Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
Filters applied. Clear all
. 2015 Dec 29;13(12):2699-714.
doi: 10.1016/j.celrep.2015.11.052. Epub 2015 Dec 17.

Cdc42EP3/BORG2 and Septin Network Enables Mechano-transduction and the Emergence of Cancer-Associated Fibroblasts

Affiliations
Free PMC article

Cdc42EP3/BORG2 and Septin Network Enables Mechano-transduction and the Emergence of Cancer-Associated Fibroblasts

Fernando Calvo et al. Cell Rep. .
Free PMC article

Abstract

Cancer-associated fibroblasts (CAFs) are non-cancerous cells found in solid tumors that remodel the tumor matrix and promote cancer invasion and angiogenesis. Here, we demonstrate that Cdc42EP3/BORG2 is required for the matrix remodeling, invasion, angiogenesis, and tumor-growth-promoting abilities of CAFs. Cdc42EP3 functions by coordinating the actin and septin networks. Furthermore, depletion of SEPT2 has similar effects to those of loss of Cdc42EP3, indicating a role for the septin network in the tumor stroma. Cdc42EP3 is upregulated early in fibroblast activation and precedes the emergence of the highly contractile phenotype characteristic of CAFs. Depletion of Cdc42EP3 in normal fibroblasts prevents their activation by cancer cells. We propose that Cdc42EP3 sensitizes fibroblasts to further cues-in particular, those activating actomyosin contractility-and thereby enables the generation of the pathological activated fibroblast state.

