Invasion of glioma cells through confined space requires membrane tension regulation and mechano-electrical coupling via Plexin-B2

Chrystian Junqueira Alves; Theodore Hannah; Sita Sadia; Christy Kolsteeg; Angela Dixon; Robert J Wiener; Ha Nguyen; Murray J Tipping; Júlia Silva Ladeira; Paula Fernandes da Costa Franklin; Nathália de Paula Dutra de Nigro; Rodrigo Alves Dias; Priscila V Zabala Capriles; José P Rodrigues Furtado de Mendonça; Paul A Slesinger; Kevin D Costa; Hongyan Zou; Roland H Friedel

doi:10.1038/s41467-024-55056-6

. 2025 Jan 2;16:272. doi: 10.1038/s41467-024-55056-6

Invasion of glioma cells through confined space requires membrane tension regulation and mechano-electrical coupling via Plexin-B2

Chrystian Junqueira Alves ^1,^✉,^#, Theodore Hannah ¹, Sita Sadia ¹, Christy Kolsteeg ¹, Angela Dixon ², Robert J Wiener ³, Ha Nguyen ¹, Murray J Tipping ⁴, Júlia Silva Ladeira ⁵, Paula Fernandes da Costa Franklin ⁵, Nathália de Paula Dutra de Nigro ⁵, Rodrigo Alves Dias ⁶, Priscila V Zabala Capriles ⁵, José P Rodrigues Furtado de Mendonça ⁶, Paul A Slesinger ¹, Kevin D Costa ³, Hongyan Zou ^1,^7,^✉,^#, Roland H Friedel ^1,^7,^✉,^#

PMCID: PMC11697315 PMID: 39747004

Abstract

Glioblastoma (GBM) is a malignant brain tumor with diffuse infiltration. Here, we demonstrate how GBM cells usurp guidance receptor Plexin-B2 for confined migration through restricted space. Using live-cell imaging to track GBM cells negotiating microchannels, we reveal endocytic vesicle accumulation at cell front and filamentous actin assembly at cell rear in a polarized manner. These processes are interconnected and require Plexin-B2 signaling. We further show that Plexin-B2 governs membrane tension and other membrane features such as endocytosis, phospholipid composition, and inner leaflet surface charge, thus providing biophysical mechanisms by which Plexin-B2 promotes GBM invasion. Together, our studies unveil how GBM cells regulate membrane tension and mechano-electrical coupling to adapt to physical constraints and achieve polarized confined migration.

Subject terms: Cell invasion, CNS cancer, Endocytosis

The biomechanical mechanisms enabling the invasive growth of brain tumors remain opaque. Here, Junqueira Alves et al. reveal that the guidance receptor Plexin-B2 controls membrane tension, facilitating confined migration of brain tumor cells.

Introduction

Cancer has the hallmarks of uncontrolled growth and high invasiveness. This is particularly pertinent for glioblastoma (GBM), the most common and lethal brain cancer that is notorious for wide dissemination in the brain^1–3. However, the underlying mechanisms of GBM invasion remain poorly understood.

Infiltrating GBM cells experience physical constraints from the tight interstitial space. Earlier research has focused on molecular drivers of cell intrinsic motility (e.g., mesenchymal shift)⁴, but how GBM cells adapt to mechanical forces to adjust migratory behavior is less clear. Physical constraints can lead to nuclear deformation, cell body distortion⁵, and plasma membrane stretching or ruffling. In intracranial patient-derived xenotransplant (PDX) models, we frequently observed GBM cells invading through the corpus callosum, with nuclei and cell bodies transformed into elongated fusiform shape along white matter tracts⁶. These observations align with reports that tumor cells can acquire in response to constriction a polarized migratory phenotype to rapidly escape crowded tissue regions⁵. Recent evidence also unveiled polarization of inner membrane surface charge and of lipid composition during cell migration⁷. How GBM cells orchestrate biophysical features in response to spatial constraints to achieve polarized migration is not fully understood.

As tumor cells frequently usurp developmental pathways during malignant transformation, we have been investigating axon guidance molecules in GBM invasion, particularly Plexin-B2, a membrane receptor initially cloned as a gene upregulated in GBM⁸. We have shown that Plexin-B2 expression inversely correlates with survival for GBM patients⁹ and that Plexin-B2 promotes GBM invasion in PDX models⁶, but the role of Plexin-B2 in confined migration and the underlying mechanisms have not been studied.

Aside from axon pathfinding, plexins also regulate diverse processes of immune activation, bone homeostasis, vascular remodeling, and cancer progression¹⁰. Plexins are transmembrane proteins that emerged more than 600 million years ago, well before the appearance of a nervous system^11,12. They have a highly conserved ring-shaped extracellular domain and an intracellular GTPase activating protein (GAP) domain that deactivates Ras/Rap small GTPases to control cytoskeleton dynamics^13,14. While canonical plexin signaling occurs through dimerization upon binding to semaphorin ligands, as exemplified in growth cone collapse^15,16, recent evidence also implicates a mechanosensory function, which was shown in endothelial cells that use Plexin-D1 to detect blood flow sheer stress via the bendable extracellular ring¹⁷ and epidermal stem cells that respond to compression forces via Plexin-B1/B2¹⁸. In echo, our own work has revealed that human embryonic stem cells (hESCs) and neural progenitor cells (hNPCs) engage Plexin-B2 to control actomyosin tension and cell stiffness for multicellular organization and neurodifferentiation¹⁹.

Here, we used microchannel migration devices to investigate how GBM cells negotiate tight space, mimicking physical constraints experienced by invading cells. Live cell imaging coupled with advanced fluorescent dyes and protein probes unveiled orchestrated biomechanical and mechano-electrical adaptations during confined migration. Our studies demonstrate the requirement of Plexin-B2 and of the associated biophysical mechanisms during confined migration.

Results

Passage through confined space enhances migratory momentum and consistency of GBM cells

To gauge the capability of GBM cells for confined migration, we conducted invasion assays in microchannels containing a glass floor (coated with laminin) and silicon walls of 10 µm height and 12 µm width with periodic 3 or 8 µm constrictions (Fig. 1a). No chemoattractants were added, thus migration through the microchannels was spontaneous, driven only by intrinsic motility. Due to the laminin coating of the channels, the migration of GBM cells occurred in adherent mode (not suspended), similar to adherence-based migration modes described for other cell types^20,21. To ascertain whether confined migration might be impacted by GBM molecular subtypes, we tested two GSC lines–SD2 (mesenchymal) and SD3 (proneural)²²—both showing wide infiltration in intracranial transplants⁶.

Fig. 1 — a Left, schematic illustration of spontaneous migration of GSCs in microchannels with periodic 3 or 8 µm constrictions. Right, still frames from videography show SD2 GSCs labeled with NucSpot invading into microchannels at 12 hr after seeding. Chevrons point to constrictions. b Diagrams and quantification of velocity across constrictions (n = 20 cells per condition) and sums of distances traveled by forward or backward movements over 17 hr. For sums of distances traveled forward: 3 µm, n = 31 cells; 8 µm, n = 38 cells. For sums of distances traveled backward: 3 µm, n = 34 cells; 8 µm, n = 40 cells. Box plots show 25–75% percentiles, minimal and maximal values (whiskers), median (line), and mean (cross). Mann–Whitney–Wilcoxon test, two-sided. c Diagram and box plots show velocity ratio when crossing 2nd vs. 1st constriction. n = 10 cells per group. Two-sided unpaired t-test. Box plots show 25–75% percentiles, minimal and maximal values (whiskers), median (line), and mean (cross). Two-sided unpaired t-test. d Still frames from live-cell fluorescent imaging of SD2 GSCs traversing constrictions (note different time frames for 3 or 8 µm). Dashed lines outline cell boundary. Arrows points to F-actin at rear zone (SPY555-actin) or MemGlow-labeled endosomes at front zone of cell. e Quantifications of fluorescence intensity ratio for SPY-actin or MemGlow at rear vs. front zones during confined migration through 3 or 8 µm constrictions. n = 10 cells per group. Two-sided unpaired t-test. Bar graphs show means ± SEM. f Left, experimental timeline for endocytosis inhibitor treatment before microchannel invasion assay. Right, still frames from videography show stalled SD2 GSCs treated by inhibitors. g Box plots of velocity through constriction, stalling time at constrictions, and sum of forward and backward movements affected by endocytosis inhibitors. Box plots show 25–75% percentiles, minimal and maximal values (whiskers), median (line), and mean (cross). For velocity across constrictions: vehicle, n = 20 cells; Dynasore, n = 16 cells; Pitstop2, n = 15 cells; EIPA, n = 16 cells. For stalling time at constriction: vehicle, n = 22 cells; Dynasore, n = 19 cells; Pitstop2, n = 26 cells; EIPA, n = 28 cells. For sums of distances traveled forward: vehicle, n = 20 cells; Dynasore, n = 16 cells; Pitstop2, n = 15 cells; EIPA, n = 16 cells. For sums of distances traveled backward: vehicle, n = 20 cells; Dynasore, n = 16 cells; Pitstop2, n = 15 cells; EIPA, n = 16 cells. Kruskal–Wallis test followed by Dunn’s multiple comparisons test. h Schematic illustration of polarized migration through constrictions, which further augments migration consistency. Note regionalized accumulation of endocytic vesicles at cell front and F-actin assembly at rear to propel cells through 3 µm constriction. Source data are provided as a Source Data file.

To visualize cell movement, GSCs were prelabeled with live cell nuclear dye NucSpot before seeding in entry ports. Time-lapse live cell imaging up to 21 hours revealed that GSCs passed more efficiently through 3 µm than 8 µm constrictions, evidenced by more cells migrating in the channels and higher migration velocity (Figs. 1a, b; S1a, b; Supplementary Movie 1). Moreover, passage through the 3 µm constrictions enhanced migration consistency (i.e., directional polarity), reflected by higher sums of distance of forward movements compared to backward movements (Figs. 1b; Supplementary Movie 1). These findings applied to both SD2 and SD3 GSCs, with SD2 traversing the 3 µm constrictions at slightly lower velocity than SD3 (~0.7 vs. 1 µm/min) (Supplementary Movie 2). Migration distance in channels with 8 µm constrictions was comparable to migration in open channels without constrictions (Figs. S1c; Supplementary Movie 3), indicating that the 8 µm constriction did not delay migration nor cause confinement for GSCs.

When following the same GBM cell traversing two consecutive 3 µm constrictions, we observed that passage through the first constriction increased the velocity for the second for both SD2 and SD3 GSCs (Figs. 1c; S1d; Supplementary Movie 1). Also of note, hNPCs derived from hESCs displayed much lower migratory capacity in microchannels than GSCs (Figs. S1e; Supplementary Movie 4), indicating cell type-specific biomechanics harnessed by GSCs for efficient migration through confined space.

Confined migration involves membrane internalization via endocytosis

Cellular adaptations to physical constraints require reorganization of cytoskeleton and plasma membrane. We thus applied in our migration assays live cell fluorogenic dyes SPY-actin and MemGlow, which revealed robust accumulation of F-actin at the rear zone and MemGlow-labeled endocytic vesicles at the front zone of GSCs traversing 3 µm constrictions (Fig. 1d, e). We also detected a smaller proportion of MemGlow⁺ vesicles at the rear zone in constricted cells, particularly after passage (Fig. 1d), while NucSpot highlighted nuclear deformation of constricted cells (Fig. 1a).

Balancing endocytosis and exocytosis are essential for adjustment of local membrane tension^23–25. We next tested if endocytosis is required for confined migration using three small molecule inhibitors targeting different endocytosis pathways: Pitstop 2 (a cell permeable clathrin inhibitor), Dynasore (a noncompetitive inhibitor of dynamin), and EIPA (an inhibitor of Na⁺/H⁺ exchange pump that blocks macropinocytosis by lowering pH near plasma membrane, thereby inhibiting actin remodeling via Rho)^26–29. All three inhibitors disrupted accumulation of MemGlow⁺ vesicles at the cell front, but also of F-actin assembly at cell rear, illustrating interconnectedness of the two processes (Figs. S2a, b). Confined migration through 3 µm constrictions was also negatively impacted by all three inhibitors, as determined by measurement of velocity through constrictions, stalling time, and migration consistence/polarity in microchannels (Figs. 1f, g; Supplementary Movie 5). In unconfined condition (i.e., 8 µm constrictions), migration velocity was also reduced by Dynasore and EIPA, while directional consistence was decreased by all three endocytosis inhibitors, reflected by reduced ratio of forward vs. backward movement (Figs. S2c; Supplementary Movie 6). By comparison, in 2D GSC cultures, endocytosis inhibitors had no major impact on F-actin pattern (phalloidin), proliferation (Ki67), or scratch wound closure (except for a modest delay by Dynasore at 24–72 hr and EIPA at 72 hr) (Fig. S2d, e). Hence, endocytosis is instrumental for polarized migration of GSCs in microchannels, and also regionalized F-actin assembly at cell rear to propel cells for migration through constrictions (Fig. 1h).

Guidance receptor Plexin-B2 controls cortical actin and membrane tension of GSCs

Our next question pertained to the molecules orchestrating the interconnected processes of endocytosis and F-actin during confined migration. Guidance receptor Plexin-B2 is an attractive candidate, as it regulates actin contractility and cell stiffness in hESCs and hNPCs¹⁹, and it is upregulated in GBM and promotes glioma invasion by modulating actomyosin and cell adhesion^6,9.

We generated SD2 and SD3 GSC sublines with lentiviral CRISPR/Cas9-mediated PLXNB2 knockout (PB2 KO), or lentiviral overexpression (PB2 OE), confirmed by Western blot and IF staining (Fig. 2a–c). Peculiarly, F-actin staining by phalloidin revealed elongated cell processes of PB2 KO cells in 2D cultures (Fig. S3a), a phenotype better visualized by live cell imaging with plasma membrane dye NR12A³⁰ (Figs. S3b; Supplementary Movie 7). This could be a sign of reduced membrane dynamics to withdraw extended cell processes in absence of Plexin-B2. Consistently, 3D surface rendering of GSCs stained with NR12A dye revealed increased membrane ruffling of PB2 KO cells, suggesting reduced membrane tension (Fig. S3c).

We next performed micro-indentation measurements with atomic force microscopy (AFM) to directly gauge cell stiffness (Fig. 2d). This revealed lower mean Young’s modulus for PB2 KO cells (about 1 kPa) than WT cells (1.6 kPa), whereas PB2 OE cells displayed higher Young’s modulus (1.8 kPa) (Fig. 2d; S3d). For comparison, human neural precursor cells (NPCs) have a mean Young’s modulus of about 5.9 kPa, while human embryonic stem cells (ESCs) have a mean Young’s modulus of 2.2 kPa, as shown in our previous study¹⁹.

To measure membrane tension, we applied optical tweezers to determine the force required to pull out plasma membrane tethers, which was lower for PB2 KO cells (mean ~12.8 pN) than control (~36 pN) or PB2 OE SD2 cells (~33 pN) (Fig. 2e). Interestingly, unlike constitutive PB2 OE, doxycycline (Dox)-induced overexpression of PB2 drastically reduced membrane tether force (Figs. S3e, f), illustrating that acute imbalance of cortical contractility can perturb membrane tension. In support, inhibition of myosin contractility with blebbistatin or actin assembly with latrunculin led to a marked reduction of membrane tether forces (phenocopying PB2 KO), but no further effect on PB2 KO cells, which already had low baseline actomyosin contractility (Fig. S3g).

