JBC Advanced Glycation Endproducts

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M307066200 on October 24, 2003

J. Biol. Chem., Vol. 279, Issue 1, 585-596, January 2, 2004
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
279/1/585    most recent
M307066200v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Leemhuis, J.
Right arrow Articles by Meyer, D. K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Leemhuis, J.
Right arrow Articles by Meyer, D. K.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Rho GTPases and Phosphoinositide 3-Kinase Organize Formation of Branched Dendrites*

Jost Leemhuis, Stephanie Boutillier, Holger Barth, Thomas J. Feuerstein{ddagger}, Carsten Brock§, Bernd Nürnberg§, Klaus Aktories, and Dieter K. Meyer

From the Institut für Experimentelle und Klinische Pharmakologie und Toxikologie, the {ddagger}Neurologische Universitätsklinik, Albert-Ludwigs-Universität, Freiburg D-79104, and the §Institut für Biochemie und Molekularbiologie II, Klinikum der Heinrich-Heine Universität, Düsseldorf D-40225, Germany

Received for publication, July 2, 2003 , and in revised form, October 20, 2003.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Neurons receive information from other neurons via their dendritic tree. Dendrites and their branches result from alternating outgrowth and retraction. The Rho GTPases Rac and Cdc42 (cell division cycle 42) facilitate the outgrowth of branches, whereas Rho attenuates it. The mechanism of neurite retraction is unknown. Because the adenylyl cyclase activator forskolin causes numerous branched extensions in NG108-15 cells, we have investigated the underlying mechanism in this cell line. In additional studies, we used cultured hippocampal neurons in which forskolin induces branched dendrites. In both cell types, forskolin enhanced the activity of Cdc42, but not that of Rac, although both GTPases were necessary for the formation of branched extensions. Time lapse microscopy showed that forskolin did not increase the rate of addition of new extensions or branches, but it reduced the rate of the retraction so that more branched extensions persisted. Inhibition of phosphoinositide 3-kinase activity by wortmannin or LY294002 also reduced the rate of retraction and thus facilitated dendritic arborization. Forskolin diminished the activity of phosphoinositide 3-kinases. Inhibitors of phosphoinositide 3-kinases not only reduced the retraction but also the addition of new dendrites and branches. This reduction was no longer present when Rho kinase was simultaneously inactivated, suggesting an interaction of phosphoinositide 3-kinases and Rho kinase. The present results show a central role of phosphoinositide 3-kinases in dendrite formation. In neuronal cells, increased levels of cAMP can support dendritic arborization by modulating the activity of the lipid kinase.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
During neuronal differentiation in the central nervous system, axons and branched dendrites grow and establish synaptic contacts. Initiation, elongation, and pathfinding of these extensions are controlled by signal transduction systems that transform cellular genetic programs and extracellular attractant and repellent cues into changes in the organization of cytoskeletal proteins (1). The small GTPases of the Rho family, i.e. Rho, Rac, and Cdc42,1 organize the actin cytoskeleton and axon and dendrite outgrowth, when they are in their active, GTP-bound form (2, 3). Together with Cdc42 Rac initiates and extends growth cone-mediated axons, whereas activated RhoA retracts axons by increasing the tension of actomyosin stress fibers. Among the effectors involved, Rho kinase plays an important role (4). Inactivation of RhoA or Rho kinase induces the formation of extensions and prevents their repellent-induced retraction in neural cell lines and in cultured neurons (5-9).

According to recent findings, dendritic branches originate during a highly dynamic process, which consists of the addition and retraction of new extensions (10). Whereas Rac and Cdc42 regulate branch addition, Rho and Rho kinase seem to inhibit branch elongation (11-13). There is no evidence that the Rho GTPases regulate the retraction of branches.

In addition, members of class I phosphoinositide 3-kinases (PI3Ks) induce or support neurite formation (14-16). The isoforms {alpha}, {beta}, and {delta} of class Ia are activated via tyrosine kinase receptors by growth factors, such as nerve growth factor and insulin-like growth factor, and also by Ras and in the case of PI3K{beta} by G{beta}{gamma} proteins (17-19). The PI3K{gamma} isoform belongs to class Ib and is activated by G{beta}{gamma} and Ras (20, 21). Upon activation, PI3Ks produce membrane-bound PI(3,4)P2 and PI(3,4,5)P3. Subsequently, proteins with a pleckstrin homology domain can bind to these phosphoinositides. One of these proteins is the serine/threonine kinase Akt/protein kinase B (22, 23). After its phosphorylation at Thr308 and Ser473 by PDK1 and PDK2, respectively, activated Akt can interact with substrates such as glycogen synthase kinase-3 (GSK3), BAD, or the forkhead transcription factor FKHR-L1 (23, 24). In peripheral nerves as well as in neurons of the central nervous system, PI3Ks are necessary for axon formation (16, 25-27). Interactions of PI3Ks and Rho GTPases play an important role in cell motility (28).

Also, protein kinase A (PKA) initiates and supports the outgrowth of extensions (29, 30). PKA phosphorylates RhoA at Ser188 and causes its inhibition, which results in neurite formation (6, 31, 32). PKA has been shown to enhance as well as reduce the activation of Akt mediated by PI3Ks (33-36). In neural NG108-15 cells, activation of PKA by forskolin induces extensions with multiple branches (37, 38). We have used NG108-15 cells and cultured hippocampal neurons to study the role of Rho GTPases and PI3Ks in dendrite and branch formation induced by forskolin.


   EXPERIMENTAL PROCEDURES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
RESULTS
 DISCUSSION
 REFERENCES
 
Materials—Y-27632 was a gift from Welfide Corporation (Osaka, Japan). H89, LY294002, forskolin, and wortmannin were from Calbiochem, Tocris (Köln, Germany), and Sigma, respectively. The C2IN-C3 fusion toxin and the C2II binding component were purified as recombinant glutathione S-transferase (GST) fusion proteins (39, 40). Toxin B from Clostridium difficile VPI 10468 was purified as described previously (41). The anti-phospho-myosin light chain (MLC) antibody was a gift from M. Aepfelbacher (München, Germany). The expression plasmid encoding the GST-PAK-Cdc42/Rac interactive binding (CRIB) domain was a gift from Dr. J. G. Collard (Amsterdam, Netherlands). The cAMP assay kit was from Amersham Biosciences.

Cell Culture—Neuroblastoma x glioma hybrid NG108-15 cells were cultured at 37 °C with Dulbecco's modified Eagle's medium, pH 7.3, containing 4.5 g/liter glucose (PAN, Aidenbach, Germany) and supplemented with 10% fetal calf serum (Biochrom, Berlin, Germany) and 100 IU/ml penicillin/100 µg/ml streptomycin (Invitrogen). To induce extensions, cells were cultured in serum-free neurobasal medium supplemented with B27 (Invitrogen).

Primary cultures of hippocampal neurons were prepared from brains of embryonic rats at day 17. The dissociated cells were seeded on coverslips coated with poly-L-lysine. The incubation medium consisted of Dulbecco's modified Eagle's medium plus 10% fetal calf serum for the first 16 h. Neuronal differentiation was induced with StartV medium (Biochrom), pH 7.3. Cultures were incubated at 37 °C in a humidified atmosphere.

