Polyacrylamide (PAA) hydrogels were prepared with stiffness ranging from 0.15, 3.43, and 80 kPa, following established protocols (Pelham and Wang, 1997; Cozzolino et al., 2016). Briefly, 12- or 20-mm round coverslips were wiped with 0.1 N NaOH and then treated with (3-aminopropyl) triethoxysilane (APTES; Sigma-Aldrich, St. Louis, MO). After washing the coverslips three times, the glass surface was treated with 0.5% glutaraldehyde for 30 min and allowed to dry at room temperature. To create PAA hydrogels with different stiffness, mixtures of 40% acrylamide/2% bisacrylamide (Sigma-Aldrich, St. Louis, MO) were prepared with sterile distilled water to a final volume of 1 mL. The solution was degassed for 8–15 min before polymerization was initiated by adding freshly prepared ammonium persulfate (APS; 1% w/v solution; Bio-Rad Laboratories, Inc., Hercules, CA) and N, N, N′, N′-tetramethylenediamine (TEMED; 0.1% w/v solution; Sigma-Aldrich, St. Louis, MO). After a brief agitation, PAA solution was added onto 12 mm × 15 mm coverslips previously coated with dichlorodimethylsilane (DCDMS; Sigma-Aldrich, St. Louis, MO). Then the 12 or 20 mm functionalized round coverslips were placed over each hydrogel droplet of PAA solution and allowed to polymerize for 30 min. Once polymerization was complete, round coverslips with the PAA hydrogel were carefully removed and placed in 24- or 12-well culture plates, washed three times with 50 mM HEPES (J.T. Baker, Thermo Fisher Scientific Inc., Phillipsburg, NJ) at pH 8.5. PAA hydrogels were photoactivated using a 1 mM solution of sulfosuccinimidyl 6-(4′-azido-2′-nitrophenylamino)hexanoate (sulfo-SANPAH; Sigma-Aldrich, St. Louis, MO) in 50 mM HEPES under UV light for 8 min, repeated twice, and then washed three times with the HEPES solution. Finally, all PAA hydrogels were coated with poly-L-lysine (PLL 500 μg/mL; Sigma-Aldrich, St. Louis, MO) for 30 min, washed three times with 1× Phosphate Buffered Saline (PBS) at pH 7.4, and incubated with RPMI 1640 growth culture media for 1 h before cell seeding.

All cell lines used in this study were obtained commercially from ATCC, including tumoral human cell lines: epithelioid cervix carcinoma cells (HeLa, ATCC CRL-CCL-2) and a metastatic prostate cancer cell line (PC-3, ATCC CRL-1435), as well as two non-tumoral human cell lines: a normal human primary prostate epithelial cell line (HPrEC, ATCC PCS-440-010) and human embryonic kidney 293 cells (HEK293, ATCC CRL-157340). Cell lines were authenticated by the National Institute of Genomic Medicine (INMEGEN, Mexico) before conducting the assays.

HEK293 is a transformed cell line derived from isolated human embryo kidney cells and is considered non-tumorigenic within 52 passages; tumors are induced in nude mice only with high passage cells, as previously reported (Shen et al., 2008). HPrEC cells are derived from normal primary prostate epithelial cells, exhibiting normal epithelial morphology, and a limited lifespan in culture (Gil et al., 2005). HeLa cells originate from cervical carcinoma and are highly tumorigenic, widely used as a model for tumor formation (Kusakawa et al., 2015). PC-3 cells, derived from bone metastasis of grade IV prostatic adenocarcinoma, lack androgen receptor and prostate-specific antigen and are considered highly metastatic and tumorigenic (Kaighn et al., 1979).

Cells were cultured in RPMI-1640 medium (Sigma-Aldrich, St. Louis, MO), or prostate epithelial cell basal medium (Prostate Epithelial cell Growth kit, ATCC PCS-440-040) for HPrEC, supplemented with 8% fetal bovine serum (FBS; Biowest, Nuaillé, France) and 1% penicillin/streptomycin (Gibco, Life Technologies Corporation, Grand Island, NY), and maintained in a 37°C/5% CO₂ atmosphere. Cells from passages 2-10 were seeded on the PAA hydrogels at a density of 1 × 10⁴ cells per cm² and incubated at 37°C/5% CO₂ for 24 or 48 h.

