Angiotensin II induction of neurite outgrowth by AT2 receptors in NG108-15 cells. Effect counteracted by the AT1 receptors.

In the present study, 3-day treatment of nondifferentiated NG108-15 cells with 100 nM angiotensin II (Ang II) induces morphological differentiation of neuronal cells characterized by the outgrowth of neurites. These morphological changes are correlated with an increase in the level of polymerized tubulin and in the level of the microtubule-associated protein, MAP2c. Mediation by the AT2 receptor may be inferred since: (a) these cells contain only AT2 receptors; (b) the effects are mimicked by CGP 42112 (an AT2 receptor agonist); (c) they are not suppressed by the addition of DUP 753 (an AT1 receptor antagonist); and (d) are abolished by co-incubation with PD 123319 (an AT2 receptor antagonist). Application of Ang II in dibutyryl cAMP-differentiated cells (which contain both types of receptors) induces neurite retraction, an effect mediated by the AT1 receptor. These results indicate that the AT2 receptor of Ang II induces neuronal differentiation, which is initiated through an increase in the levels of MAP2c associated with tubulin. Moreover, our results demonstrate that the AT1 receptor inhibit the process of differentiation induced by dibutyryl cAMP, whereas the AT2 receptors potentiate this effect, illustrating negative cross-talk interaction between the two types of Ang II receptors.

has been cloned (13,14), but a definitive physiological function has yet to be assigned. We have reported previously that Ang II decreases the T-type calcium current in nondifferentiated NG108-15 cells expressing only the angiotensin AT 2 receptor type (15,16). Kang et al. (17) also reported that activation of the AT 2 receptor increased a potassium channel activity. Moreover, Xiong and Marshall (18) reported that Ang II, via the AT 2 receptor, could inhibit membrane depolarization and action potential elicited by N-methyl-D-aspartate receptors in the locus coeruleus, a brain area containing only Ang II receptors of the AT 2 subtype. Considering the abundance of T-type Ca 2ϩ channels in neurons from fetal brain (19), the crucial role of Ca 2ϩ in neuronal differentiation (20) and the abundance of AT 2 receptors during this developmental period, it could be postulated that Ang II, via the AT 2 receptor could affect some aspects of neuronal differentiation.
Neuronal differentiation is characterized by neurite extension that involves several biochemical steps directed toward promotion of the assembly of tubulin monomers into microtubules necessary to support the growing neurites. Several molecules play crucial roles in neurite outgrowth. Among these are the microtubule-associated proteins (MAPs) (21), which include both high molecular weight proteins, termed MAP1 to MAP5, and low molecular weight proteins, including tau (22). These proteins promote tubulin polymerization as well as stabilize microtubules and occur as embryonic and adult isoforms whose differential expression during brain development correlates with the maturation of neuronal circuitry. For example, MAP2 and tau bind to distinct populations of microtubules in adult neurons: MAP2 to somatodendritic microtubules and tau to axonal microtubules (23,24). Expression of specific brain MAPs is critical for regulating neurite outgrowth and differentiation. Several lines of evidence have demonstrated a strong correlation between the pattern of expression of neuronal MAPs and the morphological differentiation of neurons (22).
In their nondifferentiated state, the hybrid cells NG108-15 (neuroblastoma ϫ glioma) are rounded and actively dividing. Chronic exposure of NG108-15 cells to dibutyryl cAMP (db-cAMP) induces a process of differentiation that includes elaboration of neurites, development of electrical excitability, formation of functional synapses, alteration of ligand-gated channel properties, and a reduced rate of cell division. Thus, differentiated NG108-15 cells exhibit a neuronal phenotype, whereas glial properties appear to be suppressed (25,26). Moreover, when nondifferentiated, NG108-15 cells express exclusively the AT 2 receptor subtype and mainly the T-type Ca 2ϩ channel (15). These cells also express tau and MAP2 and are thus useful in examining factors initiating neuronal differentiation (27).