Cytochemical Staining—Cells were fixed with 4% paraformaldehyde for 20 min, washed with PBS, and permeabilized with 0.1% (v/v) Triton X-100. Normal goat serum was used to block nonspecific binding. For staining of {beta}-tubulin III, cells were incubated with a monoclonal mouse anti-{beta}-tubulin III antibody (Sigma). The resulting immune complex was visualized with a CyTm 3-conjugated F(ab')2 fragment or CyTm 5-conjugated F(ab')2 fragment of goat anti-mouse IgG (Dianova, Hamburg, Germany). For actin staining, cells were incubated with TRITC-conjugated phalloidin (Biozol, München, Germany) and washed again with PBS.

Confocal Image Analysis—A Bio-Rad MRC 1024 (version 3.2) confocal system with a krypton-argon laser was used with an Axiovert 135TV microscope (Zeiss, Oberkochen, Germany). Images obtained with Laser sharp 2.1T software were processed with Corel Photopaint.

Time Lapse Imaging—NG108-15 cells or hippocampal neurons were plated on poly-L-lysine coated "relocate" coverslips (Eppendorf, Hamburg, Germany). After induction of neuronal differentiation, phase-contrast micrographs were made from 10 different locations on each coverslip by using an Axiovert 135/Axiocam microscopic system. Afterward, cells were stained for {beta}-tubulin III to verify that the examined cells were indeed neurons. In NG108-15 cells, extensions and their branches were analyzed at 0, 30, 60, 90, and 120 min. In hippocampal neurons, analysis was done at 0, 2, 4, 6, and 8 h. At each time point, the dendrites and branches of the selected individual cells were compared with those of the previous time point. Newly added extensions and branches were counted separately, as were extensions and branches that had been retracted. Finally, all newly added and retracted primary extensions were summed separately. The numbers of newly added and of retracted branches were calculated correspondingly. Shown are the average values of 15 randomly selected cells.

Preparation of Transfection Vectors and Cell Transfection—The coding regions of the RhoAN19, Rac1V12, and Rac1N17 genes were excised from the plasmid pGEX with BamHI/EcoRI and inserted in-frame into the BglII/EcoRI sites of pEGFP-C1 (Clontech, Heidelberg, Germany). The coding region of Cdc42wt and Cdc42N17 genes were excised from the plasmid pCDNA3 with BamHI/EcoRI and inserted in-frame into the BglII/EcoRI sites of pEGFP-C1. The mutant Cdc42V12 was obtained from the wild type form by using the QuikChange Mutagenesis kit (Stratagene, La Jolla, CA). A p110{gamma}-CAAX expression plasmid of the catalytic subunit of PI3K{gamma} was generated as follows. Human cDNA (42) encoding 1,102 amino acids of p110{gamma} was amplified by using the primers 5'-GCC ACC ATG GAG CTG GAG AAC TAT AA and 5'-GGA TCC AGC TTT CAC AAT GTC TAT TG. Subcloning into pcDNA3 was accomplished by insertion into the KpnI and XhoI sites. Two AfeI sites in p110{gamma} were removed, and the stop codon was replaced by a new AfeI site by silent mutations using the QuikChange Mutagenesis kit and appropriate primer sets. Adapter oligonucleotides encoding the 18 C-terminal amino acids of H-Ras, which contains a C-terminal isoprenylation signal and two palmitoylation sites, were inserted into the AfeI and BamHI sites to yield the desired pcDNA3-p110{gamma}-CAAX. Cells were transfected for 2.5 h at 2.5% CO2 using a calcium phosphate/DNA coprecipitation procedure (43). Afterward, the Cells were grown in serum-free neurobasal medium supplemented with B27 for 16 h prior to use.

Expression of GST-PAK-CRIB Domain—The GST fusion proteins were expressed in Escherichia coli BL21 cells grown at 37 °C. Expression was induced by adding 0.1 mM isopropyl-1-thio-{beta}-D-galactopyranoside (final concentration) at A600 1.0. Two hours after induction, the cells were collected and lysed by sonication in lysis buffer (50 mM Tris-HCl, pH 8.0, 2 mM MgCl2, 2.0 mM dithiothreitol, 10% glycerol, and 1 mM phenylmethylsulfonyl fluoride). The lysate was centrifuged at 10,000 x g, and the supernatant was used for purification of the GST-PAK-CRIB domain by affinity purification using glutathione-Sepharose beads (Amersham Biosciences). Beads loaded with the GST fusion proteins were washed twice with PBS and used immediately for GTPase pull-down experiments.

GST-PAK-CRIB Domain Pull-down Experiments—NG108-15 cells were grown in serum-free neurobasal medium supplemented with B27. 1 x 106 cells were used for the Rac and 3 x 106 cells for the Cdc42 experiments. Hippocampal neurons were cultured in serum-free StartV medium; 1 x 106 cells were used for the experiments. Cells were treated with drugs as indicated under "Results." Then, they were washed twice with PBS. After addition of 250 µl of ice-cold GST-Fish lysis buffer (50 mM Tris, pH 7.4, 100 mM NaCl, 2 mM MgCl2, 10% glycerol, 1% (v/v) Nonidet P-40, and 25 µg/ml aprotinin), the cells were harvested. The detergent-soluble supernatant was recovered after centrifugation for 15 min at 14,000 x g and 4 °C. GTP-Rac or GTP-Cdc42 proteins were precipitated for 1 h with 20 µl of GST-PAK fusion protein at 4 °C. The complexes were washed three times with ice-cold PBS, resuspended, and boiled with Laemmli buffer. Bound Rac, Cdc42, and RhoA proteins were detected by Western blotting using anti-Rac1 (BD Bioscience, Heidelberg, Germany), or anti-Cdc42 (Upstate, Milton Keynes, UK) antibodies.

Phosphorylation Assays for Akt and MLC—NG108-15 cells were incubated in serum-free Neurobasal medium supplemented with B27. After drug treatment, cells were washed twice with PBS. Then, ice-cold lysis buffer (50 mM Tris, pH 7.4, 100 mM NaCl, 2 mM MgCl2, 10% glycerol, 1% (v/v) Nonidet P-40, and 25 µg/ml aprotinin) was added, and the cells were harvested. The detergent-soluble supernatant was recovered after centrifugation for 15 min at 14,000 x g and 4 °C. Cell lysates were separated by SDS-PAGE and subjected to Western blot analysis. To determine the phosphorylation of Akt, anti-phospho-Ser473 Akt and anti-Akt antibodies were used (New England Biolabs). An anti-phospho-MLC antibody was used to measure the phosphorylation of MLC.

In Vivo cAMP Assay—Intracellular cAMP levels of hippocampal neurons grown in StartV medium supplemented with B27 were measured by using a nonacetylation EIA procedure with the Amersham cAMP assay kit (Amersham Biosciences).

Evaluation of Cell Morphology and Statistics—Cells bearing extensions were examined by using the 20-fold magnification of a Zeiss Axiophot microscope. At least 15 cells/field were counted in 10 areas. The length of the extensions was determined by confocal laser microscopy using Laser sharp 2.1T software. For statistical analysis the Mann-Whitney-U test and the Kruskal-Wallis test (for multiple comparisons) were used. If samples showed normal distribution, analysis of variance and Scheffe's test were used.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
DISCUSSION
 REFERENCES
 
Forskolin Induces Branched Extensions in NG108-15 Cells—Within 15 min after the addition of 10 µM forskolin, NG108-15 cells began to form branched extensions. After 240 min, 82% of the cells had extensions, of which 83% also had branches (Fig. 1, A and B). In untreated cells, only 28% of the NG108-15 cells had extensions, 15% of which were branched (Fig. 1, A and B). 20 µM 1,9-dideoxyforskolin, which does not activate adenylyl cyclase, did not affect the arborization (data not shown).



