Expression of Myoferlin in Human and Murine Carcinoma Tumors Cleo Leung, Carol Yu, Michelle I. Lin, Cristina Tognon, Pascal Bernatchez  The American Journal of Pathology  Volume 182, Issue 5, Pages 1900-1909 (May 2013) DOI: 10.1016/j.ajpath.2013.01.041 Copyright © 2013 American Society for Investigative Pathology Terms and Conditions

Figure 1 Myoferlin is expressed in mouse and human carcinoma cell lines and solid tumors. A: Proteins extracted from two cultured mouse tumor cell lines, LLC and B16-BL6, as well as from control and myoferlin-HA plasmid-transfected COS-7 cells, were blotted against myoferlin and showed positive staining for the 240-kDa full-length myoferlin protein and HA-tag. β-COP was used as a negative control. B: Control immortalized nontransformed murine mammary epithelial EpH4 cells and EpH4 cells expressing the transforming ETV6–NTRK3 fusion oncogene, confirmed by the anti-TrkC (NTRK3) doublet blot, showed stable myoferlin expression. HSP90 was used as a loading control. The human nonmetastatic breast cancer cell line T47D, derived from pleural effusion, the hormone-responsive nonmetastatic breast cancer cell line MCF7, and the metastatic carcinoma SUM131501, MDAMB-436, and MDAMB-231 cell lines showed positive but variable myoferlin expression. C: Stable myoferlin/HSP90 expression was found during transformation in the control immortalized human breast epithelial cell line MCF10A, transformed using H-Ras overexpression (MCF10A-Tk1) or ETV6–NTRK3 fusion oncogene overexpression, or both (MCF10A-Tk1/ETV6–NTRK3). D: Immunofluorescence shows robust myoferlin staining (red) in close proximity to the Golgi apparatus (GM130; green) and the rest of the perinuclear region (DAPI; blue). Colocalization is indicated by yellow. E: Positive myoferlin detection (brown) in solid LLC tumors (left), along with H&E stain of an adjacent section (right) reveals tumor morphology. The necrotic core is indicated by white and light blue. Inset: Control IgG condition. F: Positive myoferlin detection (brown) in human lung carcinoma tissue sections (left), along with H&E staining of an adjacent section (right). Inset: Control IgG condition. Scale bars: 1 μm (E and F, left), 5 μm (D), vand 100 μm (E and F, right). Ctrl, control; Myof, myoferlin. The American Journal of Pathology 2013 182, 1900-1909DOI: (10.1016/j.ajpath.2013.01.041) Copyright © 2013 American Society for Investigative Pathology Terms and Conditions

Figure 2 Myoferlin knockdown decreases tumor growth in vivo. A: Cultured LLC cells were pretreated with vehicle (Ctrl), control siRNA, or 75 nmol/L myoferlin siRNA for 48 or 72 hours, and then proteins were isolated and probed for myoferlin or HSP90. B: In vivo knockdown of myoferlin expression decreases tumor burden. After initial implantation and randomization, mice were intratumorally injected with DOTAP/siRNA solution every 3 days in a masked fashion. C: Before tumors reached 2500 mm3, mice were euthanized, the tumors were isolated, and proteins were solubilized and blotted against myoferlin and HSP90. The lowest value was set to 1′. Data are expressed as means ± SEM (B) or as individual ratios and means ± SEM (C). n = 6 (control) or 7 (myoferlin). Experiments were performed in duplicate (B). ∗∗P < 0.01, unpaired Student’s t-test, orthogonal comparisons; ∗∗∗P < 0.001. The American Journal of Pathology 2013 182, 1900-1909DOI: (10.1016/j.ajpath.2013.01.041) Copyright © 2013 American Society for Investigative Pathology Terms and Conditions