We have taken advantage of these properties to investigate the role of Ang II on differentiation and neurite outgrowth, with emphasis on cytoskeletal proteins. We found that Ang II induces outgrowth of neurites in nondifferentiated cells. These morphological changes are correlated with an increase in polymerized tubulin and an increase in the level of microtubuleassociated MAP2c. In contrast, during the process of differentiation with dbcAMP, where both AT 1 and AT 2 receptors are present, application of Ang II induces neurites involution. These results are the first to assign a physiological role for the AT 2 receptor in neuronal differentiation, which is to induce neurite outgrowth, by acting on tubulin polymerization and on the level of microtubule-associated MAP2c. Moreover, our results demonstrate that activation of the AT 1 receptor subtype inhibits the process of differentiation induced by dbcAMP and inhibits the effect elicited by the AT 2 receptor.  (26). The medium was replaced every 2 days. Subcultures were performed at subconfluence. Under these conditions, cells express only the AT 2 receptor subtype of Ang II (Fig. 1A). For differentiation, the cells were cultured for 24 h in the normal culture medium and thereafter replaced by DMEM low glucose (1 mg/liter glucose) containing 1% fetal calf serum (Inotech), HAT supplement, 50 mg/liter gentamycin, and 1 mM dbcAMP (Boehringer Mannheim, Montreal, Canada). The cells were allowed to differentiate for 24 h prior to Ang II or Ang II/analogs treatment. Cells were cultured for 3 subsequent days under these conditions, with the differ-entiation medium being changed every day. For all experiments, cells were plated at the same initial density. Experiments and cytoskeleton extractions were performed on the 4th day.

Cell
Binding Studies-The analog [Sar 1 ,Ile 8 ]Ang II was iodinated by the Iodogen method and separated on a 25-cm C-18 -Bondapak column (Waters, Milford, MA) with a linear gradient of 20 -60% acetonitrile in a buffer of 7% isopropanol, 0.25 M ammonium acetate, pH 5.0, at a flow rate of 1 ml/min. Carrier-free monoiodinated product was obtained as a single homogenous peak at 20 min. The specific activity was approximately 1000 Ci/mmol. Binding assays were performed on cultured cells as described previously (28). NG108-15 cells (1.0 -1.5 ϫ 10 6 cells/Petri dish) were washed with 2 ml of Hanks' buffered saline (HBS: NaCl, 130 mM; KCl, 3.5 mM; CaCl 2 , 1.8 mM; MgCl 2 , 0.5 mM; NaHCO 3 , 2.5 mM; HEPES, 5 mM, supplemented with 1 g/liter glucose and 0.5% BSA) and incubated for 15 min at 22°C in the same medium. The hormone binding reaction was initiated by quick aspiration of the HBS medium and addition to each Petri dish of 0.8 ml of HBS containing the labeled peptide alone or with analogs as described above. Incubations were performed in triplicates for 45 min at room temperature (22°C). After incubation, cells were rapidly detached by scraping them with a rubber policeman. Incubations media were filtered through Whatman GF/C filters, rinsed three times, and counted in a Beckman ␥ counter. Displacement curves illustrated in Fig. 1 show that, in nondifferentiated cells, DUP 753 did not displace the iodinated Ang II analog from the Ang II receptor (Fig. 1A), although, in differentiated cells, labeling was displaced by DUP 753 (Fig. 1B), indicating the presence of AT 1 receptors. These results show that nondifferentiated cells contain only AT 2 receptors, while cells that have initiated differentiation express both subtypes of the Ang II receptor. Three days of Ang II treatment induced the expression of AT 1 receptors, which remained in low proportions however (20% displacement) compared with the AT 2 receptors (Fig. 1C), while decreasing the number of AT 1 receptors in dbcAMP-differentiated cells (20% displacement compared with 50% in control dbcAMP-treated cells) (Fig. 1D).
Preparation of Microtubule Proteins-Preparations enriched in microtubules were obtained from cells grown in 100-mm Petri dishes as described by Solomon (29) with some modifications. The cells were pretreated with 1 M Taxol (Sigma) for 2 h before extraction of microtubules. At this concentration Taxol stabilizes microtubules without promoting polymerization. The culture medium was then aspirated and replaced by PM2G buffer (PIPES, 0.1 M; glycerol, 2 M; MgCl 2 , 5 mM; EGTA, 2 mM, aprotinin, 40 TIU/ml; phenylmethylsulfonyl fluoride, 2 mM; benzamidine, 1 mM, pH 6.9) containing Taxol (1 M). Cells were scraped from the substratum with a rubber policeman and transferred to a 15-ml conical tube and centrifuged at 1000 ϫ g for 5 min at 37°C. The cell pellet was then extracted with PM2G buffer containing 1% Nonidet P-40 and 1 M Taxol. After a 15-min incubation at 37°C the suspension was centrifuged at 1000 ϫ g for 5 min at 37°C. The pellet containing the microtubules and associated proteins was then solubilized in Tris buffer 125 mM, pH 6.8, containing 4% sodium dodecyl sulfate (SDS) (w/v), 20% glycerol (v/v), and 10% ␤-mercaptoethanol (v/v) and heated to 100°C for 5 min. After centrifugation at 10,000 ϫ g for 5 min, the supernatant was stored at Ϫ20°C until Western blot analysis. For total cell extracts, cells grown in 100-mm Petri dishes were washed twice with HBS buffer and solubilized as described above.
