growth, compared with wild-type (WT) host control animals. Contrasting with the findings of other studies, we found that Akt-dependent phosphorylation of Girdin was not a rate-limiting step in the growth of endothelial cells. In addition, Lewis lung tumors displayed limited outgrowth when cotransplanted with CAF derived from tumor-bearing SA transgenic mice, compared with CAF derived from tumor-bearing WT mice. Collectively, our results revealed a role for Akt-mediated Girdin phosphorylation in CAF during tumor progression, highlighting the need to inhibit Akt function in both tumor cells and cells that comprise the tumor microenvironment. Cancer Res; 75(5); 813–23. ©2015 AACR.

Alterations of key genes and signal transduction pathways are central to the development and progression of malignant tumors (1). Although these driver mutations events are commonly associated to the cancer's cell-of-origin, recent studies have shown that the tumor's microenvironment can also play a significant role in malignant progression (2–5). Tumor cells recruit a diverse array of cell types to their surrounding stroma, including cancer-associated fibroblasts (CAF), endothelial cells, and pericytes that constitute tumor vessels, immune cells such as tumor-associated macrophages (TAM), and adipocytes, leading to the formation of highly complex neoplastic tissues (6–10). In turn, the tumor microenvironment facilitates tumor progression and metastasis by providing a matrix for the integration of intricate interaction networks suitable to maintain and nourish tumor cells, as well as suppress the normal immunologic antitumor defenses (1, 11–13). The supportive features provided by stromal cells populating the microenvironment increase tumor tissue diversity, making it difficult to treat patients with cancer uniformly.

CAF originating from fibroblasts, bone marrow–derived fibrocytes, and mesenchymal stem cells can localize to the tumor microenvironment and promote tumorigenesis through multiple mechanisms (5, 14, 15). CAF secrete numerous oncogenic growth factors, chemokines, and cytokines and degrade extracellular matrix (ECM) proteins, thereby promoting both tumor proliferation and invasion (5, 11, 16, 17). CAF physically assist in the invasion of surrounding tissue by actively remodeling the ECM, subsequently providing routes that can be exploited by tumor cells (14, 18). On the basis of these findings, CAF gene-expression assays have been developed as useful prognostic indicators in recent clinical studies (19–23). Clearly, it is important to dissect the mechanisms mediating the localization, proliferation, and activity of CAF, as such insights may lead to the development of novel therapies that target the tumor microenvironment (24).

Recently, the mechanisms of CAF development and differentiation have attracted much attention (5). These studies have revealed that mechanical forces regulated by ECM stiffness, TGF-β1, and other soluble factors secreted from tumor cells act in synergy with TGF-β1 signaling as a critical factor in CAF development (5, 25, 26). However, relatively limited evidence addresses the question of how intracellular signaling pathways regulate the dynamic cellular activities in CAF. For instance, the activation of Janus kinase 1 by proinflammatory cytokines leads to Rho kinase–dependent signaling, which may induce ECM remodeling, and increase CAF contractility and motility (5, 27). However, other intracellular signaling pathways that define cytoskeletal control and migratory responses in CAF remain to be elucidated.

Here, we report that the actin-binding protein Girdin (girders of actin filaments; also known as Gα-interacting vesicle-associated protein) is expressed and phosphorylated in CAF in both clinical cancer tissues and tumors in animal models. Girdin is phosphorylated by Akt, which functions downstream of PI3K activation to regulate cell migration and polarization (28–32). In this study, we used a knock-in mouse line (designated SA mice) engineered to express a Girdin mutant that lacks the Akt phosphorylation site at Serine 1416 (S1416A). We found that the growth of allogeneic tumors was decreased in SA mice compared with wild-type (WT) counterparts. Notably, the cotransplantation of tumor cells with either skin fibroblasts or CAF derived from SA mice also attenuated tumor growth, further indicating the in vivo relevance of Girdin phosphorylation in formation of the tumor microenvironment. To our knowledge, although clinical relevance is yet to be clearly defined, these findings demonstrate that the PI3K–Akt pathway is important for the cellular behavior of CAF and suggest that inhibiting PI3K–Akt signaling not only in tumor cells, but also CAF, could improve the efficacy of tumor therapy.

Antibodies and reagents

Antibodies used in this study were purchased from: polyclonal rabbit anti-human Girdin (Santa Cruz Biotechnology); polyclonal rabbit anti-human Girdin and polyclonal rabbit anti-human phospho (1416-Ser)-Girdin (IBL, Gunma, Japan); polyclonal sheep anti-human Girdin (R&D Systems); monoclonal mouse anti-human α-smooth muscle actin (α-SMA) and monoclonal mouse anti-human von Willebrand Factor (vWF; Dako); monoclonal anti-mouse phospho (473-Ser)-Akt (Cell Signaling Technology); monoclonal mouse anti-human prolyl 4-hydroxylase β (P4HB; Acris Antibodies); monoclonal rat anti-mouse CD31 (Dianova); anti-mouse CD68 and CD68-APC (BioLegend); anti-mouse CD16/CD32 and 7-amino-actinomycin D (7-AAD; BD Biosciences). PE-conjugated platelet-derived growth factor receptor α (PDGFRα) and PE-conjugated isotype-matched negative control (eBioscience) were used for sorting CAF. Collagenase type II/IV and deoxyribonuclease I were purchased from Worthington.

Generation of Girdin SA knock-in mice

Conventional gene-targeting techniques were used to generate the SA knock-in mice, in which the Akt phosphorylation site of Girdin (1416-serine) was mutated to alanine, as described previously (33–35). Resultant mice homozygous for the SA mutant allele in a 129Sv background were mated with C57BL/6 mice to generate mice with a C57BL/6 genetic background.

Cell culture

LLC (Lewis lung adenocarcinoma) and GFP-tagged LLC cells (GFP-LLC) were purchased from the ATCC and AntiCancer Japan, respectively. Cells were cultured in DMEM supplemented with 10% FBS.

Immunohistochemistry, immunofluorescence, immunoprecipitation, and Western blot analysis

Tissue sections from patients with invasive ductal breast carcinoma were obtained with informed patient consent at the Nagoya Medical Center (20 cases) and Nagoya University Hospital (86 cases), as previously described (36). For IHC studies, formaldehyde-fixed, paraffin-embedded tumor tissue sections were deparaffinized, and antigens were retrieved by boiling samples in target retrieval solution at pH 9 (Dako) for 30 minutes. Subsequently, tissue sections were washed with PBS, blocked with blocking reagent (Dako), and incubated overnight at 4°C with appropriate primary antibodies diluted in antibody diluents (Dako). Tissue sections were then washed, treated with 5% hydrogen peroxide/ethanol solution for 15 minutes at room temperature, and incubated at room temperature for 30 minutes with horseradish peroxidase–conjugated anti-mouse, anti-rat, or anti-rabbit IgG secondary antibodies (EnVision System; Dako), followed by signal detection with diaminobenzidine (DAB) solution.

For immunofluorescence, tissue sections were incubated overnight at 4°C with appropriate primary antibodies diluted in antibody diluents (Dako). Sections were then washed with PBS and incubated at room temperature for 1 hour in PBS containing Alexa Fluor 488/594–conjugated secondary antibodies (Life Technologies). The sections were then mounted with PermaFluor (Thermo Fisher Scientific), and fluorescence was examined using a confocal laser-scanning microscope (LSM 700, Carl Zeiss). We followed conventional procedures for immunoprecipitation and Western blot studies, as described previously (28).

