Fig 1: Inverse tumorigenicity of ECM1 subtypes.a Detection of ECM1 mRNA levels in CLs of ovarian cancer cell lines (Hey, HeyA8, SKOV3, SKOV3ip1, OVCA433, OVCA429, and A2780) and a normal cell line (HOSE) by qRT-PCR. ACTB (encoding β-actin) was used to normalize the expression of ECM1. Data are presented as mean ± SD. n = 3 biologically independent repeats. Representative images of ECM1 staining performed by IHC in normal ovary (b), normal fallopian tube (c), and ovarian carcinoma (OC) (d) tissues. The arrows indicate typical tissues (normal or cancer epithelium) stained with ECM1. Bars with 200 or 100 μm indicate the magnification of images. Silencing efficiency of ECM1 with specific shRNAs detected by WB (e), 3D culture image showing spheroid formation (f, bars = 400 µm) and numbers (g, data are presented as mean ± SD, n = 3 biologically independent repeats, two-tailed t-test), and xenograft tumor growth (h, data are presented as mean ± SD, n = 8v8 mice, two-tailed t-test) in eight mice generated from cells expressing ECM1 shRNA or control shRNA. Functional examination of ECM1 subtypes in HeyA8-ECM1 shRNA cells (A8i) after transfection of ECM1a (A8i-A), ECM1b (A8i-B), or empty vector (A8i-V), as confirmed by WB analysis (i); 3D culture images (j, bars = 400 µm) and spheroid number (k, data are presented as mean ± SD, n = 3 biologically independent repeats, two-tailed t-test); and assessment of xenograft tumor growth (l, data are presented as mean ± SD, n = 8v8 mice, two-tailed t-test) and tumor tissues (m) in eight mice injected with cells expressing ECM1a or ECM1b. n Alterations in signaling molecules associated with AKT/FAK/Paxillin/Rac after silencing of ECM1 and overexpression of ECM1a and ECM1b. Detection of the expression of phosphorylated proteins in cells treated or not treated with an HA antibody (o, A8i-A cells) or treated/not treated with bioactive ECM1 (ECM1a) (p, OVCA429 cells). PBS and mouse IgG were used as controls. β-actin was used as a loading control for WB analysis.
Fig 2: The ECM1a GPR motif determines signaling and tumorigenesis.a Detection of HA-tagged ECM1a-MT expression in an established cell line by WB analysis. 3D cell growth (b, bars = 400 µm) and quantification (c, data are presented as mean ± SD, n = 3 biologically independent repeats, two-tailed t-test) induced by ECM1a-MT or ECM1a-WT. Xenograft tumor growth (d, data are presented as mean ± SD, n = 10v10 mice, two-tailed t-test) and tissues (e) in ten mice and cellular signaling molecules (f, the actin loading control in f left panel was also used for the same samples in Fig. 4g because these tests were parallelly performed) were inhibited by ECM1a-MT. g Loss of direct binding of integrin αX or β2 to ECM1a-MT, in contrast with the binding to ECM1a-WT, as tested by co-IP. Cellular colocalization of integrin αX or β2 with ECM1a-WT or ECM1a-MT as detected by IF (h, bars = 10 µm) and PLA (i, bars = 10 µm). Detection of integrins αX and β2 and of WT and MT ECM1 by WB analysis of cell membrane extracts (j) and detection of integrins αX and β2 (k) by co-IP/WB analysis after cell surface labeling with biotinylation reagents.
Fig 3: ABCG1 mediates ECM1a-induced cellular signaling and tumorigenesis.Detection of ABCG1 protein expression by WB analysis in ECM1-silenced/overexpressing cells (a), in ECM1a-WT/ECM1a-MT-expressing cells (b), in integrin αXβ2-silenced cells (c), and in ECM1b- or ECM1a-cross-transfected cells (d). Direct binding and colocalization between ABCG1 and AKT/pAKT, paxillin/pPaxillin, or myosin/pMyosin as detected by co-IP/WB analysis (e) and IF (f, bars = 10 μm). Stable silencing or overexpression of ABCG1 (g) inhibited or promoted the number (h, data are presented as mean ± SD, n = 3 biologically independent repeats, two-tailed t-test) and growth (i, bars = 400 μm) of 3D culture spheroids, xenograft tumor growth (j, data are presented as mean ± SD, n = 9v9 mice, two-tailed t-test) and tumor tissues (k) in nine mice injected with related cells, and cellular signaling molecules associated with ECM1a and integrin αXβ2 (l).
Fig 4: ABCG1 confers cancer cell stemness and cisplatin resistance through upregulation of CD326.Inhibition rates of various cell lines treated with cisplatin in terms of silencing or overexpression of ECM1, ABCG1, and hnRNPLL (a, data are presented as mean ± SD, n = 3 biologically independent repeats, two-tailed t-test was calculated between A8i-ABCG1 and A8i or A8i-A ) and IC50 values (b, data are presented as mean ± SD, n = 3 biologically independent repeats, two-tailed t-test) of cisplatin in these cells as detected by CCK-8 assay. c Expression of stem cell transcription factors detected by WB analysis in ECM1-, ABCG1-, and hnRNPLL-silenced or -overexpressing cells. d Numbers of CD117+ and CD326+ cells (left panel, data are presented as ± SD, n = 3 biologically independent repeats, two-tailed t-test) and cisplatin inhibition rates (right panel, data are presented as ± SD, n = 3 biologically independent repeats) after three selection-and-culture cycles from CD117 + cells. e Numbers of CD117 + and CD326 + cells (left panel, data are presented as mean ± SD, n = 3 biologically independent repeats, two-tailed t-test) and cisplatin inhibition rates (right panel, data are presented as mean ± SD, n = 3 biologically independent repeats, two-tailed t-test was calculated between CD117+ and CD326+ or CD117+/CD326+ and CD117−/CD326−) after four selection-and-culture cycles from CD326+ cells. f A low concentration of cisplatin enriches CD326+ cells to increase cancer cell stemness. Data are presented as mean ± SD, n = 3 biologically independent repeats, two-tailed t-test. g Cisplatin treatment upregulates the expression of stem cell transcription factors in CD326+ cells.
Fig 5: hnRNPLL determines ECM1a mRNA splicing and ECM1a-associated signaling and tumor growth.hnRNPLL protein expression in ECM1-silenced and ECM1a-overexpressing cells (a), in ECM1a-WT/ECM1a-MT cells (b), and in ECM1a- and ECM1b-cross-transfected cells (c). Overexpression of hnRNPLL in A8i cells (d) promotes ECM1a mRNA splicing, as tested by semiquantitative RT-PCR (e) and quantification (f, data are presented as mean ± SD, n = 3 biologically independent repeats, two-tailed t-test). Silencing of hnRNPLL in A8 cells with specific shRNAs (g) reduces ECM1a mRNA production but upregulates ECM1b mRNA production, as detected by semiquantitative RT-PCR (h) and quantification (i, data are presented as mean ± SD, n = 3 biologically independent repeats, two-tailed t-test). hnRNPLL silencing (j) or overexpression inhibits or enhances 3D cell growth as shown by spheroid quantification (k, data are presented as mean ± SD, n = 3 biologically independent repeats, two-tailed t-test) and formation (l, bars = 400 μm), xenograft tumor growth (m, data are presented as mean ± SD, n = 9v9 mice, two-tailed t-test) and tissues (n) in nine mice, and activation of ECM1-associated signaling molecules (o), respectively.
Supplier Page from R&D Systems, a Bio-Techne Brand for Recombinant Human ECM1 Protein, CF