Western Blotting-␤-Tubulin was detected with monoclonal antibody purchased from Amersham (Oakville, Ontario, Canada); monoclonal antibody that recognizes all MAP2 forms were purchased from Sigma and the monoclonal tau antibody 5E2 was kindly provided by Dr. Ken Kosik (Center for Neurologic Diseases, Brigham and Women's Hospital and Harvard Medical School, Boston, MA). Samples from equivalent number of cells were compared in each experiment. Samples were separated on 4 -15% (w/v) SDS-polyacrylamide gels. Proteins were transferred electrophoretically to polyvinylidene difluoride membranes (Immobilon-P, Millipore, Bedford, CA). Membranes were blocked with 1% gelatin, 0.05% Tween 20 in TBS buffer, pH 7.5. After three washes with TBS/Tween 20 (0.05%), membranes were incubated either with anti-tubulin (1:250), anti-MAP2 (1:500), or anti-tau (1:750) for 2 h at room temperature, followed by four washes with TBS/Tween 20. Detection was accomplished using horseradish peroxidase-conjugated antimouse antibody (Amersham, Oakville, Ontario, Canada) and enhanced chemiluminescence (ECL) detection system (Amersham). The immunoreactive bands were scanned by laser densitometry and expressed in arbitrary units.
Immunofluorescence-For immunofluorescence, cells were plated on plastic coverslips (Starsted, St. Laurent, Quebec, Canada) and were grown and treated for 3 days with appropriate stimuli. Cells were rinsed twice with HBS, fixed with formaldehyde 3.7% (v/v) (in HBS), pH 7.4, for 15 min at room temperature, and permeabilized with 0.1% Triton X-100 (v/v) in HBS. Nonspecific binding was eliminated by washing with HBS/glycine (100 mM) solution for 1 h at 4°C followed by incubation with 5% milk solution. Cells were then incubated with the anti-tubulin antibody (1:50) for 1 h at room temperature. Coverslips were washed and further incubated for 1 h at room temperature with secondary fluorescein-conjugated anti-IgG antibody (Boehringer Mannheim). The coverslips were then mounted in Vectashield (Vector Laboratories, Burlingame, CA).

Morphological Effects Induced by Ang II on NG 108-15 Cells
We first examined the effects of Ang II on the morphological changes and cytoskeletal interactions that lead to the initial neurite outgrowth from neuronal cell bodies. Cells were plated at a density of 5 ϫ 10 4 cells/dish. As shown by phase-contrast microscopy, after 3 days in culture, nontreated, nondifferentiated control cells were actively dividing and had round cell bodies, although some cells exhibited thin processes ( Fig. 2A). After a 3-day treatment with Ang II (100 nM), most cells extended one or two neurite processes with a growth cone at their tip ( Fig. 2B) along with a longer process, while the cell body retained a rounded appearance. Moreover, the number of cells was lower than in control, nontreated cells (from 4.3 Ϯ 0.8 ϫ 10 6 cells/control dishes to 3.2 Ϯ 0.6 ϫ 10 6 cells/Ang II-treated dishes, 25% decrease), indicating that the rate of cell division had decreased. Cell viability was not affected since they remained attached to the substratum, persistently excluded trypan blue, and the effect was reversible after removal of the hormone (data not shown). These morphological changes were due to AT 2 receptor activation, since cells co-incubated with Ang II plus the AT 2 receptor antagonist, PD 123319, kept their rounded appearance and formed aggregates (Fig. 2C), while co-incubation with Ang II plus DUP 753 did not alter the morphological appearance of Ang II-treated cells (data not shown). In addition, 3-day treatment with CGP 42112 also induced neurite extension and branching. These treated cells had a polygonal cell body, and the number of extensions and branchings appeared higher than in Ang II-treated cells (Fig. 2D).