Figures

None
Figure 1
Figure 1
Identification of Cdc42EP3 as a Key Regulator of CAF Functions (A) The left panels show images of F-actin (red) and paxillin (green) staining in NF#1, NF#2, CAF#1, and CAF#2. The right panels show images of F-actin (red) and pS19-MLC2 (green) staining in NF#2 and CAF#2. Scale bars represent 20 μm. (B) Histograms show focal adhesion (FA) area normalized to total cell area (left graph), an F-actin stress fibers (SF) component (middle graph), and pMLC2 associated to SF (right graph) of NF and CAF cells. Bars represent mean ± SEM. FA measurements, n = 16; measurements were performed on NF#1 and NF#2 (NF) and CAF#1 and CAF#2 (CAF). F-actin and pMLC2 measurements, n = 12 for NF and n = 15 for CAF; measurements were performed on NF#1 and CAF#1. (C) Western blot showing pY118-paxillin (pPax), pY416-Src (pSFK), and pS19-MLC2 (pMLC2) in three sets of NF and CAFs (NF#1, NF#2, NF#3, CAF#1, CAF#2, and CAF#3). A tubulin blot is also shown. (D) RNAi-based loss-of-function screen in CAFs. The indicated genes (left column) were knocked down by smart-pool siRNA, and the functional outcomes were evaluated in gel contraction (CAF#1) and organotypic invasion assays (CAF#1 and CAF#2). The values indicate the fold activity normalized to control siRNA-transfected CAFs for each experiment (red is low; blue is high). Genes are ranked on their average score (right column). (E) Western blot and quantification showing Cdc42EP3 and tubulin expression in CAF#1 after transfection with control and three different Cdc42EP3 siRNA (siCEP3). Bars represent mean ± SEM (n = 4). (F) Histogram shows gel contraction by CAF#1 following transfection with control or three different Cdc42EP3 siRNA (siCEP3). Bars represent mean ± SEM (n = 3). (G) Images and quantification of 4T1 invasion when co-cultured with CAF#1 transfected with control or three single Cdc42EP3 siRNA (siCEP3). Bars represent mean ± SEM (n = 14 organotypic assays assessed over three experiments). Scale bar, 50 μm. (H) Western blots showing Cdc42EP3, Cdc42EP5, and 14-3-3 levels in three sets of NF and CAFs (NF#1–3, CAF#1–3). The chart shows Cdc42EP3 and Cdc42EP5 fold expression (normalized to 14-3-3). See also Figure S1.
Figure 2
Figure 2
CAFs Have Enhanced Actin and Septin Cytoskeletal Networks (A) Panels show SEPT2 (green) and SEPT7 (red) staining in NF#1 and CAF#1. Zoom areas are shown for each channel and cell type. Scale bars represent 10 μm. See Figure S2A for quantification. (B) Western blot showing the fibrillar (F) and soluble (G) content of actin, SEPT2, and SEPT7 in two NF lines (NF#1-2) and four CAF lines (CAF#1-4). A 14-3-3 blot is also shown as an internal cytosolic control. See Figure S2C for quantification. (C) Representative SIM images of SEPT2-GFP (yellow) and F-actin (magenta) staining in CAF#1. White arrows indicate SEPT2-GFP filaments along actin stress fibers. The green arrows indicate SEPT2-GFP filaments connecting nearby stress fibers. Scale bars represent 1 μm. (D) Panels show actin (red), endogenous Cdc42EP3 (green), and DAPI (blue) staining of CAF#1. Cells were fixed using methanol. Zoom-up area showing actin and Cdc42EP3 is shown. Bar represents 10 μm. (E) Representative SIM image of F-actin (magenta) and Cdc42EP3-GFP (green) in CAF#1. Bar represents 1 μm. White arrows indicate Cdc42EP3-GFP filaments associated to actin stress fibers. Yellow arrows indicate Cdc42EP3-GFP wavy filaments that do not co-align with F-actin. (F) Single channel and merged confocal images of SEPT2-Cherry (red), Cdc42EP3-GFP (green), and F-actin (blue) in CAF#1. White arrows indicate SEPT2- and Cdc42EP3-containing filaments that co-align with F-actin. Yellow arrows indicate SEPT2- and Cdc42EP3-containing filaments that interconnect F-actin cables. Scale bar represents 1 μm. (G) Representative single channel and merged SIM images of SEPT2-Cherry (red), Cdc42EP3-GFP (green), and F-actin (blue) in CAF#1. Scale bars represent 1 μm. See also Figure S2 and Movies S1, S2, S3, and S4.