Cell proliferation assays

Skin fibroblasts were cultured in triplicate (2 × 103 cells/well in 96-well plates) in DMEM containing 20% FBS. Cell proliferation was monitored by measuring the conversion of a water-soluble tetrazolium salt in water-soluble tetrazolium-1 (WST-1) assays (Roche Applied Science). The WST-1 reagent (10 μL) was added to 100 μL of cell suspension and incubated for 1 hour. Absorbance in wells was measured in a microplate reader set at a wavelength of 450 nm with a reference wavelength of 620 nm.

Migration and invasion assays

Directional cell migration of fibroblasts was stimulated in monolayers using an in vitro scratch-wound assay (28, 30). Briefly, fibroblasts isolated from WT and SA mice were seeded in 35-mm tissue culture dishes, and confluent monolayers were scratched with a 200-μL disposable plastic pipette tip and allowed to migrate toward the wound. To examine PDGF-dependent migration of fibroblasts, the medium on confluent cells was replaced with fresh medium containing 0.1% FBS with or without 10 ng/mL PDGF-BB, followed by scratching. Cells were fixed at 12 hours after scratching, and migratory distances were calculated using an inverted microscope. To assess the invasion of fibroblasts through Matrigel, cells were first seeded in 24-well Transwell Matrigel invasion chambers (8.0-μm pore size; BD Biosciences). Fibroblasts (2.5 × 104/mL) suspended in serum-free medium in the upper chamber were allowed to invade for 22 hours. The remaining fibroblasts on the upper surface of filters were removed by wiping with cotton swabs, and cells that had invaded through to the lower surface were visualized by Giemsa staining. Data were obtained from three independent experiments that were performed in triplicate.

Tumor transplantation experiments

Eight- to 10-week-old WT or SA mice were anesthetized with sodium pentobarbital, and 1 × 106 LLC cells were implanted s.c. into their backs. Tumor sizes were measured at day 5, 8, 11, and 14 with calipers, and tumor volumes were calculated using the formula X2 × Y × 0.5, where X, the smaller diameter and Y, the larger diameter. Tumors were embedded in paraffin wax at day 7 to day 14 for subsequent histologic analyses. For cotransplantation studies, we used LLC cells (2.5 × 104) alone or intermixed with skin fibroblasts or CAF (2.5 × 105) from WT or SA mice. These cell populations were implanted s.c. into the backs of 8- to 10-week-old male WT or SA mice. A minimum of three to four randomly picked sections from each animal (n = 5–6 mice/group) were analyzed to measure areas of tumor vessel and quantify the populations of CAF and TAM. The numbers of WT and SA mice used for each assay are indicated in the figure legends. All animal procedures were performed in compliance with institutional ethical requirements and were approved by the Animal Care and Use Committee of Nagoya University Graduate School of Medicine.

Isolation of CAF from LLC tumor tissues

CAF from LLC allografts were purified and sorted as described previously (37), with some modifications. Briefly, GFP-LLC tumors were excised from mice at day 23 after implantation. Tumors were minced and digested for 13 minutes at 37°C with a collagenase solution containing 400 mg BSA, 100 mg collagenase II, 100 mg collagenase IV, 300 mg deoxyribonuclease I, and 80 mL PBS, and then passed through a 70-μm cell strainer. Cells were washed, red blood cells were lysed with 1× ammonium–chloride–potassium buffer for 3 minutes, and the remaining intact cells were washed again. Cell pellets were then resuspended in PBS with 1% BSA, pretreated with FcR-block (anti-CD16/32), and incubated on ice with primary antibodies (PDGFRα-PE, 1:50; CD68-APC, 1:100). Cells were washed and incubated with 7-AAD to indicate cell death. Live CAF were then sorted on a FACS Aria II flow cytometer (BD Biosciences) by gating for GFP, CD68, and PDGFRα+ cells. CAF were sorted to a purity of >90%, as verified by post-sort immunofluorescence studies. Antibody specificity was confirmed using isotype-matched control antibodies.

Data analysis

Data are presented as the means ± S.D for all quantitative analyses. Statistical significances for experimental data and the analysis on clinicopathologic features were evaluated with χ2 tests.

Expression and Akt-mediated phosphorylation of Girdin in CAF infiltrating in human breast cancers

We and others have previously shown that Girdin is expressed and phosphorylated by Akt at Serine-1416 in some, but not all, cases of breast cancer, colon cancer, and glioblastoma (30, 36, 38, 39). These studies also demonstrate that Girdin is expressed in most cancer cell lines and in freshly isolated tumor-initiating cells from patients with glioblastoma, in which it controls migration and stem cell differentiation. However, our comprehensive screening of numerous invasive ductal breast carcinoma tissues revealed that Girdin expression and phosphorylation are not limited to tumor cells. IHC and immunofluorescence studies conducted in the present study demonstrates that Girdin is also expressed and phosphorylated in cells that constitute cancer stroma, including vWF-positive endothelial cells and α-SMA–positive CAF, and is accompanied by Akt phosphorylation (Fig. 1 A–C). Dual immunolabeling indicates that most of the Girdin-positive cells are CAF, endothelial cells, and pericytes, which constitute tumor vessels (Fig. 1A and B); however, Girdin expression did not extend to CD68-positive macrophages (data not shown). In addition, we also show that Girdin expression and phosphorylation in CAF was significantly higher in the tumor tissues compared with the adjacent normal tissues (Fig. 1D). Girdin phosphorylation was found in approximately 63% of Girdin-positive CAF in tumor tissues, whereas it was less than 40% in adjacent normal tissues (Fig. 1D, right). These data imply that Girdin expression in the tumor microenvironment is important for the progression of human cancers.

Figure 1.

Girdin expression and Akt-mediated phosphorylation in the stroma of human breast cancers. A and B, sections from invasive ductal breast carcinomas were stained with anti-Girdin (A) or anti–phospho-Girdin (1416-Ser; B) antibodies. Areas highlighted with red boxes are magnified in the adjacent bottom. Yellow and white arrows indicate vessels and stromal fibroblasts, respectively. The far bottom panels show immunofluorescence staining/colocalization of Girdin or phosphorylated Girdin (green) with vWF or α-SMA (red) in endothelial cells and CAF, respectively. C, coimmunostaining analysis showed that Akt phosphorylation (pAkt) is detected in both α-SMA–positive and phospho-Girdin–positive stromal fibroblasts (arrows). D, Girdin expression and its phosphorylation in the stroma of both tumors and adjacent normal tissues were evaluated by IHC on tissue sections from patients with invasive ductal carcinoma (n = 20). The number of cells positive for Girdin expression and phosphorylation in each tissue was counted and quantified. H&E, hematoxylin and eosin stain; *, P < 0.01; **, P < 0.001.

Figure 1.

Girdin expression and Akt-mediated phosphorylation in the stroma of human breast cancers. A and B, sections from invasive ductal breast carcinomas were stained with anti-Girdin (A) or anti–phospho-Girdin (1416-Ser; B) antibodies. Areas highlighted with red boxes are magnified in the adjacent bottom. Yellow and white arrows indicate vessels and stromal fibroblasts, respectively. The far bottom panels show immunofluorescence staining/colocalization of Girdin or phosphorylated Girdin (green) with vWF or α-SMA (red) in endothelial cells and CAF, respectively. C, coimmunostaining analysis showed that Akt phosphorylation (pAkt) is detected in both α-SMA–positive and phospho-Girdin–positive stromal fibroblasts (arrows). D, Girdin expression and its phosphorylation in the stroma of both tumors and adjacent normal tissues were evaluated by IHC on tissue sections from patients with invasive ductal carcinoma (n = 20). The number of cells positive for Girdin expression and phosphorylation in each tissue was counted and quantified. H&E, hematoxylin and eosin stain; *, P < 0.01; **, P < 0.001.

