Plexin-B2 facilitates glioblastoma infiltration by modulating cell biomechanics

Infiltrative growth is a major cause of high lethality of malignant brain tumors such as glioblastoma (GBM). We show here that GBM cells upregulate guidance receptor Plexin-B2 to gain invasiveness. Deletion of Plexin-B2 in GBM stem cells limited tumor spread and shifted invasion paths from axon fiber tracts to perivascular routes. On a cellular level, Plexin-B2 adjusts cell adhesiveness, migratory responses to different matrix stiffness, and actomyosin dynamics, thus empowering GBM cells to leave stiff tumor bulk and infiltrate softer brain parenchyma. Correspondingly, gene signatures affected by Plexin-B2 were associated with locomotor regulation, matrix interactions, and cellular biomechanics. On a molecular level, the intracellular Ras-GAP domain contributed to Plexin-B2 function, while the signaling relationship with downstream effectors Rap1/2 appeared variable between GBM stem cell lines, reflecting intertumoral heterogeneity. Our studies establish Plexin-B2 as a modulator of cell biomechanics that is usurped by GBM cells to gain invasiveness.

c IF images of coronal section of brain transplanted with SD2 GSCs into right striatum (arrow indicates injection track) stained for human nuclear antigen. At 141 dpi, GBM cells had widely disseminated in the striatum and also deep into contralateral striatum along the corpus callosum (CC, outlined by dashed lines). DAPI for nuclear counterstaining.
d IF image of coronal section of the brain transplanted with SD1 GSCs at 14 dpi. Note early dissociation of tumor cells from tumor bulk and invasion into striatum and CC. Bottom: Enlarged image of boxed area, highlighting invasion of tumor cells along striatal axon fiber bundles (arrows).
e Left: Coronal section of the brain transplanted with SD4 GSCs at 84 dpi, stained with hematoxylin/eosin (H&E), showing tumor spreading into CC and striatum. Right, IF image of adjacent section (corresponding to boxed area) shows aggressive invasion of tumor cells inside striatal fiber bundles (arrows) and along corpus callosum (CC).
Supplementary Fig. 2 Control immunofluorescence staining for anti-human integrin β1. a IF images of corpus callosum area in normal adult mouse brain show positive immunosignals for vasculature (PECAM-1 + ), but absence of specific signals with human-specific anti-integrin β1 antibody (clone TS2/16).
b Left, IF staining of SD3 GSC transplant with human-specific anti-integrin β1 antibody (TS2/16) detects abundant infiltrating tumor cells in the corpus callosum (CC). Right, isotype control IF staining with purified mouse IgG serum (2 µg/ml) did not generate specific IF signals. c Gene expression levels of Plexin-B2 and Sema4 genes measured by RNA-seq as RPKM (reads-per-kb-per-million bases). Sema4C has the highest mRNA levels in GSC lines, followed by Sema4B, 4D, 4F, and 4G are expressed at lower levels. Also note that SD2 expresses lower level of Sema4C and 4B than SD3.
d Top: CRISPR/Cas9 strategy to target the second coding exon of PLXNB2 by sgRNA for cut (black triangle), which will create indel (insertion/deletion) mutations. PAM, protospacer adjacent motif. Bottom: Summary of plasmid and lentiviral CRISPR/Cas9 approaches and of possible indel mutation types.
e PLXNB2 alleles were sequenced in two clonal CRISPR/Cas9 KO lines of SD1. Line KO-1 carries bi-allelic in-frame deletion of 33 bp and line KO-2 carries one in-frame deletion of 15 bp and one frame-shift deletion of 13 bp.
f WB shows absence of mature Plexin-B2 α chain (170 kDa) in clonal PB2-KO lines of SD1 and SD4. A mutant precursor form of Plexin-B2 (240 kDa) is detectable in some KO lines due to in-frame mutations.
g IF images demonstrate loss of surface Plexin-B2 expression in KO clonal lines of SD1 and SD4 GSCs.
h WB of SD1-SD4 population lines with Plexin-B2 KO. These PB2-KO lines are polyclonal with in-frame and frame-shift mutations. In-frame mutations create a detectable mutant precursor proteins (MUTANT) at ~240 kDa that are not processed to mature α/β chain heterodimer. Note absence of mature Plexin-B2 α chain at 170 kDa. Supplementary Fig. 4 Supplementary Fig. 4 Plexin-B2 deletion does not significantly alter cell behaviors in 2D culture assays.
a Control and PB2-KO SD2 GSCs cultured on 2D laminin-coated dishes displayed no apparent differences in cell morphology revealed by staining for filamentous actin (F-actin; phalloidin staining) or for human integrin β1.
b Growth curves of SD2 and SD3 GSCs (control vs. PB2-KO), cultured on 2D laminin-coated dishes, show no significant effects of Plexin-B2 deletion on proliferation rate. Two-way ANOVA test; ns, not statistically significant.
d Limiting dilution sphere formation assay, showing that Plexin-B2 deletion did not affect stem cell frequency in either SD2 or SD3 GSCs, calculated using ELDA software. n = 3 independent replicates; p = 0.59 for SD2 and p = 0.85 for SD3; paired Student's t-test. c Representative IF images of coronal sections of brains with SD3 transplants show shift of preferred migratory path from axon fiber tracts in control GBM to along small vessels in PB2 KO GBM (arrowheads). Enlarged images of boxed area are shown on the right. Right: Quantification of invading tumor cells in association with vasculature. n = 6-9 areas from 3 independent transplants; unpaired t-test; *** p<0.001.
d IF images of coronal sections of striatum highlight tumor cell morphology and orientation as outlined by human integrin β1, and their spatial relationship with microvasculature (PECAM-1). In control SD3 transplant, invading GBM cells adhere to striatal fiber bundles surrounded by blood vessels at the periphery, with cell orientation in cross section from the coronal view. PB2 KO resulted in increased numbers of GBM cells that changed their orientation to align with vessel axes (arrowheads). Supplementary Fig. 6 Supplementary Fig. 6 Plexin-B2 reduces intercellular adhesiveness of GSC and alters durotaxis behavior.
a Cell dispersion assay. GSC aggregates were placed on laminin, and cells that detached from aggregate and dispersed on culture surface are quantified after 2 hours. Representative images and quantifications of cell dispersion from SD2 and SD3 wild-type and PB2 KO aggregates after 2 hours are shown.
b Hanging drop aggregation assay. Representative photos of aggerates of SD3 wild-type, PB2-KO, and PB2-OE GSCs at 48 hours after seeding and quantification of numbers and sizes of aggregates in hanging drops at 24 or 48 hours are shown. n = 5-11 hanging drops per group; *p<0.05, ***p<0.001; one-way ANOVA with Dunnett's multiple comparison test of each group against control.
c Differential aggregation assay. Cells were separately labeled with green or red CellTracker dyes and mixed 1:1 before seeding in hanging drops. Example images of aggregated spheres of SD3 GSC of different genotypes are shown. No differential cell distributions were detectable. This assay may not be not well-suited for SD3 GSCs (in contrast to SD2 GSCs, see Fig. 4), due to faster aggregation rates and large spheres of SD3 GSCs.
e Live-cell imaging. Top, still frames from time lapse live-imaging of SD3 GSCs labeled with CellMask membrane dye. Bottom, overlapping cell contour plots of three randomly selected cells in each group at 30 min-intervals over 90 mins. Note more dynamic movement of control GSCs, but more static locomotion for both PB2 KO and OE conditions. Supplementary Fig. 10 Uncropped membranes for Western blots that are shown in Fig. 2a, S1a, S3a, S3b, S3f, S3h.