The cells from all 4 cell lines were fixed with 4% paraformaldehyde (PFA, Sigma-Aldrich, St. Louis, MO) after being seeded on the 12 mm PAA hydrogels for 24 h. Micrographs were captured using a Nikon Eclipse TS100 phase contrast microscope (Nikon Instruments Inc, Japan) equipped with a CCD camera, using a 20x lens. The Fiji software (ImageJ2, Wane Rasband, U. S. National Institutes of Health, Bethesda, MD) was used for image analysis (Schindelin et al., 2012).

Morphological parameters quantified included cell area, aspect ratio (ratio of the major axis to the minor axis of the ellipse for each cell), circularity (expressed as 4π times the ratio of the area to the perimeter, where a value close to 1 indicates a perfect circle and close to 0 indicates greater elongation of the cell), and solidity (calculated by dividing the cell area by its convex area, assuming the inner angles are less than 180°, convex and that it contains the original region; a more branched cell results in a larger convex area and a lower solidity value) (Zanier et al., 2015). At least 12 images per coverslip were captured. Three independent experiments were performed, each with triplicates per condition.

Total RNA extraction from cultured cells on PAA hydrogels at different time points was conducted following the TRIsure RNA isolation protocol (Bioline Reagents Ltd., London, United Kinngdom). RNA concentration and purity were determined using a NanoDrop ND-1000 (Thermo Fisher Scientific, Waltham, Massachusetts, United States). cDNA was synthesized by random priming from 500 ng of total RNA using M-MLV Reverse Transcriptase (Invitrogen, ThermoFisher Scientific, Waltham, Massachusetts, United States), according to the manufacturer’s instructions.

Real-time PCR reactions were performed in triplicate with 1 µL of cDNA, 5.5 µL of 2 × SensiFast SYBR qPCR mix (Green Bioline), 5 pmol of primers, and water up to 10 µL. The sequences for each pair of primers used were: NRP1 (forward 5′-AGG​ACA​GAG​ACT​GCA​AGT​ATG​AC-3’; reverse 5′-AAC​ATT​CAG​GAC​CTC​TCT​TGA-3′) and the values of mRNA expression were normalized to that of 18S rRNA (forward 5′-GTA​ACC​CGT​TGA​ACC​CCA​TT-3′; reverse 5′-CCA​TCC​AAT​CGG​TAG​TAG​CG-3′) or YWHAZ, as it has been reported as a reliable reference gene for cancer cells (Chua et al., 2011), (forward 5′-TTG​AGC​AGA​AGA​CGG​AAG​GT-3′; reverse 5′-GAA​GCA​TTG​GGG​ATC​AAG​AA-3′). The cycling conditions for qPCR were: 2 min at 95°C, 40 cycles of 15 s at 95°C, and 1 min at 60°C in a 7500 Real-Time PCR System (Applied Biosystems, Foster City, CA). PCR efficiencies were calculated using the LinReg program (Ruijter et al., 2009), and relative gene expression was calculated with the Pfaffl equation (Pfaffl, 2001). Data were normalized according to the expression in cells cultured over the softest stiffness condition (0.15 kPa) at 12 or 24 h. Three independent experiments with triplicates were performed.

Cells cultured for 24 and 48 h on PAA hydrogels with different stiffness (0.15 kPa, 3.43 kPa, and 80 kPa) were immunostained with the anti-NRP1 antibody (APC anti-human CD304; Neuropilin-1, BioLegend, catalog 354505 or PE anti-Nrp1; R&D Systems, catalog FAB566P). After culture time, each coverslip was transferred to a clean sterile well, washed twice with 1 × PBS, and treated with trypsin/EDTA 0.25% solution (Gibco, Life Technologies Corporation, Grand Island, NY) or trypsin/EDTA for primary HPrEC cells until the cells detached from PAA hydrogel. Trypsin was then inactivated by adding an equal volume of RPMI medium with 8% FBS, or 1% trypsin neutralizing solution for HPrEC cells. The cell suspension was centrifuged for 5 min at 300 xg, washed with 1x PBS at 37°C, and the supernatant was discarded. Cells were stained for 45 min using the anti-NRP1 antibody. After the incubation, the cells were washed with 1x PBS with 1% albumin (1% PBA), centrifuged for 5 min at 300 xg and the supernatant was discarded. Cells were fixed with 4% formaldehyde (FA; ThermoFisher Scientific, catalog 28906). Flow cytometry acquisitions were performed using a BD LRS Fortessa flow cytometer (Becton-Dickinson Bioscience, San Jose, CA) and analyzed with the FlowJo software v10.6.2 (FlowJo, TreesStar, Ashland, OR). Three independent experiments were performed for each condition.

