![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
NEUROPHARMACOLOGY
Zentrum für Neurowissenschaften, Institut für Experimentelle und Klinische Pharmakologie und Toxikologie (F.H., J.L., C.F., K.L., D.K.M.), Zentrum für Neurowissenschaften, Abteilung Innere Medizin II (H.H.B.), and Sektion Klinische Neuropharmakologie der Neurochirurgischen Universitätsklinik (T.J.F.), Albert-Ludwigs-Universität, Freiburg, Germany
Received April 13, 2006; accepted July 13, 2006.
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
; Tremont-Lukats et al., 2000; Urban et al., 2005). Lately, it has been applied in the treatment of limb tremor (Zesiewicz et al., 2005).
Gabapentin-lactam (8-aza-spiro[5,4]decan-9-on; GBP-L) is spontaneously formed from gabapentin by intramolecular lactamization and has a higher lipid solubility than gabapentin (Kearney et al., 1992). The pharmacodynamic effects of GBP-L differ from those of gabapentin. In contrast to gabapentin, GBP-L does not cause sedation or ataxia. Moreover, it induces myoclonic activity and generalized clonic seizures in kindled rats (Potschka et al., 2000). GBP-L but not gabapentin exerts neuroprotective effects. GBP-L enhances neuronal survival in pressure-induced retinal ischemia in rats as well as facilitates the survival of retinal ganglion cells cultured in medium devoid of growth factors (Jehle et al., 2001; Pielen et al., 2004). In addition, GBP-L but not gabapentin can reduce protein aggregates and improve the motor performance in a transgenic mouse model of Huntington's disease (Zucker et al., 2004). In view of these neuroprotective effects, we addressed the question of whether GBP-L also affected neuronal differentiation.
The formation of dendrites and branches is a highly dynamic process that consists of the addition and partial retraction or elimination of the new extensions (Van Aelst and Cline, 2004). The Rho family of small GTPases is of special importance in the regulation of dendrite addition (Luo, 2002; Van Aelst and Cline, 2004). These molecular switches are active in the GTP-bound form and are inactivated by GTP hydrolysis (Hall, 1998). Rac and Cdc42 induce the formation of lamellipodia and filopodia, respectively (Hall, 1998). They also support the addition of new dendrites and branches by acting on the actin cytoskeleton, whereas RhoA attenuates this process via its effector Rho kinase (ROCK) (Li et al., 2000; Wong et al., 2000; Leemhuis et al., 2004). The retraction of newly formed extensions is under the control of class I phosphatidylinositol 3 (PI3) kinases, as recently shown in hippocampal neurons (Leemhuis et al., 2004).
We have studied in cultured embryonic hippocampal neurons how GBP-L affected filopodia formation as well as the formation of new dendrites and branches. Since ROCK can inhibit the formation of new extensions, we also studied the effects of GBP-L in the presence of ROCK inhibitors. Time lapse analysis was used to study the dynamics of dendrite and branch formation. The activities of Rho GTPases and PI3-kinases were measured with affinity precipitation assays and Western blot analysis, respectively.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
). Cell Culture. Primary cultures of hippocampal neurons were prepared from brains of embryonic rats at day 17. The dissociated cells were seeded on glass coverslips coated with poly-L-lysine at a density of approximately 500 cells/mm2. The incubation medium consisted in Dulbecco's modified Eagle's medium plus 10% fetal calf serum with an endotoxin level < 0.5 EU/ml. Cultures were incubated at 37°C in a humidified atmosphere with low oxygen conditions (6.5% CO2 and 9% O2) to simulate in vivo conditions. After 1 day in vitro (DIV), the incubation medium was replaced with serum-free differentiation medium, which consisted of neurobasal A, pH 7.3 (HEPES) supplemented with 2% B27. If not mentioned otherwise, the neurons were used at DIV1 for morphological studies. For biochemical studies, the cultures were incubated until DIV3 and then treated with the respective drugs. The neuronal cultures contained less than 5% astrocytes as determined by glial fibrillary acidic protein immunostaining.
Cytochemical Staining. Cultured hippocampal neurons were fixed for 15 min with 4% paraformaldehyde in cytoskeletal stabilization buffer (CSB) (Schaefer et al., 2002) containing 80 mM Pipes, 5 mM EGTA, 1 mM MgCl2, and 4% polyethylene glycol (mol. wt. 35,000). After washing with CSB, neurons were permeabilized with 0.1% (v/v) Triton X-100 in CSB. Normal goat serum in PBS was used to block nonspecific binding. For staining of 
-tubulin III, cells were incubated with a monoclonal mouse anti--tubulin III antibody (Sigma, Deisenhofen, Germany). The resulting immune complex was visualized with an Alexa 488-conjugated F(ab')2 fragment goat anti-mouse (Molecular Probes, Heidelberg, Germany). For actin staining, cells were incubated with Alexa 594-conjugated phalloidin (Molecular Probes) and washed again with CSB.
Pull-Down Experiments with the GST-PAK-CRIB Domain or GST-C21 Domain. The fusion protein GST-C21 contains the N-terminal 90 amino acids of the Rho-binding region of the Rho effector Rhotekin. The fusion protein GST-PAK-CRIB contains the Rac/Cdc42 binding domain of the common effector PAK. The GST-fusion proteins used for the pull-down assays were expressed in Escherichia coli BL21 cells grown at 37°C. Expression in the BL21 cells was induced by adding 0.1 mM isopropyl -D-thiogalactoside (final concentration) at OD600 1.0. Two hours after induction, the cells were collected and sonicated 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 10,000g. Then, the GST-PAK-CRIB domain and the GST-C21 domain, respectively, were affinity purified from the supernatant with glutathione-Sepharose beads (Pharmacia, Piscataway, NJ). Beads loaded with the GST-fusion proteins were washed twice with GST-fishing 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] at 4°C.