The morphological appearance of the cells after the 3-day treatment with Ang II was similar to those observed after a 3-day treatment with 1 mM dbcAMP, although in the latter, the cells had mainly one axon-like process and one or two dendrite-like processes (Fig. 3A). However, if Ang II was applied during the 3-day differentiating period, neurite retraction was observed, cells exhibit then a round up appearance but conserve one process (Fig. 3B). This morphological change was mediated by the AT 1 receptor activation, since cells co-incubated with DUP 753 kept their differentiated morphology (Fig. 3C), as cells treated with CGP 42112 (Fig. 3D).

Effect of Ang II on the Distribution of Microtubules in NG108-15 Cells
Related changes in the distribution of microtubules were determined by immunofluorescence microscopy with an antibody to ␤-tubulin. In nondifferentiated cells, microtubules appear as long and thin filaments, loosely distributed throughout the cell (Fig. 4A). After a 3-day treatment with Ang II, most cells exhibited neurite processes with growth cones at their tips (Fig. 4B). When examined at higher magnification, intensification of microtubule labeling was noticed at the periphery of the cell body (Fig. 4C). Moreover, microtubules became organized in parallel bundles in the neurite and remained dispersed in the central part of the growth cone (Fig. 4D).
In cells differentiated by a 3-day treatment with dbcAMP, microtubules were organized in longitudinal bundles localized within the cytoplasm of neurites and were abundant at the perinuclear region (Fig. 5A). When Ang II was added during the 3-day of differentiating period, cells became polygonal with microtubules retracted within the cell body (Fig. 5B).

Effect of Ang II on Tubulin and Microtubule-associated Proteins in Nondifferentiated Cells
Tubulin Levels-In order to quantify and to better characterize the pharmacology of the effects mediated by Ang II, the amount of polymerized tubulin was analyzed in detergentextracted cytoskeletal fractions prepared from control, Ang IIor Ang II/analog-treated cells. Since neurite outgrowth requires large amounts of tubulin, its measurement represents an index of neurite extension. Cytoskeletal fractions from equivalent numbers of cells were analyzed in parallel. Thus, the cellular contents of each of the major proteins could be compared directly. As shown in Fig. 6A Fig. 7A and show that Ang II and CGP 42112, respectively, induced a 7-and 6-fold increase of tubulin levels compared with nontreated control cells and conclusively demonstrate that the effect is mediated through the AT 2 receptor. Changes in microtubule levels may be due to an increase in total cell tubulin content or may reflect increased polymerization of the tubulin pool. In order to verify this point, total tubulin levels were measured in cellular extracts. As shown in Fig. 6B, Ang II did not induce significant changes in the level of total tubulin content.
MAP2 and Tau Levels-Since MAPs affect microtubule assembly, stability, and cross-linking in developing brain (22), MAP2 and tau levels were characterized and compared in nondifferentiated control, Ang II-and Ang II/analog-treated cells. A monoclonal antibody recognizing a common epitope in both adult (MAP2a and -b) and juvenile (MAP2c) (27,30) forms was used in Western blots to assess protein levels of MAP2c in actively dividing and in differentiated cells. Fig. 6C shows that the antibody to MAP2 revealed two bands with similar intensities of around 85 and 76 kDa. Three-day treatment with Ang II induced a 5.3-fold increase in both MAP2c bands (Fig. 6C, lane 2 versus lane 1 and Fig. 7B). This effect was reproduced with CGP 42112 (4.7-fold increase) (Fig. 6C, lane 5 versus lane 1 and Fig. 7B) and reversed when Ang II was co-incubated with PD 123319 (lane 4 versus lane 2). However, addition of DUP 753, the AT 1 receptor antagonist, also inhibited the Ang II effect (Fig. 6C, lane 3 versus lane 2 and Fig. 7B).
Tau was detected with the monoclonal antibody 5E2. In nondifferentiated cells, tau appeared as a single isoform with a molecular mass of 58 kDa (Fig. 6D). The level of tau was not significantly affected by treatment with Ang II or Ang II/analogs.