Cell migration was assessed using time-lapse microscopy in cells cultured for 48 h in PAA hydrogels (0.15 kPa, 3.43 kPa, and 80 kPa) on 20 mm diameter coverslips. The coverslips with the cells were transferred to an incubation chamber, and growth media with HEPES (final concentration of 0.01 M) was added. The chamber was then mounted on a stagetop incubator, and phase-contrast microscope images were captured at 20-s intervals, with a total recording time of 2 h 30 min. Image analysis was performed using the Fiji software (ImageJ2, Wane Rasband, U. S. National Institutes of Health, Bethesda, MD).To distinguish between migratory and non-migratory cells, a box encompassing the nucleus of each cell was delineated. Cells that displaced their nucleus at least one-third outside the box during the recording period were considered migratory (Grundy et al., 2016). Total distance and average displacement speed were quantified for each cell.

Manual tracking of the cell movement of each migratory cell was carried out using the Trackmate (Manual Tracking) plug-in of the Fiji software (Tinevez et al., 2017; Ershov et al., 2021). To visualize the cell displacement trajectories, the trajectory of each cell was transposed to a common origin using the Excel macro Diper Plot_At_Origin.txt (Gorelik and Gautreau, 2014), and plots were generated based on the x, and y coordinates of the trajectory data. Furthermore, using the Excel macro MSD.txt of the Diper program (Gorelik and Gautreau, 2014), the area explored by each cell was quantified by plotting the mean square displacement (MSD) against half the recording time (75 min). From the MSD, the slope α of the log-log curve of the MSD data was calculated for the cells in each stiffness condition. This value was used to determine the persistence of cell movement, where α = 1 represents random movement and α = 2 represents directed motion. Three independent experiments were performed for each condition.

PC-3 cells cultured for 24 h on PAA hydrogels with different stiffness (0.15 kPa, 3.43 kPa, and 80 kPa), or silenced cells cultured for 48 h, were fixed with methanol free 4% formaldehyde (Pierce™, ThermoFisher Scientific, catalog 28906), blocked with 5% horse serum in 1x PBS, and immunostained overnight with 1:500 dilution of anti-NRP1 antibody (anti-human Neuropilin 1, R&D systems, catalog AF3870) in 5% horse serum in 1x PBS. After several washes with 1x PBS, cells were incubated with a 1:800 dilution of an anti-goat IgG coupled to Alexa Fluor 488 (Invitrogen, catalog A11055), washed again in 1x PBS, and incubated with phalloidin conjugated to Alexa Fluor 546 (Invitrogen, catalog A22283). Cell cultures were washed and mounted on coverslips with a drop of 90% glycerol containing Hoechst 33342 (Invitrogen, catalog H3570). Immunostained cells were observed using confocal microscopy (Leica TCS SP8 AOBS or Leica TSP8), with images acquired using a ×63 oil immersion objective. Z-stacks of 1 µm were obtained, and maximum projection images were generated using Fiji ImageJ software.

PC-3 cells cultured on 12-well plates were transfected with either 20 or 30 pmol of Silencer^® Select Validated siRNA-NRP1 (Invitrogen) using Lipofectamine RNAiMax reagent (Invitrogen, catalog 13778075). Alternatively, cells were double transfected with a non-relevant siRNA coupled to Cy3 (Invitrogen, catalog AM4620) and siRNA-NRP1. After 24 h of transfection, cells were detached and seeded onto PAA hydrogels. They were incubated for 48 h and transferred to the incubation chamber for time-lapse migration assays, or fixed for immunostaining.

The data were assessed for normality using either the Kolmogorov-Smirnov or Shapiro-Wilk test. Normally distributed data were analyzed using the one-way ANOVA test, followed by Tukey’s multiple comparisons test or Student’s t-test. For data that were not normally distributed, Kruskal-Wallis’s test was used, followed by Dunn’s multiple comparisons test with the Bonferroni correction (Dunn Bonferroni’s test). Differences with a value of p < 0.05 were considered significant. All analyses were performed using IBM SPSS Statistics 20.0 (IBM Corp., Armonk, NY), and graphs were generated using GraphPad Prism 9.0.2 (GraphPad Software, La Jolla, CA).

