Microglia Activate Migration of Glioma Cells through a Pyk2 Intracellular Pathway

Abstract

Glioblastoma is one of the most aggressive and fatal brain cancers due to the highly invasive nature of glioma cells. Microglia infiltrate most glioma tumors and, therefore, make up an important component of the glioma microenvironment. In the tumor environment, microglia release factors that lead to the degradation of the extracellular matrix and stimulate signaling pathways to promote glioma cell invasion. In the present study, we demonstrated that microglia can promote glioma migration through a mechanism independent of extracellular matrix degradation. Using western blot analysis, we found upregulation of proline rich tyrosine kinase 2 (Pyk2) protein phosphorylated at Tyr579/580 in glioma cells treated with microglia conditioned medium. This upregulation occurred in rodent C6 and GL261 as well as in human glioma cell lines with varying levels of invasiveness (U-87MG, A172, and HS683). siRNA knock-down of Pyk2 protein and pharmacological blockade by the Pyk2/focal-adhesion kinase (FAK) inhibitor PF-562,271 reversed the stimulatory effect of microglia on glioma migration in all cell lines. A lower concentration of PF-562,271 that selectively inhibits FAK, but not Pyk2, did not have any effect on glioma cell migration. Moreover, with the use of the CD11b-HSVTK microglia ablation mouse model we demonstrated that elimination of microglia in the implanted tumors (GL261 glioma cells were used for brain implantation) by the local in-tumor administration of Ganciclovir, significantly reduced the phosphorylation of Pyk2 at Tyr579/580 in implanted tumor cells. Taken together, these data indicate that microglial cells activate glioma cell migration/dispersal through the pro-migratory Pyk2 signaling pathway in glioma cells.

Fig 1. Factors released from microglia increase glioma invasiveness and migration.

Invasion (A) and migration (B) assays were performed. Number of invading and migrating glioma cells without (control) and with microglia in the lower compartment. Results are presented as mean ± S.D. with significant differences from controls (*) shown (p < 0.05). Unpaired t-tests were used to determine significance between groups.

Fig 2. Pyk2 is mostly detected in glioma cells rather than in other cell types in mouse brain.

Immunohistochemistry was performed on C57BL/6 mice brain sections containing the tumor area. GL261 glioma cells were implanted into the brains of C57BL/6 mice and grown for 16 days. Photographs show the tumor and surrounding healthy tissue. The dash line outlines the border of tumor. Anti-GFAP antibody was used to detect glioma cells and astrocytes (red, panel A), anti-Iba 1 antibody was used to detect microglial cells (green, panel B) and Pyk2 detection is presented in blue (panel C). The merged images of anti-GFAP and anti-Pyk2, of anti-Iba1 and anti-Pyk2, and of all antibodies together can be seen in merged image boxes D, E, F correspondingly. Insert panels d, e, and f represent enlarged images of astrocytes, microglia, and invading glioma cells. Solid arrows indicate glioma cells, frame arrows indicate astrocytes, and double headed arrows indicate microglia. Scale bar: 40 μm.

Fig 3. Factors released from microglia upregulate phosphorylation of Pyk2 in glioma cells.

Western blot analysis of pPyk2 (Tyr 579/580) protein in glioma cell lines. The signal is detected at the area corresponding to molecular weight of 116 kDa. The graph shows the density of protein in MCM and AMCM treatments relative to control. Two bands in pPyk2 (Tyr 579/580) detection identify phosphorylation in one or both sites. Intensity of the chemiluminescence signal was corrected for minor changes in protein content after densitometry analysis of the India ink stained membrane. The India Ink stained membranes are provided in S1 Fig. The “relative density” axis for HS683 cells is shown in a higher grid scale compare to other cell lines due to higher relative up regulation of Pyk2 phosphorylation in this cell line. Results are presented as mean ± S.D. with significant differences from control (*) (p < 0.05). One-way ANOVA followed by the Tukey’s multiple comparison test was used to determine significance between MCM or AMCM groups compared to control. 5 repeated experiments (N = 5) for each cell line were used for statistical analysis.

Fig 4. Microglia ablation in brain tumors using the CD11b-HSVTK/GCV system.

(A) Immunocytochemical detection of microglia in tumors developed in C57BL/6 and CD11b-HSVTK mice brains after local GCV administration. The tumors were generated by intracranial implantation of GL261 glioma cells. GCV was delivered to the tumor area through mini-osmotic pumps. Image shows the significant reduction of microglia in tumors developed in CD11b-HSVTK compared to C57BL/6 mice. Anti-Iba 1 antibody was used to detect microglial cells (green) and DAPI was used to detect all cell nuclei (blue). (B) Western blot detection of Iba-1 in tumors extracted from C57BL/6 and CD11b-HSVTK mice brains after the treatment with GCV. (C) The graph shows corresponding levels of Iba-1 protein expression in C57BL/6 and CD11b-HSVTK mice brain tumors after GCV administration determined by western blot. Mean ± S.E and significant difference from control (*) are shown (p < 0.05).

Fig 5. Local microglia ablation in tumor area reduces phosphorylation of Pyk2 in implanted GL261 glioma tumors in CD11b-HSVTK transgenic mice.

(A) Western blot detection of pPyk2 (Tyr 579/580) protein in glioma cells extracted from C57BL/6 and HSVTK-CD11b mice brains after treatment with GCV. Glioma cells were purified using Percoll gradients. (B) The graph shows the quantification of corresponding levels of pPyk2(Tyr 579/580) detected by western blot. Mean ± S.E and significant difference from control (*) is shown (p < 0.05).

Fig 6. FAK is not involved in microglial stimulation of migration in glioma cells.

Data obtained from migration assays for glioma cells. Cells were treated with 5nM (concentration that effectively blocks FAK) and 16nM (concentration that effectively blocks Pyk2) of PF-562,271. Due to different migration abilities in all presented cell lines the Y-axis scales are adjusted for each cell line in order to demonstrate the absolute numbers of migrating cells. Results are presented as mean ± S.D. with significant differences from control in each group (+), from Mock without microglia on the bottom (#), or from Mock with microglia on the bottom (*) (p < 0.05). One-way ANOVA followed by the Tukey’s multiple comparison test was used to determine significance between groups.

Fig 7. siRNA knock-down and/or pharmacological blockade of Pyk2 by PF-562,271 eliminates the stimulatory effect of microglia on glioma cell migration.

Data obtained from standard migration assays for control glioma cells and cells transfected with siRNA against Pyk2 with and without additional application of PF-562,271 in the presence and absence of microglia in the lower compartment. Due to different migration abilities in all presented cell lines the Y-axis scales are adjusted for each cell line in order to demonstrate the absolute numbers of migrating cells. Results are presented as mean ± S.D. with significant difference from control in each group (+), from Mock without microglia on the bottom (#), or from Mock with microglia on the bottom (*) (p < 0.05). One-way ANOVA followed by the Tukey’s multiple comparison test was used to determine significance between groups.