View larger version (49K):
[in this window]
[in a new window]
  		FIG. 1.
Forskolin induces branched extensions in NG108-15 cells and diminishes MLC phosphorylation. A, cells were incubated in neurobasal medium. After the addition of 10 µM forskolin (FSK), cells were incubated for 240 min and analyzed with phase-contrast microscopy. Cells were treated for 4 h with 100 ng/ml C3FT or 30 ng/ml toxin B (ToxB) in the absence or presence of 10 µM forskolin; bar ~ 50 µm. B, quantification of the experiment shown in A. Upper row, the numbers show cells with extensions >= 50 µm as a percentage of the total number of cells. The diagram shows extensions with branches as a percentage of the total number of extensions. Mean ± S.E., n >=150; a indicates a significant difference (p < 0.05) from controls; b refers to a significant difference (p < 0.05) from C3FT-treated cells. C, Western blot showing the effect on MLC phosphorylation of treatment with 100 ng/ml C3FT for 4 h, 10 µM Rho kinase inhibitor Y-27632 for 4 h, or 10 µM forskolin for 4 h. The upper row shows phosphorylated MLC (P-MLC); lower row, total amount of MLC.

 
To test whether the forskolin-induced branch formation depended on Rho activity, NG108-15 cells were treated with C3 toxin of Clostridium limosum, which specifically inactivates Rho by ADP-ribosylation at Asn41 (44). For optimal cellular uptake, a C3 fusion toxin (C3FT) was used, which enters cells by using the binary C. botulinum C2 toxin as a carrier (40). In most cells, C3FT produced extensions, of which only a few had branches (Fig. 1, A and B). When forskolin and C3FT were applied together, nearly all of the resulting extensions had branches (Fig. 1, A and B), suggesting that forskolin produced branches independent of Rho.

To determine whether forskolin and C3FT affected the Rho signaling pathway, we studied the phosphorylation of MLC, which depends on Rho/Rho kinase activity (4, 45). Forskolin, as well as C3FT, reduced the amount of phosphorylated MLC, whereas the total amount of MLC was not changed by either treatment (Fig. 1C). Thus, C3FT and forskolin reduced the activity of Rho kinase.

The GTPases Rac and Cdc42 Are Necessary for the Forskolin-induced Branching in NG108-15 Cells—Next, we studied the role of other Rho GTPases in the formation of branched extensions. NG108-15 cells were treated with toxin B of C. difficile, which inactivates Rho, Rac and Cdc42 by glucosylation at Thr35/37 (41). In some cells, toxin B induced extensions, of which only a few had branches. In these cells, forskolin no longer induced branch formation. Apparently, the GTPases Rac and/or Cdc42 were necessary for this effect (Fig. 1, A and B).

To examine further the role of Rho, Rac, and Cdc42 in forskolin-induced branching, NG108-15 cells were transiently transfected with EGFP fusion proteins of dominant-negative (dn) Rho GTPases, which sequester the corresponding GDP exchange factors so that the respective endogenous GTPases remain inactive (46, 47). In the transfected cells, we also studied the effects of Rho kinase inhibitors and of C3FT. To evaluate whether the number of branches increased with the length of the extensions, we measured the longest extension and counted its branches (Fig. 2B). In cells expressing only EGFP, the Rho kinase inhibitors Y-27632 and H89 (38, 48) elongated the longest extension by ~60% without enhancing the number of branches. To confirm that Y-27632 reduced the activity of Rho kinase, we measured its effect on phosphorylation of MLC. Y-27632 indeed reduced the amount of MLC-P (Fig. 1C).



View larger version (38K):
[in this window]
[in a new window]
  		FIG. 2.
Dominant-negative Rho GTPases modulate the effects of forskolin on the morphology of NG108-15 cells. Cells were transfected with plasmids coding for EGFP alone or with EGFP/dnRhoAN19, EGFP/dnRac1N17, or EGFP/dnCdc42N17. Morphological changes were quantified 4 h after the addition of 10 µM Y-27632, 10 µM H89, or 10 µM forskolin by using EGFP fluorescence microscopy. A, micrographs of cells transfected with dn RhoA, dnCdc42, or dnRac1 in the absence or presence of forskolin; bar ~ 50 µm. B, quantification of the experiment. The numbers of cells with extensions >=50 µm are shown as a percentage of the total number of cells. The longest extension was measured by using the MRC-1024 software of the confocal microscope. The branches of the longest extension were counted. Mean ± S.E.; a refers to a significant difference (p < 0.05) from the EGFP control. b indicates a significant difference (p < 0.05) from dnRhoA, dnRac1, or dnCdc42, respectively. C, linear regression analysis of the length of the longest extension (x) and the number of its branches (y): y = 0.0046x + 0.263. {diamond}, cells transfected with EGFP; •, ~dnRhoA; {blacksquare}, ~dnRac1; +, ~dnCdc42. The 95% confidence intervals of the estimate 0.0046 were 0.0036 and 0.0057; for the estimate 0.263, they were 0.163 and 0.364. Values outside these limits were significantly different (p < 0.05). Numbers correspond to the rows of B.

 
Inactivation of Rho GTPase by C3FT also elongated the longest extension by 50% without inducing additional branches. The effect of forskolin on each parameter was different. Forskolin alone had no significant effect on elongation of the longest extension, but instead doubled the number of its branches (Fig. 2B). When forskolin was used in the presence of Y-27632 or C3FT, however, long and branched extensions were observed. In contrast, forskolin did not produce branches, when applied together with H89 (Fig. 2B). Because H89 inhibits Rho kinase as well as PKA (38), this finding suggested that forskolin induced branches by enhancing PKA activity.

Cells transfected with dnRhoAN19 had numerous extensions with few branches. Similar to cells treated with C3FT, they responded to forskolin with the production of branches (Fig. 2, A and B), confirming that forskolin induced branching in the absence of endogenous Rho activity. Compared with EGFP controls, cells transfected with dnRac1N17 had a similar number of extensions that, however, were short and had fewer branches. Neither the Rho kinase inhibitors Y-27632 and H89 nor forskolin increased the length or branching of such extensions (Fig. 2, A and B). Compared with EGFP controls, cells transfected with dnCdc42N17 had fewer and shorter extensions with only a small number of branches. Y-27632 and H89 elongated the longest extension and increased its branching, whereas forskolin had no effect on either parameter (Fig. 2, A and B).

To evaluate the relationship between extension length and number of branches, we performed a linear regression analysis (Fig. 2C). In cells not treated with forskolin, the number of branches (y) increased with the length of the extension (x) according to Equation 1.

(Eq. 1)
Forskolin significantly increased the number of branches but only in cells with active Rac and Cdc42 (Fig. 2C).