Figure 3
Figure 3
Cdc42EP3 Regulates the Actin and Septin Fibrillar Networks (A) Images show pS19-MLC2 (green), F-actin (red), and DAPI (blue) staining of CAF#1 following control and Cdc42EP3 siRNA (siCEP3, smart pool). The grayscale panels show individual channel magnifications of the indicated areas. Scale bars, 25 μm. See Figure S3A for quantification. (B) Images show SEPT2 (green), SEPT7 (red), and DAPI (blue) staining of CAF#1 following control and Cdc42EP3 siRNA (siCEP3, smart pool). The grayscale panels show individual channel magnifications of the indicated areas. Scale bars, 25 μm. See Figure S3B for quantification. (C) Western blot showing the fibrillar (F) and soluble (G) content of actin, SEPT2, and SEPT7 in CAF#1 following control and Cdc42EP3 siRNA (siCEP3, smart pool). A 14-3-3 blot is also shown as a cytosolic control. See Figure S3D for quantification. (D) Box and whisker plot showing the Young’s elastic modulus of CAF#1 cells with control and Cdc42EP3 siRNA (siCEP3, smart pool). The central box represents the lower to upper quartiles; the middle line represents the mean. The horizontal line extends from the minimum to the maximum value (n = 61 single measurements for siCtr; n = 53 single measurements for siCEP3). (E) Western blot showing the fibrillar (F) and soluble (G) content of actin, SEPT2, and SEPT7 in empty vector (−) and Flag-Cdc42EP3 (Flag-CEP3)-transfected NF#1. The histogram shows the F/G ratios of actin, SEPT2, and SEPT7. Bars represent mean ± SEM (n = 3). (F) Top left inserts show GFP (green), SEPT2 (blue), and SEPT7 (red) staining of NF#1 following transfection with GFP or GFP-tagged Cdc42EP3 (CEP3) proteins. The grayscale panels show individual channel magnifications of perinuclear areas. Scale bars, 20 μm. (G) Top left inserts show GFP (green) and actin (red) staining of NF#1 following transfection with GFP or GFP-tagged Cdc42EP3 (CEP3) proteins. The grayscale panels show individual channel magnifications of perinuclear areas. Scale bars, 20 μm. (H) Histograms show fibrillar SEPT2&7 area (top) and F-actin stress fiber (SF) component (bottom) for the indicated experimental points. Bars represent mean ± SEM (5 < n < 10 single-cell measurements). See also Figure S3.
Figure 4
Figure 4
The Septin Cytoskeleton Is Required for CAF Functions (A) Images showing F-actin (red), pS19-MLC2 (green), and DAPI (blue) staining of CAF#1 following control, Cdc42EP3, SEPT2, and DIAPH1&3 siRNA transfection (smart pools). Scale bars, 20 μm. Histograms show an F-actin stress fiber (SF) component and pMLC2 associated to SF for the indicated points. Bars represent mean ± SEM (12 < n < 15). (B) Images show F-actin (red), paxillin (green), and DAPI (blue) staining of CAF#1 following with control, Cdc42EP3, SEPT2, and DIAPH1&3 siRNA transfection (smart pools). Scale bars, 20 μm. The histogram shows the focal adhesion (FA) area normalized to total cell area in single cells. Bars represent mean ± SEM (10 < n < 12). (C) Western blot showing phosphorylated levels of pY118-paxillin (pPax), pY416-Src (pSFK), and pS19-MLC2 (pMLC2) in CAF#1 following transfection with control, Cdc42EP3 (CEP3), DIAPH1&3, and SEPT2 siRNA (smart pools). Cdc42EP3 and tubulin blots are also shown. (D) Western blots showing fibrillar (F) and soluble (G) actin and SEPT2 in CAF#1 following control, Cdc42EP3 (CEP3), SEPT2, SEPT7, and DIAPH1&3 siRNA (smart pools). The histogram shows the F/G ratios of actin and SEPT2. Bars represent mean ± SEM (n = 5). (E) Histogram shows gel contraction by CAF#1 following control, SEPT2, and SEPT7 siRNA (smart pools). Bars represent mean ± SEM (n = 8). (F) Histogram shows the elastic modulus of matrices remodeled by NF#1 or CAF#1 following control, Cdc42EP3, or SEPT2 siRNA (smart pools). Bars indicate mean ± SD (n = 167 single measurements for NF-siCtr; n = 142 for CAF-siCtr; n = 173 for CAF-siCEP3; n = 51 for siSEPT2). (G) Quantification of 4T1 invasion when co-cultured with CAF#1 transfected with control or SEPT2 and SEPT7 siRNA (smart pool). Bars represent mean ± SEM (n = 9 organotypic assays assessed over two experiments). (H) Charts show second harmonic generation (SHG) of collagen fibers and elastin-positive areas in the CAF#1 matrix plugs transfected with control, Cdc42EP3 (CEP3), and SEPT2 siRNA. Line and error bars indicate mean ± SEM (n = 4 plugs from different mice). See also Figure S4.