Effect of Ang II on Tubulin and Microtubule-associated Proteins in Differentiated Cells
Tubulin Levels-As equal amount of cells were analyzed by Western blotting, we could observe that the level of tubulin had One Ang II-treated cell is illustrated at higher magnification (C) (ϫ 1200). An example of a growth cone is shown in D (ϫ 1700). After formaldehyde fixation and permeabilization with 0.1% Triton X-100, cells were processed for immunofluorescence labeling using an anti-␤tubulin antibody and fluorescein isothiocyanate as described under "Materials and Methods." increased by 5-fold in differentiated control cells compared with nondifferentiated control cells (compared lanes 1 in Fig. 6, A and E, and Fig. 7, A and C). The addition of Ang II during the 3 days of differentiating treatment with dbcAMP induced a small, but not significant, increase in the level of tubulin. However, when the AT 1 receptor was inactivated by co-incubation with DUP 753, a net increase in the level of polymerized tubulin was observed (Fig. 6E, lane 3 versus lane 1) (2.5-fold increase over control, n ϭ 5, Fig. 7C), which appeared to be mediated by the AT 2 receptor. This observation was confirmed in cells treated with CGP 42112, which also induced an increase in tubulin content (Fig. 6E, lane 5 versus lane 1) (2.8-fold increase over control, n ϭ 5, Fig. 7C). When PD 123319 was added together with Ang II, tubulin content was similar to that seen in control cells (Fig. 6E, lane 4 versus lane 1). As seen in nondifferentiated cells, total tubulin levels in control or Ang IIor Ang II/analog-treated cells did not change over the experimental period as compared with control (Fig. 6F).
MAP2c and Tau Levels-In differentiated control cells, the 76-kDa form of MAP2c was predominant (Fig. 6G, lane 1) and had increased compared with the nondifferentiated control cells (Fig. 6C, lane 1). Addition of Ang II for 3 days increased the level of the 85-kDa form without change in the 76-kDa form (Fig. 6G, lane 2 versus lane 1 and Fig. 7D). Here again, these effects were mediated by the AT 2 receptor subtype, since coincubation with PD 123319 inhibits the increase in the level of the 85-kDa form of MAP2 (Fig. 6G, lane 4 versus lane 2) and addition of CGP 42112 increased it (Fig. 6G, lane 5 versus lane  1 and Fig. 7D). However, the addition of DUP 753 did not modify significantly the level of MAP2 that was seen in cells under Ang II treatment (Fig. 6G, lane 3 versus lane 2 and Fig.  7D). Differentiation of NG108-15 cells by dbcAMP was accompanied by an increase in the level of the 58-kDa form of tau (Fig. 6, H versus D, lane 1). A minor band of 36 kDa was also present (Fig. 6H). Treatment with Ang II or Ang II/analogs did not significantly modify the basal level of tau. DISCUSSION The most important observation of our study is that Ang II, via the AT 2 receptor, acts on the microtubule-associated-protein, MAP2, initiating neurite outgrowth in nondifferentiated, rounded, and dividing NG108-15 cells. The second point is that opposite effects were observed in cells differentiated with db-cAMP, in which both AT 1 and AT 2 receptors are present. This illustrates a control of Ang II-induced differentiation by a negative feedback mechanism whereby the process is induced by AT 2 receptor and inhibited by time-dependent expression of the AT 1 receptor. AT 2 Receptor and Induction of Neurite Extension-When cultured in the presence of 10% FBS, NG108-15 cells contain only AT 2 receptors. Although these results differ from those of Tallant et al. (31), the groups of Speth et al. (32) and Carrithers et al. (33) also found that, in nondifferentiated cells, Ang II receptor is not linked to G protein (33) and does not stimulate inositol phosphate production (32), two properties associated with the AT 1 receptor activation, thus supporting that in nondifferentiated cells, Ang II receptors are not of the AT 1 type. These nondifferentiated cells treated for 3 days with Ang II develop neurites; this stimulation of neurite extension is accompanied by an increase in tubulin polymerization. Pharmacological studies demonstrate that these effects are attributable to the AT 2 receptor, since application of CGP 42112 produces the same effects as Ang II, while co-incubation with DUP 753 does not alter the Ang II effect. Similar morphological changes were obtained in PC12 cells treated for 3 days with the well known differentiating factor, nerve growth factor (34,35). We did not observe significant changes in total tubulin content after a 3-day treatment with Ang II and/or agonist or antagonists (Fig. 6B), which indicates that the nondifferentiated cells contain a pool of tubulin sufficient to support neurite outgrowth, as reported previously by Ferreira and Caceres (36) and by Shea et al. (37). These results indicate that Ang II promotes either directly or indirectly (via MAP2c) polymerization of tubulin without modifying total tubulin content.