To compare the influence of stiffness on cell morphology, we examined non-tumoral cell lines HPrEC and HEK293, as well as tumoral HeLa and PC-3 cells, cultured on PAA hydrogels with varying stiffness, a low stiffness, non-related to the tissue of origin of the cell lines (0.15 kPa), an intermediate stiffness (3.43 kPa) akin to normal prostatic or cervical tissue (Jiang et al., 2014; Piao et al., 2017), and a higher stiffness (80 kPa), associated with prostate and cervical tumors (Porsch et al., 2015; Cui et al., 2017; Rouvière et al., 2017). Phase contrast micrographs revealed that substrate stiffness significantly influenced cell morphology across all cell lines. Specifically, all the cells appeared more rounded and smaller on softer substrates (0.15 kPa). Notably, HEK293 and HeLa cells remain mainly rounded even on the intermediate stiffness (3.43 kPa), while some HPrEC and PC-3 cells exhibited a less rounded and more extended morphology at this stiffness. At the highest stiffness (80 kPa), all cell lines displayed a similar morphology (Figure 1A).

Quantitative analysis of four morphological parameters shows that cell area increased as PAA hydrogel stiffness increased in all cell lines (Figure 1B). Moreover, cell circularity decreased with increasing substrate stiffness, and cells were less rounded as the substrate stiffness increased, being more evident in HPrEC and PC-3 cells (Figure 1C). The aspect ratio, a measure of cell elongation, increased with stiffness, albeit without significant differences observed between intermediate and higher stiffness in HeLa and PC-3 cells. Noteworthy was the higher elongation of HPrEC cells at the intermediate stiffness (3.43 kPa) compared to other cell lines (Figure 1D). Additionally, cell solidity, a descriptor of cell deformability potentially relevant to metastatic spread (Pasqualato et al., 2013), decreased (indicating increased deformability), with increasing stiffness in HEK293, HeLa, and PC-3 cells, while only HPrEC cells showed decreased solidity at the intermediate stiffness (3.43 kPa) compared to the stiffer substrate (Figure 1E).

These findings underscore the role of substrate stiffness in modulating the morphology of all the assayed cell lines, indicating that all were mechanoresponsive; however, there are discernible differences in sensitivity or response to stiffness among them, with HPrEC and PC-3 cells exhibiting more pronounced morphological changes at the intermediate and higher stiffness substrates (3.43 kPa and 80 kPa). Particularly notable is the elongated and branched morphology observed in HPrEC cells at the intermediate stiffness (3.43 kPa), in contrast to the other cell lines.

To analyze whether substrate stiffness modulates the expression of NRP1, a protein associated with cancer progression, we analyzed NRP1 mRNA expression in four different cell lines following 24 h of culture on PAA hydrogels. Our findings revealed distinct patterns of NRP1 mRNA expression among the cell lines. Specifically, non-tumoral HPrEC and HEK293 cells exhibited a significant increase in mRNA expression as stiffness increased, whereas this difference was not observed in tumoral PC-3 and HeLa cells. Notably, HPrEC cells, a non-tumoral prostate epithelial cell line, demonstrated a remarkable increase in NRP1 mRNA expression when cultured on substrates mimicking tumor-like stiffness (80 kPa) (Figure 2A).

To further explore the impact of substrate stiffness on NRP1 mRNA expression we compared the expression of the non-tumoral HPrEC and tumoral PC-3 cells at different times of incubation. We observed variations in the peak expression of NRP1 mRNA in HPrEC and PC-3 cells, with distinct responses to both time and substrate stiffness. In PC-3 cells, the peak of expression occurred at the 3.43 kPa stiffness after 12 h of incubation, remaining stable until 48 h. Conversely, HPrEC cells exhibited a significant peak in expression at 80 kPa tumor-like stiffness substrate after 24 h of incubation (Figure 2B). These results confirm the impact of stiffness on NRP1 expression and highlight differential mRNA induction dynamics between non-tumoral and tumoral cells, suggesting that non-transformed cells require stiffer conditions and longer incubation times to increase NRP1 mRNA expression. In contrast, PC-3 tumoral cells, demonstrate increased NRP1 expression at stiffness akin to normal prostate tissue, and with shorter incubation periods.