Close modal

Akt-mediated Girdin phosphorylation is crucial for CAF infiltration into tumor tissues transplanted in mice

To address the role of Akt-mediated Girdin phosphorylation in development of the tumor microenvironment, we used phospho-null Girdin knock-in mice (SA mice), previously generated in our laboratory, in which the Akt phosphorylation site (serine 1416) has been mutated to alanine (Fig. 2A; refs. 33–35). Western blot analysis of lysates prepared from WT and SA mouse brains indicated that Girdin expression levels were comparable, whereas Girdin phosphorylation was nearly abolished in SA mice (Fig. 2B). Syngeneic C57BL/6 murine LLC cells were transplanted into WT and SA mice and grew to form tumors. Engraftment ratio of the implanted tumors was 100% in both WT (12/12) and SA (10/10) groups. The growth of LLC tumor allografts was significantly retarded in SA mice compared with WT mice (Fig. 2C), suggesting that SA mice are defective in developing a supportive tumor microenvironment or inducing the host–tumor interactions that enhance LLC cell proliferation.

Figure 2.

Growth retardation of LLC tumors and limited intratumoral CAF infiltration in mice deficient for Girdin phosphorylation. A, generation of SA knock-in–mutant mice defective in Akt-mediated Girdin phosphorylation, as described previously (33). B, Western blot analysis of brain lysates isolated from WT, Girdin-deficient (KO, knockout; refs. 40, 50), and SA knock-in mice at the first postnatal (P0) day using anti-Girdin and anti–phospho-Girdin (p-Girdin) antibodies. Equivalency of protein loading is shown by Coomassie Brilliant Blue (CBB) staining of the membrane; Mr, molecular marker. C, LLC tumor volumes implanted into WT (n = 12) and SA (n = 10) mice. The diameters of implanted tumors were measured and their volumes were estimated on the days indicated. *, statistically significant difference (*, P < 0.005) was observed at day 14 after implantation. D, vasculature formation in LLC tumors. Endothelial cells in tumor sections were immunolabeled with anti-CD31 antibody and visualized by fluorescence microscopy (green fluorescence, left; WT, n = 4; SA, n = 4) or with DAB (brown, right; WT, n = 5; SA, n = 6). Quantification and representative staining images are shown; N.S., not significant. E and F, CAF and TAM infiltration in LLC tumors implanted into WT (n = 4) and SA (n = 4) mice was visualized by α-SMA and CD68 staining, respectively. The numbers of α-SMA– (E) and CD68- (F) positive cells in intratumoral regions of LLC tumors were quantified; *, P < 0.05; N.S., not significant.

Figure 2.

Growth retardation of LLC tumors and limited intratumoral CAF infiltration in mice deficient for Girdin phosphorylation. A, generation of SA knock-in–mutant mice defective in Akt-mediated Girdin phosphorylation, as described previously (33). B, Western blot analysis of brain lysates isolated from WT, Girdin-deficient (KO, knockout; refs. 40, 50), and SA knock-in mice at the first postnatal (P0) day using anti-Girdin and anti–phospho-Girdin (p-Girdin) antibodies. Equivalency of protein loading is shown by Coomassie Brilliant Blue (CBB) staining of the membrane; Mr, molecular marker. C, LLC tumor volumes implanted into WT (n = 12) and SA (n = 10) mice. The diameters of implanted tumors were measured and their volumes were estimated on the days indicated. *, statistically significant difference (*, P < 0.005) was observed at day 14 after implantation. D, vasculature formation in LLC tumors. Endothelial cells in tumor sections were immunolabeled with anti-CD31 antibody and visualized by fluorescence microscopy (green fluorescence, left; WT, n = 4; SA, n = 4) or with DAB (brown, right; WT, n = 5; SA, n = 6). Quantification and representative staining images are shown; N.S., not significant. E and F, CAF and TAM infiltration in LLC tumors implanted into WT (n = 4) and SA (n = 4) mice was visualized by α-SMA and CD68 staining, respectively. The numbers of α-SMA– (E) and CD68- (F) positive cells in intratumoral regions of LLC tumors were quantified; *, P < 0.05; N.S., not significant.

Close modal

To examine which stromal cell subpopulation(s) depend on Girdin phosphorylation to promote tumor growth, we used IHC to quantify the area of tumor vessels and the CAF and TAM populations present in LLC tumors excised from WT and SA mice (Fig. 2D–F). We have previously shown that Girdin is expressed in nascent immature endothelial cells and that its phosphorylation is important for angiogenesis in the retinal vascular plexus (34, 40). Thus, it was surprising that CD31 immunostaining revealed no significant differences in tumor vessel formation between WT and SA mice (Fig. 2D). However, the number of host-derived α-SMA–positive cells not associated with the vasculature was significantly decreased in SA mice. These cells appeared to colocalize with the CAF invading the tumor tissue (Fig. 2E). We also found that the numbers of TAM were comparable between WT and SA mice, despite their known roles in tumor progression (Fig. 2F; ref. 4). This finding suggests that the Akt–Girdin pathway specifically contributes to CAF infiltration, but not to the infiltration of vessel cells or TAM.

Recent studies have shown that CAF degrade or remodel the ECM surrounding tumors to generate tracks and passages, along which tumor cells can invade and proliferate (14, 18). Therefore, we studied CAF invasion in the peritumoral region of tumors and found that peritumoral CAF infiltration was significantly impaired in SA mice compared with that observed in WT counterparts (Fig. 3A). This finding was confirmed by staining for P4HB, another CAF marker, which showed that P4HB-positive cells had a decreased presence in the peritumoral regions in SA mice (Fig. 3B). Additional immunostaining and quantification of α-SMA and P4HB-positive cells within the peritumoral region confirmed that SA mice are likely defective exclusively in the recruitment of CAF, independent of the other host-derived cells that surround tumors (Fig. 3C). Collectively, these findings suggest that Akt-mediated Girdin phosphorylation is important for the specific recruitment of CAF, but not vessel cells or TAM, to peritumoral regions.

Figure 3.

Limited CAF infiltration in the peritumoral region of LLC tumors in mice deficient for Girdin phosphorylation. A and B, CAF infiltration in the peritumoral region of LLC tumors implanted into WT (n = 5) and SA (n = 6) mice was visualized by α-SMA (A) and P4HB (B) staining. The areas highlighted with boxes are magnified in the adjacent panels. Yellow arrows, representative infiltrated CAF. Right, quantification of α-SMA– or P4HB-positive cells in peritumoral regions of tumors. Dotted lines, interfaces between tumors and surrounding stroma; *, P < 0.05. C, the numbers of α-SMA– (left) and P4HB- (right) positive cells per field were quantified. Frequencies of α-SMA– and P4HB-positive cells in peritumoral regions of LLC tumors were expressed as a percentage of total cells, as determined by DAPI staining. Dotted lines, interfaces between tumors and surrounding stroma; *, P < 0.05; **, P < 0.005.

Figure 3.