Ang II Effects in Nondifferentiated Cells
Ang II Increases the Level of Microtubule-associated-MAP2c, but Not Microtubule-associated Tau-The addition of Ang II to nondifferentiated cells increased the level of MAP2c associated with tubulin, which may directly affect its polymerization. This effect is mediated by the AT 2 receptor, since it is reproduced by CGP 42112. Moreover, when nondifferentiated cells were coincubated with Ang II and PD 123319, they maintained their rounded appearance, without any outgrowth processes, similar to control cells. In addition, Ang II or Ang II/analogs treatment does not affect the amount of tau protein associated with microtubules. Microtubule-associated proteins play a pivotal role in the process of neurite elongation (22). For example, suppression of MAP2 expression by means of antisense oligonucleotides inhibits the outgrowth of exploratory neurites in cultured cerebellar neurons (38), whereas further neurite elongation depends upon tau (39). Therefore, our observations indicate that Ang II has a specific effect on the initial establishment of neurites, without affecting axonal elongation. Specific MAPs are also indicators of the state of brain differentiation since most exist in juvenile and mature isoforms. For instance, MAP2c is the juvenile form and MAP2a is one of the mature forms, while MAP2b is present throughout rat brain development (22). In the present experiments, the MAP2 antibody detected two polypeptides in Western blots of nondifferentiated NG108-15 cells at 76 and 85 kDa. These two bands could correspond to two different states of phosphorylation of MAP2c. Alternatively, they could also represent the two spliced isoforms of MAP2c (40). Indeed, MAP2c exists as two isoforms with either three or four tubulin binding domains, the three domain form being mainly expressed in neurones, and the four domain form being mainly expressed in glial cells (40). Both isoforms may be expressed in NG108-15 cells, due to their neuroblastoma ϫ glial hybrid origin. The specific increase of the lower molecular weight form in response to neuronal differentiation induced by dbcAMP (Fig. 6G, lane 1 versus Fig. 2C, lane 1) would favor this hypothesis and further indicate that Ang II stimulates differentiation of NG108-15 cells along a neuronal pathway. Beaman-Hall and Vallano (27) observed the presence of high molecular weight forms of MAP2, MAP2a, and MAP2b, after 9 days of differentiation, whereas we did not after 3 days. It may be that, under the experimental conditions used by these authors, cells have already achieved complete neuronal differentiation, as indicated by the presence of only one MAP2c isoform in their undifferentiated state.
Consequence of the Ang II-induced AT 1 Receptors Expression-As shown in Fig. 1C, 3-day treatment with Ang II induces expression of AT 1 receptors which anywise remain lower compared with the AT 2 receptors. The presence of these receptors may explain the discrete morphological differences observed between Ang II-and CGP 42112-treated cells as well as higher levels of MAP2 associated with microtubules. The effects observed with CGP 42112 appear stronger than those observed with Ang II. In particular, CGP 42112-treated cells exhibit several short dendrites. These differences may be due to the presence of AT 1 receptors seen after the 3-day treatment with Ang II, which may counteract the effect of the AT 2 receptor. Although DUP 753 did not affect tubulin levels, it decreased MAP2 expression as did PD 123319, suggesting that both receptors may be involved in the regulation of MAP2 associated with tubulin. Nevertheless, the expression of the AT 1 receptor is low and we cannot rule out the possibility that the inhibitory effect of DUP 753 on MAP2c expression may be nonspecific. Such nonspecific effects of DUP 753 have been shown previously in other systems (41,42). Collectively, our results support the hypothesis that Ang II, via the AT 2 receptor, induces neurite outgrowth, by acting specifically on the increased association of MAP2 with tubulin.