Approximately 1 x 106 neurons were used for the pull-down experiments on Rac or Cdc42 activity; approximately 4 x 106 neurons were used to determine RhoA activity. After the actual experiments, neurons were washed twice with PBS and harvested after addition of 250 µl of ice-cold GST-fishing buffer. The detergent-soluble supernatant was recovered after centrifugation (14,000g and 4°C) for 15 min. GTP-Rac or GTP-Cdc42 proteins were precipitated for 1 h with GST-PAK fusion protein at 4°C. GTP-RhoA was precipitated for 1 h with GST-C21 fusion protein. The complexes were washed three times with ice-cold PBS, resuspended, and boiled in Laemmli buffer. GTP-bound Rac, Cdc42, or RhoA proteins were detected by Western blotting using anti-Rac1 (BD Bioscience, Heidelberg, Germany), anti-Cdc42 (Upstate, Milton Keynes, UK), or anti-RhoA antibodies (Santa Cruz Biotechnology, Santa Cruz, CA).
Preparation of Transfection Vectors and Cell Transfection. The coding regions of the RhoAN19 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 the Cdc42N17 gene was excised from the plasmid pcDNA3 with BamHI/EcoRI and inserted in-frame into the BglII/EcoRI sites of pEGFP-C1. All plasmids were confirmed by restriction digest analysis and sequencing. Cells were transfected for 1.5 h at 2.5% CO2 using the calcium phosphate/DNA coprecipitation procedure. After transfection, neurons were grown for 16 h prior to use.
Measurement of Akt Phosphorylation. After the actual experiment, neurons were washed twice with PBS and harvested after 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] had been added. The detergent-soluble supernatant was recovered after centrifugation for 15 min at 14,000g and 4°C. Cell lysates were separated by SDS-polyacrylamide gel electrophoresis and subjected to Western blot analysis. To determine the phosphorylation of Akt, anti-phosphoSer473 Akt and anti-Akt antibodies were used (Cell Signaling, Beverly, MA).
Time Lapse Imaging. Hippocampal neurons were plated and incubated on poly-L-lysine-coated glass-bottom microwell dishes (MatTek, Ashland, MA). For the imaging procedure, the dishes were placed in a chamber that provided a humidified atmosphere (6.5% CO2, 9% O2) at 37°C. It was staged on a Zeiss Axiovert 200M microscope (Carl Zeiss GmbH, Jena, Germany) equipped with a digital camera (Coolsnap HQ; Roper Scientific, Tucson, AZ). DIC images were acquired every 60 s with a Zeiss x40; A: 1.4 oil immersion lens for 8 h. The Metamorph 6.25 software (Universal Imaging, Downingtown, PA) was used to acquire and process the resulting stacks of images. To determine the number of added and retracted branches, processes > 5 µm were counted that persisted longer than 30 min. In this way, dendritic filopodia were not included in the analysis. Finally, all newly added and retracted branches were summed up. Shown are the average values of at least 10 randomly selected cells. After acquisition of the time lapse movies, in some experiments, neurons were fixed and stained for -tubulin III and F-actin to secure that dendrites had been analyzed. For the analysis of dendrite elongation, DIC photographs were used that had been taken at 20-min intervals during a period of 8 h. The position of the growth cone center was traced with the help of Metamorph Software (Visitron Systems, Puchheim, Germany). For each time point, the distance was calculated, which had been covered by the growth cone during the respective period.
Morphometry of Fixed Cells. After staining for -tubulin III and F-actin, cells were imaged with a Zeiss Axiovert 200 microscope equipped with a digital camera (Coolsnap HQ; Roper Scientific). For data processing, Metamorph 6.5.2 software (Universal Imaging) was used. All images were thresholded. After subtraction of the cell body, the remaining dendrites were analyzed. The Integrated Morphometry Analysis tool was used to measure the length of dendrites or branches. Only structures positive for both -tubulin III and F-actin were considered as dendrites or branches. Extensions were taken for filopodia if they contained only F-actin and were <10 µm. Dendrites were defined as structures that arose from the soma and were longer than 10 µm. Structures originating from dendrites were considered as branches. Only primary branches were counted. To quantify dendritic filopodia formation, we did not only count the filopodia per 10 µm but also measured the F-actin contained in filopodia. For this purpose, Metamorph was used, and the F-actin content of all filopodia was related to that of the total dendrite. The relative tubulin content of dendrites was measured by relating the area that showed a -tubulin III signal to the area positive for F-actin.
Statistics. Whenever applicable, analysis of variance was used followed by Scheffe's test. Student's t test was used when two groups had to be compared. The Kruskal-Wallis and Mann-Whitney U or Wilcoxon test were applied for processed values.
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
-tubulin III immunostaining for microtubules, neurites containing -tubulin III immunoreactivity up to their tips were considered as axon-like. Neurites that contained -tubulin III immunoreactivity predominantly in their proximal part and F-actin in their tips were counted as dendrites. A dendritic branch was defined as a -tubulin III-positive process originating from a dendrite. In contrast, dendritic filopodia contained F-actin but not -tubulin III immunoreactivity and were shorter than 10 µm.
