Spatial Patterns of Ca2+ Signals Define Intracellular Distribution of a Signaling by Ca2+/Calmodulin-dependent Protein Kinase II*

Ca2+ plays a central role in cell signaling, and Ca2+/calmodulin-dependent protein kinase II (CaMKII) is a major mediator of Ca2+actions. The spatial distribution of intracellular Ca2+signaling is not homogenous, rather it is dynamically organized, and it has been speculated that spatial patterns of Ca2+ signals may function as a form of cellular information transmitted to downstream molecules. To address this issue, we studied the intracellular distributions of the signalings by CaMKII and Ca2+ in the same astrocytes. The former was visualized by monitoring site-specific phosphorylation of a cytoskeletal protein vimentin, using site- and phosphorylation-specific antibodies, while the latter was examined by fura-2-based Ca2+microscopy. Local Ca2+ signals induced vimentin phosphorylation by CaMKII localized in the same area. On the other hand, Ca2+ waves in astrocytes induced global phosphorylation of vimentin by CaMKII. A small population of vimentin filaments highly phosphorylated by CaMKII underwent structural alteration into short filaments at electron microscopic level. These results indicate that CaMKII transmits spatial patterns of Ca2+ signals to vimentin as cellular information. The possibility is discussed that spatial patterns of vimentin phosphorylation may be important for intracellular organization of vimentin filament networks.

Cell signaling is the fundamental strategy by which cells respond to extracellular stimuli. Intracellular distribution of cell signaling is considered to be an important factor affecting the manner in which cells respond to extracellular stimuli with spatial specificity (1,2). Although little is known of the spatial aspect of cell signaling, that of Ca 2ϩ signaling visualized by Ca 2ϩ microscopy is presently the best characterized example. Numbers of reports have shown that intracellular Ca 2ϩ signals occur locally and globally (3)(4)(5)(6). The intracellular distribution of Ca 2ϩ signaling in various types of cells was defined by the amplitude and direction of extracellular stimuli (7)(8)(9)(10). Therefore, it has been speculated that spatial patterns of Ca 2ϩ signals might be transmitted, as a form of cellular information, by a downstream molecule that induces Ca 2ϩ -dependent cellular responses (1-2, 6 -10).
To address this issue, we visualized site-specific phosphorylation of vimentin by CaMKII 1 and Ca 2ϩ signaling in the same astrocytes. CaMKII is located downstream of Ca 2ϩ signaling and is thought to regulate various cellular responses (11,12). Vimentin is an intermediate filament protein distributed widely in the cytoplasm (13,14) and is phosphorylated by several protein kinases, including CaMKII, in vivo (15,16). Therefore, vimentin can serve as a substrate for the examination of the cytoplasmic distribution of protein kinase activities (17,18). Here we report that vimentin phosphorylation by CaMKII was induced locally and globally by Ca 2ϩ signaling. The intracellular area of the phosphorylation was precisely defined by that of Ca 2ϩ signaling.
[Ca 2ϩ ] i Measurements-The [Ca 2ϩ ] i of cultured astrocytes was measured as described elsewhere (8,24). Briefly, the cells were incubated * This research was supported in part by Grants-in-Aid for Scientific Research and Cancer Research from the Ministry of Education, Science, Sports, and Culture of Japan, and special coordination funds from the Science and Technology Agency of the Government of Japan. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
§ To whom correspondence should be addressed. Tel.: 81-52-762-6111 (ext. 8825); Fax: 81-52-763-5233; E-mail: ninagaki@aichi-cc.pref.aichi.jp. with 10 M fura-2/AM in the Hepes-buffered Krebs-Ringer solution for 1 h and washed with the solution for 30 min. Cells on a coverslip were placed on the stage of an Olympus IMT-2 inverted microscope. Fluorescence images were obtained by a Hamamatsu CCD camera C2400 and stored in a digital image processor Argus-50. [Ca 2ϩ ] i was calculated from the ratio of the fluorescence intensities obtained with excitations at 340 nm and 380 nm on a pixel basis.
Electron Microscopy-Immunogold localization using MO82 was done as described previously (19). For standard electron microscopy, cells were fixed in 2% glutaraldehyde and 1 mM MgCl 2 in 0.1 M cacodylate buffer for 30 min followed by further fixation in 0.15% tannic acid in the same buffer at room temperature for 5 min. They were fixed again with 1% glutaraldehyde and 0.5% tannic acid in 0.1 M cacodylate buffer (pH 7.4) for 30 min, followed by postfixation with 1% OsO 4 in the same buffer on ice for 1 h. The cells were dehydrated with ethanol and embedded in Epon 812. Thin sections were mounted on grids, doubly stained with uranyl acetate and lead citrate, and observed under an electron microscope (JEM1200EX).