Limited CAF infiltration in the peritumoral region of LLC tumors in mice deficient for Girdin phosphorylation. A and B, CAF infiltration in the peritumoral region of LLC tumors implanted into WT (n = 5) and SA (n = 6) mice was visualized by α-SMA (A) and P4HB (B) staining. The areas highlighted with boxes are magnified in the adjacent panels. Yellow arrows, representative infiltrated CAF. Right, quantification of α-SMA– or P4HB-positive cells in peritumoral regions of tumors. Dotted lines, interfaces between tumors and surrounding stroma; *, P < 0.05. C, the numbers of α-SMA– (left) and P4HB- (right) positive cells per field were quantified. Frequencies of α-SMA– and P4HB-positive cells in peritumoral regions of LLC tumors were expressed as a percentage of total cells, as determined by DAPI staining. Dotted lines, interfaces between tumors and surrounding stroma; *, P < 0.05; **, P < 0.005.

Close modal

Roles of Girdin phosphorylation in fibroblast proliferation and mobilization

To identify the cellular processes and activities of fibroblasts regulated by Akt-mediated Girdin phosphorylation, we first examined the properties of cultured primary skin-derived fibroblasts isolated from WT and SA mice (Fig. 4). Basal Girdin phosphorylation was detected in primary fibroblast lysates from WT mice but not SA mice, as shown by Western blot analysis (Fig. 4A, left). WST-1 proliferation assays also demonstrate that, compared with WT fibroblasts, SA fibroblasts grow at a slower rate, particularly during the later phase of subconfluent culture (after 5 days; Fig. 4A, right). In addition, we found that serum- and PDGF-dependent directional migrations in culture dishes are significantly impaired in SA fibroblasts (Fig. 4B). Considering that Girdin associates with the actin cytoskeleton (28), it is possible that Girdin incorporates into the machinery that regulates α-SMA–mediated contractile activity in fibroblasts, in which its phosphorylation by Akt promotes cell migration.

Figure 4.

Proliferation and migration of skin fibroblasts isolated from WT and SA mice. A, skin fibroblasts isolated from WT and SA mice were expanded in culture dishes. Total Girdin in fibroblast lysates was immunoprecipitated (IP) with an anti-Girdin antibody, followed by immunoblot analysis (IB) by using the indicated antibodies. Two independent lots of fibroblasts in each group (WT1, WT2, SA1, and SA2) were tested. Right, proliferation rates of WT and SA fibroblasts on culture, as measured in WST1 assays (n = 4 for each group); *, P < 0.05. B, directional cell migration of WT and SA fibroblasts was monitored in monolayers by using an in vitro scratch-wound assay. Cells were seeded on tissue culture dishes. Confluent cells were scratched with 200 μL disposable plastic pipette tips and were allowed to migrate toward the wound for 12 hours in the presence of FBS or PDGF (10 ng/mL). Dotted lines, the leading front of migrating cells. The graphs show quantification of migration distances covered by WT (n = 3–6) and SA (n = 3–6) fibroblasts; **, P < 0.05. C, in vitro Matrigel invasion assays. The numbers of WT (n = 4) and SA (n = 4) fibroblasts that invaded through Matrigel for 22 hours were quantified; N.S., not significant.

Figure 4.

Proliferation and migration of skin fibroblasts isolated from WT and SA mice. A, skin fibroblasts isolated from WT and SA mice were expanded in culture dishes. Total Girdin in fibroblast lysates was immunoprecipitated (IP) with an anti-Girdin antibody, followed by immunoblot analysis (IB) by using the indicated antibodies. Two independent lots of fibroblasts in each group (WT1, WT2, SA1, and SA2) were tested. Right, proliferation rates of WT and SA fibroblasts on culture, as measured in WST1 assays (n = 4 for each group); *, P < 0.05. B, directional cell migration of WT and SA fibroblasts was monitored in monolayers by using an in vitro scratch-wound assay. Cells were seeded on tissue culture dishes. Confluent cells were scratched with 200 μL disposable plastic pipette tips and were allowed to migrate toward the wound for 12 hours in the presence of FBS or PDGF (10 ng/mL). Dotted lines, the leading front of migrating cells. The graphs show quantification of migration distances covered by WT (n = 3–6) and SA (n = 3–6) fibroblasts; **, P < 0.05. C, in vitro Matrigel invasion assays. The numbers of WT (n = 4) and SA (n = 4) fibroblasts that invaded through Matrigel for 22 hours were quantified; N.S., not significant.

Close modal

Interestingly, but for reasons presently unclear, WT and SA fibroblasts show no difference in invasiveness through the ECM in response to multiple growth factors in sera, as determined by Matrigel invasion assays (Fig. 4C). This finding contradicts an earlier study showing that Girdin is crucial for cancer cell invasion (30). A potential explanation for these discordant results is that Girdin phosphorylation may not exert a potent effect on fibroblast ECM invasion, a property that may diverge from that of cancer cell invasion.

Significance of Girdin phosphorylation in fibroblasts in promoting tumor growth

To test whether Girdin phosphorylation affects protumorigenic properties of fibroblasts, we next grafted LLC cells intermixed with WT or SA skin fibroblasts (henceforth referred to as LLC/WT and LLC/SA, respectively) into WT C57BL/6 mice and monitored tumor growth (Fig. 5). As a negative control experiment, only LLC cells were injected into WT mice (Fig. 5A, left). Engraftment ratio of the implanted tumors was comparable among the three groups (LLC/WT, 21/21; LLC/SA, 12/12; LLC only, 16/17). Consistent with the findings of previous studies reporting tumor-supportive functions of fibroblasts (11, 41), we found that LLC/WT tumors grew at a faster rate than those derived from LLC cells alone (Fig. 5A, right). Significantly, LLC/SA tumors failed to grow as well as those coimplanted with WT fibroblasts, implicating a tumor-supportive role of Girdin phosphorylation in fibroblasts.

Figure 5.

Importance of Girdin phosphorylation in fibroblasts in promoting LLC tumor growth in vivo. A and B, LLC cells (2.5 × 104) intermixed with WT or SA skin fibroblasts (2.5 × 105) were implanted into either WT (A) or SA (B) mice, and the volume of each tumor was measured at the indicated days after implantation (right). The sample number for each group is indicated in the graph legends; *, P < 0.05; **, P < 0.005. C, sections from the allograft tumors developed in the experiments shown in A, which comprised LLC cells and either WT or SA fibroblasts, were stained for α-SMA to visualize CAF. The graph on the right shows the number of α-SMA–positive CAF in each field of the peripheral region of tumors in each group (WT; n = 6, SA; n = 5); **, P < 0.005.

Figure 5.

Importance of Girdin phosphorylation in fibroblasts in promoting LLC tumor growth in vivo. A and B, LLC cells (2.5 × 104) intermixed with WT or SA skin fibroblasts (2.5 × 105) were implanted into either WT (A) or SA (B) mice, and the volume of each tumor was measured at the indicated days after implantation (right). The sample number for each group is indicated in the graph legends; *, P < 0.05; **, P < 0.005. C, sections from the allograft tumors developed in the experiments shown in A, which comprised LLC cells and either WT or SA fibroblasts, were stained for α-SMA to visualize CAF. The graph on the right shows the number of α-SMA–positive CAF in each field of the peripheral region of tumors in each group (WT; n = 6, SA; n = 5); **, P < 0.005.