The GTPases Rac and Cdc42 Are Not Sufficient for the Forskolin-induced Branching in NG108-15 Cells—Because Rac and Cdc42 were essential for forskolin-induced branching, we studied how forskolin affected the activity of each GTPase by using a GST-PAK-CRIB pull-down assay (49). Forskolin increased the binding of Cdc42 to GST-PAK-CRIB in a time-dependent manner, indicating that it enhanced the activity of the GTPase (Fig. 3, A and B). After 15 min, the binding of Cdc42 to GST-PAK-CRIB was increased nearly 3-fold compared with controls. In contrast, forskolin did not alter the binding of Rac1 to GST-PAK-CRIB (Fig. 3A).



View larger version (44K):
[in this window]
[in a new window]
  		FIG. 3.
A, in NG108-15 cells, forskolin enhances the GTP binding of Cdc42 but not of Rac. GTP-Rac1 and GTP-Cdc42, as determined by PAK-CRIB pull-down assay, are shown in the upper lanes. Respective inputs are shown in lower lanes; the figure is a representative blot of three independent experiments. FSK, forskolin. B, quantification of three PAK-CRIB pull-down experiments for Cdc42. The results of the separate experiments were normalized by setting the controls at 100%. Mean ± S.E.; a shows a significant difference (p < 0.05) from lysates of cells treated for 0 min. C, effects of forskolin in NG108-15 cells transfected with constitutively active EGFP/Rac1V12 (caRac1; upper row) or EGFP/Cdc42V12 (caCdc42; lower row). 10 µM Y-27632 and/or 10 µM forskolin was added 16 h after transfection for 4 h; bar ~ 50 µm. EGFP fluorescence is shown.

 
To test whether active Cdc42 or Rac1 induced branches, NG108-15 cells were transfected with expression plasmids coding for fusion proteins consisting of EGFP and the constitutively active (ca) GTPases. Cells transfected with caCdc42V12 produced filopodia-like extensions (Fig. 3C). When used alone, forskolin had no effect, whereas Y-27632 induced extensions with numerous branches in a number of cells. Apparently, caCdc42V12 was sufficient to produce branches, if Rho kinase was inhibited. Combined application of forskolin and Y-27632 resulted in cells displaying a dense web of branches, suggesting that forskolin had an additional effect on branch formation (Fig. 3C). In NG108-15 cells transfected with caRac1V12, Y-27632 produced extensions with some branches, whereas forskolin alone formed thick extensions with some branches. Combined application of Y-27632 and forskolin produced thin extensions with numerous branches (Fig. 3C). Taken together, these results showed that Rac itself produced no branches, whereas Cdc42 was able to do so. Forskolin strongly induced branching, but this effect seemed to be additive to that of Cdc42.

Inhibition of PI-3K Induces Branching in NG108-15 Cells—Next, we addressed the question of how forskolin caused branching. Because cAMP can regulate the activity of PI3Ks (33-36), we studied the role of the lipid kinases in branch formation. As a first approach, the PI3K inhibitors LY294002 and wortmannin were used (50, 51). 20 µM LY294002 alone did not affect the outgrowth of extensions or branches (Fig. 4A). In NG108-15 cells treated with Y-27632, however, LY294002 strongly enhanced the number of branches. This increase was more pronounced than that caused by forskolin alone but similar to that produced by LY294002 plus forskolin. The PI3K inhibitor wortmannin (100 nM) had similar effects (data not shown).



View larger version (26K):
[in this window]
[in a new window]
  		FIG. 4.
The PI3K inhibitor LY294002 enhances the number of branches in Y-27632-treated NG108-15 cells but does not affect the GTP binding of Rac1 or Cdc42. A, 20 µM LY294002, 10 µM forskolin (FSK), or 10 µM Y-27632 alone or in combination were added for 4 h. Upper row, numbers show cells with extensions >=50 µm as a percentage of the total cell number. The diagram shows extensions with branches as a percentage of the total number of extensions. Mean ± S.E., n >= 150; a indicates a significant difference (p < 0.05) from controls; b refers to a significant difference (p < 0.05) from forskolin or Y-27632 treatment alone. B, GTP binding of Rac1 and Cdc42 was measured by PAK-CRIB pull-down assay. NG108-15 cells were incubated with 10 µM Y-27632, 100 nM wortmannin, and 20 µM LY294002, alone or in combination, for 4 h. For further explanation, see Fig. 3A. C, GTP binding of Rac1 and Cdc42 was measured by PAK-CRIB pull-down assay. NG108-15 cells were incubated with 10 µM forskolin (FSK) for 15 min and 20 µM LY294002, alone or in combination, for 30 min. D, Western blot showing the effect on MLC phosphorylation of treatment with 10 µM Y-27632 and 20 µM LY294002, alone or in combination, for 30 min. The upper row shows phosphorylated MLC (MLC-P); lower lane, total amount of MLC. E, in NG108-15 cells transfected with p110{gamma} CAAX, forskolin no longer induces branched extensions. NG108-15 cells were transfected with a plasmid encoding EGFP/p110{gamma} CAAX. 10 µM forskolin was added for 4 h; bar ~ 50 µm. Upper panels show EGFP fluorescence of transfected cells. These cells do not have branches. Lower panels show phase-contrast micrographs. Arrows indicate transfected cells. Untransfected cells have branches.

 
To confirm that the PI3K inhibitors did not act via Rac and Cdc42, we examined their effects on the activities of the GTPases. Neither LY294002 nor wortmannin significantly changed Rac or Cdc42 activity (Fig. 4B). Because LY294002 produced branches when used together with Y-27632, we also studied the effects of the Rho kinase inhibitor on Rac and Cdc42 activity. Y-27632 strongly increased the activity of Rac but not that of Cdc42 (Fig. 4B). When applied together with Y-27632, neither LY294002 nor wortmannin affected the Rac activation caused by the Rho kinase inhibitor (Fig. 4B). In view of the finding that LY294002 further enhanced branching when combined with FSK (see Fig. 4A), we also investigated whether the PI3K inhibitor activated Rac or Cdc42 under these conditions. However, this was not the case (Fig. 4C). Thus, we obtained no evidence that the PI3K inhibitors changed the activities of Rac or Cdc42.

To test further the hypothesis that PI3Ks were involved in forskolin-induced branching, we overexpressed the catalytic subunit p110{gamma} of PI3K{gamma} in NG108-15 cells. In contrast to the catalytic subunits of class Ia PI3Ks, the catalytic subunit p110{gamma} of PI3K{gamma} can be stably expressed in the absence of its adaptor subunit. This recombinant protein contained a CAAX box at its C terminus to facilitate membrane attachment. In cells transfected with this membrane-bound, constitutively active p110{gamma}-CAAX protein, forskolin no longer produced branched extensions (Fig. 4E), supporting an inhibitory effect of PI3Ks on branching.

Because forskolin and the PI3K inhibitors induced branching, we studied whether forskolin inhibited the activity of PI3Ks in NG108-15 cells. For this purpose, we analyzed the effect of forskolin on the phosphorylation of Akt at Ser473 by pyruvate dehydrogenase kinase-2 (PDK2). Forskolin indeed diminished the phosphorylation of Akt in a time-dependent manner (Fig. 5, A and B). After 15 min, a significant reduction of ~60% was observed with forskolin, whereas LY294002 abolished the phosphorylation of Ser473 (Fig. 5, A and B). Because H89 prevented the forskolin-induced branching (see Fig. 2B), its effect on Akt phosphorylation was also examined. It indeed blocked the forskolin-induced decrease in Akt phosphorylation (Fig. 5A).