Flow cytometry analysis of NRP1 protein expression revealed differences in the percentage of NRP1-expressing cells among the different cell lines. HEK293 and HeLa cells exhibited less than 20% NRP1-positive cells, with even lower percentages on stiffer gels, while HPrEC and PC-3 showed higher percentages, exceeding 20% and 50%, respectively, on softer gels, and about 40% and 80% on stiffer substrates (Figure 2C). Similar to mRNA findings, substrate stiffness exerts differential effects on NRP1 protein expression across cell lines. As stiffness increases, HEK293 cells at 24 h and PC-3 cells after 48 h of incubation displayed a higher percentage of positive cells. Conversely, NRP1 expression decreased with increasing stiffness in HEK293 cells at 48 h, as well as in HeLa cells at both 24 and 48 h (Figure 2C). Although HPrEC cells exhibited no changes in NRP1-positive cells across when cultured on the different substrate stiffnesses; however, a tendency toward increased expression on stiffer substrates was observed (Figure 2C). Mean fluorescent intensity (MFI) remains unchanged in HPrEC, HEK293, and HeLa cells under all conditions, suggesting that stiffness does not regulate the amount of protein at the cell membrane in these cell lines (Figure 2D). In contrast, in PC-3 cells, NRP1 MFI increased on the substrate with intermediate stiffness (3.43 kPa) compared to softer and stiffer substrates, indicating that this stiffness favored NRP1 protein expression in these cells.

Cell migration is a fundamental aspect related to tumor progression and metastasis. Thus, we examined whether substrate stiffness has differential effects on cell migration in both non-tumoral and tumoral cells. Specifically, we investigated the impact of substrate stiffness on the migration of HPrEC and PC-3 cells, given their sensitivity to stiffness, using time-lapse microscopy 48 h post culture on soft (0.15 kPa), intermediate (3.43 kPa), and stiff (80 kPa) PAA hydrogels.

Our first observation revealed different cell protrusion modes at the different substrate stiffnesses; cell membranes exhibited blebs, pseudopodia, or transitions between blebs and pseudopodia. In the softest PAA hydrogels, blebs were the predominant membrane protrusions in both HPrEC and PC-3 cells, with other protrusion modes emerging as the substrate stiffened (Figure 3A; Supplementary Videos SV1-S6). HPrEC cells show alternating blebs and lamellipodia, with a minority of cells maintaining solely blebs (Figure 3A). PC-3 cells exhibit blebs at 3.4 kPa, but an increasing number of cells displayed bleb-pseudopodia combinations; in some instances, small lamellipodia or filopodia were formed. Notably, at 80 kPa, certain PC-3 cells adopted an elongated morphology with small and active pseudopodia that abruptly retracted and impulse forward cell displacement (Figure 3A; Supplementary Video SV6). Quantitative analysis of the percentage of cells exhibiting each protrusion mode confirmed that PAA hydrogel stiffness influences protrusive activity. The proportion of PC-3 cells displaying only pseudopodia increases at 80 kPa, while decreasing the number of blebbing cells, more prevalent at 3.43 kPa. Conversely, HPrEC cells showed no significant changes in the percentage of protrusions on substrates with stiffnesses of 3.43 and 80 kP (Figure 3B).

Analysis of cell migration revealed that HPrEC cells exhibit high migratory behavior, with over 90% of cells migrating on both, intermediate and stiff substrates, but not on the softest gels, where only about 2% of the cells migrated. Similarly, PC-3 cell migration was influenced by substrate stiffness; a significantly higher percentage of migrating cells (78%) was observed on stiffer substrates compared to softer ones, where the percentages were 20% and 35%, respectively (Figure 4B). In migratory PC-3 cells, both total displacement distance and average velocity increased with the stiffness of PAA hydrogels, indicating that cells traveled further and at higher average velocity on stiffer substrates (Figures 4C,D). In contrast, HPrEC migratory cells traveled greater distances when cultured on 80 kPa stiffer substrate compared to cells cultured on 3.43 kPa substrate; however, there was no significant difference in average cell velocity among the substrate stiffnesses, although there was a tendency for increased velocity with increasing stiffness (Figures 4C,D).