Close modal

To exclude the contribution of Girdin phosphorylation in other cell types derived from host mice, LLC/WT and LLC/SA cell mixtures were implanted into SA mice (Fig. 5B, left), so as to disregard the role of Girdin phosphorylation in host-derived vessel and immune cells in development of the tumor stroma. Again, we found that LLC/SA exhibited a retarded growth rate compared with LLC/WT counterparts (Fig. 5B, right). IHC on tumor sections revealed that infiltration of α-SMA–positive CAF in the peritumoral region was significantly impaired in LLC/SA tumors compared with LLC/WT counterparts (Fig. 5C). These data suggest Girdin phosphorylation in fibroblasts not only enhances their proliferative and migratory responses, but also facilitates tumor growth in vivo. Taking together, these data suggest that Girdin phosphorylation promotes tumor growth by several potential direct and/or indirect mechanisms, such as regulating ECM remodeling or growth factor/cytokine networks.

CAF defective of Girdin phosphorylation have limited effects on tumor growth in vivo

To recapitulate the in vivo features of tumor growth and further validate the function of Girdin phosphorylation in CAF, CAF isolated from WT and SA mice harboring LLC tumors were subsequently used in our tumor transplantation model (Fig. 6). To this end, LLC cells engineered to express GFP (termed GFP-LLC) were allografted into WT and SA mice. Three weeks after implantation, tumors were harvested and digested with collagenase to form single-cell suspensions, which were then studied by FACS analysis. Cells were sorted into tumor cell, CAF, and TAM subpopulation based on expression of GFP, PDGFRα, and CD68 expression, respectively (Fig. 6A and B). The purity of sorted CAF was monitored by immunofluorescence staining, indicating that contamination with GFP-LLC cells was less than 10% (data not shown). The role of Girdin phosphorylation in the tumor growth–promoting ability imparted by CAF was assessed by coimplanting CAF isolated from LLC tumors grafted in WT and SA mice (henceforth referred to as LLC/WT-CAF and LLC/SA-CAF, respectively; Fig. 6C). The engraftment ratio of the implanted tumors was comparable between the two groups (LLC/WT-CAF, 6/6; LLC/SA-CAF, 6/6). Similar to previous experiments, the growth and infiltration of α-SMA–positive CAF in peritumoral tissues were attenuated in LLC/SA-CAF tumors compared with LLC/WT-CAF counterparts (Fig. 6C and D). These data indicate that Girdin phosphorylation in CAF participates in the formation of the tumor microenvironment, which facilitates tumor growth.

Figure 6.

Importance of Girdin phosphorylation in CAF in promoting LLC tumor growth in vivo. A and B, schematic illustration of the procedure used for isolating CAF from tumor-bearing WT and SA mice. GFP-LLC cells were transplanted into WT and SA mice, and the developed tumors were digested with collagenase. Cells in suspension were labeled with the indicated antibodies and analyzed by flow cytometry. Shown in B is an example of expression profiles observed following sequential gating based on CD68, PDGFRα, and GFP expressions. Apoptotic or dead cells were excluded by staining with 7-AAD, a marker of cell death and apoptosis. C, LLC cells (2.5 × 104) intermixed with CAF (2.5 × 105) derived from WT (n = 6) or SA (n = 6) mice were implanted into WT mice, and the volume of each tumor was measured at the indicated days after implantation (bottom); **, P < 0.005. D, sections from allograft tumors were stained for α-SMA to visualize CAF. The graph on the right shows the number of α-SMA–positive CAF observed in each field, examining peripheral regions of tumors from each group (WT; n = 8, SA; n = 8); *, P < 0.05. E, the migratory response of WT- and SA-CAF to recombinant TGF-β1, PDGF, and SDF-1 was examined by in vitro scratch-wound assays, and the migration distances covered by WT (n = 6) and SA (n = 6) CAF were quantified; *, P < 0.05.

Figure 6.

Importance of Girdin phosphorylation in CAF in promoting LLC tumor growth in vivo. A and B, schematic illustration of the procedure used for isolating CAF from tumor-bearing WT and SA mice. GFP-LLC cells were transplanted into WT and SA mice, and the developed tumors were digested with collagenase. Cells in suspension were labeled with the indicated antibodies and analyzed by flow cytometry. Shown in B is an example of expression profiles observed following sequential gating based on CD68, PDGFRα, and GFP expressions. Apoptotic or dead cells were excluded by staining with 7-AAD, a marker of cell death and apoptosis. C, LLC cells (2.5 × 104) intermixed with CAF (2.5 × 105) derived from WT (n = 6) or SA (n = 6) mice were implanted into WT mice, and the volume of each tumor was measured at the indicated days after implantation (bottom); **, P < 0.005. D, sections from allograft tumors were stained for α-SMA to visualize CAF. The graph on the right shows the number of α-SMA–positive CAF observed in each field, examining peripheral regions of tumors from each group (WT; n = 8, SA; n = 8); *, P < 0.05. E, the migratory response of WT- and SA-CAF to recombinant TGF-β1, PDGF, and SDF-1 was examined by in vitro scratch-wound assays, and the migration distances covered by WT (n = 6) and SA (n = 6) CAF were quantified; *, P < 0.05.

Close modal

As observed with skin fibroblasts (Fig. 4), these findings might support a role for the Girdin phosphorylation–mediated regulation of CAF migration in response to certain growth factors and cytokines (Fig. 6E); however, this was not conclusively determined in this study. SA-CAF exhibited defective migratory responses following stimulation with TGF-β1 and PDGF, but not with stromal cell–derived factor 1 (SDF-1), when analyzed by in vitro migration assays (although PDGF data were not statistically significant). Our observation that Girdin functions downstream of TGF-β1 is consistent with a recent study describing the role of Girdin in liver fibrosis (42). However, we also found that Girdin phosphorylation seems not to be involved in α-SMA expression regulated downstream of TGFβ signaling (Supplementary Fig. S1).

Survival of coimplanted CAF in the tumor transplantation model

Considering controversy over the fate of coimplanted CAF in the mouse tumor transplantation model and giving implications for further research, we assessed the survival of coimplanted CAF within the developed tumors. To this end, GFP-LLC cells were allografted into the Rosa26–tdTomato reporter mice and isolated tdTomato-positive CAF from the developed tumors. The isolated CAF were then coimplanted with GFP-LLC cells into WT mice (Fig. 7A). Immunofluorescent analyses showed that the tdTomato-positive CAF populating the intratumoral region significantly decreased from days 14 to 21 after transplantation. Notably, the ratios of tdTomato-positive CAF to total CAF (both implanted and host-derived) were 56% and 16% at day 14 and 21, respectively (Fig. 7B and C). These data indicate that the implanted CAF did not survive in the tumor tissues for an extended period, and leads to the speculation that the implanted tdTomato-positive CAF are involved in early, but not later, stages of tumor progression in tumor transplantation models.

Figure 7.

The fate of coimplanted CAF during tumor progression in mice. A, GFP-LLC cells were implanted into the Rosa26-tdTomato reporter mice to develop tumors. tdTomato-positive CAF were sorted by flow cytometer, and TAM were eliminated with CD11b microbeads. The isolated CAF were then coimplanted with GFP-LLC cells into WT mice, and the developed tumors were resected for further analysis. B and C, the sections prepared from the developed tumors at the indicated days after implantation, which harbor tdTomato-positive CAF (red), were stained for α-SMA (blue). Regions indicated by the white boxes were magnified in the bottom. The merged images show the number of cells double-positive for both tdTomato and α-SMA (yellow arrows) decreased after implantation (B, day 14; C, day 21 after implantation), indicating that the implanted CAF did not survive at the late stage of tumor development. Arrowheads, newly infiltrated α-SMA–positive CAF that are negative for tdTomato. The ratios of tdTomato-positive CAF to total infiltrating CAF (both tdTomato-positive implanted and tdTomato-negative host derived) were 56% and 16% at days 14 and 21, respectively.