Ang II Effects in dbcAMP-differentiated Cells
NG108-15 cells differentiated with dbcAMP exhibit a neuronal phenotype characterized by a long axon-like process and one or two short dendrite-like processes. These morphological observations are correlated with the increased expression of tau proteins (Fig. 6, H versus D, lane 1) and MAP2 (Fig. 6, G versus C, lane 1). Application of Ang II during this differentiating period induced a rounding appearance of the cells, which still maintained their long axon-like process (Figs. 3B versus 2C). These observations further indicate that the target of Ang II effect is MAP2 rather than tau. This inhibitory effect is mediated by activation of the AT 1 receptor, since it is not modified by co-incubation with PD 123319 (data not shown), but abolished if cells are co-incubated with DUP 753 (Fig. 3, C  versus 3A). Moreover, if dbcAMP-treated cells are incubated with CGP 42112, an effect mediated by the AT 2 receptor is observed (cells exhibit several processes). Western blot experiments support these morphological data, since co-incubation with Ang II ϩ DUP 753 (Fig. 6E, lane 3 versus lane 1) induces an important increase in tubulin level similar to that observed in nondifferentiated cells treated with Ang II (Fig. 6A, lane 2  versus 1). These results indicate that activation of the AT 1 receptor inhibits the process of differentiation induced by db-cAMP (at least on the short processes). Similar observations have been reported for dopamine receptors. Both D 1 and D 2 receptors are present in fetal brain. While stimulation of the D 2 receptor increased branching and extension of neurites (43), stimulation of the D 1 receptor reduced neurite outgrowth (44), suggesting that hormones and neurotransmitters may be capable of controlling the development of specific types of neurones. Moreover, Ang II treatment of dbcAMP-differentiated cells decreases the number of AT 1 receptors. This observation that Ang II down-regulates its own receptor has been previously described in other cell types (45). This indicates that Ang II induced a time-dependent expression of the AT 1 receptor which obviously exerts a fine modulation of AT 2 receptor-induced neurite outgrowth.

What Could Be the Mechanism of Action of Ang II on MAP2 and Tubulin Polymerization?
The mechanism by which Ang II, via the AT 2 receptor, increases the amount of MAP2c associated with tubulin and, as a likely consequence, tubulin polymerization, remains to be elucidated. However, two types of mechanisms can be considered. 1) A possible effect of Ang II on MAP2 and tubulin state of phosphorylation. Accumulating data demonstrate that the AT 2 receptor is associated with stimulation of a tyrosine phosphatase activity (46,47). Preliminary results by our group (16,48) have shown that the AT 2 receptor modulates phosphorylation of tyrosine residues of several proteins, with a predominant effect on proteins with molecular masses of 110 -120, 72-75, and 20 -22 kDa. A reduction in MAP kinase activity by the AT 2 receptor has also been recently described in AT 2 receptortransfected cells (49). Moreover, we found that the AT 2 receptor inhibits p21 ras activity (48), which could be one of the first steps in the aforementioned reduction of MAP kinase activity (49). Activation of AT 2 receptor by modulating the state of phosphorylation of MAP2c could provide an additional level of control of tubulin polymerization. Several studies have shown that phosphorylation of tubulin and MAP2 inhibited microtubule assembly (50), while dephosphorylation increased it (51). It could be suggested that, in control nondifferentiated cells, growth factors present in the culture medium may stimulate the phosphorylation of MAP2c, thus preventing its binding to tubulin. Ang II, via the AT 2 receptor, could inhibit a phosphorylation cascade, in turn leading to MAP2c dephosphorylation, hence increasing its ability to interact with tubulin to promote microtubule assembly. 2) Involvement of calcium in cytoskeleton dynamics has also been largely reported (20). Based on our recent observations that the AT 2 receptor inhibits the T-type Ca 2ϩ current (15,16) and thereby modulates intracellular calcium concentration, we can postulate that this modulation could play a pivotal role in the process of AT 2 -induced polymerization-depolymerization of tubulin. This hypothesis is reinforced by the observation that the expression of T-type Ca 2ϩ channels are developmentally regulated, indicating their involvement in brain development (19).
The negative cross-talk between AT 2 and AT 1 receptors was observed at different levels. It has been demonstrated recently in vascular smooth muscle cells transfected with the AT 2 receptor (52), and in endothelial cells (53), that the AT 2 receptor has an antiproliferative effect and inhibits the growth action of the AT 1 receptor. Ichiki et al. (54), using transgenic mice lacking the AT 2 receptor, have shown that this receptor mediates a depressor effect and antagonizes the AT 1 -mediated pressor action of Ang II. All these data suggest that the AT 1 receptor may be implicated in cell growth, in part via p21 ras and MAP kinase activation (55), and that the AT 2 receptor could initiate differentiation by inhibiting the same pathways as well.
Taken together, our data suggest that Ang II, via the AT 2 receptor, promotes neurite outgrowth by increasing the amount of polymerized tubulin which likely results from an increased association of MAP2c with tubulin. Moreover, the presence of the AT 1 receptor inhibits the process of differentiation induced by dbcAMP. The precise nature of this interaction remains to be elucidated.