Figure 5
Figure 5
Cdc42EP3 Is Required for the Pro-tumorigenic Activity of CAFs In Vivo (A) mRNA levels of angiogenic factors Vegfa, Vegfc, and Tgfb1 in NF#1 and CAF#1 following transfection with control and Cdc42EP3 siRNA (CEP3, smart pool). Bars represent mean ± SEM (n = 12). (B) Secreted levels of VEGF-A on CAF#1 following transfection with control, Cdc42EP3 (CEP3), SEPT2, or SEPT7 siRNA (smart pool). Bars represent mean ± SEM (n = 6). (C) Representative images showing endomucin and vimentin staining of matrix plugs with NF#1 or CAF#1 transfected with control, Cdc42EP3 (CEP3), and SEPT2 siRNA (smart pools) and injected sub-cut in mice. Scale bars, 100 μm. (D) Chart shows quantification of endomucin staining relative to fibroblast number (vimentin staining). Line and error bars indicate mean ± SEM (n = 7, except for CAF#1-siCtr that n = 9 and CAF#1-siSEPT2 that n = 5; each experimental point is an individual plug from a different mouse). (E) Box and whisker plot showing volumes of orthotopic tumors generated by co-injection of TS1 murine MMTV-PyMT breast cancer cells with CAF#3 stably expressing control or two different shRNA targeting Cdc42EP3. The central box represents the lower to upper quartiles; the middle line represents the mean; and the horizontal line extends from the minimum to the maximum value (n = 7, except for CAF-shCtr that n = 6). Tumors were measured at day 24 post-injection. (F) Survival curves for orthotopic tumors generated by co-injection of TS1 murine MMTV-PyMT breast cancer cells with CAF#3 stably expressing control or two different shRNA targeting Cdc42EP3. The survival curve represents the percentage of animals alive at the indicated time point after injection. Survival curves were estimated by the Kaplan-Meier method and compared among subsets using the log rank test. (G) Representative images of orthotopic tumors generated by co-injection of TS1 murine MMTV-PyMT breast cancer cells with CAF#3 stably expressing control or two different shRNA targeting Cdc42EP3 and stained for H&E or αSMA. Scale bar, 500 μm. (H) Representative images of orthotopic tumors generated by co-injection of TS1 murine MMTV-PyMT breast cancer cells with CAF#3 stably expressing control or two different shRNA targeting Cdc42EP3 and stained for H&E or endomucin. Scale bar, 200 μm. See also Figure S5.
Figure 6
Figure 6
Cdc42EP3 Is Required for Mechano-sensing (A) Images show single and merged channels of SEPT2-GFP (yellow) and F-actin (magenta) staining of NF#1 and CAF#1 when seeded on soft (∼500 Pa) and stiff (∼12 kPa) matrices. Scale bar represents 10 μm. (B) Images show F-actin (red) and pY118-Paxillin (pPax) staining of NF#1 when seeded on stiff (∼12 kPa) matrices following transfection with GFP-CAAX and Cdc42EP3-GFP. Inserts show GFP signal (grayscale). Scale bar represents 10 μm. (C) Images show GFP (green) and YAP (red) staining of NF#1 when seeded stiff (∼12 kPa) matrices following transfection with Cdc42EP3-GFP. Scale bar represents 10 μm. (D) Histogram shows nuclear YAP fluorescent intensity/cytoplasmic YAP fluorescent intensity of NF#1 when seeded on soft (∼500 Pa) and stiff (∼12 kPa) matrices following transfection with GFP-CAAX or Cdc42EP3-GFP. The mean and 95% confidence interval are shown (23 < n < 80 cells). (E) Fold mRNA levels of YAP target genes Ankrd1 and Ctgf in stable cell lines NF#1 null, NF#1-Cdc42EP3low, and NF#1-Cdc42EP3high maintained in 1% FBS for 6 hr. Bars represent mean ± SEM (n = 5). (F) Images show F-actin (red) and paxillin (green) staining of CAF#1 when seeded on soft (∼500 Pa) and stiff (∼12 kPa) matrices