GBP-L Enhances Filopodia Formation. When added to the neurons at DIV1 for 8 h, GBP-L (10 µM) increased the average number of dendritic filopodia from 0.55 ± 0.07/10 µm in controls to 0.92 ± 0.08/10 µm (n = 50; p < 0.0018; see also Fig. 1A). To further quantify filopodia formation, we calculated the dendritic F-actin contained in filopodia. In controls, 43% of the dendritic F-actin content was found in filopodia. GBP-L increased this value to 63% (Fig. 1B). Next, we determined whether the generated filopodia were able to make contact with neighboring neurons. To assist interactions between neurites of different neurons, we plated the neurons not at the usual density of approximately 500 cells/1 mm2 but at approximately 1250 cells/1 mm2. In such cultures, GBP-L increased the dendritic F-actin contained in filopodia from 37.6 ± 3.85% to 62 ± 4.76% (p < 0.05; n = 20), when applied for 8 h. In addition, GBP-L caused the formation of a neurite network (Fig. 1C). Contact points were formed by extensions that contained only F-actin and extensions that also contained -tubulin III (Fig. 1C). GBP-L increased the number of contacts formed by extensions which contained only F-actin by a factor of 6.25 but did not change the number of contacts formed by -tubulin III-containing extensions. Taken together, these data showed that GBP-L not only mediated the formation of filopodia but also of filopodial networks.
|
|
). When applied for 8 h, Y-27632 did not increase filopodia activity (Fig. 1, A and B) or the numbers of dendrites and branches (Fig. 1, A and D). Compared with controls, the combination of Y-27632 and GBP-L increased the average number of dendrites by 38% and that of branches even 12-fold (Fig. 1, A and D). In contrast, Y-27632 did not change the effect of GBP-L on filopodia formation (Fig. 1, A and B). When we measured the GTP binding of the Rho GTPases, Y-27632 applied alone had no effect on the level of GTP-bound Cdc42 but significantly raised the level of GTP-bound Rac1 (Fig. 2, A and B). In combination with GBP-L, Y-27632 did not further enhance the levels of GTP-bound Cdc42 or GTP-bound Rac1 (Fig. 2, A and B). In the presence of Y-27632, we tested several concentrations of GBP-L on dendrite and branch formation. Although 1 µM GBP-L was ineffective, 100 µM GBP-L was more effective by 46% than 10 µM (data not shown). However, 100 µM GBP-L caused cytotoxicity as indicated by pearl string-like neurites. Therefore, 10 µM GBP-L was used for the subsequent experiments.
HI-1152 is 10-fold more potent as a ROCK inhibitor than Y-27632 (Ikenoya et al., 2002). At a concentration of 1 µM, HI-1152 increased the number of dendrites and branches per neuron, when applied for 8 h together with GBP-L (Table 1A). In addition to ROCK, Y-27632 inhibits with similar potencies other serine/threonine kinases, such as MSK1, MAPKAP-K1b, and PRK2 (Davies et al., 2000). HI-1152 has not been tested for such additional effects. Ro 318220 inhibits MSK1 and MAPKAP-K1b but not ROCK, when used at nanomolar concentrations (Davies et al., 2000). In contrast to Y-27632, Ro 318220 (200 nM) did not facilitate the dendrite or branch formation induced by GBP-L (Table 1B). In addition, at a concentration of 2 µM, Ro 318220 was without effect (data not shown).
|
As shown in Fig. 1, GBP-L increased filopodia formation in neurons cultured for 1 or 3 days. Next, we studied whether the effects of GBP-L plus Y-27632 on dendrite and branch formation were influenced by the cultivation time or the seeding density (Table 2A). In cultures seeded at low or high density, GBP-L + Y-27632 increased the number of dendrites per neuron by approximately 30%, when used at DIV1 (Table 2A). At DIV3, however, the neurons had already spontaneously formed a larger number of dendrites so that GBP-L plus Y-27632 had no additional effect (Table 2A). In addition, the number of spontaneously formed branches increased from DIV1 to 3. Again, GBP-L plus Y-27632 had a smaller effect at DIV3 than at DIV1 (Table 2A).
|
Next, we studied whether the dendrites and branches produced by the combination of GBP-L and Y-27632 persisted after removal of both drugs. In one group of cultured neurons, we added the drug combination for 1 day and then replaced it with vehicle during the following day. In the second group, both agents were present during both days. Compared with controls treated with vehicle for 2 days, both types of treatment with GBP-L plus Y-27632 enhanced the average number of dendrites and branches (Table 2B).
Since the neurotrophic effects of GBP-L can be prevented by 100 µM5-hydroxydecanoate (5-HD) (Pielen et al., 2004), we studied whether this agent also blocked the effects of GBP-L on dendrite formation (Table 1A). Used alone for 8 h, 5-HD (100 µM) had no effect on the average number of dendrites or branches. However, it prevented the dendrite and branch formation induced by GBP-L plus Y-27632 (Table 1A). Lower 5-HD concentrations of 1 and 10 µM proved to be ineffective (data not shown).