As another way to gauge plasma membrane tension in live cells, we utilized Flipper-TR, a membrane dye that changes fluorescence lifetime according to membrane tension³¹. Fluorescence lifetime imaging (FLIM) revealed shorter Flipper-TR lifetimes (i.e., lower membrane tension) in SD2 and SD3 PB2 KO or SD2 with Dox-inducible shRNA knockdown of PB2 (Figs. 2f; S4a-c). Echoing the optical tweezers results, acute induction of PB2 OE transiently led to shorter Flipper-TR lifetimes at 6 hr after Dox administration, which were recovered at 24 and 48 hr after Dox, but constitutive PB2 OE had no major effect (Figs. S4a, d). Live-cell imaging with SPY-actin revealed also reduced cortical F-actin in PB2 KO relative to control GSCs (Fig. 2g).

To understand the connection between cortical actin and membrane tension, we examined the expression of phosphorylated ezrin-radixin-moesin (pERM), which functions as a linker between cortical actin and membrane proteins³². IF staining revealed co-localization of Plexin-B2, pERM, and F-actin near the plasma membrane of control GSCs, a pattern perturbed by PB2 KO (reduced pERM) or PB2 OE (aberrant pERM aggregation, particularly in SD3 cells) (Figs. S5a, b). WB showed a moderate decrease of pERM in PB2 KO cells, while total ezrin levels were comparable (Figs. S5c, d). In sum, biomechanical measurements by AFM, optical tweezers, and Flipper-TR FLIM converge on the model that Plexin-B2 governs membrane tension of GSCs via cortical actin and pERM (Fig. 2h).

Plexin-B2 affects endocytosis and membrane internalization in GSCs

To further examine the impact of Plexin-B2 on endocytosis, we performed a dextran uptake assay (used previously to show endocytosis in semaphorin stimulated growth cone collapse^15,16). After 40 min of exposure to dextran (10 kDa) conjugated with Alexa 488, PB2 KO cells contained less intracellular fluorescent puncta than control GSCs, indicating reduced endocytosis (Fig. 3a). We observed also diffuse intracellular dextran fluorescence in some PB2 KO cells, and confocal z-stack imaging confirmed the increased cytosolic uptake of dextran³³ (Fig. 3a, b). The cytosolic dextran uptake was not associated with decreased viability of cells, as confirmed by cell viability assay using calcein violet staining (see Fig. S16b).

Fig. 3 — a Dextran uptake assay of SD2 GSCs labeled with SPY-Actin. Enlarged images of boxed areas are shown below. Box plots show areas of dextran⁺ puncta per cell, with 25–75% quartiles, median (line), and mean (plus sign). n = 85 cells for Ctrl, n = 44 cells for PB2 KO. Mann–Whitney–Wilcoxon test, two-sided. b Top, live cell confocal plane images of SD2 GSCs with side views of z-stacks show intracellular localization of diffuse dextran-Alexa 488 in PB2 KO cells, whereas Ctrl cells contained only dextran⁺ endocytic vesicles. Bottom, histograms of fluorescence profiles show bimodal distribution of dextran-Alexa 488 fluorescence intensities in PB2 KO GSCs (blue and brown arrows). n = 177 cells for Ctrl, n = 161 cells for PB2 KO. Mann–Whitney–Wilcoxon test, two-sided. c, d Left, schematic of myr-palm-GFP or -CFP attached to inner membrane leaflet. Right, live cell images at 72 hr after transfection show fluorescent probes on endomembranes (arrow) in control GSCs, but membrane retention of the probes (arrowhead) in PB2 KO GSCs. For myr-palm-GFP: n = 21 cell membranes for WT; n = 13 cell membranes for PB2 KO. For myr-palm-CFP: n = 6 cell membranes for WT; n = 5 cell membranes for PB2 KO. Two-sided unpaired t-test. Bar graphs show means ± SEM. e Left, schematic of TauSTED super-resolution microscopy of GSCs labeled with MemGlow and NucSpot. Middle, TauSTED live-cell images show reduced endocytic vesicles (arrowheads) in PB2 KO cells compared to Ctrl. Right, box plots show areas of MemGlow⁺ clusters per cell. Box plots show 25–75% percentiles, minimal and maximal values (whiskers), median (line), and mean (cross). n = 26 cells for Ctrl, n = 13 cells for PB2 KO. Two-sided unpaired t-test. f Working model of regulation of cortical and membrane tension by Plexin-B2, affecting endocytosis in GSCs. Source data are provided as a Source Data file.

Concordantly, fluorescence intensity histograms revealed two types of dextran fluorescence in PB2 KO cells: endosomal puncta and bright cytosolic signals (Fig. 3b). Lysotracker dye staining overlapped with the dextran puncta (confirming endocytic uptake), but not with the cytosolic dextran signals (Fig. S6a). By contrast, PB2 OE led to a modest increase in dextran puncta, but also bright diffuse cytosolic fluorescence in some cells (Fig. S6a). Similar results were also obtained with SD3 PB2 KO cells (Fig. S6b).

Actomyosin inhibition by blebbistatin or endocytosis inhibitors also led to increased cytosolic dextran uptake in GSCs, mirroring the PB2 KO phenotype (Figs. S6c, d), while the effect of PB2 OE was attenuated by blebbistatin (Fig. S6c). Cooling the cells to 4 °C, which slows down all thermally active processes, markedly reduced cytosolic dextran in PB2 KO and PB2 OE cells (Fig. S6e).

To characterize membrane internalization in dependence of Plexin-B2, we introduced into GSCs expression plasmids encoding membrane anchored fluorescent proteins (myr-palm-GFP or -CFP). Live-cell imaging at 72 hr after transfection revealed that myr-palm-GFP or -CFP were mostly localized on intracellular vesicles in control GSCs, but retained on the plasma membrane of PB2 KO cells (Fig. 3c, d). Super-resolution microscopy further demonstrated reduced intracellular MemGlow⁺ vesicles in PB2 KO cells (Fig. 3e).

Plexin signals via its GAP domain to inactivate Ras/Rap GTPases. To explore downstream effectors, we introduced constitutively active (CA) or dominant-negative (DN) isoforms of RAP1B or RAP2A into GSCs (Fig. S7). Live-cell imaging revealed that CA RAP1B, but not RAP2A, led to membrane retention of myr-palm-GFP (phenocopying PB2 KO), whereas DN RAP1B had no overt effect. These data support that Plexin-B2 signals through RAP1B deactivation to control membrane internalization in GSCs. Altogether, Plexin-B2 manipulations affected membrane tension and membrane internalization (Fig. 3f).

Plexin-B2-deficient GSCs display compromised confined migration

Given that membrane tension and endocytosis were reduced in PB2 KO GSCs, we next investigated their capability for confined migration. Live cell imaging revealed a significant difference between PB2 KO (SD2 or SD3) and control GSCs when navigating 3 µm constrictions, with fewer KO cells invading into microchannels, lower velocity at constrictions points, longer stalling times, and reduced migration consistency (i.e., increased backward movements) (Fig. 4a, b; Figs. S8a–c; Supplementary Movies 8, 9). Spatial organization of F-actin assembly at cell rear and endocytic vesicles at cell front were also disrupted in constricted KO cells (Fig. 4c, d). PB2 OE cells performed better than PB2 KO cells, but worse than control cells; F-actin and endosomes were also disorganized (Figs. S8b–d).

Fig. 4 — a Still frames from live-cell videography show compromised migration of PB2 KO SD3 (visualized by NucSpot) into microchannels with 3 µm constrictions (denoted by orange chevron signs) as compared to Ctrl cells. b Quantifications of velocity through constrictions (Ctrl, n = 11 cells; PB2 KO, n = 12 cells), stalling time at constriction (Ctrl, n = 25 cells; PB2 KO, n = 19 cells), sums of distances traveled forward (n = 11 cells per group), and sums of distances traveled backward (n = 12 cells per group). Box plots show 25–75% quantiles, minimal and maximal values (whiskers), median (line), and mean (cross). Two-sided unpaired t-test. c Still frames from live-cell imaging of SD3 traversing 3 µm constrictions reveal F-actin assembly at rear zone (SPY555-actin, arrowhead) and MemGlow-labeled endosomes at front (arrow) in Ctrl but not in stalled KO cells. Note different time stamps for Ctrl and PB2 KO. Dashed lines delineate cell boundary. d Right, quantifications of the ratio of fluorescence intensity for SPY-actin or MemGlow at rear vs. front zones during confined migration. Bar graphs show means ± SEM. n = 10 cells per group. Two-sided unpaired t-test. e Left, still frames from live-cell imaging show a long tether (arrows) connecting PB2 KO cells in microchannel with 8 µm constriction. f Box plot quantifications of SPY-actin tether length. Box plots show 25–75% percentiles, minimal and maximal values (whiskers), median (line), and mean (cross). n = 24 cells for Ctrl; n = 20 cells for PB2 KO. Two-sided unpaired t-test. Source data are provided as a Source Data file.

PB2 KO did not affect the velocity or forward movement of SD2 or SD3 GSCs migrating in open channels near entry ports or in channels with 8 µm restrictions, but directional consistency was still affected as shown by increased backward movement (Figs. 9–c; Supplementary Movies 10, 11). Also of interest was the observation of long tethers connecting PB2 KO GSCs in microchannels that contained F-actin, reflective of failed dissociation from neighboring cells (Figs. 4e, f; Supplementary Movie 12). These findings are in agreement with increased adhesiveness of Plexin-B2 KO GBM cells⁶.

To extend biomechanical assays of confined migration, we designed a microdevice with central chambers connected by pairs of tunnels of 3 or 8 µm width and 50 μm length (Fig. 5a). This design allows GBM cells to choose different exit routes, while the long tunnels present continuous spatial confinement. Cells were seeded into entry ports from which they migrated through 8 µm tunnels into central chambers. Live-cell imaging showed that control SD3 cells were adept at exploring either 3 or 8 µm outlets and passing through the long tunnels; after passage, cells then altered direction in central chambers and continued to migrate through the same or neighboring tunnels (Figs. 5b; Supplementary Movie 13). In contrast, PB2 KO cells were frequently stalled in tunnels, showing reduced speed, less completed passages and lower propensity to explore exit outlets or alter direction in central chambers (Fig. 5b, c; S10a; Supplementary Movie 13).

Fig. 5 — a Schematic of microdevice with central chambers connected by pairs of narrow tunnels of 50 μm length and 3 or 8 µm width. b Representative still images from videography show successful passage of Ctrl SD3 through 3 µm tunnel in 60 min timeframe, but stalled PB2 KO cell after 90 min. Long arrows on left denote migration direction. Note SPY-actin at rear zone (arrowhead) and diffuse MemGlow-labeled endocytic vesicles (arrow) in Ctrl, both reduced in PB2 KO GSC. Dashed lines delineate cell boundary. c Box plots show the number of successful migration events per cell through the 3 µm tunnel in 17 hr timeframe and the speed in the tunnel. Box plots show 25–75% percentiles, minimal and maximal values (whiskers), median (line), and mean (cross). n = 9 cells per group. Two-sided unpaired t-test. d Left, still images show SPY-actin at rear zone of Ctrl GSCs traversing through tunnels, more prominent in 3 µm than 8 µm tunnels, and reduced in PB2 KO cells. Arrows denote direction of migration. Right, bar graphs show ratios of SPY-actin fluorescence intensity at rear vs. front during passage through tunnels. For 3 µm: n = 21 cells for Ctrl; n = 20 cells for PB2 KO. For 8 µm: n = 17 cells for Ctrl; n = 16 cells for PB2 KO. One-way ANOVA followed by Tukey’s multiple comparison test. Data represent mean ± SEM. Source data are provided as a Source Data file.

For control SD3 GSCs traversing micro-tunnels, live-cell imaging also revealed accumulation of F-actin at rear zone and endocytic vesicles at cell front, more pronounced in 3 µm than in 8 µm tunnels (Fig. 5d); PB2 KO cells did not display such regionalized patterns. The nuclei of GSCs in micro-tunnels were compressed into elongated shapes in both WT and PB2 KO (Fig. 5d), indicating intact nuclear response to physical constraints in absence of Plexin-B2.

Recent studies showed that exposure to high fluid viscosity can increase cytoskeletal contractility through focal adhesions, leading to increased membrane tension³⁴, and we wondered if Plexin-B2 would affect such responses. Confocal microscopy revealed that after 72 hours of exposure to high viscosity media, both control and PB2 KO GSCs showed increased cortical F-actin (Figs. S10c, d). Interestingly, viscosity induced spreading of plasma membrane beyond cortical actin, which was farther in PB2 KO cells, likely signifying a larger membrane reservoir and/or less stringent attachment of plasma membrane to cortical actin (Figs. S10c, d).

Molecular dynamics modeling predicts coupling of actin cytoskeleton and membrane involution during confined migration

To help understand the mechanical coupling between cytoskeleton and cell membrane, we performed molecular dynamics simulations of confined cell migration. First, we developed a coarse-grained bead model with cell plasma membrane, actin filaments, and nuclear envelope represented by discrete bead elements connected by springs (Fig. S11a). Simulated cells were then placed between two opposing virtual walls, one fixed and adhesive (simulating extracellular matrix attachment), and the other movable (simulating increasing confinement in microchannels) (Fig. S11a).

We first simulated confined cells with fixed membrane elastic constant κ₀₀ but variable membrane-cytoskeleton energy U₃ (approximating control vs. PB2 KO conditions) (Fig. S11b). At t = 0 without compression, cells with either high or low U₃ assumed rounded shape; but with constriction (t = 1), simulations predicated distinct patterns of actin stretching and membrane ruffling. When constriction was further increased to mimic narrow tunnels (t = 2 and 3), spontaneous involution of cell membrane occurred (resembling endocytic vesicles), but only in cells with high membrane-cytoskeleton energy (Fig. S11b; Supplementary Movie 14). Thus, molecular dynamics modeling predicted that membrane internalization triggered by cell confinement requires adequate actin-membrane interaction (e.g., U₃ = 15), resembling the control condition.

The molecular dynamics simulations allowed us to vary both membrane-cytoskeleton energy (U₃) and cell membrane elastic constant (κ₀₀) to calculate their impact on cell motility (measured as velocity of the center of cell mass) and endocytosis (measured as plasma membrane involution or negative area A^-) (Fig. S11c). Simulations predicted high motility of confined cells when the membrane-cytoskeleton energy was in the medium range with cell membrane elastic constant in medium to high range (κ₀₀ = 50-5,000), indicating that optimal cell motility requires balanced forces (Fig. S11c). Interestingly, very high membrane-cytoskeleton energy (U₃ = 30) did not confer higher motility, resembling PB2 OE condition (Fig. S11c). The involution rate of cell membrane (simulating endocytosis) followed a more direct correlation with intracellular forces, i.e., highest rates of involution with membrane-cytoskeleton energy U₃ ≥ 15 and cell membrane elastic constant κ₀₀ ≥ 50 (Fig. S11c).