following transfection with control, Cdc42EP3, or SEPT2 siRNA (smart pools). Scale bar represents 10 μm. (G) Images show F-actin (red) and YAP (green) staining of CAF#1 when seeded on soft (∼500 Pa) and stiff (∼12 kPa) matrices following transfection with control, Cdc42EP3, or SEPT2 siRNA (smart pools). Scale bar is 10 μm. (H) Histogram shows nuclear YAP fluorescent intensity/cytoplasmic YAP fluorescent intensity for the indicated conditions. The mean and 95% confidence interval are shown (23 < n < 80 cells). (I) Relative mRNA levels of YAP target genes Ankrd1 and Ctgf in NF#1 and CAF#1 following transfection with control, Cdc42EP3 (CEP3), and YAP siRNA (smart pools). Bars represent mean ± SEM (n = 6). See also Figure S6.
Figure 7
Figure 7
Cdc42EP3 Is Induced and Required Early in Fibroblast Activation (A) Western blots of Cdc42EP3 in fibroblasts isolated from different disease stages (NF, HpAF, AdAF, and CAF). (B) Relative levels of Cdc42ep3 mRNA in fibroblasts from different disease stages. Bars indicate mean ± SEM (n = 4). (C) Western blot showing the fibrillar (F) and soluble (G) actin, SEPT2, and SEPT7 in fibroblasts from different disease stages. A 14-3-3 blot is also shown as an internal cytosolic control. (D) Western blots showing pY118-paxillin (pPax), pY416-Src (pSFK), and pS19-MLC2 (pMLC2) and tubulin levels in fibroblasts isolated from different disease stages. (E) GSEA-derived heatmap showing the relative levels of mRNA expression of YAP target genes (Cyr61, Amotl2, Ankrd1, and Ctgf) in fibroblasts isolated from different disease stages on duplicate. Colors range from dark red to dark blue, representing, respectively, the highest and lowest normalized expression of the gene. (F) Western blots showing pY118-paxillin (pPax), pY416-Src (pSFK), and pS19-MLC2 (pMLC2) in NF#1 subjected to 1% FBS (NM), TGFβ (2 ng/ml), and breast cancer cell conditioned media (CM) for 48 hr. Cdc42EP3 and actin blots are also shown. (G) Western blots of fibrillar (F) and soluble (G) fractions of actin and SEPT2 in NF#1 subjected to 1% FBS (NM) or breast cancer cell conditioned media (CM) following transfection with control and two single Cdc42EP3 (CEP3) siRNA. (H) Western blots showing tubulin and pY118-paxillin (pPax), pY416-Src (pSFK), and pS19-MLC2 (pMLC2) in NF#1 subjected to 1% FBS (NM) or breast cancer cell conditioned media (CM) following transfection with control and two single Cdc42EP3 (CEP3) siRNA. (I) Histogram showing gel contraction of NF#1 subjected to normal media (NM) or breast cancer cell conditioned media (CM) following transfection with control and two single Cdc42EP3 (CEP3) siRNA. Bars represent mean ± SEM (n = 5). (J) Histogram showing organotypic cancer cell invasion of NF#1 subjected to normal media (NM) or breast cancer cell conditioned media (CM) following transfection with control two single Cdc42EP3 (CEP3) siRNA. Bars represent mean ± SEM (n = 5). See also Figure S7.

Similar articles

See all similar articles

Cited by 32 articles

See all "Cited by" articles

References

    1. Bahou W.F., Campbell A.D., Wicha M.S. cDNA cloning and molecular characterization of MSE55, a novel human serum constituent protein that displays bone marrow stromal/endothelial cell-specific expression. J. Biol. Chem. 1992;267:13986–13992. - PubMed
    1. Bhowmick N.A., Moses H.L. Tumor-stroma interactions. Curr. Opin. Genet. Dev. 2005;15:97–101. - PMC - PubMed
    1. Calvo F., Sahai E. Cell communication networks in cancer invasion. Curr. Opin. Cell Biol. 2011;23:621–629. - PubMed
    1. Calvo F., Ege N., Grande-Garcia A., Hooper S., Jenkins R.P., Chaudhry S.I., Harrington K., Williamson P., Moeendarbary E., Charras G., Sahai E. Mechanotransduction and YAP-dependent matrix remodelling is required for the generation and maintenance of cancer-associated fibroblasts. Nat. Cell Biol. 2013;15:637–646. - PMC - PubMed
    1. Deakin N.O., Turner C.E. Paxillin comes of age. J. Cell Sci. 2008;121:2435–2444. - PMC - PubMed

Publication types

Feedback