Role of Rho Proteins in the Induction of Dendrites and Branches by GBP-L. To find out whether Rac and Cdc42 were necessary for the GBP-L-induced dendrite formation, we transfected hippocampal neurons with EGFP fusion proteins of dominant-negative (dn) Rho GTPases (Fig. 3A). These recombinant isoforms sequester the respective GEFs so that the endogenous GTPases become inactive (Diekmann et al., 1991; Ridley and Hall, 1992). To increase the transfection rates, neurons were used at DIV4 for the calcium phosphate/DNA coprecipitation procedure. In our hippocampal neurons, the transfection rate was 3 to 8%. Positive cells were mostly found in groups of several neurons. The neurites formed a dense network so that the analysis of single dendrites by -tubulin III immunostaining was impossible. Therefore, we counted the EGFP-positive extensions, not all of which may have been dendrites.
|
EGFP control neurons showed an average number of 5.8 extensions (Fig. 3, A and B). The number increased to 7.5, when the neurons were incubated for 8 h with GBP-L plus Y-27632. In neurons transfected with dnRac1 or dnCdc42, however, GBP-L plus Y-27632 did not change the number of extensions (Fig. 3, A and B). Neurons transfected with dnRhoA had a significantly higher average number of extensions than EGFP controls, i.e., 8.1 (Fig. 3B). In such neurons, treatment with GBP-L alone for 8 h further increased the average number of extensions by 17%. According to these results, Rac1 and Cdc42 seemed to be necessary for the GBP-L-induced formation of extensions, whereas RhoA had an inhibitory effect.
To confirm the involvement of Rho proteins in dendrite formation, we inactivated the GTPases with clostridial toxins (Fig. 3, C and D). Toxin B of C. botulinum inactivates RhoA, Rac1, and Cdc42 (Aktories and Barbieri, 2005). When added to the neuronal cultures at DIV1 for 8 h at a concentration of 10 ng/ml, it did not affect the number of dendrites and branches but prevented their increase induced by GBP-L plus Y-27632 (Fig. 3D). In contrast, C3 toxin of C. botulinum selectively inactivates RhoA (Aktories and Barbieri, 2005). Its application for8hata concentration of 100 ng/ml slightly enhanced the formation of dendrites and branches (Fig. 3, C and D). In the presence of C3 toxin, GBP-L as strongly increased the number of dendrites and branches as in the presence of Y-27632 (Fig. 3, C and D). When the two components of C3 toxin (see Materials and Methods) were used alone, they had no effects of their own and did not facilitate the effects of GBP-L (data not shown).
Gabapentin. In contrast to GBP-L, the anticonvulsant gabapentin (10 µM) did not change the levels of GTP-bound Cdc42, Rac1, or RhoA (Fig. 2, A-C). Gabapentin did not increase the filopodial content of dendritic F-actin [controls, 42.9 ± 2.3% (n = 46) versus 49.1 ± 2.1 (n = 37); p = 0.085], when applied for 8 h. Neither alone nor together with Y-27632 (10 µM), gabapentin (10 µM) changed the average number of dendrites (controls, 3.3 ± 0.5; gabapentin, 3.3 ± 0.3; gabapentin plus Y-27632, 3.0 ± 0.2; n = 60) or branches (controls, 0.65 ± 0.15; gabapentin, 0.85 ± 0.13; gabapentin plus Y-27632, 0.3 ± 0.1). Also at concentrations of 100 or 1000 µM, gabapentin did not increase dendrite and branch formation (data not shown).
|
PI3-Kinases Are Involved in the GBP-L-Induced Dendrite and Branch Formation. Since PI3-kinases can mediate the elimination of dendrites and branches (Leemhuis et al., 2004), we next studied the effect of GBP-L on PI3-kinase activity. We measured the phosphorylation of Akt at Ser473, which depends on PI3-kinase activity. Control neurons showed a low amount of phosphorylated Akt protein (P-Akt). Treatment of the cultures for 30 min with the PI3-kinase inhibitor LY294002 (20 µM) reduced the level of P-Akt (Fig. 4B), confirming that it was due to active PI3-kinase. GBP-L increased the amount of P-Akt, whereas the ROCK inhibitor Y-27632 had no effect. However, simultaneous application of Y-27632 prevented the increase in P-Akt induced by GBP-L (Fig. 4B).
To study whether the increase in PI3-kinase activity was responsible for the elimination of the GBP-L-induced branches, we used a PI3-kinase inhibitor (Fig. 4A). Since the experiments lasted for 8 h, we did not use the light-sensitive wortmannin but LY294002. Compared with controls, treatment of the neurons with 20 µM LY294002 reduced the average number of added and eliminated branches by 50 and 66%, respectively. The rise in the ratio of added to eliminated branches from 1.23 in controls to 1.8 after treatment with LY294002 explained the resulting average net formation of 0.7 branches (Fig. 4A). Y-27632 prevented the LY294004-induced decrease in branch addition but not that in elimination. This divergence was reflected in a ratio of 3.1 and resulted in the average net formation of 2.7 branches. When LY294002 was combined with GBP-L, approximately five branches were added. This value was slightly lower than that observed with GBP-L alone. In this combination, LY294002 reduced the average number of eliminated branches to 2.1. The resulting ratio of 2.83 was again higher than that of controls or GBP-L alone so that more persistent branches were formed (Fig. 4A). The additional application of Y-27632 only slightly enhanced branch addition but did not change branch elimination. In independent experiments, we confirmed that LY294002 applied together with GBP-L for 8 h increased the number of dendrites and branches (data not shown). To find out whether the PI3-kinase inhibitor affected dendrite and branch formation by changing the activities of the Rho GTPases, we studied its effect on GTP binding of Rac1 and Cdc42. LY294002 prevented the increase in GTP-bound Rac1 caused by GBP-L but did not change the respective effect on Cdc42 (Fig. 4C). Our additional finding that LY294002 inhibited the increase in AKT phosphorylation caused by GBP-L was in agreement with the observed morphological effects (Fig. 4C). Taken together, these data suggested that PI3-kinases played a role in both branch addition and elimination.