View larger version (19K):
[in this window]
[in a new window]
  		FIG. 5.
Forskolin reduces the phosphorylation of Akt-Ser473. A, 20 µM LY294002 was added to the incubation medium for 15 min. 10 µM H89 was added 30 min prior to 10 µM forskolin, and incubation continued for 15 min. Phosphorylation of Akt was detected using an anti-phospho-Ser473-specific antibody. Total Akt was detected using an Akt-specific antibody. B, forskolin reduces the phosphorylation of Akt-Ser473 in a time-dependent manner. NG108-15 cells were treated with 10 µM forskolin for the times indicated. Phosphorylation of Akt in controls was taken as 100%. Shown is quantitation of the results from four experiments. a indicates a significant difference from control of p < 0.05.

 
Inhibition of PI3Ks Changes the Kinetics of Extension and Branch Formation—In neurons, dendritic branches result from a highly dynamic process of alternating outgrowth and retraction (10). To study how inhibition of PI3Ks affected the dynamics of branching in NG108-15 cells, the addition and retraction of extensions and branches were analyzed with time lapse microscopy. Because pronounced changes in arborization occurred during the first 2 h, cells were analyzed during this period. Within 2 h, control cells produced on average 1.9 extensions and 6.3 branches and retracted 1.0 extension and 4.1 branches (Fig. 6). Cells treated with Y-27632 added slightly more extensions and branches than controls but retracted the same number of extensions and branches. To estimate the dynamics of the process, we calculated the ratio of retracted to added extensions. Compared with controls, Y-27632 did not change the ratios of retracted to added extensions or branches. In contrast, the PI3K inhibitor wortmannin nearly abolished the addition and retraction of extensions and branches (Fig. 6). When combined with wortmannin, Y-27632 prevented the reduction in the addition of extensions and branches and reduced the decrease in retraction (Fig. 6). The ratios of retracted to newly added extensions or branches were significantly smaller than those of controls or of cells treated with Y-27632 alone (Fig. 6). Compared with controls or cells treated only with Y-27632, significantly more branches persisted after 2 h in cells treated with a combination of wortmannin and Y-27632 (p < 0.05). Similar results were obtained when LY294002 was combined with Y-27632 (data not shown). To investigate further the antagonistic interaction of PI3K inhibitors and Y-27632, we measured the effect of LY294002 on phosphorylation of MLC. Used alone, LY294002 enhanced MLC phosphorylation (Fig. 4D). This effect was blocked by Y-27632, confirming the opposite effects of the inhibitors.



View larger version (18K):
[in this window]
[in a new window]
  		FIG. 6.
Time lapse analysis of the effect of wortmannin on the initiation and retraction of extensions and branches in NG108-15 cells. The kinetics of the formation of extensions and branches are shown in the upper and lower panel, respectively. Micrographs of the same cells were made at 0, 30, 60, 90, and 120 min. By comparing the micrographs of consecutive time points newly added primary extensions and branches were counted, as were primary extensions and branches, which were no longer present. Black columns, sums of the number of extensions or branches newly detected at consecutive time points; white columns, sums of the number of extensions or branches no longer detected at consecutive time points. The ratios of retracted to newly added dendrites or branches are shown at the top of each panel. Mean ± S.E., n = 15; a indicates a significant difference (p < 0.05) from controls.

 
Also, forskolin reduced the ratios of retracted to newly added extensions or branches without changing the number of newly added extensions or branches (Fig. 6). Its effect on retraction was similar to that of the combination of Y-27632 and wortmannin.

Forskolin and Inhibitors of PI3Ks Induce Branched Dendrites in Cultured Hippocampal Neurons—In view of their effects in NG108-15 cells, we studied whether forskolin and PI3K inhibitors also induced the formation of extensions and branches in cultured hippocampal neurons prepared from brains of embryonic rats at day 17. After 1 day in culture, the neurons already had several extensions. Developing dendrites contain F-actin up to their tips (52, 53). Therefore, we stained the cellular F-actin with phalloidin to identify dendrites and branches. Most dendrites also showed {beta}-tubulin III immunoreactivity in their proximal regions (Fig. 7, A and C). In contrast, not more than 26% of the branches showed {beta}-tubulin III immunoreactivity (Fig. 7, A and C).



View larger version (34K):
[in this window]
[in a new window]
  		FIG. 7.
Forskolin, Y-27632, and wortmannin affect dendrite and branch formation in hippocampal neurons. A, F-actin and {beta}-tubulin III in dendrites and branches of control neuron; bar ~ 10 µm. B, 100 nM wortmannin induces branches in the presence of 10 µM Rho kinase inhibitor Y-27632, whereas 100 nM toxin B (ToxB) prevents the effect. Hippocampal neurons in StartV medium were treated with the drugs for 8 h. Immunostaining for {beta}-tubulin III is shown; bar ~ 10 µm. C, forskolin, Y-27632, and wortmannin on dendrite and branch formation. Dendrites and branches containing only F-actin (a) or also {beta}-tubulin (t) were counted. The longest extension that contained {beta}-tubulin III up to its tip was considered an axon. Axon was measured separately. Mean ± S.E., n >= 25; a following a number indicates a significant difference (p < 0.05) compared with controls.

 
First, we tested whether dendrites and branches positive for {beta}-tubulin, called t-dendrites and t-branches from now on, behaved like dendrites and branches which contained only F-actin, i.e. a-dendrites and a-branches (Fig. 7C). Control neurons had 3.2 t-dendrites and 3.7 a-dendrites/cell, as well as 2.3 t-branches and 12.1 a-branches. Within 8 h, 10 µM Y-27632 alone slightly enhanced the number of a-branches but had no effect on t-branches or dendrites (Fig. 7C). 100 nM wortmannin did not change the number of dendrites or branches, irrespective of the {beta}-tubulin III content (Fig. 7C). However, it significantly reduced the length of the axon, confirming the role of PI3Ks in axon development. Compared with controls, the combination of Y-27632 and wortmannin enhanced the numbers of a- and t-dendrites and of a- and t-branches (Fig. 7, B and C). In the presence of Y-27632, wortmannin no longer reduced the axon length (Fig. 7C). LY294002 acted like wortmannin on dendrite and branch formation (data not shown).

Following the protocol used for the NG108-15 cells, the cultured hippocampal neurons were treated with 10 µM forskolin, but the adenylyl cyclase activator had no effect on dendrite or branch formation (Fig. 7C). Because neurons can have high phosphodiesterase 4 activity that could degrade cAMP (54), the phosphodiesterase 4 inhibitor rolipram was used next. Whereas 10 µM rolipram alone did not affect dendrite or branch formation (Fig. 7C), its combined application with 10 µM forskolin caused profound changes. Compared with controls, t- and a-dendrites as well as t- and a-branches increased in number (Fig. 7C). Next, we measured the cellular levels of cAMP to determine whether rolipram indeed enhanced the effect of forskolin. Compared with controls, forskolin increased the neuronal level of cAMP by 182% in the absence, but by 402% in the presence of rolipram. Rolipram alone increased cAMP levels by 45% (data not shown).

