The alternatively-included 11a sequence modifies the effects of Mena on actin cytoskeletal organization and cell behavior

During tumor progression, alternative splicing gives rise to different Mena protein isoforms. We analyzed how Mena11a, an isoform enriched in epithelia and epithelial-like cells, affects Mena-dependent regulation of actin dynamics and cell behavior. While other Mena isoforms promote actin polymerization and drive membrane protrusion, we find that Mena11a decreases actin polymerization and growth factor-stimulated membrane protrusion at lamellipodia. Ectopic Mena11a expression slows mesenchymal-like cell motility, while isoform-specific depletion of endogenous Mena11a in epithelial-like tumor cells perturbs cell:cell junctions and increases membrane protrusion and overall cell motility. Mena11a can dampen membrane protrusion and reduce actin polymerization in the absence of other Mena isoforms, indicating that it is not simply an inactive Mena isoform. We identify a phosphorylation site within 11a that is required for some Mena11a-specific functions. RNA-seq data analysis from patient cohorts demonstrates that the difference between mRNAs encoding constitutive Mena sequences and those containing the 11a exon correlates with metastasis in colorectal cancer, suggesting that 11a exon exclusion contributes to invasive phenotypes and leads to poor clinical outcomes.


G-actin purification and labeling
G-actin was extracted from rabbit muscle acetone powder and gel-filtered over a Superdex-200 gel filtration column, using standard techniques. Gel filtered G-actin was polymerized to F-actin in F-actin buffer (1 mM ATP, 5 mM MgCl 2 , 50mM KCl, 50mM Tris/HCl, pH 8.0) and labeled with Rhodamine-X succinimidyl ester (Invitrogen) following manufacturer's instructions. F-actin was depolymerized in G-actin buffer (0.2 mM ATP, 0.5 mM DTT, 0.2 mM CaCl 2 , 2 mM Tris/HCl, pH 8.0) to G-actin, and passed through PD-10 columns (GE Healthcare) to get rid of free rhodamine.

Barbed ends assay
MTLn3 cells were starved for 4 hours in L15 medium supplemented with 0.35% BSA.
For stimulation, cells were treated with bath application of 0.5 nM or 5 nM EGF at 37°C, and 60 or 180 seconds later were permeabilized with 0.125 mg/ml saponin (Sigma) in the presence of 0.5 mM rhodamine-conjugated G-actin. After 1 minute of labeling, samples were fixed in 0.5% glutaraldehyde in cytoskeleton buffer, permeabilized with 0.5% Triton X-100 in cytoskeleton buffer, quenched in 100mM Na-Borohydride in PBS, and blocked in the presence of CF405-phalloidin (Biotium). Images were taken with a deconvolution microscope. The ratio of barbed end intensity to phalloidin intensity along the edge was quantified as described in Image Quantification below.

Kymography
DIC time-lapse sequence movies of MTLn3 cells were 5 minutes long; frames were taken every 3 seconds with a 40X DIC oil immersion objective. Kymographs were produced and analyzed using Metamorph or ImageJ. Kymographs were generated along 1-pixel-wide line regions oriented along individual protrusions. For quantitative analysis, straight lines were drawn on kymographs from the beginning to the end of individual protrusion events, and slopes were used to calculate velocities; line projections along the x-axis (time) were used to calculate the persistence of protrusions.
The protrusion time is the total time that the membrane is engaged in a protrusion, over the time of imaging.
Images were captured on film using a TEM JEOL 200CX. Films were scanned and an unsharp mask filter was applied to the pictures in Adobe Photoshop.

Spreading assay
Spreading assays were performed as described 7 with modifications. MV D7 EGFP and EGFP-Mena isoform expressing cells were trypsinized, re-plated on coverslips coated with laminin (20 µg/ml, Southern Biotech), allowed to spread for 20 minutes, and fixed in cold methanol for 15 minutes at -20°C. Coverslips were washed twice with PBS, cells were permeabilized for 3 minutes at room temperature with 0.2% Triton-X 100 in PBS, washed twice in PBS, and incubated with the mouse monoclonal Fascin antibody to stain filopodia. Cell phenotypes were analyzed and categorized in cells with or without filopodia.