GBP-L Does Not Affect Dendrite Elongation. The dendritic tree does not only grow by adding new dendrites and branches but also by elongating existing extensions. Next, we studied whether GBP-L changed the dynamics of dendrite elongation. For this purpose, we measured with time lapse microscopy the total outgrowth and the total retraction distance over an 8-h period. Subtraction of both values corresponded to the actual elongation (Fig. 5, A and B). Total average dendrite outgrowth amounted to 15 µm in control neurons. Since 13 µm was again retracted, only 2 µm persisted (Fig. 5, A and B). During treatment with Y-27632, average outgrowth was 22 µm, of which only 6 µm was again retracted, so that net elongation amounted to 16 µm. In contrast, GBP-L neither affected outgrowth nor retraction compared with controls. When Y-27632 and GBP-L were used together, the average outgrowth and retraction distances did not differ from those induced by Y-27632 alone (Fig. 5, A and B). According to these results, GBP-L had no effect on dendrite elongation.
|
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
|
; Luo, 2002; Leemhuis et al., 2004; Van Aelst and Cline, 2004). In contrast, filopodia are generated by active Cdc42 in the presence of active endogenous RhoA (Kozma et al., 1995; Yuan et al., 2003). GBP-L increased the GTP-binding and thus the activity of the three Rho proteins. We propose that GBP-L exerted its multiple effects on dendrite arborization via these actions (Fig. 6). Most likely, GBP-L induced filopodia by activating Cdc42. In densely seeded cultures, where contacts between cells were more likely, GBP-L caused the formation of a network. The network forming extensions contained F-actin but not -tubulin III, suggesting that the filopodia induced by GBP-L attached to neighboring neurons. Presently, we cannot exclude that GBP-L also affected the process of neurite attachment.
Although GBP-L activated Rac and Cdc42 and induced the formation of filopodia, it did not generate more dendrites and branches. Our data indicate that the simultaneous activation of RhoA blocked the GBP-L-induced dendritic changes. Thus, GBP-L was able to generate extensions in neurons transfected with dnRhoA. In neurons treated with C3 toxin from C. botulinum, which selectively inactivates RhoA (Aktories and Barbieri, 2005), GBP-L induced dendrites and branches. To study the involvement of the RhoA effector ROCK, we used Y-27632 and HI-1152. Both agents are widely used as ROCK inhibitors but also inactivate the serine/threonine kinases PRK2, MSK1, and MAPKAP-K1b (Davies et al., 2000). The MSK1 and MAPKAP-K1b inhibitor Ro 318220 did not facilitate the GBP-L-induced formation of dendrites and branches, suggesting that the kinases were not involved. Like ROCK, PRK2 is a RhoA effector that may organize stress fibers (Vincent and Settleman, 1997). Thus, we cannot exclude that Y-27632 and HI-1152 facilitated the GBP-L-induced dendrite and branch formation by inhibiting ROCK as well as PRK2. In contrast to Y-27632 or HI-1152, dnRhoA and C3 toxin slightly enhanced the number of neuronal extensions, suggesting that additional RhoA effectors may be involved. The essential roles of Rac and Cdc42 in neurite formation were confirmed in experiments with neurons transfected with dnRac or dnCdc42 as well as after treatment with toxin B, which inactivates all GTPases of the Rho family (Aktories and Barbieri, 2005). Under all these conditions, the combination of GBP-L plus Y-27632 no longer increased dendrite formation.
Time lapse microscopy shows the dynamics of branch formation. In retinal neurons, Rac and RhoA enhance and reduce, respectively, addition and elimination of new extensions (Wong et al., 2000). In rat retinal neurons as well as tectal neurons from Xenopus laevis, RhoA inhibits the outgrowth of new dendrites and branches induced by active Rac and/or Cdc42 but has no effect on retraction or elimination of newly added branches (Li et al., 2000; Wong et al., 2000; Leemhuis et al., 2004). In our cultured hippocampal neurons, RhoA was involved in branch addition. However, this action became only apparent when branch addition was diminished by PI3-kinase inhibitors. Under these conditions, Y-27632 prevented the decrease (see also Leemhuis et al., 2004). GBP-L modulated this complex regulation. It activated the three Rho GTPases as well as PI3-kinases, thereby enhancing branch addition as well as elimination (Fig. 6). We will first discuss the data concerning branch addition.
GBP-L enhanced the addition of branches, although it increased the activity of RhoA. Moreover, Y-27632 did not further increase the branch addition caused by GBP-L. GBP-L even increased branch addition in the presence of the PI3-kinase inhibitor LY-294002, although to a slightly lower extent, which was prevented by Y-27632. Taken together, these results showed that RhoA had only a negligible inhibitory effect on the branch addition caused by GBP-L.
In contrast, Rac and Cdc42 were essential for branch formation as shown by the effects of dnRac, dnCdc42, and toxin B (Fig. 6). It is difficult to evaluate which of the two GTPases was more important for branch addition. Y-27632 increased the level of GTP-bound Rac over that of controls but had no effect on GTP-bound Cdc42, confirming previous data (Yamaguchi et al., 2001; Tsuji et al., 2002; Leemhuis et al., 2004). However, Y-27632 did not enhance branch addition, suggesting that Rac alone was not sufficient for this action. In contrast, the selective activation of Cdc42 by forskolin can increase branch addition (Leemhuis et al., 2004).