Figure 7.

The fate of coimplanted CAF during tumor progression in mice. A, GFP-LLC cells were implanted into the Rosa26-tdTomato reporter mice to develop tumors. tdTomato-positive CAF were sorted by flow cytometer, and TAM were eliminated with CD11b microbeads. The isolated CAF were then coimplanted with GFP-LLC cells into WT mice, and the developed tumors were resected for further analysis. B and C, the sections prepared from the developed tumors at the indicated days after implantation, which harbor tdTomato-positive CAF (red), were stained for α-SMA (blue). Regions indicated by the white boxes were magnified in the bottom. The merged images show the number of cells double-positive for both tdTomato and α-SMA (yellow arrows) decreased after implantation (B, day 14; C, day 21 after implantation), indicating that the implanted CAF did not survive at the late stage of tumor development. Arrowheads, newly infiltrated α-SMA–positive CAF that are negative for tdTomato. The ratios of tdTomato-positive CAF to total infiltrating CAF (both tdTomato-positive implanted and tdTomato-negative host derived) were 56% and 16% at days 14 and 21, respectively.

Close modal

Relation between Girdin phosphorylation in CAF and clinicopathologic features in invasive breast cancer

Finally, we performed immunohistochemical analysis by using tissue sections from 86 patients with invasive ductal breast carcinoma. However, the data failed to show any significant correlation between Girdin phosphorylation in CAF and tumor stages and grades, histologic features, metastatic rate, or tumor recurrence, although Girdin phosphorylation tended to be associated with HER2-positive cases (Supplementary Fig. S2; Supplementary Tables S1 and S2). No significant correlation may be partially attributed to the difference between transplanted animal models and progressive human cancers. Because a variety of microenvironmental factors are involved in the progression of human cancer, synergistic or additive effects of Girdin phosphorylation with those intrinsic factors could influence the clinical outcome of the patients.

Here, we report that the Akt substrate Girdin is expressed and phosphorylated in CAF, and exerts tumor growth–promoting effects with in vivo animal tumor models. Our findings suggest a novel mechanism, whereby Akt signaling is central to both tumor cells and cells that constitute the tumor microenvironment. We demonstrate that the Akt-mediated phosphorylation of Girdin is important for CAF infiltration and tumor cell growth in mice. Given that Girdin is an actin-binding protein and its phosphorylation regulates actin remodeling (28, 30), it is plausible that Girdin integrates into thick actin cables predominantly composed of α-SMA to regulate the contractility of CAF, which could promote ECM remodeling required for tumor growth. Although the clinical relevance of our findings are not provided in this present study and should await further clinical analysis, our study provide a foundation for the potential targeting of the PI3K–Akt signaling pathway within the tumor microenvironment, with the goal of developing novel therapies against human malignancies.

One limitation of this study is that it does not address the potential upstream regulator(s) that may induce Girdin phosphorylation in CAF. Although speculative, it is interesting to consider that Girdin phosphorylation in CAF may result from signals derived from proliferating tumor cells. Recent studies have shown that tumor cells secrete many soluble growth factors, cytokines, and chemokines, including TGF-β1, hepatocyte growth factor, and SDF-1, which act upon their cognate receptors expressed on CAF to modulate their ability to promote an aggressive tumor phenotype (3, 5, 8, 43). Migration assays performed in this study implicate PDGF and TGF-β1 (42, 44) as candidate growth factors that may govern the function of Girdin, as fibroblasts and CAF isolated from SA mice were defective in PDGF- and TGF-β1–induced migration (Figs. 4B and 6E). Consistent with this hypothesis, cumulative evidence has suggested that tumor cells overexpress and secrete PDGF and TGF-β1, and that CAF express the cognate receptors for these factors (22, 25, 26, 45, 46). These effects, combined with the known autocrine effects of these factors on tumor cells, suggest that CAF may play an integral part in a signaling network regulated by synergy or extensive crosstalk with tumor cells to provide for the supportive nature of the tumor microenvironment.

Another intriguing issue in this study is the role of Girdin phosphorylation in tumor vessel formation. We have previously shown that Girdin is expressed in nascent small capillaries and cultured endothelial cells, in which it is phosphorylated in response to vascular endothelial growth factor stimulation through the PI3K–Akt signaling pathway (34, 40). However, our quantitative analysis of tumor vessel development in allografted LLC tumors revealed no significant differences in vessel area or branching morphogenesis between tumors grown in WT and SA host mice (Fig. 2D). One potential explanation for this discrepancy might be that Akt-mediated Girdin phosphorylation is dispensable for the recruitment of endothelial cells in tumor tissues, but pivotal for the maintenance of vascular integrity or permeability. A definitive conclusion regarding this point requires further investigation in tumor tissues and sophisticated imaging techniques.

A challenging view of tumor stroma is that the distribution of CAF is not homogeneous and the composition of the tumor stroma is rich in diversity, even within the tumor derived from the same tissue (47). The tumor stroma is inconspicuous in some cases, whereas in others it is desmoplastic. Our histologic experiments on human breast cancer tissues also indicate the heterogeneity of Girdin expression and its phosphorylation in both CAF and tumor cells (Fig. 1D). In LLC implanted tumors, α-SMA–positive CAF distributed into both the intratumoral and peritumoral regions. Although the infiltration of CAF in both regions were prone to regulation by Girdin phosphorylation (Figs. 2E and 3), the significance of their distribution patterns are not yet determined. Given that clinical investigation and biologic evaluation of several PI3K and Akt inhibitors are underway (48, 49), it would be necessary to investigate the effects of these inhibitors on the distribution of Girdin expression and its phosphorylation in CAF and their relevance to clinical outcomes.

In the present study, the clinical importance of Girdin-dependent Akt signaling has not been determined. Our study did not show any correlation between Girdin phosphorylation in CAF and tumor stages, malignancy grades, and pathologic features of invasive ductal breast carcinomas, although Girdin phosphorylation tended to be associated with HER2-positive cases (Supplementary Table S2). Because a variety of microenvironmental factors are involved in the progression of human cancer, synergistic or additive effects of Girdin phosphorylation with those factors may be important for clinical outcome of the patients. In addition, the gene-expression profiles of CAF in individual cancers need to be considered (19–23). Therefore, future studies should concentrate on determining the importance of Akt–Girdin signaling in the tumor microenvironment for the evaluation of clinical outcomes of the patients with various stages of malignant tumors on an individualized basis.

No potential conflicts of interest were disclosed.

Conception and design: Y. Yamamura, A. Enomoto, T. Murohara, M. Takahashi

Development of methodology: Y. Yamamura, N. Asai, T. Kato

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): Y. Yamamura, N. Asai, S. Mii, Y. Kondo, N. Tsunoda, M. Nagino, S. Ichihara

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): Y. Yamamura, K. Niimi, T. Murohara, M. Takahashi

Writing, review, and/or revision of the manuscript: Y. Yamamura, A. Enomoto, K. Maeda, M. Takahashi

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): S. Mii, K. Ushida, M. Nagino, S. Ichihara, T. Murohara

Study supervision: K. Furukawa, K. Maeda, T. Murohara, M. Takahashi

The authors thank Katsuhiro Kato for helpful discussions, Mayu Isotani-Sakakibara for Western blot analysis, and Minoru Tanaka for FACS and sorting analysis.