RESULTS AND DISCUSSION
Visualization of a Signaling by CaMKII-For visualization of CaMKII signaling, we monitored the site-specific phosphorylation of the cytoskeletal protein vimentin. Ser 38 and Ser 82 of vimentin are identified as the two major in vitro phosphorylation sites of vimentin by CaMKII, while Ser 6 , Ser 33 , Ser 50 , and Ser 55 are phosphorylated not by CaMKII but by other kinases (15) (Table I). We recently developed monoclonal antibodies YT33, TM50, 4A4, and MO82 that recognize the site-specific phosphorylation of vimentin at Ser 33 , Ser 50 , Ser 55 , and Ser 82 , respectively (19 -21, 27) (Table I). In addition, we produced here a monoclonal antibody MO6 and a polyclonal antibody GK38 that recognize the phosphorylation of vimentin at Ser 6 and Ser 38 , respectively (Table I), as described under "Experimental Procedures." Consistent with the in vitro CaMKII phos-phorylation sites, Western blotting analysis showed that GK38 and MO82 reacted with vimentin phosphorylated by CaMKII but not with nonphosphorylated vimentin (Fig. 1, A and B). On the other hand, MO6, YT33, TM50, and 4A4 did not recognize vimentin phosphorylated by CaMKII (data not shown).  Cultured astrocytes differentiated by dibutyryl cAMP were used to detect CaMKII activity. Previous studies have demonstrated the existence of CaMKII in astrocytes (26,28), and CaMKII activated by Ca 2ϩ was shown to phosphorylate vimentin in these cells (19,26). Furthermore, they display local and global Ca 2ϩ signaling in response to neurotransmitters (8,29). In vivo phosphorylation of vimentin at Ser 6 , Ser 33 , Ser 38 , Ser 50 , Ser 55 , and Ser 82 were immunocytochemically visualized using antibodies MO6, YT33, GK38, TM50, 4A4 and MO82, respectively. When [Ca 2ϩ ] i of astrocytes was elevated by incubation of the cells with 1 M ionomycin for 10 min, the phosphorylation of vimentin at Ser 38 and Ser 82 remarkably increased (Fig. 1, C-F) but those of Ser 6 , Ser 33 , Ser 50 , and Ser 55 did not (Fig. 1, G-J). Elevations in the levels of phosphorylation at Ser 38 and Ser 82 were further confirmed by Western blotting analysis using GK38 and MO82 (Fig. 2). Thus, the sites of vimentin phosphorylated by [Ca 2ϩ ] i elevation completely overlapped with the in vitro phosphorylation sites by CaMKII (Table I).
These results indicate that the phosphorylation of vimentin at Ser 38 and Ser 82 detected the vimentin phosphorylation by CaMKII.
We also located CaMKII in differentiated astrocytes. The affinity-purified antibody specific for 50-and 60-kDa subunits of CaMKII (23) immunostained astrocytes as described previously (26). Both the cell bodies and processes showed diffuse immunoreactivity, indicating that CaMKII is distributed throughout the cytoplasm of differentiated astrocytes (Fig. 1, K  and L).
Local and Global Signaling of CaMKII Induced by Ca 2ϩ Signals-Ser 82 is at present the only known in vitro phosphorylation site specific to CaMKII (Table I) and the Ca 2ϩ -induced vimentin phosphorylation at Ser 82 was inhibited by a specific inhibitor of CaMKII, KN-62 (19). Therefore in the following studies, we monitored the phosphorylation of Ser 82 to visualize CaMKII signaling. Ca 2ϩ signaling in astrocytes was induced by PGF 2␣ ; PGF 2␣ binds to FP-receptors on astrocytes and induces phosphatidylinositol 4,5-bisphosphate hydrolysis and intracel-lular Ca 2ϩ mobilization (30). [Ca 2ϩ ] i of astrocytes was measured using fura-2-based digital imaging Ca 2ϩ microscopy, then they were fixed and immunostained with MO82.