Figure 3 Myoferlin knockdown does not decrease blood vessel density in tumors. A: Immunofluorescence images from tumor sections (25 μm thick) treated with anti–PECAM-1/CD31 antibodies coupled to Alexa Fluor 488 secondary antibodies (green) and myoferlin antibody no. 1 coupled to Alexa Fluor 568 antibodies (red). The merged image (right) includes DAPI nuclear counterstain (blue). Insets show a portion of the same image at higher magnification. B: Immunohistochemistry against PECAM-1/CD31 and Mayer’s hematoxylin counterstain using tumor sections (6 μm thick) treated with control or myoferlin siRNA sequences. PECAM-1/CD31–positive structures are indicated by arrows. Insets show a portion of the same image at higher magnification. C: Average (bars) and individual (symbols) quantification of vessel density (PECAM-1/CD31–positive structures) in tumors with control or myoferlin siRNA sequences. n = 6. D: Average (bars) and individual (symbols) quantification of pericyte-covered PECAM-1/CD31–positive vessels at day 22 in treated LLC tumors. Sections from tumors shown in panel A were costained with PECAM-1 and smooth muscle cell actin antisera. Data are expressed as means ± SEM and are representative of two independent experiments with similar results. n = 5. n = 5. Scale bars: 100 μm (A); 75 μm (B). The American Journal of Pathology 2013 182, 1900-1909DOI: (10.1016/j.ajpath.2013.01.041) Copyright © 2013 American Society for Investigative Pathology Terms and Conditions

Figure 4 Myoferlin knockdown decreases proliferation in vivo. A: Representative tumor section stained by immunohistochemistry against KI-67 expression, a proliferation marker, in tumor sections after treatment with control or myoferlin siRNA sequences until day 22. B and C: Average (bars) and individual (symbols) quantification of total KI-67-positive cells (B) in the tumor sections shown in panel A, and KI-67–positive/KI-67–negative cell ratio (C) at day 14 and 22. D: Left panel: Representative histological sections of tumors with control or myoferlin siRNA treatment subjected to TUNEL assay after isolation at day 22. TUNEL-positive cells show robust brown staining. Right panel: Average (bars) and individual (symbols) quantification of TUNEL-positive cells at days 14 and 22. Data are expressed as means ± SEM and are representative of two independent experiments with similar results. n = 5 (myoferlin); n = 6 (control). ∗∗P < 0.01. Original magnification, ×20 μm (A and D). The American Journal of Pathology 2013 182, 1900-1909DOI: (10.1016/j.ajpath.2013.01.041) Copyright © 2013 American Society for Investigative Pathology Terms and Conditions

Figure 5 Myoferlin knockdown decreases proliferation. A: LLC cells were seeded in 12-well plates and transfected with control or myoferlin siRNA sequences at 15 or 75 nmol/L. After starvation, cells were allowed to proliferate for 48 hours or 72 hours and then were counted using a hemocytometer. B: The antiproliferative effect of myoferlin knockdown is not triggered by apoptosis. Treatment of LLC cells with control or myoferlin siRNA does not increase caspase-8 activation, a marker of apoptosis. Fas ligand (FasL) and ceramide (Cer) stimulation were used as positive controls. C and D: Loss of myoferlin drastically increases membrane damage after laser injury in proliferating LLC cells. A small (3 μm × 3 μm) area of membrane damage was induced by confocal laser in the presence of FM1-43, a dye that fluoresces only in lipid membrane structures, and images were captured every minute from time 0 (before injury) to 9 minutes. Quantification (C) and visualization (D) of local fluorescence was performed in LLC cells after siRNA treatment (48 hours), under starvation or proliferation conditions. The effect of myoferlin gene knockdown on the damaged area in starved LLC cells (C) is indicated by a short arrow and in proliferating LLC cells by a long arrow. FM1-43 fluorescence (D) increased in both control and myoferlin siRNA–treated LLC cells under proliferating conditions; The injury site before and 9 minutes after damage is indicated by a white arrow in each image. Data are expressed as means ± SEM. Data in B are representative of three independent experiments with similar results. n = 6. ∗P < 0.05. Original magnification, ×63. The American Journal of Pathology 2013 182, 1900-1909DOI: (10.1016/j.ajpath.2013.01.041) Copyright © 2013 American Society for Investigative Pathology Terms and Conditions