Accordingly, molecular dynamics simulation of cell migration demonstrated that intermediate values (e.g., U₃ = 6 and κ₀₀ = 100) promoted sustained unidirectional confined migration over time (consistency/directional polarity) (Fig. S11d). Conversely, for low values (U₃ = 0.03 and κ₀₀ = 10), simulated cells moved for a short period in one direction but frequently changed directionality (mimicking PB2 KO), whereas for highest values (U₃ = 30 and κ₀₀ = 5000), simulated cells displayed a hypercontractile state with severely reduced motility (mimicking PB2 OE) (Fig. S11d; Supplementary Movie 15).

Plexin-B2 function affects membrane phospholipid composition

The lipid composition of the plasma membrane may critically affect polarized migration⁷. Anionic phospholipids such as phosphatidylinositol phosphates (PIPs) are constituents of the inner plasma membrane and crucial for signaling events³⁵. We wondered whether reduced membrane internalization/endocytosis in PB2 KO cells might impact lipid composition of the inner plasma membrane.

We first interrogated the localization and levels of phosphatidylinositol 4,5-bisphosphate (PIP2), an essential lipid of the cell membrane and a second messenger in diverse signaling pathways^36,37. Using the fluorescent protein probe PH(PLCD1)-GFP, i.e., GFP fused to a domain that binds to PIP2, we observed that at 72 hr after transfection, the majority of GFP signals (PIP2) were internalized in control GSCs, but largely retained on plasma membrane in PB2 KO cells (Fig. 6a). This may reflect reduced endocytosis in PB2 KO cells or suppressed activity of phospholipase C (PLC), which hydrolyzes PIP2 into IP3 and DAG (diacylglycerol) (Fig. S12a). Time course analysis revealed high PH(PLCD1)-GFP levels (a readout of abundant PIP2) on the plasma membrane of control GSCs at 18 hr after transfection, which was internalized or hydrolyzed by 24 and 48 hr; by contrast, both in PB2 KO and OE cells, the majority of PH(PLCD1)-GFP was retained on the plasma membrane, reflecting reduced PIP2 internalization/suppressed PLC activity (Fig. S12b). In comparison to PIP2, the membrane content PIP3 appeared low in both control and PB2 KO SD2 cells, shown by the PH(Btk)-GFP probe that binds to PIP3³⁷ (Fig. S12c), consistent with PIP3 being a minor phospholipid constituent of the plasma membrane³⁸.

Fig. 6 — a Left, schematic of PH(PLCD1)-GFP PIP2 probe. Right, live-cell imaging at 72 hr post transfection reveals that the PH(PLCD1)-GFP probes were largely internalized in control GSCs (arrow), but retained on membrane of PB2 KO GSCs (arrowhead). b Left, still images from videography show accumulation of PH(PLCD1)-GFP probes (arrow) in front of the nucleus (NucSpot) of migrating control SD2 in tunnels, more so in 3 than 8 µm tunnel, but not in PB2 KO cells. Dashed lines delineate cell boundary. Long arrow denotes direction of migration. Right, quantifications of the ratio of PH(PLCD1)-GFP fluorescence intensity at front vs. rear of GSCs during passage. For 3 µm: n = 15 cells for WT, n = 13 cells for PB2 KO. For 8 µm: n = 13 cells for WT, n = 16 cells for PB2 KO. One-way ANOVA followed by Tukey’s multiple comparison test. Data represent mean ± SEM. c Left, schematic of R( + 8)-prenyl-GFP probe for negative surface charge of inner leaflet of plasma membrane. Right, live-cell imaging at 72 hr post-transfection shows internalization of the probes (arrow) in control GSCs, in contrast to the predominant membrane localization in PB2 KO GSCs (arrowhead). d Left, still images from videography show accumulation of the R( + 8)-pre-GFP probe (arrow) at front zone of control GSC traversing 3 µm tunnel, but not in PB2 KO cells. Right, bar graphs show the ratio of R-pre-GFP fluorescence intensity at rear vs. front of GSC when passing through tunnels. n = 22 cells for WT, n = 27 cells for PB2 KO. Mann–Whitney–Wilcoxon test, two-sided. Data represent mean ± SEM. Source data are provided as a Source Data file.

We next conducted live-cell imaging of GSCs expressing the PH(PLCD1)-GFP probe during confined migration in micro tunnels. We detected PH(PLCD1)-GFP/PIP2 accumulated at the front of migrating SD2 GSCs traversing 3 µm but not 8 µm tunnels, whereas PB2 KO cells that stalled in the tunnels displayed no discernable patterns of PIP2 distribution (Fig. 6b).

Plexin-B2 deletion alters inner membrane surface charge

PIP2 is a negatively charged phospholipid that contributes to the negative charge of the inner leaflet of the plasma membrane³⁹. To assess if the observed PIP2 accumulation on the inner plasma membrane of PB2 KO cells might alter the inner leaflet surface charge (zeta potential)⁴⁰, we transfected GSCs with a plasmid expressing R( + 8)-prenyl-GFP, a probe containing a series of eight positively charged arginine residues that preferentially locates to negatively charged membrane surface⁴¹. Indeed, we detected a distinctive distribution of R( + 8)-prenyl-GFP on the plasma membrane of PB2 KO cells, in contrast to control SD2 GSCs that harbored mostly GFP signals on endomembrane (Fig. 6c). In confined migration assay, R( + 8)-prenyl-GFP accumulated at cell front of migrating SD2 in 3 µm tunnels, but not in stalled PB2 KO cells, concordant with the signals from the PIP2 probe (Fig. 6d). Hence, during polarized confined migration, the front of GSCs contained more PIP2 and more negative inner leaflet surface charge, a process that required Plexin-B2 function.

As PIP2 accumulation/negative surface charge in inner leaflet could impact local electric field across the plasma membrane, we labeled GSCs with FluoVolt, a voltage-sensitive fluorescent membrane dye⁴² (Fig. 7a). This revealed a clear difference between control and PB2 KO cells, with KO cells displaying lower FluoVolt signals (Fig. 7b; S13a, b). Of note the transmembrane potential of GSCs was not affected, as patch clamping recording showed comparable resting potential and membrane conductivity between PB2 KO and control GSCs (Fig. S13c). Thus, the changes of FluoVolt signals likely correspond to local changes of inner membrane surface charges.

Fig. 7 — a Illustration of voltage sensitive membrane dye FluoVolt, with fluorescence quenched by electron transfer from electron-rich donor (high inner membrane surface charge), mediated by “molecular wire” in plasma membrane (figure modified after ref. ⁴²). b Left, live-cell images show reduced FluoVolt fluorescence in plasma membrane of PB2 KO cells (higher negative charges of inner membrane). Right, box plots of membrane FluoVolt intensity. Box plots show 25–75% percentiles, minimal and maximal values (whiskers), median (line), and mean (cross). n = 25 cells for Ctrl, n = 27 cells for PB2 KO. Two-sided unpaired t-test. Data represent mean ± SEM. c Top, still images from live-cell imaging show higher FluoVolt fluorescence at rear zone (arrowhead) in Ctrl GSCs when traversing tunnels, more so in 3 than 8 µm tunnel, but not in PB2 KO cells. Calcium chelator BAPTA-AM disrupted polarized FluoVolt pattern. White arrow denotes migration direction. Bottom, bar graphs show the ratio of FluoVolt intensity at rear vs. front during confined migration. n = 15 cells per group for Ctrl. vs KO, one-way ANOVA followed by Tukey’s multiple comparison test. n = 21 cells for vehicle and n = 16 cells for BAPTA-AM, two-sided unpaired t-test. Bar graphs represent mean ± SEM. d Live-cell images and quantifications show the effects of constitutive active (CA) RAP1B-V12 or dominant-negative (DN) RAP1B-N17 on FluoVolt intensity in Ctrl or PB2 KO GSCs. Box plots show 25–75% percentiles, minimal and maximal values (whiskers), median (line), and mean (cross). n = 25 cells per group. Kruskal–Wallis test followed by Dunn’s multiple comparisons test. e Model of Plexin-B2 signaling affecting membrane surface charge and electric field during polarized confined migration, with PIP2 enrichment at cell front and Ca²⁺ at rear zone, and asymmetry of FluoVolt and R( +8)-pre-GFP. Source data are provided as a Source Data file.

In micro-tunnels, migrating GSCs harbored higher FluoVolt signals at the rear (i.e., less negative charge), more evident in 3 µm than 8 µm tunnels, but not in stalled PB2 KO cells (Fig. 7c; Supplementary Movie 16). The FluoVolt patterns at cell rear paralleled that of F-actin, but opposite the signals from probes for PIP2 or negative membrane surface charge. These processes are interconnected, as endocytosis inhibitors reduced FluoVolt signals at cell rear (Fig. S13d). In addition, calcium imaging using Fluo-4AM revealed calcium accumulation at the rear of migrating GSCs, more evident in the 3 µm than 8 µm tunnels, but not in stalled PB2 KO cells (Fig. S13e). Also calcium chelation with BAPTA-AM led to reduced FluoVolt signals at cell rear and compromised confined migration (Figs. 7c; Supplementary Movie 17).

The introduction of RAP1 mutants also altered FluoVolt signals in GSCs. The constitutively active (CA) form of RAP1B attenuated FluoVolt signals, phenocopying PB2 KO, whereas dominant negative (DN) RAP1B reversed the KO phenotype (Fig. 7d).

We next examined the role of class 4 semaphorins, which are membrane-bound Plexin-B2 ligands, although Sema4D can also be cleaved⁴³. We applied genetic disruption of all six Sema4 genes – SEMA4A, B, C, D, E, G – (Sema4 KO) by lentiviral CRISPR/Cas9 in SD2 GSCs (Fig. S14a) and confirmed the reduction of protein levels for Sema4B and Sema4C, the two highest expressed Sema4s in this GSC line⁶ (Figs. S14b, c). We observed in Sema4 KO SD2 GSCs traversing 3 µm tunnels the attenuation of the regionalized FluoVolt signals at cell rear, similar to Plexin-B2 KO, as well as compromised migration consistency and speed (Figs. S14d, e). The KO of both Plexin-B2 and Sema4s did not further augment the phenotypes (Fig. S14e), suggesting ligand-receptor communication in the same pathway. Since GSCs pass in micro-tunnels as individual cells through confinement, Sema4s may bind to Plexin-B2 expressed in the same cell via membrane ruffling, in line with previous reports of cell-autonomous signaling of semaphorins to co-expressed plexins^44,45. Interestingly, we observed that Sema4B levels were strongly reduced by KO of Plexin-B2 (Fig. S14b), perhaps in relation to a mechanism of cis-interaction and co-regulation of a Plexin/Sema complex. It should also be noted that we confirmed the Sema4 KO on protein level only for Sema4B and Sema4C, and one has to consider the possibility that the KO of Sema4B and Sema4C alone may be sufficient to recapitulate the effects of Plexin-B2 KO.

Altogether, these results illustrate that Plexin-B2 orchestrates mechano-electrical coupling of regionalized accumulation of endocytic vesicles/PIP2/higher inner leaflet negative surface charge at cell front, and F-actin assembly and calcium at rear zone to propel polarized migration of GSCs through confined space (Fig. 7e).

Flexible ring of Plexin-B2 extracellular domain is required for efficient confined migration

Our next question pertained to how migrating GSCs sense physical constraints and convey mechanosignals to reorganize cytoskeleton and plasma membrane. Intriguingly, a recent study showed that endothelial cells use Plexin-D1 to detect sheer stress from blood flow via the flexible ring-like extracellular domain, independent of activation by semaphorins¹⁷. This was demonstrated with “lock” mutants that contain a cysteine bond that prevents bending of the ring, thus ablating mechanosensitivity while preserving responsiveness to exogenous semaphorin. To explore the importance of bendability of the Plexin-B2 ring domain for confined migration, we generated two lock mutants by introducing cysteine residues into the extracellular domain of Plexin-B2: I436C and S993C for lock1; I436C and T1051C for lock2 (Fig. 8a; Figs. S15a, b). Control experiments using hESCs showed that the two lock mutants reintroduced sensitivity to exogenous semaphorin stimulation, but failed to rescue the Plexin-B2 KO phenotype of disrupted hESC colony geometry¹⁹ (Figs. S15c, d).

Fig. 8 — a Structure model of the extracellular domain of human Plexin-B2 show the locations of lock1 and lock2 mutations predicted to form disulfide bridges that lock the ring structure. b Western blots show absence of mature Plexin-B2 (170 kDa) in PB2 KO GSC, and expression of lock mutants in PB2 KO for both SD2 and SD3. β-actin serves as a loading control. c Still images from videography show passage of GSCs (nuclei visualized by NucSpot) through microchannels by wildtype PB2 rescue construct, but not PB2 lock mutants, nor PB2 with deletion of extracellular domain (dECTO). Chevrons point to 3 µm constrictions. d Box plots show velocity through constrictions (Rescue, n = 20 cells; dECTO, n = 20 cells; Lock1, n = 21 cells; Lock2, n = 17 cells), stalling time at constrictions (Rescue, n = 20 cells; dECTO, n = 14 cells; Lock1, n = 28 cells; Lock2, n = 22 cells), and sum of forward and backward movements constrictions (Rescue, n = 20 cells; dECTO, n = 20 cells; Lock1, n = 21 cells; Lock2, n = 17 cells), with 25–75% quartiles, minimal and maximal values (whiskers), median (line), and mean (cross). One-way ANOVA followed by Dunnett’s multiple comparisons test. e Still images from videography show wildtype but not PB2 mutants restored localization of SPY-actin (arrowhead) at cell rear and MemGlow⁺ endocytic vesicles (arrow) at cell front in GSC traversing 3 µm constrictions (chevrons). f Bar graphs show fluorescence intensity ratio of SPY-actin and MemGlow at rear vs. front of GSCs during confined migration. For SPY-actin: n = 11 cells for Recue; n = 15 cells for dECTO; n = 10 cells for Lock1; n = 10 cells for Lock2. For MemGlow: n = 15 cells for Recue; n = 18 cells for dECTO; n = 10 cells for Lock1; n = 10 cells for Lock2. Kruskal–Wallis test followed by Dunn’s multiple comparisons test. Data represent mean ± SEM. g Model of Plexin-B2 signaling in regulating mechano-electrical coupling of membrane tension and membrane surface charge during polarized confined migration. Note regionalized enrichment of endocytosis/PIP2 at cell front and F-actin/Ca²⁺ at rear zone and asymmetry of FluoVolt and R( + 8)-pre-GFP membrane probes. Source data are provided as a Source Data file.

The lock mutants were then introduced into PB2 KO GSCs, and expression was confirmed by IF staining and WB in both SD2 and SD3 KO cells (Figs. 8b; Fig. S16a). The morphology of PB2 KO cells rescued with wildtype or lock mutants of Plexin-B2 appeared comparable in 2D culture (Fig. S16a). However, both lock mutants failed to rescue aberrant membrane phenotypes of PB2 KO, including reduced endocytosis, increased dextran uptake, less internalization of myr-palm-CFP membrane probes, and higher plasma membrane retention of PIP2 (Figs. S16b–d). Additionally, the lock mutation also resulted in lower pERM levels at cell processes, similar to PB2 KO (Fig. S16e).

