Astrocyte stellation, a process dependent on Rac1 is sustained by the regulated exocytosis of enlargeosomes

Abstract

Cultured astrocytes exhibit a flat/epitelioid phenotype much different from the star-like phenotype of tissue astrocytes. Upon exposure to treatments that affect the small GTPase Rho and/or its effector ROCK, however, flat astrocytes undergo stellation, with restructuring of cytoskeleton and outgrowth of processes with lamellipodia, assuming a phenotype closer to that exhibited in situ. The mechanisms of this change are known only in part. Using the ROCK blocker drug Y27632, which induces rapid (tens of min), dose-dependent and reversible stellations, we focused on two specific aspects of the process: its dependence on small GTPases and the large surface expansion of the cells. Contrary to previous reports, we found stellation to be governed by the small G protein Rac1, up to disappearance of the process when Rac1 was downregulated or blocked by a specific drug. In contrast cdc42, the other G-protein often involved in phenotype changes, appeared not involved. The surface expansion concomitant to cytoskeleton restructuring, also dependent on Rac1, was found to be at least partially sustained by the exocytosis of enlargeosomes, small vesicles distinct from classical cell organelles, which are abundant in astrocytes. Exhaustion of stellation induced by repeated administrations of Y27632 correlated with the decrease of the enlargeosome pool. A whole-cell process like stellation of cultured astrocytes might be irrelevant in the brain tissue. However, local restructuring of the cytoskeleton coordinate with surface expansion, occurring at critical cell sites and sustained by mechanisms analogous to those of stellation, might be of importance in both astrocyte physiology and pathology.

Fig. 1

Astrocyte stellation induced by 8Br-cGMP and Y27632. (A) Astrocytes of the brain tissue (left) and cultured in vitro (right) immunolabeled for GFAP. The cells in B-E were dually immunolabeled for β-tubulin (green) and α-actin (red). In the microscopy panels of this and the following figures nuclei were labeled blue by DAPI. (B) astrocytes preincubated in complete medium (left) or pre-starved overnight in low (1%) serum medium (right) were treated with 8Br-cGMP (100 μM, 16 h). (C–E) Dose dependence and time-course of the phenotype changes induced by treatment with Y-27632 in astrocytes. In (C), from left to right, 1 h treatment with 0, 1, 10, and 25 μM Y27632; in (D), from left to right, 25 μM Y27632 administered for 15, 30, and 60 min. (E) Enlargement of a 60 min-treated cells showing lamellipodia emerging from processes. (F, G) Distribution of ezrin and GFAP in astrocytes incubated for 0 and 60 min with 25 μM Y27632. Bars in (A) (valid in B–D and F), (E), and (G) are 10 μm. (H) Number/cell and length of the major processes induced in populations of at least 25 astrocytes by 30 and 60 min treatment with 25 μM Y27632.

Fig. 2

Effects of nocodazol and latrunculin A on cultured astrocytes treated or not with Y27632. The left panels of (A) and (B) show cells treated with nocodazol (A, 30 μM, 1 h) and latrunculin A (B, 5 μM, 15 min). The central and right panels show cells that, upon treatment with nocodazol (A) and latrunculin A (B) were further incubated with Y27632 (25 μM, 60 min) together with the cytoskeleton-addressed drugs. All cells were immunolabeled for β-tubulin (green) and α-actin (red). Bar in (A), valid in (B), is 10 μm.

Fig. 3

Astrocyte stellation depends on Rac1. (A, B) Cultured astrocytes 48 hr after transient transfection of shRNA, scrambled (A) or Rac1-specific (B), followed by treatment with Y27632 (25 μM, 60 min). (C, D) Subsequent frames of astrocytes transfected (GFP-positive, green) or not (GFP-negative) with the scrambled (C) or the Rac1-specific (D) shRNA, analyzed by spinning disc time-lapse imaging upon administration of Y27632 (25 μM). The numbers on the frames of C (valid also for D) mark the 4 min upon addition of the drug. (E–G) Astrocytes preincubated with the specific blocker of Rac1, EHT 1864 (5 μM, overnight, E), the specific blocker of cdc42 secramine (dissolved in 1% DMSO-0.5% BSA; 15 μM, 1 h, F) or the PI3 kinase blocker, wortmannin (0.3 μM, 1 h, G), followed by 1 h incubation with the same drugs without (left) or with (right; in G also the middle panel) Y27632 (25 μM). The cells of (A), (B), and (E–G) were immunolabeled for β-tubulin (green) and α-actin (red). Bars in (A) (valid in B), (C) (valid in D), and (E) (valid in F, G) are10 μm.

Fig. 4

Enlargeosomes participate in the surface enlargement of stellation. (A) Western blot results showing that astrocytes are rich of enlargeosomes (marker Ahnak, expressed in the range of enlargeosome-rich neural cells, PC12-27 and SH-SY5Y). (B) Astrocytes before (left) and after (right) stellation, immunolabeled for GFAP (green) and Ahnak (red). (C) Flat cultured astrocytes dually labeled for Ahnak (red) and markers of cytoplasmic organelles (green): calnexin for endoplasmic reticulum (left panel), 58 K for Golgi complex (arrows, middle panel), TGN38 for transGolgi network (arrows, right panel). (D, E) astrocytes before (left panels) and after Y27632-induced (25 μM, 60 min) stellation (middle panels) dually immuno-labeled for Ahnak (red) and the vesicular glutamate transporters vGLUT1 (D) and vGLUT2 (E) (green). The right panels of (D, E) are higher magnification, deconvolved images of the corresponding middle panels. Bars in (B) (valid in C and left/middle panels of D, E), and in (D, E) right panels, are 10 μm.

Fig. 5

Enlargeosome exocytosis induced by Y27632 depends on Rac1. (A) A resting flat astrocyte surface-negative for the enlargeosome marker, Ahnak. (B) Ahnak surface-labeling in astrocyte stimulated with ionomycin (1μM, 10 min). (C) Surface Ahnak labeling in astrocytes stellated by Y27632 (25 μM, 1 h). (D) Quantitative immuno-labeling of surface Ahnak labelling in groups of at least 25 cells, treated with Y27632 (25 μM) for the indicated times. (E) Ahnak surface labelling in astrocytes transfected (see Fig. 3) with the scrambled (left) or the Rac1-specific (right) shRNA, and then treated with Y27632 (25 μM, 1 h). Bars in (A), valid for (B, C) left panel; (C) right panel and (E) are 10 μm. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Fig. 6

Repeated Y27632 application/washout cycles. Effects on the phenotype and Ahnak levels of cultured astrocytes. (A–F) Cultured astrocytes immuno-labeled for β-tubulin (green) and α-actin (red). (A) Before any treatment; (B) after 1 h of Y27632 (25 μM); (C) as (B) but after 2 h washout; (D) as in (C) but after 1 h reapplication of Y27632; (E) as in D but after a second 2 h washout; (F) as in (E) but after a third 1 h application of Y27632. Bar in (A), valid in (B–F), is 10 μm. (A′–F′): Levels of Ahnak assayed by quantitative immunolabeling in groups of at least 25 cells each treated as in (A–F).