Class I PI3-kinases contribute to branch addition and elimination (Leemhuis et al., 2004). Indeed, GBP-L increased PI3-kinase activity and induced the elimination of newly added branches. Y-27632 did not reduce the basal activity of PI3-kinases but inhibited the stimulatory effect of GBP-L. Apparently, RhoA acted synergistically with another mechanism induced by GBP-L (Fig. 6). We suggest that Y-27632 reduced branch elimination by preventing the GBP-L-induced increase in PI3-kinase activity. The direct inhibition of PI3-kinase activity with LY294002 always strongly reduced branch elimination. LY294002 also allowed GBP-L to form numerous new branches (Fig. 6). LY294002 did not only increase the formation of branches (Fig. 4) but also of dendrites (data not shown). We assume that similar effects were involved.
The GBP-L-induced increase in PI3-kinase activity is not only of interest in the context of dendrite arborization. Since class I PI3-kinases facilitate the survival of various cells (Rameh and Cantley, 1999; Vanhaesebroeck et al., 2001), the GBP-L-induced increase in kinase activity may explain its neuroprotective effects (Jehle et al., 2001; Pielen et al., 2004).
Although Y-27632 had no effect on dendrite and branch formation, it caused the elongation of dendrites. This finding confirmed previous observations (Wong et al., 2000; Sin et al., 2002) and pointed out that different mechanisms regulate the formation of new dendrites and their elongation. For elongation, active Rac and ROCK inhibition seem to be essential, whereas Cdc42 is mainly involved in the increased formation of dendrites and branches. The finding that GBP-L not only activated Rac and but also RhoA explains why the agent did not cause elongation. Y-27632 facilitated the extension of microtubules into the dendrites, suggesting that these structures were involved in elongation. Thus, ROCK and possibly PRK2 seem to restrain the outgrowth of microtubules during dendrite formation. Since RhoA can affect microtubule organization in polarized cells via its effector mDIA (Ishizaki et al., 2001; Palazzo et al., 2001), this finding indicates that RhoA may regulate microtubules via different effectors.
5-HD blocks the neuroprotective effects of GBP-L (Pielen et al., 2004). In our neurons, it prevented the effects of GBP-L on dendrite and branch formation. 5-HD affects the formation of reactive oxygen species as well as blocks mitochondrial KATP channels (Szewczyk et al., 1993; Dzeja et al., 2001; Das et al., 2003). We did not address the question of whether GBP-L and 5-HD acted on one of these mechanisms when they exerted their morphological effects. Instead, we focused our analysis on the pathways involved in neurite formation. Consequently, the question remained unanswered how GBP-L activates the Rho GTPases. The GBP-L-induced Rac1 activation was abolished by LY294002, suggesting that PI3-kinases activate Rac1 as has been shown previously (Reif et al., 1996; Sarner et al., 2000). Cdc42 and RhoA share GEFs such as Dbl and Dbs (Schmidt and Hall, 2002). The dominant-negative isoforms of the Rho proteins used in our study scavenge GEFs. However, our finding that dnRhoA produced numerous neurites indicates that Cdc42 was still active. We consider this as indirect evidence that GBP-L-regulated RhoA and Cdc42 independently. The GEFs involved have to be elucidated in future studies.
GBP-L has pronounced neuroprotective effects (Jehle et al., 2001; Pielen et al., 2004). Now, we report that it also has neurotrophic effects. In contrast, the antiepileptic agent gabapentin did not generate filopodia as well as dendrites or branches. Gabapentin also did not change the activities of Rho GTPases or PI3-kinases. Only GBP-L induced dendritic filopodia. Since filopodia are an important early step in synapse formation (Ziv and Smith, 1996), GBP-L may facilitate synaptic interactions. GBP-L also caused the formation of neurite networks. In addition, GBP-L enhanced the outgrowth of new dendrites and branches but also their retraction. This increased motility was turned into net formation of dendrites and branches when GBP-L was combined with inhibitors of RhoA or PI3-kinases. The observed effects of GBP-L on dendritic arborization are of potential therapeutic interest, which will be further investigated. In addition, the combined effects of GBP-L on the GTPases of the Rho family and PI3-kinases make it an interesting tool for further investigations on the interactions of these proteins.
|
Footnotes |
|---|
F.H. and J.L. contributed equally to this work.
Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.
ABBREVIATIONS: GBP-L, gabapentin-lactam; ROCK, Rho kinase; PI3, phosphatidylinositol 3; Y-27632, (+)-(R)-trans-4-(1-aminoethyl)-N-(4-pyridyl) cyclohexanecarboxamide; HI-1152, (S)-(+)-2-methyl-1-[(4-methyl-5-isoquinolinyl)sulfonyl]-homopiperazine; LY294002, 2-(4-morpholinyl)-8-phenyl-1(4H)-benzopyran-4-one hydrochloride; Ro 318220, 2-{1-[3-(amidinothio)propyl]-1H-indol-3-yl}-3-(1-methylindol-3-yl) maleimide methanesulfonate salt; DIV, day(s) in vitro; CSB, cytoskeletal stabilization buffer; Pipes, 1,4-piperazinediethanesulfonic acid; PBS, phosphate-buffered saline; GST, glutathione S-transferase; PAK, p21-activated kinase; DIC, differential interference contrast; 5-HD, 5-hydroxydecanoate; EGFP, enhanced green fluorescent protein; dn, dominant-negative; P-Akt, phosphorylated Akt protein; GEF, guanine nucleotide exchange factor.
