*2.8. 3D Cell Invasion Assay*

The 3D invasion assay was performed as previously described [25]. Briefly, for the generation of spheroids, we combined 1/4 of the final volume of methylcellulose (6 mg/mL), 10% (*w*/*v*) FBS, 20 to 40 μg/mL collagen, 1 μg/mL mitomycin in DMEM medium. Five hundred cancer cells were then suspended in 50 μL of this spheroid formation medium and plated into a non-adherent 384-well plate. Cells were incubated for 24 h to allow the formation of an individual spheroid per well. The medium was then replaced by 50 μL of collagen matrix composed of two parts of collagen (3 mg/mL), 15% (*w*/*v*) FBS, 1 μg/mL mitomycin, 2% (*w*/*v*) NaOH (1 M) and DMEM. Deflamin, at a concentration of 50 μg/mL and 100 μg/mL, was added to the treated cells. After 1 h incubation, imaging of the spheroids was performed by fluorescence microscopy in a Zeiss LSM710 confocal microscope. This time point was accounted as 0 h. The next images were taken at the time points of 24, 48, and 72 h. Images were analyzed using the FIJI/ImageJ software. Invasive cells were counted between the inner perimeter (spheroid external border at 0 h) and the outer perimeter (the furthest invasive cell at each time point).

### *2.9. Cell Labeling*

CRC cell lines at 70% confluence were stained with lipophilic dye Vybrant™ DiI cell labeling solution (V22885 from Invitrogen) at 4 μL/mL in PBS 1X for 10 min at 37 ◦C.

### *2.10. In Vivo Zebrafish Xenograft Model*

Wild-type and transgenic Tg(kdrl:EGFP) zebrafish (*Danio rerio*) [26] embryos were provided by the zebrafish facility of Instituto de Medicina Molecular (iMM). For husbandry, adult zebrafish were maintained at 28.5 ◦C in a 10/14 h dark-light cycle, according to standard protocols of the European Animal Welfare Legislation, Directive 2010/63/EU (European Commission, 2016), following the Federation of European Laboratory Animal Science Associations (FELASA) guidelines and recommendations. All procedures in this study were performed in early life forms of zebrafish development, with embryos up to 120 h post-fertilization (hpf), that do not yet show the ability to feed themselves and are, therefore, considered unprotected under the European Animal Welfare Legislation, Directive 2010/63/EU (European Commission, 2016).

For zebrafish injection, HCT116 cells expressing dTomato fluorescent protein or labeled with DiI were used as indicated. 2.5 × <sup>10</sup><sup>5</sup> cells/mL (approximately 800 cells per injection) were microinjected into the perivitelline space (PVS) of 48 hpf larvae previously anesthetized with tricaine 1.5% (*w*/*v*; Pharmaq), as described before [27]. Microinjections were performed under a stereo microscope (Leica S8 APO) using borosilicate glass microcapillaries attached to a microinjector (World Precision Instruments, Pneumatic Pico pump PV820) coupled with a micromanipulator (Narishige MN-153). At 24 h post-injection, successfully injected xenografts were treated with 100 μg/mL deflamin in E3 medium or E3 with PBS (controls) and incubated at 34 ◦C. The medium was replaced daily for three days. After this period, animals were fixed in 4% (*w*/*v*) paraformaldehyde overnight and stored at −20 ◦C in 100% (*v*/*v*) methanol.

#### *2.11. Immunofluorescence*

Frozen xenografts were re-hydrated in methanol series (75% > 50% > 25%) and then permeabilized in acetone at −20 ◦C. Xenografts were then washed in a buffer containing 1x PBS, 0.5% (*v*/*v*) Tween 20, 0.5% TritonX-100, and 100 mM glycine for 1 h at room temperature (RT), followed by blocking in 1x PBS, 1% *w*/*v* BSA, 1% *v*/*v* DMSO, 1% *v*/*v* TritonX-100 and 1.5% *w*/*v* FBS, for 1 h at RT. Next, xenografts were incubated with primary antibodies: rabbit monoclonal anti-cleaved caspase-3 (Asp175) (1:100, #9661 from Cell Signaling); and mouse monoclonal anti-GFP (1:100, #11814460001 from Roche) for 1 h at RT and overnight at 4 ◦C. The following day, secondary antibody incubation with goat Anti-rabbit IgG H&L Alexa Fluor® 488 (1:400, #A-11008 from Invitrogen) and 50 μg/mL DAPI was performed for 1 h at RT and then overnight at 4 ◦C. Xenografts were mounted with Vectashield® mounting media between two coverslips and stored at 4 ◦C for subsequent analysis.

Xenografts were imaged in a confocal microscope Zeiss LSM 710 with a Z-stack interval of 5 μm. Images were analyzed in FIJI/ImageJ software using the plugin cell counter. For tumor size calculation, three representative slices of the tumor from the top (Zfirst), the middle (Zmidle), and the bottom (Zlast) were analyzed and the number of cells calculated as the sum of cells in Zfirst, Zmidle, Zlast/total number of Z stacks × 1.5. The 1.5 correction number was estimated for these CRC cells that have nuclei with an average of 10–12 μm of diameter. The number of mitotic figures and activated caspase-3 were counted in every slice of the tumor (from Zfirst to Zlast) and the percentage was obtained by dividing the value by the total number of cells in tumor size.

Metastasis was counted in the caudal hematopoietic tissue and in the gills of zebrafish larvae.

For the analysis of angiogenesis, all images were obtained with a 7 μm interval Z-stack and two parameters were measured: vessel density (VD) and vessel infiltration (VI). Vessel density and vessel infiltration were assessed through Z-projections of corresponding images using the ImageJ Z-Projection tool and the area of eGFP fluorescent signal per tumor was quantified. To analyze the vessel infiltration, the superficial slices of the images were not considered (about 20% of total stacks).

$$\text{VD} = \frac{\text{eGFP area}}{\text{Turnour area}} \times 100, \text{ VI} = \frac{\text{eGFP area in tumour core}}{\text{core of the tumour area}} \times 100$$

For degraded collagen analysis, the 5-FAM fluorescence area corresponding to collagen degradation points per tumor was measured, using Z-projections of corresponding images with ImageJ Z-projection tool. The value was normalized by the total tumor area.

$$\text{Degraded collagen} = \frac{\text{eGFP area}}{\text{Turnour area}} \times 100^{\circ}$$

#### *2.12. Statistical Analysis*

GraphPad Prism 8 software (Dotmatics, San Diego, USA) was used for the statistical analysis of in vivo experiments. Pared, unpaired t-test and one-way ANOVA were used to analyze in vitro and in vivo data, as indicated in figure legends. Results are presented as average ± standard error of the mean (SEM). The level of statistical significance was set as non-significant (NS); \*, *p* < 0.05; \*\*, *p* < 0.01; \*\*\*, *p* < 0.001; and \*\*\*\*, *p* < 0.0001. For all the statistical analysis *p* value as a confidence interval of 95%.