In microchannels, both lock mutants and a PB2 mutant form lacking the extracellular domain (dECTO) failed to rescue confined migration through 3 µm constrictions, in contrast to rescue with wildtype PB2 (Figs. 8c, d; Fig. S16f). The regionalization of F-actin in rear zone and endocytic vesicles at cell front were also disrupted in PB2 lock mutants (Fig. 8e, f).

Discussion

Invasion contributes to the high lethality of GBM, thus understanding the cellular and molecular underpinnings is critical. Here, we studied how GBM cells adapt to physical constraints to achieve polarized migration through confined space. We show that regulation of membrane tension is a key process during confined migration, which affects membrane internalization/endocytosis, phospholipid composition, and inner leaflet membrane surface charge. These results shed light into the biophysical mechanisms by which guidance receptor Plexin-B2 promotes GBM invasion (Fig. 8g).

By utilizing microchannel devices that mimic the physical constraints experienced by invading GBM cells, we demonstrate that GSCs are adept at squeezing through confined space, with a stimulatory effect on migration speed and directional consistency. In echo, in polyacrylamide hydrogels, narrow channels have been shown to accelerate the migration of glioma cells⁴⁶. Our live cell imaging captured the dynamic extension of cell processes of GSCs to explore channels and then rapidly adjust plasma membrane, cytoskeleton, and nuclear shape to squeeze through constrictions. During the process, GSCs accumulated endocytic vesicles at cell front and F-actin at cell rear, in line with molecular dynamics simulations highlighting the importance of balanced membrane tension and membrane-cytoskeleton interactions for optimal migration in confined space. These data extended previous studies on the involvement of endocytosis in cancer cell migration^47,48, and also provide insights into why Plexin-B2 KO and OE both perturbed confined migration.

The finding that Plexin-B2 affects membrane internalization during confined migration also resonates with axon growth cone studies in which endocytosis played a key role in semaphorin triggered collapse^{15,16,49–51}. Genetic studies in C. elegans also showed that semaphorin/plexin phenotypes can be suppressed by mutations of genes controlling vesicle transport⁵². Our observation that Plexin-B2 KO GSCs showed reduced directional consistence in micro-channels parallels the phenotypes of Plxnb2-KO mice with Plxnb2-deficient cerebellar granule cell precursors displaying reduced migration consistence in culture⁵³ and aberrant cerebellar development⁵⁴. These findings demonstrate how GBM cells usurp the plexin developmental pathway to gain invasiveness.

The findings on the impact of Plexin-B2 on membrane properties also extend the model that negative surface charge on the inner plasma membrane acts a hub for cytoskeletal organization in migrating cells^7,40,55. The mechanisms by which Plexin-B2 influences inner membrane charges warrant further study but may involve regulation of phospholipid composition via either membrane internalization or activity of PLC, a phospholipase that hydrolyzes PIP2 to DAG and IP3, leading to calcium influx and other signaling cascades. Additionally, levels of anionic lipids like PIP2 can impact local concentration of inorganic (e.g., Ca²⁺) and organic (e.g., spermine) cations, as well as accumulation of net positively charged proteins on the inner leaflet^56,57. Since calcium chelation by BAPTA reduced FluoVolt signals at rear zones of confined cells, local calcium levels may also influence the membrane electric field. Whether regionalized PIP2 accumulation or inner leaflet surface charge are causal or consequences of confined migration needs further study. The cytosolic uptake of dextran that was observed in Plexin-B2 KO and OE GSCs was unexpected. Several mechanisms of cytosolic uptake of mid-sized and large biomolecules have been described that could underly these observations, e.g., disordered lipid packing enabling membrane translocation⁵⁸, transient permeabilization by membrane ruffling⁵⁹, or passage through mechanochannels⁶⁰. How membrane tension perturbation by Plexin manipulations might be linked to these phenomena awaits further investigation.

Plexin-B2 lock mutants with rigid extracellular domain failed to rescue Plexin-B2 KO phenotypes in microchannel migration assays, suggesting that flexibility of the extracellular ring may be key to sense mechanical forces, akin to the functions of Plexin-D1 in endothelial cells and Plexin-B1 in epithelial cells^17,18. Whether Plexin-B2 is activated by physical force-induced conformational change needs further study, and it does not rule out ligand-induced conformational changes of the extracellular ring. Indeed, deleting all six Sema4s also compromised confined migration through 3 µm tunnels, while KO of both Plexin-B2 and Sema4s did not add further severity, thus supporting ligand-receptor signaling in the same pathway. It will be worthwhile to test if physical force can bend the ring in the absence of Sema4s. Currently no mutational strategy is known to abolish Sema binding surface without compromising the bendable ring. In hESCs, exogenous Sema4C (coated on culture dish together with laminin, not by adding to culture media) altered the colony geometry, and PB2 lock mutants could restore Sema4C responsiveness in PB2 KO cells, suggesting that Sema4 can signal through Plexin-B2 without bending the ring in certain contexts.

The conserved GAP domain is essential for Plexin-B2 function^19,61 and our FluoVolt and myr-Palm-GFP data with Rap mutants suggest RAP1B as a major effector, but other related Ras family small GTPases such as R-Ras, M-Ras, TC21, or Rin may also be involved⁶². Rap GTPases can affect a plethora of downstream effectors to control cytoskeleton and cell adhesion⁶³. The precise mechanisms by which Plexin-B2-Rap controls PIP2 and membrane surface charge await further studies. In this context, it is noteworthy that Rap can directly bind to and regulate the activity of PLCε, which catalyzes hydrolysis of PIP2 to DAG and IP3⁶⁴. Rap can also signal through RalGDS⁶⁵, an activating GEF for Ral, a Ras family member associated with exocytosis⁶⁶, thus linking plexin function with membrane homeostasis.

GBM is known for high intertumoral heterogeneity, and transcriptomic profiling has identified different molecular subtypes⁶⁷, with the mesenchymal subtype linked to the highest levels of infiltration and high malignant potency⁴. In our assays, both SD2 (mesenchymal) and SD3 (proneural) are proficient at invading microchannels, and overall, the functional impact of Plexin-B2 KO did not substantially differ between SD2 and SD3. It is noteworthy that the microchannel assay does not fully equate to invasion in vivo where tumor cells need to clear extracellular matrix and interact with tumor microenvironment², nevertheless the microchannel system provides a useful platform for dissecting biomechanical processes during confined migration. The geometry of constrictions and local cell density also influence migration consistency⁶⁸, and constriction from all sides as in 5 × 5 µm channels can impact migratory behavior⁶⁹. The microchannels used in this study impose constrictions only laterally, as the height of the channel was 10 µm. Future studies with alternative channel designs will explore these parameters.

While Plexin-B2 deletion compromised confined migration, nuclear deformation in narrow tunnels still occurred, indicating intact nuclear compression responses⁵. Plexin-B2 KO also did not affect viscosity-induced cortical actin contractions, which involves focal adhesions in response to mechanical stretch forces³⁴. Hence, nuclear deformation, cortical actomyosin contraction, and membrane tension appear to be regulated by different pathways in different contexts. Additional molecular dynamics simulations can help predict the performance of GSCs at constrictions wider than 3 μm (about the lowest limit for nucleus to pass through), and how nuclear compression functioning as a mechanosensitive ruler⁷⁰ is linked to Plexin-B2-mediated biophysical mechanisms.

In summary, our studies establish a primary role of Plexin-B2 in mediating membrane mechanics and mechano-electrical coupling that enable GBM cells to adapt to physical constraints to achieve polarized migration. These conceptual advances on the mechanisms of GBM invasion will open avenues to investigate mechanosignaling as translational opportunities to curb GBM spread.

Methods

Human GBM cell lines

De-identified human GBM stem cell (GSC) lines SD2 and SD3 had been established from resected tumor tissue of GBM patients at the University of California, San Diego, and have been characterized by their transcriptomes as mesenchymal and proneural subtype, respectively⁶. GSCs were propagated in neural stem cell media (Neurocult NS-A proliferation kit (human), Stemcell Technologies), supplemented with bFGF (10 ng/ml; Peprotech), EGF (20 ng/ml; Peprotech), heparin (0.0002%; Stemcell Technologies), and penicillin–streptomycin (1:100; Gibco). Cells were propagated in adherent conditions on culture dishes coated with laminin (10 µg/ml in PBS; 1 hour at 37 °C; Gibco). Passaging of GSCs was performed by dissociating cells with Accutase (BD Biosciences).

hESC culture

To test the effect flexible ring of Plexin-B2 extracellular domain during colony formation and response to semaphorin, we utilized WA09 (H9) hESC line obtained from the University of Wisconsin and validated at the Mount Sinai Stem Cell core facility¹⁹. Briefly, H9 hESC were seeded on matrigel-coated cell culture dishes (1:100, Corning) and grown in mTeSR1 medium (STEMCELL Technologies). Cells were passaged around every 5 days by dissociation with Accutase (BD Biosciences) and addition of 2.5 μM Rho-associated protein kinase (ROCK) inhibitor Thiazovivin (Millipore). The study was approved by the Embryonic Stem Cell Research Oversight Committee (ESCRO) at Icahn School of Medicine at Mount Sinai.

Neural differentiation

Human neural progenitor cells (hNPCs) were generated by a single-cell monolayer protocol using STEMDiff Neural Induction Medium (StemCell Technologies)¹⁹. Briefly, hESCs were plated at a density of 2.5 × 10⁵ cells/cm² in STEMDiff medium, and on day 6 cells were replated at the high confluence. After day 11, cells were plated on glass coverslips coated with Matrigel for analysis by immunocytochemistry to confirm NPC identity and were then used for subsequent experiments using microchannels devices.

Microchannel migration devices

To mimic the passage of GBM cells through narrow pores, microchannel devices consisting of polydimethylsiloxane (PDMS) polymer structures on 35 mm glass bottom Petri dishes with parallel rows of microchannels of 10 µm height, 12 µm width, and constrictions of 3 or 8 µm were used (4D Cell, #MC011).

To simulate the passage of GBM cells through narrow tunnels, we created microdevices containing central chambers with outflow tunnels of 3 or 8 μm width and 50 µm length. This device allows cells to choose different exit tunnels from the chambers. These devices were designed using CAD software to create a photolithographic chrome mask for groove and cell culture compartment regions. Photolithography was then used to print mask features onto a silicon wafer, coated with negative SU-8 photoresist. The final devices were fabricated in a replica molding process by casting a PDMS prepolymer mixture against the positive relief master mold to obtain a negative device replica. Well openings were punched into devices, which were then bonded to the glass bottom of a cell culture dish using corona discharge treatment.

Before seeding of cells, the devices were immersed in 70% ethanol and left for 5 min inside a benchtop vacuum desiccator (Southern Labware) to remove air bubbles. After washes with PBS, the microchannels were coated with laminin (100 µg/ml in PBS; Invitrogen) for 1 hour at 37 °C. All wash and coating steps included 5 min placement inside a vacuum desiccator.

To initiate the microchannel culture, 3 × 10⁴ − 10⁵ GSCs cells were seeded into the entry ports. After incubation for 4 − 24 h, cells were labeled by adding the dyes MemGlow 488 (Cytoskeleton; 1:200), SPY555-Actin (Cytoskeleton; 1:1000), and NucSpot Live 650 (Biotium; 1:1000). Migrating GBM cells in the microchannels were imaged every 5 min on a LSM 780 (Zeiss) confocal microscope (heated stage, 37 °C, with 5% CO₂) for up to 17 hr. Image analysis was performed with ImageJ (MTrackJ plugin) to determine values as described below.

For the devices containing microchannels with constrictions, velocity of cells through constrictions was calculated by tracking cells from the time the front of the cell reached the constriction (position 1) until the rear of the cell completely exited the constriction (position 2). The time and distance between the two positions were used to calculate velocity. Stalling time was calculated as the time span a migrating cell would remain stalled in front of a constriction before entering the constriction.

To calculate the ratio of intensities of different fluorescent labels, including SPY-actin, Fluo-4, FluoVolt, and PH-PLCD1-GFP, annexin V, and MemGlow, the different front and rear fluorescence intensities of a migrating cell were measured when the nucleus was in the middle of a tunnel or constriction. The mean intensities were measured within each compartment (using the nucleus as boundary between the front or rear compartments).

Treatment with endocytosis inhibitors

Treatment with small molecule endocytosis inhibitors for live-cell imaging experiments was carried out on cells plated into the microchannel devices. Cells were incubated for 10 min with Pitstop 2 (30 µM; Abcam)²⁶, 30 min with Dynasore (80 µM; Sigma)⁷¹, or 90 min with 5-(N-ethyl-N-isopropyl)-amiloride (EIPA) (50 µM; Cayman Chemical)⁷². After drug or vehicle treatment, the GBM cells were washed with PBS and stained with SPY555-Actin, NucSpot, and FluoVolt or MemGlow as described above and imaged immediately on a LSM 780 confocal microscope (Zeiss).

Scratch assay

GBM cells were seeded in six-well plates at a density of 10⁶ cells per well. After one day of culture, cells were treated with endocytosis inhibitors (Pitstop 2, Dynasore, and EIPA) as described above. The cells were washed with PBS and the media was replenished, followed by a longitudinal scratch made at the center of the dish with a 200-µl size micropipette tip (t = 0 hr), and wound closure speed was determined by micrographs of identical areas at t = 24 hr, 48 hr, and 72 hr after scratch.

Immunocytochemistry

For immunocytochemistry (ICC) of cells in culture, cells were fixed with 3.7% formaldehyde in PBS at room temperature for 10 min, followed by three washes with PBS, and incubation with blocking buffer (5% donkey serum and 0.3% Triton X-100 in PBS) for 1 hr. Primary antibodies were added in dilution buffer (PBS with 1% BSA and 0.3% Triton X-100) at 4 °C overnight. Secondary antibodies were incubated in dilution buffer at room temperature for 1 hr together with nuclear stain (DAPI, 1:1000; Thermo Fisher). For most preparations, the cells were also counterstained for F-actin with Alexa-594 phalloidin (Thermo Fisher).

Western blotting

For Western blotting (WB), cells were lysed with RIPA buffer (Sigma) containing protease and phosphatase inhibitors. For phospho-ERM WB, cells were lysed using Cell Lysis Buffer (Cell Signaling, 9803). Protein concentrations were determined using a BCA assay (Thermo Scientific). Proteins were resolved by SDS-PAGE on 4–12% polyacrylamide NuPAGE gels (Invitrogen) and transferred onto nitrocellulose or PVDF membranes (Li-Cor Biosciences) with the XCell transfer system (Invitrogen). Membranes were incubated at 4 °C overnight with primary antibodies and then for 1 hr with secondary donkey antibodies coupled to IRDye 680 or 800 (Li-Cor). The fluorescent bands were detected with an Odyssey infrared imaging system (Li-Cor Biosciences).