Differential interference contrast (DIC) microscopy
For live cell imaging experiments, cells were plated on glass bottom dishes (MatTek Corporation), treated with 1M HCl for 5 minutes, followed by 70% ethanol and PBS washes. Cells were imaged with an ORCA-ER camera (Hammamatsu) attached to a Nikon TE300 microscope, using either 10x DIC/0.30NA or 40xDIC/1.3NA Nikon objectives. During time lapse, MTLn3 cells were kept at 37°C with aid of a Solent Incubator chamber (Solent Inc.) fitted for the microscope. All images were collected, measured and compiled with Metamorph imaging software (Molecular Devices) and ImageJ.

Deconvolution wide field fluorescent microscopy
Cells were plated on glass coverslips coated with 100 μg/ml rat-tail Collagen type I (BD bioscience), or 10 μg/ml bovine plasma fibronectin (for MV D7 cells) (Sigma), fixed in 4% For tissue staining, 5 μm sections were deparaffinized in xylene, treated with a graded series of alcohol, rehydrated in PBS and subjected to heat-induced antigen retrieval in 10 mM citrate buffer (pH 6.0). Sections were preincubated with 10% normal donkey or goat serum in 0.5% Tween-20 for 2 hours at room temperature, incubated with primary antibodies in 1% donkey or goat serum and 0.5% Tween-20 buffer over night at 4°C, washed 3 times in PBS and incubated in fluorescently labeled secondary antibodies (AlexaFluor, Molecular Probes) for 2 hours at room temperature, and in Hoechst to label the DNA.
z-series of cells and tissues were imaged using a Deltavision microscope using SoftWoRx acquisition software (Applied Precision) or a Nikon Ti inverted microscope using NIS Elements acquisition software (Nikon), a 40X and 60X 1.4 NA Plan-Apochromat objective lens (Olympus) or a 40X 1.15 NA Plan-Apochromat objective lens (Nikon), and a camera (CoolSNAP HQ; Photometrics or a Zyla4.2 sCMOS; Andor, respectively). Images taken with the Deltavision microscope were deconvolved using Deltavision SoftWoRx software and objective-specific point spread function. Images were combined in Adobe Photoshop for presentation.

3-Dimensional Structural Illumination Microscopy.
Cells were imaged with an OMX-3D Super-resolution microscope (Applied Precision/GE) equipped with 405 nm, 488 nm, 594 nm lasers and 3 Photometrics Cascade II, EMCCD cameras. Images were acquired with a 100X, NA 1.4 oil objective, at 0.125 μm z step, using 1.512 immersion oil. All images were acquired under the same illumination settings (405 nm laser at 19% strength, for 100 msec, 488 nm laser at 1% strength for 150 msec, and 594 nm laser at 50% strength for 100 msec) and then processed with OMX softWoRx software (Applied Precision). Images were saved as .tiff of maximum projections of 8 x 0.125 micron z section stack.

Image quantification
Signal intensities from antibodies or rhodamine-labeled barbed ends along the cell edge were quantified with a published contour-based ImageJ macro 8 . We measured the distribution of signal along the membrane plotted as a function of distance from the cell edge (mean ± SEM) and the sum of the intensities in the first 0.65 μm from the cell edge.

Supplementary Movie Legends
Supplementary Movie S1

Mena cells.
MTLn3 cells stably expressing GFP, GFP-Mena11a, and GFP-Mena were stimulated with 5nM EGF for 300 seconds (starting at 60 seconds). Images were analyzed by time-lapse differential interference contrast microscopy (TE300, Nikon). Frames were taken every 3 seconds for 360 seconds.

Mena11aS>A cells.
MTLn3 cells stably expressing GFP-Mena11a and GFP-Mena11aS>A were stimulated with 5nM EGF for 300 seconds (starting at 60 seconds). Images were analyzed by time-lapse differential interference contrast microscopy (TE300, Nikon). Frames were taken every 3 seconds for 360 seconds.