This work was supported by grant-in-aid for Scientific Research on Innovative Areas (22117005), Scientific Research (A) (23249020) and Scientific Research (S) (26221304; M. Takahashi), grant-in-aid for Scientific Research (C) (24390095; N. Asai), and grant-in-aid for Young Scientists (A) (20432255; A. Enomoto) from the Ministry of Education, Culture, Sports, Science, and Technology of Japan.

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.

1.
Hanahan
D
,
Weinberg
RA
. 
Hallmarks of cancer: the next generation
.
Cell
2011
;
144
:
646
74
.
2.
Bergers
G
,
Benjamin
LE
. 
Tumorigenesis and the anigiogenic switch
.
Nat Rev Cancer
2003
;
3
:
401
10
.
3.
Bhowmick
NA
,
Neilson
EG
,
Moses
HL
. 
Stromal fibroblasts in cancer initiation and progression
.
Nature
2004
;
432
:
332
7
.
4.
Mantovani
A
,
Allavena
P
,
Sica
A
,
Balkwill
F
. 
Cancer-related inflammation
.
Nature
2008
;
454
:
436
44
.
5.
Hinz
B
,
Darby
IA
,
Gabbiani
G
,
Desmoulière
. 
The role of the myofibroblast in fibrosis and cancer progression
. In:
Mueller
MM
,
Fusenig
NE
, editors. 
Tumor-associated fibroblasts and their matrix
.
Springer
; 
2011
. p.
37
74
.
6.
Polyak
K
,
Haviv
I
,
Campbell
IG
. 
Co-evolution of tumor cells and their microenvironment
.
Trends Genet
2009
;
25
:
30
8
.
7.
Egeblad
M
,
Nakasone
ES
,
Werb
Z
. 
Tumors as organs: complex tissues that interface with the entire organism
.
Dev Cell
2010
;
18
:
884
901
.
8.
Chung
LWK
. 
Critical roles of stromal fibroblasts in the cancer microenvironments
. In:
Mueller
MM
,
Fusenig
NE
, editors. 
Tumor-associated fibroblasts and their matrix
.
Springer
; 
2011
. p.
3
22
.
9.
Eikesdal
HP
,
Kalluri
R
. 
The multifaceted role of cancer associated fibroblasts in tumor progression
. In:
Mueller
MM
,
Fusenig
NE
, editors. 
Tumor-associated fibroblasts and their matrix
.
Springer
; 
2011
. p.
361
82
.
10.
Polyak
K
,
Kalluri
R
The role of the microenvironment in mammary gland development and cancer
. In:
Bissell
MJ
,
Polyak
K
,
Rosen
JM
, editors. 
The mammary gland as an experimental model
.
Cold Spring Harbor
; 
2011
.
p163
76
.
11.
Orimo
A
,
Gupta
PB
,
Sgroi
DC
,
Arenzana-Seisdedos
F
,
Delaunay
T
,
Naeem
R
, et al
Stroma fibroblasts present in invasive human breast carcinomas promote tumor growth and angiogenesis through elevated SDF-1/CXCL12 secretion
.
Cell
2005
;
121
:
335
48
.
12.
Schreiber
RD
,
Old
LJ
,
Smyth
MJ
. 
Cancer immunoediting: integrating immunity's roles in cancer suppression and promotion
.
Science
2011
;
331
:
1565
70
.
13.
Zhang
XH
,
Jin
X
,
Malladi
S
,
Zou
Y
,
Wen
YH
,
Brogi
E
, et al
Selection of bone metastasis seeds by mesenchymal signals in the primary tumor stroma
.
Cell
2013
;
154
:
1060
73
.
14.
De Wever
O
,
Demetter
P
,
Mareel
M
,
Bracke
M
. 
Stromal myofibroblasts are drivers of invasive cancer growth
.
Int J Cancer
2008
;
123
:
2229
38
.
15.
Worthley
DL
,
Si
Y
,
Quante
M
,
Churchill
M
,
Mukherjee
S
,
Wang
TC
. 
Bone marrow cells as precursors of the tumor stroma
.
Exp Cell Res
2013
;
319
:
1650
6
.
16.
Lazennec
G
,
Richmond
A
. 
Chemokines and chemokine receptors: new insights into cancer-related inflammation
.
Trends Mol Med
2010
;
16
:
133
44
.
17.
Goetz
JG
,
Minguet
S
,
Navarro-Lérida
I
,
Lazcano
JJ
,
Samaniego
R
,
Calvo
E
, et al
Biomechanical remodeling of the microenvironment by stromal caveolin-1 favors tumor invasion and metastasis
.
Cell
2011
;
146
:
148
63
.
18.
Gaggioli
C
,
Hooper
S
,
Hidalgo-Carcedo
C
,
Grosse
R
,
Marshall
JF
,
Harrington
K
, et al
Fibroblast-led collective invasion of carcinoma cells with differing roles for Rho GTPases in leading and following cells
.
Nat Cell Biol
2007
;
9
:
1392
400
.
19.
Ono
S
,
Ishii
G
,
Nagai
K
,
Takuwa
T
,
Yoshida
J
,
Nishimura
M
, et al
Podoplanin-positive cancer-associated fibroblasts could have prognostic value independent of cancer cell phenotype in stage I lung squamous cell carcinoma: usefulness of combining analysis of both cancer cell phenotype and cancer-associated fibroblast phenotype
.
Chest
2013
;
143
:
963
70
.
20.
Herrera
M
,
Herrera
A
,
Dominquez
G
,
Silva
J
,
Garcia
V
,
Garcia
JM
, et al
Cancer-associated fibroblast and M2 macrophage markers together predict outcome in colorectal cancer patients
.
Cancer Sci
2013
;
104
:
437
44
.
21.
Herrera
M
,
Islam
AB
,
Herrera
A
,
Martín
P
,
García
V
,
Silva
J
, et al
Functional heterogeneity of cancer-associated fibroblasts from human colon tumors shows specific prognostic gene expression signature
.
Clin Cancer Res
2013
;
19
:
5914
26
.
22.
Busch
S
,
Acar
A
,
Magnusson
Y
,
Gregersson
P
,
Rydén
L
,
Landberg
G
. 
TGF-beta receptor type-2 expression in cancer-associated fibroblasts regulates breast cancer cell growth and survival and is a prognostic marker in pre-menopausal breast cancer
.
Oncogene
2015
;
34
:
27
38
.
23.
Peña
C
,
Céspedes
MV
,
Lindh
MB
,
Kiflemariam
S
,
Mezheyeuski
A
,
Edqvist
PH
, et al
STC1 expression by cancer-associated fibroblasts drives metastasis of colorectal cancer
.
Cancer Res
2013
;
73
:
1287
97
.
24.
Rupp
C
,
Dolznig
H
,
Haslinger
C
,
Schweifer
N
,
Garin-Chesa
P
. 
Cancer associated fibroblasts as therapeutic targets
. In:
Mueller
MM
,
Fusenig
NE
, editors. 
Tumor-associated fibroblasts and their matrix
.
Springer
; 
2011
. p.
383
402
.
25.
Bierie
BR
,
Moses
HL
TGF-β signaling in fiboblasts regulates tumor initiation and progression in adjacent epithelia