Antibodies

Primary antibodies for immunocytochemistry (ICC):

anti-Plexin-B2, extracellular domain, R&D systems AF5329 (sheep), dilution 1:300;

anti-Ezrin, Sigma SAB4200806 (mouse), dilution 1:1,000;

anti-phospho-ERM (Ezrin (Thr567)/Radixin (Thr564)/Moesin (Thr558)) (48G2), Cell Signaling 3726S (rabbit), dilution 1:1,000;

anti-Ki67, Abcam ab15580 (rabbit), dilution 1:200;

anti- β-catenin, BD Bioscience 610153 (mouse), dilution 1:200.

Secondary antibodies for ICC:

Alexa Fluor 488, 594, or 647-conjugated donkey anti-sheep, -rabbit, or -mouse IgG (Jackson ImmunoResearch Laboratories, dilution 1:300).

Primary antibodies for WB:

anti-Plexin-B2, extracellular domain, R&D systems AF5329 (sheep), dilution 1:500;

anti-Ezrin, Sigma SAB4200806 (mouse), dilution 1:1,000;

anti-phospho-ERM (Ezrin (Thr567)/Radixin (Thr564)/Moesin (Thr558)), Cell Signaling 3141S (rabbit), dilution 1:500;

anti-β-actin, Sigma A1978 (mouse), dilution 1:10,000;

Secondary antibodies for WB:

IRDye 800CW anti-mouse (donkey, Li-Cor Biosciences 926-32212,1:10,000), IRDye 680LT anti-goat (donkey, Li-Cor Biosciences 926-68024, 1:10,000), and IRDye 680RD anti-rabbit (donkey, Li-Cor Biosciences 926-68073, 1:10,000).

CRISPR KO of PLXNB2

The deletion of Plexin-B2 in human cell lines by lentiviral CRISPR-Cas9 vectors has been described in detail in our previous studies^6,19. Briefly, GSCs were stably transduced with lentivirus expressing Cas9 and a short guide RNA against the second coding exon of PLXNB2 or a guide RNA against the EGFP coding sequence (used as control). The plasmids for the lenti-CRISPR vectors have been deposited at Addgene (#86152 and #86153, respectively).

Lentiviral Plexin-B2 cDNA vectors

For overexpression and rescue experiments, human PLXNB2 cDNA lentiviral vectors^6,19 were used to transduce GSCs. The plasmids for lenti-PLXNB2 expression have been deposited at Addgene (pLV-PLXNB2: #86237 and pLV-PLXNB2-dECTO).

The locations of the Plexin-B2 lock mutations were determined by structural prediction of Plexin-B2 as described below. The mutations I436C & S993C (“Lock1”) and I436C & T1051C (“Lock2”) were introduced into a CRISPR-resistant PLXNB2 cDNA plasmid by site directed mutagenesis (Thermo). The mutant cDNAs were transferred by Gateway reaction into pLENTI PGK Neo DEST (Addgene #19067; deposited by Eric Campeau & Paul Kaufman). The lock mutant plasmids have been deposited at Addgene (lock1: #182879 and lock2: #182880). The Plexin-B2 lock mutant expressing lentiviruses were transduced into GBM cells carrying CRISPR/Cas9 Plexin-B2 KO and were selected with 200 µg/ml G418 (Gibco) for 7 days before confirmation of expression by WB and ICC.

CRISPR KO of SEMA4 genes

To investigate the role of Sema4s on GBM cell migration, we designed two lentiviral vectors, LV-sgSEMA4BCG-Cas9-Hygro and LV-sgSEMA4ADF-Cas9-Neo, and the respective plasmids were generated by VectorBuilder, Inc. The sgRNAs of these vectors target sequences in the structurally conserved Sema domains (see Fig. S14). The “single cut” CRISPR/Cas9 KO approach creates indel mutations, which in some cases do not cause frameshift mutations, with some remaining mutant protein being detectable. As the Sema4 indels are placed in the highly structured Sema domains, we predict that such remaining protein is non-functional, in analogy to our previous observations on Plexin-B2 Cas9 KO⁶. Lentiviral particles were generated following standard procedures and used to consecutively transduce WT or Plexin-B2 KO GSC lines.

Prediction of Plexin-B2 structure for locked ring mutations

We first predicted the 3D structure of human Plexin-B2 by mapping it onto multiple templates by following our previously described approach¹². To model the extracellular part of HsPLXB2, we used the PDB protein structure 5L56 as template, and to model the intracellular part, we used the PDB structures 5E6P, 3IG3, and 3RYT as templates. We identified amino acid positions in the N-terminal Sema domain and the juxta-membrane IPT5 domain where newly introduced cysteines are prone to form a disulfide to lock the ring. The mutations in I436C, S993C, and T1051C were obtained using the mutagenesis tool of Pymol and optimized through Swiss PDB Viewer. To determine these mutations, we analyzed the consensus results of the following predictors: (i) Disulfide by Design (cptweb.cpt.wayne.edu/DbD2/index.php), (ii) CYSPRED (gpcr.biocomp.unibo.it/cgi/predictors/cyspred/pred_cyspredcgi.cgi), (iii) Protein Interactions Calculator (pic.mbu.iisc.ernet.in/job.html), and (iv) PDB2PQR and PROPKA programs (server.poissonboltzmann.org/pdb2pqr).

Doxycycline-inducible Plexin-B2 knockdown or overexpression

Temporally controlled knockdown of Plexin-B2 was achieved with TET-ON lentiviral vectors expressing doxycycline (Dox)-inducible shRNA targeting PLXNB2 (pLKO-Tet-On-PLXNB2–shRNA2; deposited as Addgene #98400). Puromycin selection at 1 µg/ml was used to establish transduced cell lines.

The lentiviral vector for Dox-inducible Plexin-B2 overexpression was generated by inserting human PLXNB2 cDNA into a Dox controlled expression vector (pLenti-CMVtight-PLXNB2 iOE; deposited as Addgene #176849)¹⁹. Target cells were coinfected with a lentivirus expressing Tet-On 3 G transactivator protein with hygromycin resistance (VectorBuilder). Stable PLXNB2 Tet-On GSC lines were established by successive puromycin (1 μg/ml) and hygromycin (200 μg/ml) selection steps. The expression of PLXNB2 was induced by addition of 1 μg/ml Dox (MP Biomedicals) to the culture medium.

GBM cell stiffness measured by atomic force microscopy

Atomic force microscopy (AFM) measurements were conducted using an Asylum MFP-3D-BIO instrument (Asylum Research), coupled to an Olympus IX-80 inverted microscope¹⁹. Briefly, for AFM experiments, 10⁵ GBM cells were plated on 60 mm laminin-coated tissue culture dishes. The AFM setup included a gold-coated silicon nitride probe with triangular shaped body, blunted pyramidal tip (39° half-angle), and nominal spring constant of 0.09 N/m (TR400PB, Asylum Research). A 3 × 3 indentation array was conducted over a 5 µm square region of each cell body, avoiding the nucleus and cell edge. The AFM was set to perform indentations at a rate of 1 Hz and trigger at 20 nm of cantilever deflection to ensure consistency across measurements. Each indentation was fitted using a Hertzian model (assuming cell Poisson’s ratio of 0.45) to calculate the Young’s modulus. Poorly fitted curves were removed from downstream analysis.

Measurement of membrane tension with flipper-TR dye

Membrane tension measurements with the dye Flipper-TR (Spirochrome SC020)³¹ were performed by labelling GSC lines with Flipper-TR (1:1000) in neural stem cell media. After incubation at 37 °C for 15 min, cells were washed with PBS and once with fresh media, and cells were imaged live at a Leica TCS SP8 STED 3X super-resolution microscope on a heated and CO₂ and humidity-controlled stage. Z-stack images (1 µm total height, 0.1 µm intervals) were acquired from cell membrane regions using a 93x Plan-Apochromat 1.3 NA glycerin immersion objective. Excitation was performed using a pulsed 488 nm laser operating at 20 MHz with emission collected through bandpass 565/610 nm filter. Lifetimes of Flipper-TR were extracted from fluorescence lifetime microscopy (FLIM) images and fitted to a dual exponential model using FLIM Wizard LAS AF software (Leica). The longest lifetime with the higher fit amplitude was used to quantify membrane tension (lifetime (τ2) at the Leica LAS software).

Measurement of membrane tension with optical tweezers

The measurement of membrane tension with optical tweezers was performed using a C-Trap device (Lumicks). To measure the mechanical properties of the cell membrane, carboxyl polystyrene particles (1.5 – 1.9 µm diameter, Spherotec) were coated with the lectin concanavalin A (eBioscience) and added to the culture medium. A laser beam (10 W, 1,064 nm) was focused through a series of Keplerian beam expanders and a high numerical aperture objective (×63/1.45, oil immersion, Nikon) to trap a particle, which was then moved towards a cell membrane at a speed of 1 µm/s. After momentary contact between bead and plasma membrane, beads were moved away again to extend tethers from cells. The displacement of the bead from the center of the optical trap was recorded by a position-sensitive detector to calculate the tension force. Cell membrane stiffness was determined as the stable force value that was obtained during tether extrusion. Data collection and post-processing were performed using Lumicks Bluelake software and Python.

Optical tweezers measurements after treatment with drugs targeting the cytoskeleton were carried out directly in live GSCs after 3 hr of addition of blebbistatin (10 μM), latrunculin (5 μM), or vehicle. After the experiment, GSCs were fixed and stained with phalloidin for F-actin and DAPI for nucleus and imaged using LSM 710 confocal microscope (Zeiss).

Whole-cell patch-clamp recordings

Cells were plated onto poly-D-lysine (100 mg/ml; SigmaAldrich) coated 12-mm glass coverslips in 24-well plates. Whole-cell patch-clamp recordings were made following our established protocol⁷³. Borosilicate glass electrodes (Warner Instruments) of 3–6 MΩ were filled with an intracellular solution containing 130 mM KCl, 20 mM NaCl, 5 mM EGTA, 5.46 mM MgCl₂, 2.56 mM K₂ATP, 0.3 mM Li₂GTP and 10 mM HEPES (pH 7.4, ~320 mOsm). The extracellular solution contained 5 mM KCl, 155 mM NaCl, 0.5 mM CaCl₂, 2 mM MgCl₂, and 10 mM HEPES (pH 7.4, ~325 mOsm). For voltage clamp recordings, currents were elicited at 0.5 Hz with voltage steps to −100 mV from a holding potential of −70 mV, then repeated to the next voltage steps until + 80 mV (20 mV difference in each step). For current clamp recordings, currents were held at 0 A, and the changes in membrane potential (Vm) were measured.

Fluorescent protein localization assays

Plasmids for expression of lipid anchored fluorescent proteins were obtained from the Addgene repository: MyrPalm-CFP (Addgene #14867) and MyrPalm-GFP (#21037), as well as plasmids for expression of a PI(4,5)P2 binding fluorescent protein: PH-PLCD1-GFP (#51407) and a PIP3 binding protein: PH-Btk-GFP (#51463). Also obtained from Addgene was a plasmid expressing a surface charge probe containing a series of arginine residues linked to a prenylation signal fused with GFP R( + 8)-prenyl-GFP (#17274).

For plasmid transfection, a cell suspension of GBM cells was electroporated using a Neon transfection system (Invitrogen) with the following parameters: 2 × 10⁵ cells, 15 µg of plasmid DNA, 1 pulse for 30 ms at 1300 V. Cells were seeded after electroporation on laminin-coated four-chamber glass-bottom dishes (Cellvis) at a density of 5 × 10⁴ cells per chamber and live imaged using a LSM 710 confocal microscope (Zeiss).

Dextran-Alexa 488 endocytosis assay

To visualize the endocytic activity of GBM cells, cells were incubated for 40 min with 5 mg/ml dextran Alexa488, MW 10,000 (Invitrogen D22910). After a wash with PBS, live cells were imaged at a confocal microscope. Dextran was excited at 488 nm and emitted fluorescence was recorded with an LSM 710 confocal microscope (Zeiss).

Low temperature effect on membrane permeability was assessed by keeping cells at 4 °C for 10 min, followed by dextran treatment also at 4 °C for 10 min, and then immediate imaging at a LSM 710 confocal microscope (Zeiss).

To visualize the effects of cytoskeletal drug treatment on dextran Alexa488 GSC permeability, cells were incubated for 1 hr with blebbistatin (20 μM). After wash with PBS, GSCs were incubated with dextran Alexa488 as described above, stained with nuclear dye Hoechst 33342 (1:2,000; Thermo 62249) and imaged using LSM 710 confocal microscope (Zeiss).

Staining of cells for cytoskeleton, lysosomes, and membrane components

To visualize filamentous actin in live cells, we stained cells with SPY555-actin (1:1,000; Cytoskeleton CY-SC202) at 37 °C for 1 hr. To visualize lysosomes, cultured cells were stained with LysoTracker Red DND-99 (Invitrogen L7528; 75 nM) at 37 °C for 10 min. Colocalization of LysoTracker and dextran Alexa488 in individual pixels was analyzed with the ImageJ plugin JACoP.

To assess cell viability, cultured cells were stained with membrane permeable Calcein Violet 450 AM (Thermo; 5 µM) at 37 °C for 20 min.

Membrane dynamics and endocytosis assessment with MemGlow, NR12A, and Flipper-TR

For the assessment of membrane dynamics and endocytosis, GBM stem cells were stained with MemGlow or NR12A for 20 min at 37 °C then imaged on a LSM 710 or 780 confocal microscope (Zeiss). For temporal analysis of endocytosis, GSCs were stained with Flipper-TR for 20 min at 37 °C and cells were analyzed at 2 −5 hr time points using Leica TCS SP8 STED 3X super-resolution microscope at 37 °C and 5% CO₂. No wash step was used for these experiments.

Three-dimensional surface rendering of GSCs

Z-stack images (20 µm total height, 0.21 µm intervals) were acquired from live-cell membranes stained with Nile red (NR12A) dye using a Plan-Neofluar 40x/1.30 oil immersion objective on a LSM 710 confocal microscope (Zeiss). Three-dimensional surface rendering was created in surpass 3D mode from CZI files (Zeiss software) using Imaris v10.0 (Bitplane).

FluoVolt live staining

To assess membrane potential of cells in 2D cell culture, FluoVolt (Thermo, 1:1000) membrane dye was diluted together with Powerload supplement (1:100) in neural stem cell media according to manufacturer’s instructions. Cells were stained for 30 min at 37 °C in a cell culture incubator, washed twice with PBS, and then imaged 15 min later with 488 nm excitation with an LSM 710 or 780 inverted confocal microscope (Zeiss).

To assess membrane potential of cells migrating in microchannels, cells were seeded in laminin-coated microchannel devices at a density of 3 × 10⁴ cells per chamber. After one day, cells were incubated with FluoVolt dye diluted together with Powerload supplement as described above. Additionally, NucSpot Live 650 (1:1,000; Biotium) dye was used to label nuclei. Cells were imaged on a Zeiss LSM 780 inverted confocal microscope.