Address correspondence to: Dr. Dieter K. Meyer, Zentrum für Neurowissenschaften, Institut für Experimentelle und Klinische Pharmakologie und Toxikologie, Albert-Ludwigs-Universität, Albert-Strasse 25, D-79104 Freiburg, Germany. E-mail: dieter.meyer{at}pharmakol.uni-freiburg.de
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Aktories K and Barbieri JT (2005) Bacterial cytotoxins: targeting eukaryotic switches. Nat Rev Microbiol 3: 397-410.[CrossRef][Medline]
Andrews CO and Fischer JH (1994) Gabapentin: a new agent for the management of epilepsy. Ann Pharmacother 28: 1188-1196.[Abstract]
Barth H, Hofmann F, Olenik C, Just I, and Aktories K (1998) The N-terminal part of the enzyme component (C2I) of the binary Clostridium botulinum C2 toxin interacts with the binding component C2II and functions as a carrier system for a Rho ADP-ribosylating C3-like fusion toxin. Infect Immun 66: 1364-1369.
Das M, Parker JE, and Halestrap AP (2003) Matrix volume measurements challenge the existence of diazoxide/glibencamide-sensitive KATP channels in rat mitochondria. J Physiol 547: 893-902.
Davies SP, Reddy H, Caivano M, and Cohen P (2000) Specificity and mechanism of action of some commonly used protein kinase inhibitors. Biochem J 351: 95-105.[CrossRef][Medline]
Diekmann D, Brill S, Garrett MD, Totty N, Hsuan J, Monfries C, Hall C, Lim L, and Hall A (1991) Bcr encodes a GTPase-activating protein for p21rac. Nature (Lond) 351: 400-402.[CrossRef][Medline]
Dzeja PP, Holmuhamedov EL, Ozcan C, Pucar D, Jahangir A, and Terzic A (2001) Mitochondria: gateway for cytoprotection. Circ Res 89: 744-746.
Hall A (1998) Rho GTPases and the actin cytoskeleton. Science (Wash DC) 279: 509-514.
Ikenoya M, Hidaka H, Hosoya T, Suzuki M, Yamamoto N, and Sasaki Y (2002) Inhibition of rho-kinase-induced myristoylated alanine-rich C kinase substrate (MARCKS) phosphorylation in human neuronal cells by H-1152, a novel and specific Rho-kinase inhibitor. J Neurochem 81: 9-16.[CrossRef][Medline]
Ishizaki T, Morishima Y, Okamoto M, Furuyashiki T, Kato T, and Narumiya S (2001) Coordination of microtubules and the actin cytoskeleton by the Rho effector mDia1. Nat Cell Biol 3: 8-14.[CrossRef][Medline]
Jehle T, Feuerstein TJ, and Lagreze WA (2001) The effect of gabapentin and gabapentin-lactam on retinal ganglion cell survival. Situation after acute retinal ischemia in animal models. Ophthalmologe 98: 237-241.[CrossRef][Medline]
Kearney AS, Mehta SC, and Radebaugh GW (1992) The effect of cyclodextrins on the rate of intramolecular lactamization of gabapentin in aqueous solution. Int J Pharm 78: 25-34.[CrossRef]
Kozma R, Ahmed S, Best A, and Lim L (1995) The Ras-related protein Cdc42Hs and bradykinin promote formation of peripheral actin microspikes and filopodia in Swiss 3T3 fibroblasts. Mol Cell Biol 15: 1942-1952.[Abstract]
Leemhuis J, Boutillier S, Barth H, Feuerstein TJ, Brock C, Nurnberg B, Aktories K, and Meyer DK (2004) Rho GTPases and Phosphoinositide 3-Kinase Organize Formation of Branched Dendrites. J Biol Chem 279: 585-596.
Leemhuis J, Boutillier S, Schmidt G, and Meyer DK (2002) The protein kinase A inhibitor H89 acts on cell morphology by inhibiting Rho kinase. J Pharmacol Exp Ther 300: 1000-1007.
Li Z, Van Aelst L, and Cline HT (2000) Rho GTPases regulate distinct aspects of dendritic arbor growth in Xenopus central neurons in vivo. Nat Neurosci 3: 217-225.[CrossRef][Medline]
Luo L (2002) Actin cytoskeleton regulation in neuronal morphogenesis and structural plasticity. Annu Rev Cell Dev Biol 18: 601-635.[CrossRef][Medline]
Palazzo AF, Cook TA, Alberts AS, and Gundersen GG (2001) mDia mediates Rho-regulated formation and orientation of stable microtubules. Nat Cell Biol 3: 723-729.[CrossRef][Medline]
Pielen A, Kirsch M, Hofmann HD, Feuerstein TJ, and Lagreze WA (2004) Retinal ganglion cell survival is enhanced by gabapentin-lactam in vitro: evidence for involvement of mitochondrial KATP channels. Graefes Arch Clin Exp Ophthalmol 242: 240-244.[CrossRef][Medline]
Potschka H, Feuerstein TJ, and Loscher W (2000) Gabapentin-lactam, a close analogue of the anticonvulsant gabapentin, exerts convulsant activity in amygdala kindled rats. Naunyn-Schmiedeberg's Arch Pharmacol 361: 200-205.[CrossRef][Medline]
Rameh LE and Cantley LC (1999) The role of phosphoinositide 3-kinase lipid products in cell function. J Biol Chem 274: 8347-8350.