To determine whether Rho GTPases were involved in branch formation, we treated the neurons with toxin B. When the GTPases were inhibited by toxin B, the neurons still had extensions that were short and without branches (Fig. 7B) (see also Refs. 2 and 55). However, neither wortmannin nor Y-27632 was able to induce branches, when used alone or in combination (Fig. 7B), indicating the essential role of Rho GTPases.

Therefore, we studied whether the agents acted on Rac and Cdc42 in the hippocampal neurons. By using a PAK-CRIB pull-down assay, we found that their effects corresponded to those in NG108-15 cells. Y-27632 enhanced the activity of Rac1, whereas wortmannin and LY294002 did not. None of the agents changed the activity of Cdc42 (Fig. 8A). Forskolin increased Cdc42 activity only in the presence of rolipram, confirming that inhibition of phosphodiesterase 4 was necessary for forskolin to have a maximal effect in the neurons (Fig. 8B). As was observed for NG108-15 cells, the effect of forskolin was also transient in hippocampal neurons, with a maximum occurring 15 min after application. In contrast, forskolin had no effect on the activity of Rac1 (Fig. 8C).



View larger version (27K):
[in this window]
[in a new window]
  		FIG. 8.
GST-PAK-CRIB pull-down assays for Rac1 and Cdc42 in hippocampal neurons. A, Y-27632 increases the GTP binding of Rac1. 10 µM Y-27632, 100 nM wortmannin, and 20 µM LY294002 were added for 1 h. B and C, 10 µM forskolin (FSK) enhances the GTP binding of Cdc42, but only in the presence of 10 µM rolipram. For further explanation, see Fig. 3A.

 
To investigate how inhibition of PI3Ks and the combination of forskolin and rolipram induced dendrite and branch formation in hippocampal neurons, time lapse analysis based on phase-contrast microscopy was used (Fig. 9A). The number of branches detected by phase-contrast microscopy corresponded to that of t-branches, suggesting that this subpopulation was analyzed. In control neurons, an average of 3.3 dendrites and 2.6 branches was added/cell during the 8-h observation period, whereas 2.7 dendrites and 1.7 branches were retracted (Fig. 9B). Treatment of the neurons with Y-27632 did not change the addition or retraction of dendrites or branches. Consequently, the ratios of retracted to added extensions or branches did not differ from those of controls (Fig. 9B). Wortmannin, however, reduced the addition of dendrites and branches by ~50% and nearly abolished retraction (Fig. 9B). The ratios of retracted to added extensions or branches were significantly smaller than those of controls or of neurons treated with Y-27632 (Fig. 9B). When Y-27632 and wortmannin were applied in combination, Y-27632 prevented the inhibitory effect of wortmannin on dendrite and branch addition and reduced that on the retraction of branches (Fig. 9B). Again, the ratios of retracted to newly added extensions or branches were smaller than in controls (Fig. 9B). The combined application of forskolin and rolipram reduced the ratios of retracted to newly added extensions or branches (Fig. 9B).



View larger version (14K):
[in this window]
[in a new window]
  		FIG. 9.
In hippocampal neurons, forskolin and the PI3K inhibitor wortmannin reduce the retraction of newly formed dendrites and branches. A, the neurons were treated for 8 h with 100 nM wortmannin, 10 µM Y-27632, 10 µM forskolin, and 10 µM rolipram. At the times indicated, phase-contrast micrographs were made. Drawings prepared from micrographs show typical morphological changes; bar ~ 50 µm. B, the kinetics of the formation of dendrites and branches are shown in the upper and lower panel, respectively. Micrographs of the same cells were made at 0, 2, 4, 6, and 8 h. By comparing the micrographs of consecutive time points newly added dendrites and branches were counted, as were dendrites and branches, which were no longer present. Black columns, sums of dendrites or branches newly detected at the consecutive time points; white columns, sums of the dendrites or branches no longer detected at the consecutive time points. The ratios of retracted to newly added dendrites or branches are shown at the top of each panel. Mean ± S.E., n = 15; a indicates a significant difference (p < 0.05) compared with controls.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
REFERENCES
 
In neural NG108-15 cells and in cultured hippocampal neurons, Rho GTPases and PI3K were found to play an important role in the formation of dendrites and branches. Rac and Cdc42 were necessary for formation of dendrites and branches, whereas Rho exerted a negative effect via Rho kinase. Inhibition of PI3Ks reduced the addition of new dendrites and branches. This reduction was no longer present when Rho kinase was inactivated simultaneously. Inhibition of PI3Ks also reduced the retraction of dendrites and branches. Consequently, the simultaneous inhibition of PI3Ks and Rho kinase caused the rapid growth of the dendritic arbor. Forskolin increased dendrite and branch formation by activating Cdc42 and inhibiting Rho kinase and PI3Ks.

Dendrites and branches result from outgrowth and elongation and subsequent partial retraction. Changes in the rate of addition or retraction have been shown to cause the dendritic arbor to grow or regress (10). The Rho GTPases Rac and Cdc42 selectively enhance the rate of addition (11-13). Our results provide additional information on the roles of both GTPases in neurite and branch formation. In control NG108-15 cells, the number of branches was related to the length of the primary extensions with an approximate ratio of 1 branch/160 µm. Expression of dnRac or dnCdc42 did not change this ratio. However, the negative isoforms decreased the total number of branches by reducing the length of the primary extensions. NG108-15 cells expressing dnRacN17 had short extensions, which were no longer elongated in response to forskolin or Rho kinase inhibitors. Thus, the number of branches remained unchanged. After transfection with dnCdc42N17, NG108-15 cells had few and very short extensions. They were still able to form and elongate extensions, when treated with Rho kinase inhibitors, but they no longer responded to forskolin. Apparently, Rac and Cdc42 were essential for the formation of extensions and their elongation, although the contribution of Cdc42 seems to be stimulus-specific. A direct action of Cdc42 on branching became evident in experiments with the constitutively active isoforms of the GTPases. caCdc42V12 but not caRacV12 enhanced the number of branches in the presence of Y-27632, indicating that Cdc42 is indeed able to induce branches and that this effect is antagonized by Rho kinase. A similar effect of Cdc42 on branch formation has been shown in axons (56).

Rho can negatively affect dendrite and axon formation as well as elongation (11, 23, 57). In our experiments, Rho indeed inhibited dendrite and branch formation but in a cell type-dependent manner. In NG108-15 cells, inhibition of Rho kinase by Y-27632 and inactivation of Rho by C3FT or dnRhoAN19 enhanced the number of primary extensions/cell but did not change the number of branches. In hippocampus neurons, however, Y-27632 did not change the number of dendrites, confirming previous findings with neurons (11-13). Y-27632 enhanced the number of small branches containing only F-actin, but did not change the number of branches containing F-actin and {beta}-tubulin III.

