The Unfolded Protein Response Sensor PERK Mediates Stiffness-Dependent Adaptation in Glioblastoma Cells

Abstract

Glioblastoma multiforme (GBM) is the most aggressive brain tumor in adults. In addition to genetic causes, the tumor microenvironment (TME), including stiffening of the extracellular matrix (ECM), is a main driver of GBM progression. Mechano-transduction and the unfolded protein response (UPR) are essential for tumor-cell adaptation to harsh TME conditions. Here, we studied the effect of a variable stiff ECM on the morphology and malignant properties of GBM stem cells (GSCs) and, moreover, examined the possible involvement of the UPR sensor PERK herein. For this, stiffness-tunable human blood plasma (HBP)/alginate hydrogels were generated to mimic ECM stiffening. GSCs showed stiffness-dependent adaptation characterized by elongated morphology, increased proliferation, and motility which was accompanied by F-Actin cytoskeletal remodeling. Interestingly, in PERK-deficient GSCs, stiffness adaptation was severely impaired, which was evidenced by low F-Actin levels, the absence of F-Actin remodeling, and decreased cell proliferation and migration. This impairment could be linked with Filamin-A (FLN-A) expression, a known interactor of PERK, which was strongly reduced in PERK-deficient GSCs. In conclusion, we identified a novel PERK/FLNA/F-Actin mechano-adaptive mechanism and found a new function for PERK in the cellular adaptation to ECM stiffening.

Keywords: PERK; extracellular matrix stiffening; glioblastoma; mechanical stress; tumor microenvironment; unfolded protein response.

Figure 1

Preparation and characterization of human blood plasma (HBP)/alginate hydrogels with tunable stiffness. (A) CaCl₂ was used as an ionic crosslinker for both alginate and HBP, and NaCl was added to minimize the free radical effect on gel degradation. Tranexamic acid was added to inhibit HBP decomposition. (B) Fabricated PDMS cell culture mold was used for hydrogel preparations. (C) By varying the alginate concentration, the stiffness was tuned in a range from soft (1.4 kPa) to stiff (40 kPa). (D) The water content (%) of the hydrogels was determined in relation to stiffness. (E) Hydrogel structure was determined using scanning electron microscopy (SEM) images from the hydrogels with different stiffnesses.

Figure 2

Adaptation of GG16-LVGFP cells to an increasingly stiff matrix involves changes in cell morphology, F-Actin expression, and proliferation. (A) GG16-LVGFP were grown on hydrogels at the indicated stiffness, and changes in cell morphology were detected using laser scanning confocal microscopy (40×). (B) Microscopic analyses of GG16 cells cultured for 6 days on hydrogels with increasing stiffness. Nuclei stained with Dapi^TM (Blue), and F-Actin stained with Alexa fluor^TM 546-Phalloidin (Red). F-Actin expression and cell elongation were quantified. (C) Magnified image of F-Actin remodeling in GG16 cells showing plasma membrane localization at highest stiffness. (D) GG16-LVGFP cell proliferation was determined at increasing hydrogel stiffnesses, showing elevated proliferation rates with increasing matrix stiffnesses. * p ≤ 0.05; ** p ≤ 0.01; *** p ≤ 0.001.

Figure 3

PERK-deficient GG16 cells are impaired in cellular adaptation to increasing stiffness which is linked with aberrant FLNA expression. (A,E) GG16-WT and PERK-KO cells were cultured for 6 days on hydrogels with different stiffnesses and stained with Dapi^TM (Blue), Alexa fluor^TM 546-Phalloidin (Red), and FLNA—Alexa fluor^TM 488 (Green). Cell morphology, F-Actin, and FLNA expression was quantified and is depicted in (B–D) and (F–H) for GG16-WT and PERK-KO cells, respectively. Cell morphology (from round to elongated), F-Actin, and FLNA expression altered gradually in a stiffness-dependent manner in GG16-WT cells, which was not seen in PERK-deficient cells. * p ≤ 0.05; ** p ≤ 0.01; *** p ≤ 0.001.

Figure 4

Inhibition of F-Actin polymerization mimics phenotype of PERK-deficient cells by impairing cellular adaption to matrix stiffness. (A,E) GG16-WT cells treated with latrunculin B were cultured for 6 days on hydrogels of varying stiffness. Cells were stained for F-Actin, FLNA, and PERK together with the nuclei and imaged with confocal microscopy. Specific fluorescence was quantified for FLNA (B), F-Actin (C), cell elongation (D), and PERK expression (F). ** p ≤ 0.01; *** p ≤ 0.001.

Figure 5

Inhibition of PERK kinase activity partially impairs stiffness-dependent cellular adaptation. GG16-WT cells were treated with GSK414 for 6 days while growing on hydrogels with increasing stiffness. (A) Cells were stained for FLNA and F-Actin and imaged with confocal microscopy. (B–D) Quantified FLNA, F-Actin expression, and cell elongation. * p ≤ 0.05; *** p ≤ 0.001.

Figure 6

ATF4 is not involved in cellular adaption to matrix stiffness. GG16-WT and GG16-ATF4-KO cells were cultured on hydrogels with increasing stiffness. (A,E) Cells were stained for FLNA, F-Actin, and nuclei and imaged with confocal microscopy. Quantified expression of FLNA and F-Actin (B,C,F,G) and cell elongation (D,H). * p ≤ 0.05; ** p ≤ 0.01; *** p ≤ 0.001.

Figure 7

PERK deficiency reduces stiffness-dependent cell motility and proliferation. GG16-WT and PERK-KO spheroids were seeded on three hydrogels with increasing stiffness, as indicated (A). Cell migration was monitored during 48 h using brightfield microscopy. The surface area of the expanded spheroids was calculated and normalized to the surface area at 0 h (B,C). (D) Cell proliferation of GG16-WT and PERK-KO cells was determined after 6 days of growth on indicated stiff matrixes with Cell Tracker™ Green CMFDA Dye. Cell numbers were normalized to the 24 h point. * p ≤ 0.05; ** p ≤ 0.01; *** p ≤ 0.001.

Figure 8

Model depicting the role of PERK in mediating cellular adaptation of GSC to increasing matrix stiffness. In PERK-proficient cells, increasing matrix stiffness results in increased PERK and FLNA expression. PERK-FLNA interactions are required for F-Actin remodeling which is essential for cellular adaptation to stiffness. Phosphorylation of PERK likely facilitates FLNA interaction. This is accompanied by a change in cell morphology from round to an elongated phenotype and increased cell proliferation and migration. In PERK-deficient cells, this mechanism regulating stiffness-dependent F-Actin remodeling is disrupted, resulting in lack of morphological change and no stimulation of proliferation and migration.