When 10 M PGF 2␣ was locally applied using a micropipette for 15 s near the end of a process of an astrocyte, [Ca 2ϩ ] i was elevated from the basal level (about 100 nM) to about 600 nM in the process but not in the cell body or in other processes (Fig.  3A, a and b). [Ca 2ϩ ] i then decreased to the basal level within 4 min (Fig. 3A, c). The [Ca 2ϩ ] i increase did not appreciably spread beyond the boundary seen in Fig. 3A, b, throughout the period. Activation of CaMKII monitored by the phosphorylation at Ser 82 localized only in the process where [Ca 2ϩ ] i had been elevated (Fig. 3A, d, arrowheads). We also observed local CaMKII activations that were similarly defined by the area of Ca 2ϩ signals in five other experiments. Propagation of intracellular Ca 2ϩ waves has been observed in astrocytes (8,29). Consistent with reports that Ca 2ϩ waves often initiate when cells receive stimuli strong enough to induce sustained [Ca 2ϩ ] i elevation in a localized area (8,31), sustained PGF 2␣ -induced [Ca 2ϩ ] i elevation propagated from a process to the cell body and then to the rest of the cell in the form of waves (Fig. 3B, a-c). In this case, vimentin phosphorylation by CaMKII was evoked throughout the cell (Fig. 3B, d). The data above show a good spatial correlation between Ca 2ϩ signaling and vimentin phosphorylation by CaMKII. Next, astrocytes were double-immunostained by MO82 and an anti-vimentin antibody, then examined by confocal microscopy. Vimentin phosphorylation by CaMKII occurred locally and globally, defined by the area of Ca 2ϩ signaling (Fig. 3C, a, b, d, and e). On the other hand, vimentin was localized diffusely throughout the cells (Fig. 3C,  c and f). Furthermore, CaMKII immunoreactivity was observed diffusely throughout the cells (Fig. 1, K and L). These data demonstrate that local and global phosphorylation of vimentin by CaMKII was not due to local and global intracellular distribution of vimentin or CaMKII, thereby indicating that the spatial patterns of Ca 2ϩ signaling were indeed transmitted by CaMKII to vimentin.
Electron Microscopic Analysis of the Vimentin Filaments Phosphorylated by CaMKII-We noted here that a small population of vimentin filaments in the processes of ionomycin-or PGF 2␣ -stimulated astrocytes underwent structural alteration into partial granular aggregates, but not in unstimulated cells. Fig. 4 is an electron microscopic analysis of the CaMKII-phosphorylated vimentin filaments in astrocytes. Vimentin filaments in most of glial processes formed thick bundles running along glial processes (Fig. 4A), as reported previously (32). On the other hand, vimentin filaments in the aggregates were fragmented into short filaments running in random directions and thin glial processes appeared to form a varicosity there (Fig. 4B). When examined by immunoelectron microscopy, the density of MO82 immunoreactive gold particles was higher on the aggregates of vimentin filaments (Fig. 4, C and D) compared with those on the filaments in other regions (Fig. 4, C  and E). We counted the number of the gold particles per micrometer of vimentin filaments in Fig. 4C. The mean density of the particles in the aggregates of vimentin filaments was 11.7 particles/m of filament, while that in the other regions was 2.9 particles/m of filament. Similar data were obtained in three other samples. These data suggest that the filament reorganization occurs when the level of phosphorylation by CaMKII is very high. It is unclear whether the structural alteration observed here is a typical change of filament structure under control of cell signaling. Because the population of the fragmented filaments was very low, more minute and coordinated alteration of the filament dynamics not detectable by microscopy may predominate. However, these findings are consistent with in vitro data that vimentin filaments disassembled when phosphorylated by CaMKII (15). The possibility that organization of intracellular vimentin filament networks is regulated by local and global phosphorylation by CaMKII would need to be considered.
In conclusion, we visualized CaMKII signaling by monitoring the site-specific phosphorylation of vimentin and showed that the spatial patterns of Ca 2ϩ signaling defined the intracellular distribution of vimentin phosphorylated by CaMKII. These results suggest that the spatial patterns of Ca 2ϩ signaling were transmitted via CaMKII to vimentin, as a type of spatial information. Although the population was very low, the structural change of vimentin filaments observed here raises the possibility that spatial signaling from Ca 2ϩ via CaMKII to vimentin may regulate the dynamics of vimentin filaments in astrocytes with spatial specificity. CaMKII phosphorylates a wide range of cellular proteins as well as vimentin (11,12), therefore a population of CaMKII activity that could not be monitored by phosphorylation of vimentin might exist. Spatial signaling from Ca 2ϩ via CaMKII to other substrates needs to be addressed in the future studies.