Fluo-4 live staining

To visualize Ca²⁺ signals, cells were seeded in laminin-coated microchannel devices at a density of 3 × 10⁴ cells per chamber. After one day of culture, cells were incubated with NucSpot Live 650 (1:1,000; Biotium), the fluorescent Ca²⁺ indicator Fluo-4-AM (1:500; Invitrogen), and 20% pluronic acid F-127 (1:500; Thermo P3000MP) for 1 hr at 37 °C. Cells were kept in medium and imaging was performed at 488 nm excitation for Fluo-4-AM and 647 nm excitation for NucSpot and emitted fluorescence was recorded with an LSM 780 inverted confocal microscope (Zeiss).

Viscous media assay

Viscous media was prepared by adding methylcellulose 65 kDa (Sigma) at a concentration of 0.6% (92 μM) to neural stem cell media to obtain a viscosity of 8 centipoise (cP). Notably, this is in the maximum range of physiological viscosity⁷⁴. To assess the effects of viscosity on the cytoskeleton and membrane dynamics of GBM cells, the cells were plated in regular neural stem cell media for 24 hr and exposed to 8 cP viscous media for 72 hr. The control group was kept in regular neural stem cell media. The cells were labeled with the dyes SPY555-Actin (Cytoskeleton; 1:1,000) and NucSpot Live 650 (Biotium; 1:1,000) for 1 hr. The media was replaced with fresh media containing MemGlow 488 (Cytoskeleton; 1:200) and cells were incubated for 20 min at 4 °C. GBM cells were then imaged on a LSM 710 (Zeiss) confocal microscope. The areas of plasma membrane (MemGlow) and of actin filaments (SPYactin) were quantified using ImageJ, and the ratio of membrane vs. actin areas was reported as membrane spreading.

Relative SPY555-actin and MemGlow intensities were measured using the corrected total cell fluorescence quantification (CTCF) method with ImageJ. Briefly, cells were manually selected on micrographs, and CTCF for SPY555-actin and MemGlow were calculated for each cell as (integrated density of fluorescence signal) − (area of cell × mean fluorescence background reading).

Treatment with oleoyl-L-α-lysophosphatidic acid

Treatment with oleoyl-L-α-lysophosphatidic acid (LPA; 5 µM; Sigma, L7260) was carried out by addition to media of live cells for 5 min. After treatment, GSCs were washed with PBS and fixed for ICC or lysed in RIPA buffer containing protease and phosphatase inhibitors for subsequent Western blotting analysis.

Molecular dynamics simulations

Cell model: Coarse-grained molecular dynamics modelling was performed using a cell model composed of beads representing the cell membrane, the nuclear membrane, actin filaments, and heads of actin filaments¹⁹. The nuclear membrane is connected to the cell membrane through actin filaments. Connections between the beads are modelled by springs of elastic constant κ. The total number of beads per cell in the model was n = 700.

Potentials: Each bead was associated with potentials, which produce an interaction force with neighbor beads. Thus, these potentials are responsible for generating the dynamics of the system. As we are differentiating cell membrane (particle 0), actin (particle 1), and heads of actin filaments (particle 3), the elastic constants associated with each particle may be different. Thus, κ₀₀ is the elastic constant that connects the membrane (particles 0 with 0), κ₃₀ is the constant that connects particle 3 with particle 0.

The model allows determining the position vector ${\vec{r}}_{i} (t)$ of each bead i at time t in relation to a coordinate system. With the position vector it is possible to find the position vector of the center of mass in the x direction, x_CM(t). The model also allows, through the position vectors, to determine the area vector ${\vec{A}}_{i} (t)$ of the cell at each time point. The negative areas correlated with membrane involution.

To compute the forces of interactions between a given particle i and its neighbors we chose some common potentials in the literature in the study of polymers. The potentials used in the model are Lennard-Jones Potentials (LJ), Weeks-Chandler-Anderson Potentials (WCA), Finite Extensible Nonlinear Elastic (FENE), and Bend Potential (BP). The potential BP allows the twisting of membranes and actin filaments. Thus, the total potential to which each particle i is subjected is given by:

U_{t o t} (\vec{r_{i}}, t) = U_{L J} ({\vec{r_{i}}}, t) + U_{W C A} ({\vec{r_{i}}}, t) + U_{F E N E} ({\vec{r_{i}}}, t) + U_{B P} (θ_{i})

where θ_i is the angle formed between 3 particles, the middle particle being particle i and $\{{\vec{r}}_{i}\}$ represents the set of position vectors of particle i with all other particles. With the expression of the potential, it is possible to obtain the force that is subject to particle i is given by

\vec{F_{i}} = - \vec{\nabla} U (\vec{r_{i}}, t)

Constrictions simulations: To perform constriction simulations, we added two walls to the model, one mobile and the other fixed. The fixed wall is at the top and there is no friction. The bottom wall is movable and has friction. The potential associated with friction was modeled by a cosine function in the x direction, that is,

U_{f r i c} (x_{i}, y_{i}) = U_{0} (y_{i}) [1 + \cos (G_{x} x_{i})]

where G_x = 2π/a is the lattice parameter and a being the distance between two beads that make up the bottom wall. The potential U₀(y_i) is of the type:

U_{0} (y_{i}) = \frac{b}{{[y_{i} - y_{w a l l}]}^{6}}

where y_i is the vertical position of a given particle i, y_wall is the vertical distance from the wall to that particle and b is a constant. This potential (Eq. 3) must be added to the total potential of Eq. 1. The frictional force generated by this potential produces an action-reaction pair that allows, under certain conditions, the cell to move between the walls. Initially (t = 0) the cell has a circular shape with radius R_C and the walls are farther apart. The system evolves and close to time t = 2 (250 arb. units) the lower wall stops moving, remaining at a distance d < R_C of the top wall. At this instant, the cell is constricted between the upper and lower walls.

For the constriction simulation, the variables N, ϒ (representing the ratio between nucleus and cell radius), R_C, U_fric and d were fixed. The variables of interest were the elastic constant of the membrane (κ₀₀) representing the membrane tension and the potential for the head of the actin filament to bind to the cell membrane beads, which were defined as U₃.

Thus, we used this model to analyze the movement of the cell’s center of mass (CM) and area (A) as a function of time for a given set of parameters of cell membrane elastic constant κ₀₀, and the potential at the head of the actin filament that binds to the cell membrane (U₃).

For a given set of parameters defining a cell, initial velocities were randomly drawn from a Maxwell-Boltzmann distribution at a specific temperature T. For most parameter choices, the cell exhibited random movement. However, once a direction was chosen, the movement became progressive, with acceleration tending toward a constant limiting velocity. Of note, in our simulation, the cell is immersed in a thermal reservoir at a constant temperature, thus the energy required to maintain progressive movement comes from this thermal reservoir. In other words, the total energy of the system remains constant.

Statistics and reproducibility

Data are presented as mean ± SEM unless otherwise indicated, with sample sizes and statistical details provided in the figure legends. All cell culture experiments in this manuscript were independently repeated at least three times with similar results, and no data were excluded from statistical analysis. Data collection experiments were not randomized, and investigators were not blinded. Statistical tests were chosen based on data distribution, following normality tests. No statistical methods were used to predetermine sample sizes, but our sample sizes are consistent with those in prior publications^6,19.

For comparisons between two groups, an unpaired two-sided t-test or Mann-Whitney test was applied. For data involving more than two groups, one-way analysis of variance (ANOVA) with Tukey’s or Dunnett’s post-hoc test, or Kruskal-Wallis test with Dunn’s multiple comparisons test, was used. For studies with repeated measures (RM), two-way RM ANOVA with Bonferroni post-hoc test was performed. Statistical analyses were conducted using GraphPad Prism 9, applying the NEJM (New England Journal of Medicine) style setting for reporting P values. Statistical significance was considered as P < 0.05 (*); P < 0.01 (**); and P < 0.001 (***), with P values less than 0.001 reported as “<0.001.”

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Supplementary information

Supplementary Information^{(26.8MB, pdf)}

41467_2024_55056_MOESM2_ESM.docx^{(15.8KB, docx)}

Description of Additional Supplementary Information

Supplementary Movie 1^{(20.7MB, mp4)}

Supplementary Movie 2^{(47.9MB, mp4)}

Supplementary Movie 3^{(15.5MB, mp4)}

Supplementary Movie 4^{(26.4MB, mp4)}

Supplementary Movie 5^{(10.5MB, mp4)}

Supplementary Movie 6^{(11.7MB, mp4)}

Supplementary Movie 7^{(52.7MB, mp4)}

Supplementary Movie 8^{(32.2MB, mp4)}

Supplementary Movie 9^{(20.2MB, mp4)}

Supplementary Movie 10^{(11.3MB, mp4)}

Supplementary Movie 11^{(9.1MB, mp4)}

Supplementary Movie 12^{(12.9MB, mp4)}

Supplementary Movie 13^{(53.2MB, mp4)}

Supplementary Movie 14^{(10.1MB, mp4)}

Supplementary Movie 15^{(61.4MB, mp4)}

Supplementary Movie 16^{(34.8MB, mp4)}

Supplementary Movie 17^{(10.8MB, mp4)}

Reporting Summary^{(88.4KB, pdf)}

Transparent Peer Review file^{(873.5KB, pdf)}

Source data

Source Data^{(7.3MB, xlsx)}

Acknowledgements

We thank Yvonne Jones and Vitul Jain, University of Oxford, for advice on Plexin structure predictions and the Mount Sinai AFM Departmental Core Facility, funded by a shared instrument grant from the National Institutes of Health/National Center for Research Resources to K.D.C. (S10RR027609). We also thank the microscopy core at Icahn School of Medicine for support with super-resolution microscopy and fluorescence lifetime imaging (NIH instrumentation grant S10OD021838). This work was supported by a National Institutes of Health K01 career development award to C.J.A (NS127948) and research awards to R.H.F (NS092735, NS125700) and H.Z. (NS107462, NS134159). Additional support was provided by the New York State Department of Health awards to H.Z. (C38330GG and C39068GG) and CAPES and CNPq funding to R.A.D, J.P.M, and P.V.Z.C. Fellowship support was provided by FAPEMIG and UFJF to J.S.L. and P.F.

Author contributions

C.J.A., R.H.F., and H.Z. designed the overall study, analyzed, and compiled data, and wrote the manuscript. C.J.A., T.H., S.S., and C.K. conducted experiments and performed data analyses. A.D. engineered microdevices. R.J.W. and K.D.C. performed AFM measurements and analyzed the data. H.N. and P.A.S. performed and analyzed the electrophysiological measurements. M.J.T. performed the optical tweezers experiment. J.S.L., P.F.C.F, N.P.D.N., and P.V.Z.C. performed structural predictions of Plexin proteins. R.A.D. and J.P.R.F.M. performed molecular dynamics simulations. All authors discussed manuscript preparation.

Peer review

Peer review information

Nature Communications thanks Jean-François Berret, Yuan Ma, and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Data availability

All data that has been generated for this study are available from the corresponding author R.H.F. upon request. Source data are provided with this paper.

Code availability

Molecular dynamics (MD) source codes have been deposited at Github [https://github.com/diasrodri/SimCellMD-1] and Zenodo [10.5281/zenodo.4977916].

Competing interests

K.D. Costa discloses his role as scientific co-founder and Chief Scientific Officer of Novoheart, Medera Inc. NovoHeart did not play any role in the design, conduct, or funding of this study. The other authors declare no competing interests.

Footnotes

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

These authors jointly supervised this work: Chrystian Junqueira Alves, Hongyan Zou, Roland H. Friedel.

Contributor Information

Chrystian Junqueira Alves, Email: chrystian.junqueira-alves@mssm.edu.

Hongyan Zou, Email: hongyan.zou@mssm.edu.

Roland H. Friedel, Email: roland.friedel@mssm.edu

Supplementary information

The online version contains supplementary material available at 10.1038/s41467-024-55056-6.