Confirming the importance of PI3Ks in axon formation (16, 25-27), wortmannin reduced the axon length in our hippocampal neurons. In addition, time lapse analysis showed that inhibition of PI3Ks with wortmannin or LY294002 strongly reduced the addition of new extensions and branches in NG108-15 cells and in hippocampal neurons, whereas Y-27632 prevented the effects. Y-27632 and LY294002 also had opposite effects on MLC phosphorylation, which is enhanced by the action of Rho/Rho kinase. Whereas the PI3K inhibitor enhanced MLC phosphorylation above the control level, the Rho kinase inhibitor prevented this effect. One may speculate that active PI3Ks diminish Rho or Rho kinase activity and thus facilitate outgrowth and elongation of extensions. Accordingly, inhibition of PI3Ks increases the activity of Rho kinase which can be prevented by Y-27632 (Fig. 10). Taken together, our findings suggest that PI3Ks can act synergistically with Rac and Cdc42 by reducing the inhibitory effect of Rho/Rho kinase on extension/branch addition and elongation (Fig. 10).



View larger version (16K):
[in this window]
[in a new window]
  		FIG. 10.
Proposed model showing regulation of the formation of branched dendrites by Rho GTPases and PI3Ks (arrow indicates stimulation, {perp} shows inhibition). Active Rac and Cdc42 lead to dendrite/branch addition. PI3Ks facilitate dendrite/branch addition by inhibiting Rho or Rho kinase.? indicates that the mode of this inhibition remains unclear. Dendrites and branches can be added, if the Rho/Rho kinase pathway is inhibited. PI3Ks also facilitate dendrite and branch retraction. Inhibition of PI3Ks by wortmannin or LY294002 (wo/LY) reduce retraction of dendrites and branches. In NG108-15 cells, forskolin (FSK) activates Cdc42 and inhibits Rho and PI3Ks.

 
In PC12 cells, PI3Ks have been shown to inhibit Rho by activating Rac (58). However, in NG108-15 cells transfected with caRac1V12, the Rho kinase inhibitor Y-27632 still induced extensions, indicating that Rho/Rho kinase were still active. Moreover, in NG108-15 cells and in hippocampal neurons, Y-27632 enhanced the activity of Rac1, as measured by PAK-CRIB pull-down assay. This finding confirmed that Rho kinase can inhibit Rac1 (59, 60).

In NG108-15 cells and hippocampal neurons, wortmannin or LY294002 also reduced the retraction of primary extensions and branches. Thus, PI3Ks simultaneously mediated the addition and retraction of dendrites and branches. The possibility that the same isoform caused these functionally opposite effects cannot be excluded.

We can only speculate on the PI3K-dependent effectors involved in retraction. Of the pleckstrin homology domain proteins presently known, those of the cytohesin family may be of particular importance. These proteins are GDP exchange factors for members of the family of ADP-ribosylation factors (ARF) (23, 24). Cytohesin-2/ARNO is present in the embryonic hippocampus, as are ARF1/3 and ARF6. Moreover, overexpression of inactive ARNO or of inactive ARF6 leads to dendritic branches in cultured hippocampal neurons (61). Thus, it is possible that inactivation of PI3Ks reduces the activity of ARNO and subsequently that of ARF6.

Rac and Cdc42 can act upstream, as well as downstream, of PI3Ks (62-64). Neither wortmannin nor LY294002 reduced the activities of Rac or Cdc42, as measured by GST-PAK-CRIB pull-down assays in NG108-15 cells or hippocampal neurons. Moreover, LY249002 did not affect the activation of Rac by Y-27632 nor the activation of Cdc42 by forskolin, indicating that the PI3Ks did not regulate the activities of the GTPases. Because LY249002 further enhanced branching, when combined with forskolin or Y-27632, it seems unlikely that PI3Ks acted downstream of Rac and Cdc42. Apparently, PI3Ks acted independently of the GTPases.

In NG108-15 cells and in hippocampal neurons, the adenylyl cyclase activator forskolin strongly enhanced the formation of primary extensions and branches. We have observed previously that forskolin does not induce branched extensions in NG108-15 cells transfected with the selective PKA inhibitor PKI (38). In the present study, the PKA inhibitor H89 abolished the effect of forskolin on branching, confirming the mediator role of PKA.

As already shown for macrophages (65), forskolin enhanced the activity of Cdc42 in NG108-15 cells and in hippocampal neurons. Active Cdc42 can induce the formation of branches, as became evident in NG108-15 cells transfected with caCdc42V12 and treated with Y-27632. Because forskolin further enhanced the number of branches in such cells, it is unlikely that it produced branches just by activating Cdc42 and inhibiting Rho/Rho kinase.

The additional mechanism involved turned out to be inhibition of PI3Ks. Forskolin and the PI3K inhibitors had similar effects on dendrite and branch dynamics. They reduced the retraction of newly added extensions and branches in NG108-15 cells, as well as in hippocampal neurons. Expression of the isoprenylated and therefore membrane-bound, and functionally constitutively active p110{gamma} mutant of PI3K{gamma} (p110{gamma}-CAAX) prevented the branching induced by forskolin in NG108-15 cells. A direct link between forskolin and PI3Ks is supported by our finding that forskolin decreased the phosphorylation of the PI3K effector Akt by acting through PKA.

Activation of PKA has been shown to inhibit RhoA (31, 32, 66, 67). In NG108-15 cells, forskolin diminished the Rho kinase-dependent phosphorylation of MLC, confirming that it reduced the activity of Rho and/or Rho kinase. Compared with NG108-15 cells treated with forskolin alone, cells had longer extensions, when forskolin and Y-27632 were applied together. The increase in Rac activity caused by Y-27632 but not by forskolin may have been responsible for this additional effect.

Taken together, our results show that the rise in cAMP levels caused by forskolin induced branched extensions via three synergistic mechanisms (Fig. 10). Forskolin reduced the activity of Rho/Rho kinase and their inhibitory effect on outgrowth. In addition, forskolin reduced the activity of PI3Ks. The resulting inhibitory effect on dendrite and branch outgrowth was overcome by the simultaneous reduction in Rho/Rho kinase activity. Finally, the forskolin-induced inhibition of PI3Ks reduced the retraction of newly formed extensions and thus enhanced the growth of the dendritic tree. Forskolin/cAMP exerted a third effect on Cdc42. The activation of the GTPase, which like Rac was necessary for branch formation, may have further supported the outgrowth of extensions.

cAMP has been shown to modulate axonal guidance in Xenopus (57, 68). According to our findings, PKA may also modulate dendritic arborization in the brain by regulating the PI3Ks. Alternatively, growth factors may regulate PI3Ks. Their localized release and action could influence branch formation in individual neurons or in distinct parts of their dendritic trees. In this context, the phosphoinositide 3-phosphatase PTEN, which dephosphorylates the products of PI3Ks, may also be of importance (69, 70).

In the present study, it was shown for the first time that Rho GTPases and PI3Ks control dendrite and branch formation. The effectors involved and their interactions will be studied in future investigations.


   FOOTNOTES
 
* This work was supported by the Deutsche Forschungsgemeinschaft and Fonds der Chemischen Industrie. 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. Back

To whom correspondence should be addressed: Institut für Experimentelle und Klinische Pharmakologie und Toxikologie, Albert-Ludwigs-Universität, Albert-Strasse 23 (Zentrum für Neurowissenschaften), D-79104 Freiburg, Germany. Tel.: 761-203-5327; E-mail: Dieter.Meyer{at}pharmakol.uni-freiburg.de.