. In:
Mueller
MM
,
Fusenig
NE
, editors. 
Tumor-associated fibroblasts and their matrix
.
Springer
; 
2011
. p.
223
44
.
26.
Calon
A
,
Espinet
E
,
Palomo-Ponce
S
,
Tauriello
DVF
,
Iglesias
M
,
Cé spedes
MV
, et al
Dependency of colorectal cancer on a TGF-β–driven program in stromal cells for metastasis initiation
.
Cancer Cell
2012
;
22
:
571
84
.
27.
Sanz-Moreno
V
,
Gaggioli
C
,
Yeo
M
,
Albrengues
J
,
Wallberg
F
,
Viros
A
, et al
ROCK and JAK1 signaling cooperate to control actomyosin contractility in tumor cells and stroma
.
Cancer Cell
2011
;
20
:
229
45
.
28.
Enomoto
A
,
Murakami
H
,
Asai
N
,
Morone
N
,
Watanabe
T
,
Kawai
K
, et al
Akt/PKB regulates actin organization and cell motility via Girdin/APE
.
Dev Cell
2005
;
9
:
389
402
.
29.
Le-Niculescu
H
,
Niesman
I
,
Fischer
T
,
DeVries
L
,
Farquhar
MG
. 
Identification and characterization of GIV, a novel Galpha i/s-interacting protein found on COPI, endoplasmic reticulum-Golgi transport vesicles
.
J Biol Chem
2005
;
280
:
22012
20
.
30.
Jiang
P
,
Enomoto
A
,
Jijiwa
M
,
Kato
T
,
Hasegawa
T
,
Ishida
M
, et al
An actin-binding protein Girdin regulates the motility of breast cancer cells
.
Cancer Res
2008
;
68
:
1310
8
.
31.
Ghosh
P
,
Garcia-Marcos
M
,
Bornheimer
SJ
,
Farquhar
MG
. 
Activation of Galphai3 triggers cell migration via regulation of GIV
.
J Cell Biol
2008
;
182
:
381
93
.
32.
Ohara
K
,
Enomoto
A
,
Kato
T
,
Hashimoto
T
,
Isotani-Sakakibara
M
,
Asai
N
, et al
Involvement of Girdin in the determination of cell polarity during cell migration
.
PloS ONE
2012
;
7
:
e36681
.
33.
Wang
Y
,
Kaneko
N
,
Asai
N
,
Enomoto
A
,
Isotani-Sakakibara
M
,
Kato
T
, et al
Girdin is an intrinsic regulator of neuroblast chain migration in the rostral migratory stream of the postnatal brain
.
J Neurosci
2011
;
31
:
8109
22
.
34.
Ito
T
,
Komeima
K
,
Yasuma
T
,
Enomoto
A
,
Asai
N
,
Asai
M
, et al
Girdin and its phosphorylation dynamically regulate neonatal vascular development and pathological neovascularization in the retina
.
Am J Pathol
2013
;
182
:
586
96
.
35.
Miyake
H
,
Maeda
K
,
Asai
N
,
Shibata
R
,
Ichimiya
H
,
Isotani-Sakakibara
M
, et al
The actin-binding protein Girdin and its Akt-mediated phosphorylation regulate neointima formation after vascular injury
.
Circ Res
2011
;
108
:
1170
9
.
36.
Nishimae
K
,
Tsunoda
N
,
Yokoyama
Y
,
Kokuryo
T
,
Iwakoshi
A
,
Takahashi
M
, et al
The impact of Girdin expression on recurrence-free survival in patients with luminal-type breast cancer
.
Breast Cancer
. 
2013 Oct 24
.
[Epub ahead of print]
.
37.
Erez
N
,
Truitt
M
,
Olson
P
,
Arron
ST
,
Hanahan
D
. 
Cancer-associated fibroblasts are activated in incipient neoplasia to orchestrate tumor-promoting inflammation in an NF-κB-dependent manner
.
Cancer Cell
2010
;
17
:
135
47
.
38.
Garcia-Marcos
M
,
Jung
BH
,
Ear
J
,
Cabrera
B
,
Carethers
JM
,
Ghosh
P
. 
Expression of GIV/Girdin, a metastasis-related protein, predicts patient survival in colon cancer
.
FASEB J
2011
;
25
:
590
9
.
39.
Natsume
A
,
Kato
T
,
Kinjo
S
,
Enomoto
A
,
Toda
H
,
Shimato
S
, et al
Girdin maintains the stemness of glioblastoma stem cells
.
Oncogene
2012
;
31
:
2715
24
.
40.
Kitamura
T
,
Asai
N
,
Enomoto
A
,
Maeda
K
,
Kato
T
,
Ishida
M
, et al
Regulation of VEGF-mediated angiogenesis by the Akt/PKB substrate Girdin
.
Nat Cell Biol
2008
;
10
:
329
37
.
41.
Quante
M
,
Tu
SP
,
Tomita
H
,
Gonda
T
,
Wang
SSW
,
Takashi
S
, et al
Bone marrow-derived myofibroblasts contribute to the mesenchymal stem cell niche and promote tumor growth
.
Cancer Cell
2011
;
19
:
257
72
.
42.
Lopez-Sanchez
I
,
Dunkel
Y
,
Roh
YS
,
Mittal
Y
,
De Minicis
S
,
Muranyi
A
, et al
GIV/Girdin is a central hub for profibrogenic signalling networks during liver fibrosis
.
Nat Commun
2014
;
5
:
4451
.
43.
Mahadevan
D
,
Von Hoff
DD
. 
Tumor-stroma interactions in pancreatic ductal adenocarcinoma
.
Mol Cancer Ther
2007
;
6
:
1186
97
.
44.
Miyachi
H
,
Mii
S
,
Enomoto
A
,
Murakumo
Y
,
Kato
T
,
Asai
N
, et al
Role of Girdin in intimal hyperplasia in vein grafts and efficacy of atelocollagen-mediated application of small interfering RNA for vein graft failure
.
J Vasc Surg
2013
;
60
:
479
89
.
45.
Anderberg
C
,
Li
H
,
Fredriksson
L
,
Andrae
J
,
Betsholtz
C
,
Li
X
, et al
Paracrine signaling by platelet-derived growth factor-CC promotes tumor growth by recruitment of cancer-associated fibroblasts
.
Cancer Res
2009
;
69
:
369
78
.
46.
Hellberg
C
,
Heldin
CH
. 
Role of PDGF in tumor-stroma interactions
. In:
Mueller
MM
,
Fusenig
NE
, editors. 
Tumor-associated fibroblasts and their matrix
.
Springer
; 
2011
. p.
257
65
.
47.
Tripathi
M
,
Billet
S
,
Bhowmick
NA
. 
Understanding the role of stromal fibroblasts in cancer progression
.
Cell Adh Migr
2012
;
6
:
231
5
.
48.
Rodon
J
,
Dienstmann
R
,
Serra
V
,
Tabernero
J
. 
Development of PI3K inhibitors: lessons learned from early clinical trials
.
Nat Rev Clin Oncol
2013
;
10
:
143
53
.
49.
Pal
I
,
Mandal
M
. 
PI3K and Akt as molecular targets for cancer therapy: current clinical outcomes
.
Acta Pharmacologica Sinica
2012
;
33
:
1441
58
.
50.
Enomoto
A
,
Asai
N
,
Namba
T
,
Wang
Y
,
Kato
T
,
Tanaka
M
, et al
Roles of disrupted-in-schizophrenia 1-interacting protein girdin in postnatal development of the dentate gyrus
.
Neuron
2009
;
63
:
774
87
.

Supplementary data