References

1.Cuddapah, V. A., Robel, S., Watkins, S. & Sontheimer, H. A neurocentric perspective on glioma invasion. Nat. Rev. Neurosci.15, 455–465 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
2.Mair, D. B., Ames, H. M. & Li, R. Mechanisms of invasion and motility of high-grade gliomas in the brain. Mol. Biol. Cell29, 2509–2515 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Beeghly, G. F., Amofa, K. Y., Fischbach, C. & Kumar, S. Regulation of tumor invasion by the physical microenvironment: lessons from breast and brain cancer. Annu Rev. Biomed. Eng.24, 29–59 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Kim, Y. et al. Perspective of mesenchymal transformation in glioblastoma. Acta Neuropathol. Commun.9, 50 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Venturini, V. et al. The nucleus measures shape changes for cellular proprioception to control dynamic cell behavior. Science370, eaba2644 (2020). [DOI] [PubMed] [Google Scholar]
6.Huang, Y. et al. Plexin-B2 facilitates glioblastoma infiltration by modulating cell biomechanics. Commun. Biol.4, 145 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Banerjee, T. et al. Spatiotemporal dynamics of membrane surface charge regulates cell polarity and migration. Nat. Cell Biol.24, 1499–1515 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Shinoura, N. et al. Identification and partial sequence of a cDNA that is differentially expressed in human brain tumors. Cancer Lett.89, 215–221 (1995). [DOI] [PubMed] [Google Scholar]
9.Le, A. P. et al. Plexin-B2 promotes invasive growth of malignant glioma. Oncotarget6, 7293–7304 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Buck, M. Membrane Proteins | The Plexin Family of Transmembrane Receptors. in Encyclopedia of Biological Chemistry III (Third Edition) (ed. J. Joseph) 594-610 (Elsevier, Oxford, 2021).
11.Junqueira Alves, C., Yotoko, K., Zou, H. & Friedel, R. H. Origin and evolution of plexins, semaphorins, and Met receptor tyrosine kinases. Sci. Rep.9, 1970 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Junqueira Alves, C. et al. Evolution and diversity of semaphorins and plexins in choanoflagellates. Genome Biol. Evol.13, evab035 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Kong, Y. et al. Structural Basis for Plexin Activation and Regulation. Neuron91, 548–560 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Wang, Y., Pascoe, H. G., Brautigam, C. A., He, H. & Zhang, X. Structural basis for activation and non-canonical catalysis of the Rap GTPase activating protein domain of plexin. Elife2, e01279 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Fournier, A. E. et al. Semaphorin3A enhances endocytosis at sites of receptor-F-actin colocalization during growth cone collapse. J. Cell Biol.149, 411–422 (2000). [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Kabayama, H. et al. Ca2+ induces macropinocytosis via F-actin depolymerization during growth cone collapse. Mol. Cell Neurosci.40, 27–38 (2009). [DOI] [PubMed] [Google Scholar]
17.Mehta, V. et al. The guidance receptor plexin D1 is a mechanosensor in endothelial cells. Nature578, 290–295 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Jiang, C. et al. Mechanochemical control of epidermal stem cell divisions by B-plexins. Nat. Commun.12, 1308 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Junqueira Alves, C. et al. Plexin-B2 orchestrates collective stem cell dynamics via actomyosin contractility, cytoskeletal tension and adhesion. Nat. Commun.12, 6019 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Fletcher, D. A. & Mullins, R. D. Cell mechanics and the cytoskeleton. Nature463, 485–492 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Friedl, P. & Wolf, K. Tumour-cell invasion and migration: diversity and escape mechanisms. Nat. Rev. Cancer3, 362–374 (2003). [DOI] [PubMed] [Google Scholar]
22.Tejero, R. et al. Gene signatures of quiescent glioblastoma cells reveal mesenchymal shift and interactions with niche microenvironment. EBioMedicine42, 252–269 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Wang, G. & Galli, T. Reciprocal link between cell biomechanics and exocytosis. Traffic19, 741–749 (2018). [DOI] [PubMed] [Google Scholar]
24.Djakbarova, U., Madraki, Y., Chan, E. T. & Kural, C. Dynamic interplay between cell membrane tension and clathrin-mediated endocytosis. Biol. Cell113, 344–373 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Gauthier, N. C., Masters, T. A. & Sheetz, M. P. Mechanical feedback between membrane tension and dynamics. Trends Cell Biol.22, 527–535 (2012). [DOI] [PubMed] [Google Scholar]
26.Dutta, D., Williamson, C. D., Cole, N. B. & Donaldson, J. G. Pitstop 2 is a potent inhibitor of clathrin-independent endocytosis. PLoS One7, e45799 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Macia, E. et al. Dynasore, a cell-permeable inhibitor of dynamin. Dev. Cell10, 839–850 (2006). [DOI] [PubMed] [Google Scholar]
28.Masereel, B., Pochet, L. & Laeckmann, D. An overview of inhibitors of Na(+)/H(+) exchanger. Eur. J. Med Chem.38, 547–554 (2003). [DOI] [PubMed] [Google Scholar]
29.Koivusalo, M. et al. Amiloride inhibits macropinocytosis by lowering submembranous pH and preventing Rac1 and Cdc42 signaling. J. Cell Biol.188, 547–563 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Danylchuk, D. I., Moon, S., Xu, K. & Klymchenko, A. S. Switchable Solvatochromic Probes for Live-Cell Super-resolution Imaging of Plasma Membrane Organization. Angew. Chem. Int Ed. Engl.58, 14920–14924 (2019). [DOI] [PubMed] [Google Scholar]
31.Colom, A. et al. A fluorescent membrane tension probe. Nat. Chem.10, 1118–1125 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Fehon, R. G., McClatchey, A. I. & Bretscher, A. Organizing the cell cortex: the role of ERM proteins. Nat. Rev. Mol. Cell Biol.11, 276–287 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Chyan, W., Kilgore, H. R. & Raines, R. T. Cytosolic uptake of large monofunctionalized dextrans. Bioconjug Chem.29, 1942–1949 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Bera, K. et al. Extracellular fluid viscosity enhances cell migration and cancer dissemination. Nature611, 365–373 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Harayama, T. & Riezman, H. Understanding the diversity of membrane lipid composition. Nat. Rev. Mol. Cell Biol.19, 281–296 (2018). [DOI] [PubMed] [Google Scholar]
36.Balla, T. & Várnai, P. Visualization of cellular phosphoinositide pools with GFP-fused protein-domains. Curr Protoc Cell BiolChapter24, Unit 24.24 (2009). [DOI] [PMC free article] [PubMed]
37.Várnai, P. & Balla, T. Visualization of phosphoinositides that bind pleckstrin homology domains: calcium- and agonist-induced dynamic changes and relationship to myo-[3H]inositol-labeled phosphoinositide pools. J. Cell Biol.143, 501–510 (1998). [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Dickson, E. J. & Hille, B. Understanding phosphoinositides: rare, dynamic, and essential membrane phospholipids. Biochem J.476, 1–23 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
39.McLaughlin, S., Wang, J., Gambhir, A. & Murray, D. PIP(2) and proteins: interactions, organization, and information flow. Annu Rev. Biophys. Biomol. Struct.31, 151–175 (2002). [DOI] [PubMed] [Google Scholar]
40.Ma, Y., Poole, K., Goyette, J. & Gaus, K. Introducing membrane charge and membrane potential to T cell signaling. Front Immunol.8, 1513 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Yeung, T. et al. Receptor activation alters inner surface potential during phagocytosis. Science313, 347–351 (2006). [DOI] [PubMed] [Google Scholar]
42.Miller, E. W. et al. Optically monitoring voltage in neurons by photo-induced electron transfer through molecular wires. Proc. Natl Acad. Sci. USA109, 2114–2119 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Wang, X. et al. Functional soluble CD100/Sema4D released from activated lymphocytes: possible role in normal and pathologic immune responses. Blood97, 3498–3504 (2001). [DOI] [PubMed] [Google Scholar]
44.Rozbesky, D. et al. Structural basis of semaphorin-plexin cis interaction. EMBO J.39, e102926 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
45.Haklai-Topper, L., Mlechkovich, G., Savariego, D., Gokhman, I. & Yaron, A. Cis interaction between Semaphorin6A and Plexin-A4 modulates the repulsive response to Sema6A. EMBO J.29, 2635–2645 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
46.Pathak, A. & Kumar, S. Independent regulation of tumor cell migration by matrix stiffness and confinement. Proc. Natl Acad. Sci. USA109, 10334–10339 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
47.Lanzetti, L. & Di Fiore, P. P. Endocytosis and cancer: an ‘insider’ network with dangerous liaisons. Traffic9, 2011–2021 (2008). [DOI] [PubMed] [Google Scholar]
48.Khan, I. & Steeg, P. S. Endocytosis: a pivotal pathway for regulating metastasis. Br. J. Cancer124, 66–75 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Tojima, T., Itofusa, R. & Kamiguchi, H. Asymmetric clathrin-mediated endocytosis drives repulsive growth cone guidance. Neuron66, 370–377 (2010). [DOI] [PubMed] [Google Scholar]
50.Jurney, W. M., Gallo, G., Letourneau, P. C. & McLoon, S. C. Rac1-mediated endocytosis during ephrin-A2- and semaphorin 3A-induced growth cone collapse. J. Neurosci.22, 6019–6028 (2002). [DOI] [PMC free article] [PubMed] [Google Scholar]
51.Wu, K. Y. et al. Semaphorin 3A activates the guanosine triphosphatase Rab5 to promote growth cone collapse and organize callosal axon projections. Sci. Signal7, ra81 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
52.Tanaka, H., Kanatome, A. & Takagi, S. Involvement of the synaptotagmin/stonin2 system in vesicular transport regulated by semaphorins in Caenorhabditis elegans epidermal cells. Genes Cells25, 391–401 (2020). [DOI] [PubMed] [Google Scholar]
53.Van Battum, E. et al. Plexin-B2 controls the timing of differentiation and the motility of cerebellar granule neurons. Elife10, e60554 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
54.Friedel, R. H. et al. Plexin-B2 controls the development of cerebellar granule cells. J. Neurosci.27, 3921–3932 (2007). [DOI] [PMC free article] [PubMed] [Google Scholar]
55.Goldenberg, N. M. & Steinberg, B. E. Surface charge: a key determinant of protein localization and function. Cancer Res70, 1277–1280 (2010). [DOI] [PubMed] [Google Scholar]
56.Heo, W. D. et al. PI(3,4,5)P3 and PI(4,5)P2 lipids target proteins with polybasic clusters to the plasma membrane. Science314, 1458–1461 (2006). [DOI] [PMC free article] [PubMed] [Google Scholar]
57.Suh, B. C. & Hille, B. Electrostatic interaction of internal Mg2+ with membrane PIP2 Seen with KCNQ K+ channels. J. Gen. Physiol.130, 241–256 (2007). [DOI] [PMC free article] [PubMed] [Google Scholar]
58.Murayama, T. et al. Loosening of Lipid Packing Promotes Oligoarginine Entry into Cells. Angew. Chem. Int Ed. Engl.56, 7644–7647 (2017). [DOI] [PubMed] [Google Scholar]
59.Akishiba, M. & Futaki, S. Inducible Membrane Permeabilization by Attenuated Lytic Peptides: A New Concept for Accessing Cell Interiors through Ruffled Membranes. Mol. Pharm.16, 2540–2548 (2019). [DOI] [PubMed] [Google Scholar]
60.Ballesteros, A., Fenollar-Ferrer, C. & Swartz, K. J. Structural relationship between the putative hair cell mechanotransduction channel TMC1 and TMEM16 proteins. Elife7, e38433 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
61.Worzfeld, T. et al. Genetic dissection of plexin signaling in vivo. Proc. Natl Acad. Sci. USA111, 2194–2199 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
62.Hota, P. K. & Buck, M. Plexin structures are coming: opportunities for multilevel investigations of semaphorin guidance receptors, their cell signaling mechanisms, and functions. Cell Mol. Life Sci.69, 3765–3805 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
63.Kooistra, M. R., Dubé, N. & Bos, J. L. Rap1: a key regulator in cell-cell junction formation. J. Cell Sci.120, 17–22 (2007). [DOI] [PubMed] [Google Scholar]
64.Bunney, T. D. & Katan, M. Phospholipase C epsilon: linking second messengers and small GTPases. Trends Cell Biol.16, 640–648 (2006). [DOI] [PubMed] [Google Scholar]
65.Bos, J. L., de Rooij, J. & Reedquist, K. A. Rap1 signalling: adhering to new models. Nat. Rev. Mol. Cell Biol.2, 369–377 (2001). [DOI] [PubMed] [Google Scholar]
66.Lepore, D. M., Martínez-Núñez, L. & Munson, M. Exposing the Elusive Exocyst Structure. Trends Biochem Sci.43, 714–725 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
67.Wang, Q. et al. Tumor Evolution of Glioma-Intrinsic Gene Expression Subtypes Associates with Immunological Changes in the Microenvironment. Cancer Cell32, 42–56.e46 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
68.Mahmud, G. et al. Directing cell motions on micropatterned ratchets. Nat. Phys.5, 606–612 (2009). [Google Scholar]
69.Chabaud, M. et al. Cell migration and antigen capture are antagonistic processes coupled by myosin II in dendritic cells. Nat. Commun.6, 7526 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
70.Lomakin, A. J. et al. The nucleus acts as a ruler tailoring cell responses to spatial constraints. Science370, eaba2894 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
71.Burk, K. et al. Post-endocytic sorting of Plexin-D1 controls signal transduction and development of axonal and vascular circuits. Nat. Commun.8, 14508 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
72.Jayashankar, V. & Edinger, A. L. Macropinocytosis confers resistance to therapies targeting cancer anabolism. Nat. Commun.11, 1121 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
73.Bodhinathan, K. & Slesinger, P. A. Molecular mechanism underlying ethanol activation of G-protein-gated inwardly rectifying potassium channels. Proc. Natl Acad. Sci. USA110, 18309–18314 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
74.Rosenson, R. S., McCormick, A. & Uretz, E. F. Distribution of blood viscosity values and biochemical correlates in healthy adults. Clin. Chem.42, 1189–1195 (1996). [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Information^{(26.8MB, pdf)}

41467_2024_55056_MOESM2_ESM.docx^{(15.8KB, docx)}

Description of Additional Supplementary Information

Supplementary Movie 1^{(20.7MB, mp4)}

Supplementary Movie 2^{(47.9MB, mp4)}

Supplementary Movie 3^{(15.5MB, mp4)}

Supplementary Movie 4^{(26.4MB, mp4)}

Supplementary Movie 5^{(10.5MB, mp4)}

Supplementary Movie 6^{(11.7MB, mp4)}

Supplementary Movie 7^{(52.7MB, mp4)}

Supplementary Movie 8^{(32.2MB, mp4)}

Supplementary Movie 9^{(20.2MB, mp4)}

Supplementary Movie 10^{(11.3MB, mp4)}

Supplementary Movie 11^{(9.1MB, mp4)}

Supplementary Movie 12^{(12.9MB, mp4)}

Supplementary Movie 13^{(53.2MB, mp4)}

Supplementary Movie 14^{(10.1MB, mp4)}

Supplementary Movie 15^{(61.4MB, mp4)}

Supplementary Movie 16^{(34.8MB, mp4)}

Supplementary Movie 17^{(10.8MB, mp4)}

Reporting Summary^{(88.4KB, pdf)}

Transparent Peer Review file^{(873.5KB, pdf)}

Source Data^{(7.3MB, xlsx)}

Data Availability Statement

All data that has been generated for this study are available from the corresponding author R.H.F. upon request. Source data are provided with this paper.

Molecular dynamics (MD) source codes have been deposited at Github [https://github.com/diasrodri/SimCellMD-1] and Zenodo [10.5281/zenodo.4977916].

[CR1] 1.Cuddapah, V. A., Robel, S., Watkins, S. & Sontheimer, H. A neurocentric perspective on glioma invasion. Nat. Rev. Neurosci.15, 455–465 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR2] 2.Mair, D. B., Ames, H. M. & Li, R. Mechanisms of invasion and motility of high-grade gliomas in the brain. Mol. Biol. Cell29, 2509–2515 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR3] 3.Beeghly, G. F., Amofa, K. Y., Fischbach, C. & Kumar, S. Regulation of tumor invasion by the physical microenvironment: lessons from breast and brain cancer. Annu Rev. Biomed. Eng.24, 29–59 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR4] 4.Kim, Y. et al. Perspective of mesenchymal transformation in glioblastoma. Acta Neuropathol. Commun.9, 50 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR5] 5.Venturini, V. et al. The nucleus measures shape changes for cellular proprioception to control dynamic cell behavior. Science370, eaba2644 (2020). [DOI] [PubMed] [Google Scholar]

[CR6] 6.Huang, Y. et al. Plexin-B2 facilitates glioblastoma infiltration by modulating cell biomechanics. Commun. Biol.4, 145 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR7] 7.Banerjee, T. et al. Spatiotemporal dynamics of membrane surface charge regulates cell polarity and migration. Nat. Cell Biol.24, 1499–1515 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR8] 8.Shinoura, N. et al. Identification and partial sequence of a cDNA that is differentially expressed in human brain tumors. Cancer Lett.89, 215–221 (1995). [DOI] [PubMed] [Google Scholar]

[CR9] 9.Le, A. P. et al. Plexin-B2 promotes invasive growth of malignant glioma. Oncotarget6, 7293–7304 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR10] 10.Buck, M. Membrane Proteins | The Plexin Family of Transmembrane Receptors. in Encyclopedia of Biological Chemistry III (Third Edition) (ed. J. Joseph) 594-610 (Elsevier, Oxford, 2021).