1 The abbreviations used are: Cdc42, cell division cycle 42; a-, actin; ARF, ADP-ribosylation factor; C3FT, C3 fusion toxin; ca, constitutively active; CRIB, Cdc42/Rac interactive binding; dn, dominant negative; EGFP, enhanced green fluorescent protein; GST, glutathione S-transferase; MLC, myosin light chain; PAK, p21-activated kinase; PBS, phosphate-buffered saline; PI3K(s), phosphoinositide 3-kinase(s); PKA, protein kinase A; Rac1, Ras-related C3 botulinum toxin substrate 1; RhoA, Ras homologous member A; t-, tubuliin; TRITC, tetramethylrhodamine isothiocyanate; wt, wild type. Back


   ACKNOWLEDGMENTS
 
We thank B. Wilson for critical reading of the manuscript.



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Mueller, B. K. (1999) Annu. Rev. Neurosci. 22, 351-388[CrossRef][Medline] [Order article via Infotrieve]
  2. Hall, A. (1998) Science 279, 509-514[Abstract/Free Full Text]
  3. Luo, L. (2000) Nat. Rev. Neurosci. 1, 173-180[Medline] [Order article via Infotrieve]
  4. Kimura, K., Ito, M., Amano, M., Chihara, K., Fukata, Y., Nakafuku, M., Yamamori, B., Feng, J., Nakano, T., Okawa, K., Iwamatsu, A., and Kaibuchi, K. (1996) Science 273, 245-248[Abstract]
  5. Jalink, K., van Corven, E. J., Hengeveld, T., Morii, N., Narumiya, S., and Moolenaar, W. H. (1994) J. Cell Biol. 126, 801-810[Abstract/Free Full Text]
  6. Tigyi, G., Fischer, D. J., Sebok, A., Marshall, F., Dyer, D. L., and Miledi, R. (1996) J. Neurochem. 66, 549-558[Medline] [Order article via Infotrieve]
  7. Bito, H., Furuyashiki, T., Ishihara, H., Shibasaki, Y., Ohashi, K., Mizuno, K., Maekawa, M., Ishizaki, T., and Narumiya, S. (2000) Neuron 26, 431-441[CrossRef][Medline] [Order article via Infotrieve]
  8. Wahl, S., Barth, H., Ciossek, T., Aktories, K., and Mueller, B. K. (2000) J. Cell Biol. 149, 263-270[Abstract/Free Full Text]
  9. Neumann, H., Schweigreiter, R., Yamashita, T., Rosenkranz, K., Wekerle, H., and Barde, Y. A. (2002) J. Neurosci. 22, 854-862[Abstract/Free Full Text]
  10. Wu, G. Y., and Cline, H. T. (1998) Science 279, 222-226[Abstract/Free Full Text]
  11. Wong, W. T., Faulkner-Jones, B. E., Sanes, J. R., and Wong, R. O. (2000) J. Neurosci. 20, 5024-5036[Abstract/Free Full Text]
  12. Li, Z., Van Aelst, L., and Cline, H. T. (2000) Nat. Neurosci. 3, 217-225[CrossRef][Medline] [Order article via Infotrieve]
  13. Li, Z., Aizenman, C. D., and Cline, H. T. (2002) Neuron 33, 741-750[CrossRef][Medline] [Order article via Infotrieve]
  14. Kimura, K., Hattori, S., Kabuyama, Y., Shizawa, Y., Takayanagi, J., Nakamura, S., Toki, S., Matsuda, Y., Onodera, K., and Fukui, Y. (1994) J. Biol. Chem. 269, 18961-18967[Abstract/Free Full Text]
  15. Kobayashi, M., Nagata, S., Kita, Y., Nakatsu, N., Ihara, S., Kaibuchi, K., Kuroda, S., Ui, M., Iba, H., Konishi, H., Kikkawa, U., Saitoh, I., and Fukui, Y. (1997) J. Biol. Chem. 272, 16089-16092[Abstract/Free Full Text]
  16. Atwal, J. K., Massie, B., Miller, F. D., and Kaplan, D. R. (2000) Neuron 27, 265-277[CrossRef][Medline] [Order article via Infotrieve]
  17. Rameh, L. E., and Cantley, L. C. (1999) J. Biol. Chem. 274, 8347-8350[Free Full Text]
  18. Vanhaesebroeck, B., Leevers, S. J., Ahmadi, K., Timms, J., Katso, R., Driscoll, P. C., Woscholski, R., Parker, P. J., and Waterfield, M. D. (2001) Annu. Rev. Biochem. 70, 535-602[CrossRef][Medline] [Order article via Infotrieve]
  19. Maier, U., Babich, A., and Nurnberg, B. (1999) J. Biol. Chem. 274, 29311-29317[Abstract/Free Full Text]
  20. Leopoldt, D., Hanck, T., Exner, T., Maier, U., Wetzker, R., and Nurnberg, B. (1998) J. Biol. Chem. 273, 7024-7029[Abstract/Free Full Text]
  21. Brock, C., Schaefer, M., Reusch, H. P., Czupalla, C., Michalke, M., Spicher, K., Schultz, G., and Nurnberg, B. (2003) J. Cell Biol. 160, 89-99[Abstract/Free Full Text]
  22. Balla, T., Bondeva, T., and Varnai, P. (2000) Trends Pharmacol. Sci. 21, 238-241[CrossRef][Medline] [Order article via Infotrieve]
  23. Cantley, L. C. (2002) Science 296, 1655-1657[Abstract/Free Full Text]
  24. Vanhaesebroeck, B., and Waterfield, M. D. (1999) Exp. Cell Res. 253, 239-254[CrossRef][Medline] [Order article via Infotrieve]
  25. Markus, A., Zhong, J., and Snider, W. D. (2002) Neuron 35, 65-76[CrossRef][Medline] [Order article via Infotrieve]
  26. Yang, H., Shaw, G., and Raizada, M. K. (2002) Am. J. Physiol. 283, R107-R114
  27. Shi, S. H., Jan, L. Y., and Jan, Y. N. (2003) Cell 112, 63-75[CrossRef][Medline] [Order article via Infotrieve]
  28. Ridley, A. J. (2001) FEBS Lett. 498, 168-171[CrossRef][Medline] [Order article via Infotrieve]
  29. Gunning, P. W., Landreth, G. E., Bothwell, M. A., and Shooter, E. M. (1981) J. Cell Biol. 89, 240-245[Abstract/Free Full Text]
  30. Goldberg, J. L., Espinosa, J. S., Xu, Y., Davidson, N., Kovacs, G. T., and Barres, B. A. (2002) Neuron 33, 689-702[CrossRef][Medline] [Order article via Infotrieve]
  31. Lang, P., Gesbert, F., Delespine-Carmagnat, M., Stancou, R., Pouchelet, M., and Bertoglio, J. (1996) EMBO J. 15, 510-519[Medline] [Order article via Infotrieve]
  32. Dong, J. M., Leung, T., Manser, E., and Lim, L. (1998) J. Biol. Chem. 273, 22554-22562[Abstract/Free Full Text]
  33. Monfar, M., Lemon, K. P., Grammer, T. C., Cheatham, L., Chung, J., Vlahos, C. J., and Blenis, J. (1995) Mol. Cell. Biol. 15, 326-337[Abstract]
  34. Cass, L. A., Summers, S. A., Prendergast, G. V., Backer, J. M., Birnbaum, M. J., and Meinkoth, J. L. (1999) Mol. Cell. Biol. 19, 5882-5891