Reif K, Nobes CD, Thomas G, Hall A, and Cantrell DA (1996) Phosphatidylinositol 3-kinase signals activate a selective subset of Rac/Rho-dependent effector pathways. Curr Biol 6: 1445-1455.[CrossRef][Medline]
Ridley AJ and Hall A (1992) Distinct patterns of actin organization regulated by the small GTP-binding proteins Rac and Rho. Cold Spring Harb Symp Quant Biol 57: 661-671.
Sarner S, Kozma R, Ahmed S, and Lim L (2000) Phosphatidylinositol 3-kinase, Cdc42, and Rac1 act downstream of Ras in integrin-dependent neurite outgrowth in N1E-115 neuroblastoma cells. Mol Cell Biol 20: 158-172.
Schaefer AW, Kabir N, and Forscher P (2002) Filopodia and actin arcs guide the assembly and transport of two populations of microtubules with unique dynamic parameters in neuronal growth cones. J Cell Biol 158: 139-152.
Schmidt A and Hall A (2002) Guanine nucleotide exchange factors for Rho GTPases: turning on the switch. Genes Dev 16: 1587-1609.
Sin WC, Haas K, Ruthazer ES, and Cline HT (2002) Dendrite growth increased by visual activity requires NMDA receptor and Rho GTPases. Nature (Lond) 419: 475-480.[CrossRef][Medline]
Szewczyk A, Mikolajek B, Pikula S, and Nalecz MJ (1993) Potassium channel openers induce mitochondrial matrix volume changes via activation of ATP-sensitive K+ channel. Pol J Pharmacol 45: 437-443.[Medline]
Tremont-Lukats IW, Megeff C, and Backonja MM (2000) Anticonvulsants for neuropathic pain syndromes: mechanisms of action and place in therapy. Drugs 60: 1029-1052.[CrossRef][Medline]
Tsuji T, Ishizaki T, Okamoto M, Higashida C, Kimura K, Furuyashiki T, Arakawa Y, Birge RB, Nakamoto T, Hirai H, et al. (2002) ROCK and mDia1 antagonize in Rho-dependent Rac activation in Swiss 3T3 fibroblasts. J Cell Biol 157: 819-830.
Urban MO, Ren K, Park KT, Campbell B, Anker N, Stearns B, Aiyar J, Belley M, Cohen C, and Bristow L (2005) Comparison of the antinociceptive profiles of gabapentin and 3-methylgabapentin in rat models of acute and persistent pain: implications for mechanism of action. J Pharmacol Exp Ther 313: 1209-1216.
Van Aelst L and Cline HT (2004) Rho GTPases and activity-dependent dendrite development. Curr Opin Neurobiol 14: 297-304.[CrossRef][Medline]
Vanhaesebroeck B, Leevers SJ, Ahmadi K, Timms J, Katso R, Driscoll PC, Woscholski R, Parker PJ, and Waterfield MD (2001) Synthesis and function of 3-phosphorylated inositol lipids. Annu Rev Biochem 70: 535-602.[CrossRef][Medline]
Vincent S and Settleman J (1997) The PRK2 kinase is a potential effector target of both Rho and Rac GTPases and regulates actin cytoskeletal organization. Mol Cell Biol 17: 2247-2256.[Abstract]
Wong WT, Faulkner-Jones BE, Sanes JR, and Wong RO (2000) Rapid dendritic remodeling in the developing retina: dependence on neurotransmission and reciprocal regulation by Rac and Rho. J Neurosci 20: 5024-5036.
Yamaguchi Y, Katoh H, Yasui H, Mori K, and Negishi M (2001) RhoA inhibits the nerve growth factor-induced Rac1 activation through Rho-associated kinase-dependent pathway. J Biol Chem 276: 18977-18983.
Yuan XB, Jin M, Xu X, Song YQ, Wu CP, Poo MM, and Duan S (2003) Signalling and crosstalk of Rho GTPases in mediating axon guidance. Nat Cell Biol 5: 38-45.[CrossRef][Medline]
Zesiewicz TA, Elble R, Louis ED, Hauser RA, Sullivan KL, Dewey RB Jr, Ondo WG, Gronseth GS, and Weiner WJ (2005) Practice parameter: therapies for essential tremor: report of the Quality Standards Subcommittee of the American Academy of Neurology. Neurology 64: 2008-2020.
Ziv NE and Smith SJ (1996) Evidence for a role of dendritic filopodia in synapto-genesis and spine formation. Neuron 17: 91-102.[CrossRef][Medline]
Zucker B, Ludin DE, Gerds TA, Lucking CH, Landwehrmeyer GB, and Feuerstein TJ (2004) Gabapentin-lactam, but not gabapentin, reduces protein aggregates and improves motor performance in a transgenic mouse model of Huntington's disease. Naunyn-Schmiedeberg's Arch Pharmacol 370: 131-139.[CrossRef][Medline]
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
10 µm. Bottom, higher magnification of dendrites with filopodia and microtubules in tips. B, quantification of filopodial F-actin. All protrusions containing F-actin but not 
, significant difference (p < 0.05) from dnRhoA. C, at DIV1, neurons were treated with C3 fusion toxin (C3FT; 100 ng/ml) from C. botulinum or C3 toxin plus GBP-L. Cells were seeded at a density of approximately 500/mm2. In one panel, the same neuron is shown after phalloidin staining for F-actin and 