[CR11] 11.Junqueira Alves, C., Yotoko, K., Zou, H. & Friedel, R. H. Origin and evolution of plexins, semaphorins, and Met receptor tyrosine kinases. Sci. Rep.9, 1970 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR12] 12.Junqueira Alves, C. et al. Evolution and diversity of semaphorins and plexins in choanoflagellates. Genome Biol. Evol.13, evab035 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR13] 13.Kong, Y. et al. Structural Basis for Plexin Activation and Regulation. Neuron91, 548–560 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR14] 14.Wang, Y., Pascoe, H. G., Brautigam, C. A., He, H. & Zhang, X. Structural basis for activation and non-canonical catalysis of the Rap GTPase activating protein domain of plexin. Elife2, e01279 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR15] 15.Fournier, A. E. et al. Semaphorin3A enhances endocytosis at sites of receptor-F-actin colocalization during growth cone collapse. J. Cell Biol.149, 411–422 (2000). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR16] 16.Kabayama, H. et al. Ca2+ induces macropinocytosis via F-actin depolymerization during growth cone collapse. Mol. Cell Neurosci.40, 27–38 (2009). [DOI] [PubMed] [Google Scholar]

[CR17] 17.Mehta, V. et al. The guidance receptor plexin D1 is a mechanosensor in endothelial cells. Nature578, 290–295 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR18] 18.Jiang, C. et al. Mechanochemical control of epidermal stem cell divisions by B-plexins. Nat. Commun.12, 1308 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR19] 19.Junqueira Alves, C. et al. Plexin-B2 orchestrates collective stem cell dynamics via actomyosin contractility, cytoskeletal tension and adhesion. Nat. Commun.12, 6019 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR20] 20.Fletcher, D. A. & Mullins, R. D. Cell mechanics and the cytoskeleton. Nature463, 485–492 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR21] 21.Friedl, P. & Wolf, K. Tumour-cell invasion and migration: diversity and escape mechanisms. Nat. Rev. Cancer3, 362–374 (2003). [DOI] [PubMed] [Google Scholar]

[CR22] 22.Tejero, R. et al. Gene signatures of quiescent glioblastoma cells reveal mesenchymal shift and interactions with niche microenvironment. EBioMedicine42, 252–269 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR23] 23.Wang, G. & Galli, T. Reciprocal link between cell biomechanics and exocytosis. Traffic19, 741–749 (2018). [DOI] [PubMed] [Google Scholar]

[CR24] 24.Djakbarova, U., Madraki, Y., Chan, E. T. & Kural, C. Dynamic interplay between cell membrane tension and clathrin-mediated endocytosis. Biol. Cell113, 344–373 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR25] 25.Gauthier, N. C., Masters, T. A. & Sheetz, M. P. Mechanical feedback between membrane tension and dynamics. Trends Cell Biol.22, 527–535 (2012). [DOI] [PubMed] [Google Scholar]

[CR26] 26.Dutta, D., Williamson, C. D., Cole, N. B. & Donaldson, J. G. Pitstop 2 is a potent inhibitor of clathrin-independent endocytosis. PLoS One7, e45799 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR27] 27.Macia, E. et al. Dynasore, a cell-permeable inhibitor of dynamin. Dev. Cell10, 839–850 (2006). [DOI] [PubMed] [Google Scholar]

[CR28] 28.Masereel, B., Pochet, L. & Laeckmann, D. An overview of inhibitors of Na(+)/H(+) exchanger. Eur. J. Med Chem.38, 547–554 (2003). [DOI] [PubMed] [Google Scholar]

[CR29] 29.Koivusalo, M. et al. Amiloride inhibits macropinocytosis by lowering submembranous pH and preventing Rac1 and Cdc42 signaling. J. Cell Biol.188, 547–563 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR30] 30.Danylchuk, D. I., Moon, S., Xu, K. & Klymchenko, A. S. Switchable Solvatochromic Probes for Live-Cell Super-resolution Imaging of Plasma Membrane Organization. Angew. Chem. Int Ed. Engl.58, 14920–14924 (2019). [DOI] [PubMed] [Google Scholar]

[CR31] 31.Colom, A. et al. A fluorescent membrane tension probe. Nat. Chem.10, 1118–1125 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR32] 32.Fehon, R. G., McClatchey, A. I. & Bretscher, A. Organizing the cell cortex: the role of ERM proteins. Nat. Rev. Mol. Cell Biol.11, 276–287 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR33] 33.Chyan, W., Kilgore, H. R. & Raines, R. T. Cytosolic uptake of large monofunctionalized dextrans. Bioconjug Chem.29, 1942–1949 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR34] 34.Bera, K. et al. Extracellular fluid viscosity enhances cell migration and cancer dissemination. Nature611, 365–373 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR35] 35.Harayama, T. & Riezman, H. Understanding the diversity of membrane lipid composition. Nat. Rev. Mol. Cell Biol.19, 281–296 (2018). [DOI] [PubMed] [Google Scholar]

[CR36] 36.Balla, T. & Várnai, P. Visualization of cellular phosphoinositide pools with GFP-fused protein-domains. Curr Protoc Cell BiolChapter24, Unit 24.24 (2009). [DOI] [PMC free article] [PubMed]

[CR37] 37.Várnai, P. & Balla, T. Visualization of phosphoinositides that bind pleckstrin homology domains: calcium- and agonist-induced dynamic changes and relationship to myo-[3H]inositol-labeled phosphoinositide pools. J. Cell Biol.143, 501–510 (1998). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR38] 38.Dickson, E. J. & Hille, B. Understanding phosphoinositides: rare, dynamic, and essential membrane phospholipids. Biochem J.476, 1–23 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR39] 39.McLaughlin, S., Wang, J., Gambhir, A. & Murray, D. PIP(2) and proteins: interactions, organization, and information flow. Annu Rev. Biophys. Biomol. Struct.31, 151–175 (2002). [DOI] [PubMed] [Google Scholar]

[CR40] 40.Ma, Y., Poole, K., Goyette, J. & Gaus, K. Introducing membrane charge and membrane potential to T cell signaling. Front Immunol.8, 1513 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR41] 41.Yeung, T. et al. Receptor activation alters inner surface potential during phagocytosis. Science313, 347–351 (2006). [DOI] [PubMed] [Google Scholar]

[CR42] 42.Miller, E. W. et al. Optically monitoring voltage in neurons by photo-induced electron transfer through molecular wires. Proc. Natl Acad. Sci. USA109, 2114–2119 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR43] 43.Wang, X. et al. Functional soluble CD100/Sema4D released from activated lymphocytes: possible role in normal and pathologic immune responses. Blood97, 3498–3504 (2001). [DOI] [PubMed] [Google Scholar]

[CR44] 44.Rozbesky, D. et al. Structural basis of semaphorin-plexin cis interaction. EMBO J.39, e102926 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR45] 45.Haklai-Topper, L., Mlechkovich, G., Savariego, D., Gokhman, I. & Yaron, A. Cis interaction between Semaphorin6A and Plexin-A4 modulates the repulsive response to Sema6A. EMBO J.29, 2635–2645 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR46] 46.Pathak, A. & Kumar, S. Independent regulation of tumor cell migration by matrix stiffness and confinement. Proc. Natl Acad. Sci. USA109, 10334–10339 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR47] 47.Lanzetti, L. & Di Fiore, P. P. Endocytosis and cancer: an ‘insider’ network with dangerous liaisons. Traffic9, 2011–2021 (2008). [DOI] [PubMed] [Google Scholar]

[CR48] 48.Khan, I. & Steeg, P. S. Endocytosis: a pivotal pathway for regulating metastasis. Br. J. Cancer124, 66–75 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR49] 49.Tojima, T., Itofusa, R. & Kamiguchi, H. Asymmetric clathrin-mediated endocytosis drives repulsive growth cone guidance. Neuron66, 370–377 (2010). [DOI] [PubMed] [Google Scholar]

[CR50] 50.Jurney, W. M., Gallo, G., Letourneau, P. C. & McLoon, S. C. Rac1-mediated endocytosis during ephrin-A2- and semaphorin 3A-induced growth cone collapse. J. Neurosci.22, 6019–6028 (2002). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR51] 51.Wu, K. Y. et al. Semaphorin 3A activates the guanosine triphosphatase Rab5 to promote growth cone collapse and organize callosal axon projections. Sci. Signal7, ra81 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR52] 52.Tanaka, H., Kanatome, A. & Takagi, S. Involvement of the synaptotagmin/stonin2 system in vesicular transport regulated by semaphorins in Caenorhabditis elegans epidermal cells. Genes Cells25, 391–401 (2020). [DOI] [PubMed] [Google Scholar]

[CR53] 53.Van Battum, E. et al. Plexin-B2 controls the timing of differentiation and the motility of cerebellar granule neurons. Elife10, e60554 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR54] 54.Friedel, R. H. et al. Plexin-B2 controls the development of cerebellar granule cells. J. Neurosci.27, 3921–3932 (2007). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR55] 55.Goldenberg, N. M. & Steinberg, B. E. Surface charge: a key determinant of protein localization and function. Cancer Res70, 1277–1280 (2010). [DOI] [PubMed] [Google Scholar]

[CR56] 56.Heo, W. D. et al. PI(3,4,5)P3 and PI(4,5)P2 lipids target proteins with polybasic clusters to the plasma membrane. Science314, 1458–1461 (2006). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR57] 57.Suh, B. C. & Hille, B. Electrostatic interaction of internal Mg2+ with membrane PIP2 Seen with KCNQ K+ channels. J. Gen. Physiol.130, 241–256 (2007). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR58] 58.Murayama, T. et al. Loosening of Lipid Packing Promotes Oligoarginine Entry into Cells. Angew. Chem. Int Ed. Engl.56, 7644–7647 (2017). [DOI] [PubMed] [Google Scholar]

[CR59] 59.Akishiba, M. & Futaki, S. Inducible Membrane Permeabilization by Attenuated Lytic Peptides: A New Concept for Accessing Cell Interiors through Ruffled Membranes. Mol. Pharm.16, 2540–2548 (2019). [DOI] [PubMed] [Google Scholar]

[CR60] 60.Ballesteros, A., Fenollar-Ferrer, C. & Swartz, K. J. Structural relationship between the putative hair cell mechanotransduction channel TMC1 and TMEM16 proteins. Elife7, e38433 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR61] 61.Worzfeld, T. et al. Genetic dissection of plexin signaling in vivo. Proc. Natl Acad. Sci. USA111, 2194–2199 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR62] 62.Hota, P. K. & Buck, M. Plexin structures are coming: opportunities for multilevel investigations of semaphorin guidance receptors, their cell signaling mechanisms, and functions. Cell Mol. Life Sci.69, 3765–3805 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR63] 63.Kooistra, M. R., Dubé, N. & Bos, J. L. Rap1: a key regulator in cell-cell junction formation. J. Cell Sci.120, 17–22 (2007). [DOI] [PubMed] [Google Scholar]

[CR64] 64.Bunney, T. D. & Katan, M. Phospholipase C epsilon: linking second messengers and small GTPases. Trends Cell Biol.16, 640–648 (2006). [DOI] [PubMed] [Google Scholar]

[CR65] 65.Bos, J. L., de Rooij, J. & Reedquist, K. A. Rap1 signalling: adhering to new models. Nat. Rev. Mol. Cell Biol.2, 369–377 (2001). [DOI] [PubMed] [Google Scholar]

[CR66] 66.Lepore, D. M., Martínez-Núñez, L. & Munson, M. Exposing the Elusive Exocyst Structure. Trends Biochem Sci.43, 714–725 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR67] 67.Wang, Q. et al. Tumor Evolution of Glioma-Intrinsic Gene Expression Subtypes Associates with Immunological Changes in the Microenvironment. Cancer Cell32, 42–56.e46 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR68] 68.Mahmud, G. et al. Directing cell motions on micropatterned ratchets. Nat. Phys.5, 606–612 (2009). [Google Scholar]

[CR69] 69.Chabaud, M. et al. Cell migration and antigen capture are antagonistic processes coupled by myosin II in dendritic cells. Nat. Commun.6, 7526 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR70] 70.Lomakin, A. J. et al. The nucleus acts as a ruler tailoring cell responses to spatial constraints. Science370, eaba2894 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR71] 71.Burk, K. et al. Post-endocytic sorting of Plexin-D1 controls signal transduction and development of axonal and vascular circuits. Nat. Commun.8, 14508 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR72] 72.Jayashankar, V. & Edinger, A. L. Macropinocytosis confers resistance to therapies targeting cancer anabolism. Nat. Commun.11, 1121 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR73] 73.Bodhinathan, K. & Slesinger, P. A. Molecular mechanism underlying ethanol activation of G-protein-gated inwardly rectifying potassium channels. Proc. Natl Acad. Sci. USA110, 18309–18314 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR74] 74.Rosenson, R. S., McCormick, A. & Uretz, E. F. Distribution of blood viscosity values and biochemical correlates in healthy adults. Clin. Chem.42, 1189–1195 (1996). [PubMed] [Google Scholar]

PERMALINK

Invasion of glioma cells through confined space requires membrane tension regulation and mechano-electrical coupling via Plexin-B2

Chrystian Junqueira Alves

Theodore Hannah

Sita Sadia

Christy Kolsteeg

Angela Dixon

Robert J Wiener

Ha Nguyen

Murray J Tipping

Júlia Silva Ladeira

Paula Fernandes da Costa Franklin

Nathália de Paula Dutra de Nigro

Rodrigo Alves Dias

Priscila V Zabala Capriles

José P Rodrigues Furtado de Mendonça

Paul A Slesinger

Kevin D Costa

Hongyan Zou

Roland H Friedel

Abstract

Introduction

Results

Passage through confined space enhances migratory momentum and consistency of GBM cells

Fig. 1. Confined migration of GSCs requires endocytosis.

Confined migration involves membrane internalization via endocytosis

Guidance receptor Plexin-B2 controls cortical actin and membrane tension of GSCs

Fig. 2. Plexin-B2 regulates cytoskeletal and membrane tension in GSCs.

Plexin-B2 affects endocytosis and membrane internalization in GSCs

Fig. 3. Plexin-B2 affects membrane internalization in GSCs.

Plexin-B2-deficient GSCs display compromised confined migration

Fig. 4. Plexin-B2 is required for migration of GSCs through confined space.

Fig. 5. Confined migration in long tunnels.

Molecular dynamics modeling predicts coupling of actin cytoskeleton and membrane involution during confined migration

Plexin-B2 function affects membrane phospholipid composition

Fig. 6. Plexin-B2 KO affects PIP2 localization and inner membrane surface charge in GSCs.

Plexin-B2 deletion alters inner membrane surface charge

Fig. 7. Plexin-B2 function regulates inner membrane surface charge.

Flexible ring of Plexin-B2 extracellular domain is required for efficient confined migration

Fig. 8. Confined migration of GSCs requires the flexible extracellular ring of Plexin-B2.

Discussion

Methods

Human GBM cell lines

hESC culture

Neural differentiation

Microchannel migration devices

Treatment with endocytosis inhibitors

Scratch assay

Immunocytochemistry

Western blotting

Antibodies

CRISPR KO of PLXNB2

Lentiviral Plexin-B2 cDNA vectors

CRISPR KO of SEMA4 genes

Prediction of Plexin-B2 structure for locked ring mutations

Doxycycline-inducible Plexin-B2 knockdown or overexpression

GBM cell stiffness measured by atomic force microscopy

Measurement of membrane tension with flipper-TR dye

Measurement of membrane tension with optical tweezers

Whole-cell patch-clamp recordings

Fluorescent protein localization assays

Dextran-Alexa 488 endocytosis assay

Staining of cells for cytoskeleton, lysosomes, and membrane components

Membrane dynamics and endocytosis assessment with MemGlow, NR12A, and Flipper-TR

Three-dimensional surface rendering of GSCs

FluoVolt live staining

Fluo-4 live staining

Viscous media assay

Treatment with oleoyl-L-α-lysophosphatidic acid

Molecular dynamics simulations

Statistics and reproducibility

Reporting summary

Supplementary information

Source data

Acknowledgements

Author contributions

Peer review

Peer review information

Data availability

Code availability