Interestingly, trajectory plots of migrating cells confirmed that both HPrEC and PC-3 cells explore greater distances as stiffness increases. However, HPrEC cells explored a larger area on both intermediate and stiffer substrates compared to PC-3 cells, which explored a smaller area when cultured on the 3.43 kPa substrate (Figure 5A). Analysis of cell displacement using mean square displacement (MSD) demonstrated that HPrEC cells exhibited significantly larger displacements at 3.43 and 80 kPa, whereas for PC-3 cells, the difference was only observed at 80 kPa (Figure 5B). Notably, α values indicated that higher stiffness led to greater persistence of migration in PC-3 cells, whereas HPrEC cells did not show significant differences in α values across different substrate stiffnesses. However, the persistence values of HPrEC movement on intermediate and stiffer substrates were higher than those of PC-3 cells, suggesting that HPrEC cells maintained a straight, persistent movement (<1) unaffected by the stiffness, unlike PC-3 cells, which exhibited a more random movement (−1) on softer substrates (Figure 5C).

The distribution and remodeling of the actin cytoskeleton play a crucial role in determining cell shape and migration. To characterize the distribution of actin and NRP1 in tumoral cells cultured on substrates with varying stiffness, we performed phalloidin staining and NRP1 immunostaining in PC-3 cells. We observed changes in polymerized actin as stiffness increased. On 0.15 kPa substrates, PC-3 cells exhibited few polymerized actin, mainly distributed as small dots and as subcortical actin. In contrast, in cells cultured on 3.43 and 80 kPa PAA hydrogels actin is distributed at the cells’ leading edges, in actin arcs and stress fibers, predominantly in cells on the stiffer substrate (Figure 6A). In the softest 0.15 kPa gels, NRP1 appeared as small dots distributed throughout the cells. However, on 3.43 and 80 kPa substrates, NRP1 localization was predominantly concentrated at the cells’ leading edge. Particularly noteworthy was the formation of larger NRP1 aggregates on the stiffer substrate (Figure 6A, arrowheads). Interestingly, NRP1 colocalized with actin at the polarized regions of the cells’ leading edges on 3.43 and 80 kPa substrates, as well as with the actin stress fibers under stiffer conditions (Figure 6A, arrows).

To investigate the role of NRP1 in cell morphology and cytoskeleton, we used siRNA targeting NRP1 to knock down its expression in tumoral PC-3 cells (Figures 6B–D). Upon transfection with siRNA-NRP1 and subsequent culturing on coverslips, we observed a decrease in cell spreading and an increase in rounded cell morphology (Figure 6E, arrowheads), as compared with cells with no relevant siRNA coupled to Cy3 or those lacking siRNA-NRP1. Evaluation of the impact of NRP1 knockdown on actin cytoskeleton distribution in cells cultured on coverslips revealed that non-silenced cells exhibited a spread morphology, characterized by lamellipodia and filopodia, along with strong cortical actin staining distributed in arcs and actin fibers (Figure 6F, arrows). In contrast, NRP1 knockdown cells displayed reduced spreading, assuming a rounded shape devoid of lamellipodia, and filopodia in many instances. Actin fibers were absent, with only scattered actin dots observed (Figure 6F, arrowheads). When we analyzed the effect of NRP1 knockdown on the cells cultured on the PAA hydrogels we observed similar trends, albeit less pronounced due to the inherently less spread morphology of cells compared to those on coverslips. On 80 kPa PAA hydrogels, silenced cells exhibited a more rounded morphology devoid of actin fibers, with diminished cortical actin and actin arcs. Some cells displayed only small aggregates of actin (Figure 7C). In contrast, effects were less evident in cells cultured on 3.43 and 0.15 kPa PAA hydrogels due to their less spread morphology and the lack of stress fibers, even in the non-silenced cells (Figures 7A,B).

To identify whether NRP1 plays a role in the migration of tumoral cells, we performed time-lapse imaging of silenced PC-3 cells cultured on 80 kPa gels. Surprisingly, NRP1 knockdown did not appear to hinder the number of migrating cells; the percentages of migratory versus non-migratory cells were similar between non-transfected and siRNA-NRP1 cultures (Figure 8A). Interestingly when we examined the modes of cell protrusion during migration. We found significant differences between silenced and control cells. siRNA-NRP1 led to a decrease in the number of cells exhibiting pseudopodia protrusions, while concurrently increasing the number of cells displaying a bleb-pseudopodia protrusive mode (Figure 8B, Supplementary videos SV-9, SV-10).