Thiazolidinediones Promote Axonal Growth through the Activation of the JNK Pathway

The axon is a neuronal process involved in protein transport, synaptic plasticity, and neural regeneration. It has been suggested that their structure and function are profoundly impaired in neurodegenerative diseases. Previous evidence suggest that Peroxisome Proliferator-Activated Receptors-γ (PPARγ promote neuronal differentiation on various neuronal cell types. In addition, we demonstrated that activation of PPARγby thiazolidinediones (TZDs) drugs that selectively activate PPARγ prevent neurite loss and axonal damage induced by amyloid-β (Aβ). However, the potential role of TZDs in axonal elongation and neuronal polarity has not been explored. We report here that the activation of PPARγ by TZDs promoted axon elongation in primary hippocampal neurons. Treatments with different TZDs significantly increased axonal growth and branching area, but no significant effects were observed in neurite elongation compared to untreated neurons. Treatment with PPARγ antagonist (GW 9662) prevented TZDs-induced axonal growth. Recently, it has been suggested that the c-Jun N-terminal kinase (JNK) plays an important role regulating axonal growth and neuronal polarity. Interestingly, in our studies, treatment with TZDs induced activation of the JNK pathway, and the pharmacological blockage of this pathway prevented axon elongation induced by TZDs. Altogether, these results indicate that activation of JNK induced by PPARγactivators stimulates axonal growth and accelerates neuronal polarity. These novel findings may contribute to the understanding of the effects of PPARγ on neuronal differentiation and validate the use of PPARγ activators as therapeutic agents in neurodegenerative diseases.


Introduction
Neurons are one of the most highly polarized cell types, their processes being divided morphologically and functionally into two distinct parts, the axon and dendrites [1,2]. Axon and dendrites are distinguished from each other by their different membrane and protein composition, length, and function [3,4]. Interestingly, it has been shown that the shortening and loss of axons are common pathological features of neurodegenerative diseases [5,6]. Growing evidence suggest that axonal impairment may be involved in the neuronal dysfunction reported in neurodegenerative diseases, including Alzheimer's disease (AD), Parkinson, and Huntington's disease (HD) [5].
Peroxisome Proliferator-Activated Receptor-c (PPARc) is a member of the family of transcription factor of PPARs. It has been demonstrated to play an important role in the regulation of cell differentiation in several cells, such as adipocytes and macrophages [7,8]. An important role of PPARc in the differentiation of rat mesangial, human trophoblast, and clonal neuronal cells has been demonstrated [9,10]. PPARc is expressed in the central nervous system [11,12], and 15-deoxy-PGJ2, a natural PPARc ligand stimulates differentiation of pheochromocytoma 12 (PC12) and human neuroblastoma cells [13]. Interestingly, significant defects in brain development have been reported in PPARc 2/2 and PPARc +/2 mice, indicating the important role of PPARc in neuronal development [14]. Previously, we reported that PPARc is present in rat hippocampal neurons and that its activation by thiazolidinediones (TZDs), including rosiglitazone (RGZ), ciglitazone (CGZ), and troglitazone (TGZ), PPARc activators that have been routinely used for treatment of diabetes type 2 [15], prevented axon degeneration, neurite loss, and mitochondrial impairment induced by Ab [11,12]. More importantly, previous studies showed that treatment with PPARc agonists induced neurite elongation in PC12 cells, and this event was produced by the activation of Mitogen activated kinase-c-Jun N-terminal kinase (MAPK-JNK) pathway [16]. However, the possible role of PPARc pathway and JNK on axonal elongation is unknown.
JNK is a member of the mitogen-activated protein (MAP) kinase family [17]. Because of its activation during cellular stress, JNK has been studied extensively as a stress-activated protein kinase. However, it is clear that JNK plays other important roles in neuronal development [17,18]. JNK signaling has been implicated in the development of cerebellar granule neurons [19]. Mice null for the Jnk1 gene exhibit abnormalities in axonal tracts [18]. Furthermore, mice null for both Jnk1 and Jnk2 exhibit severe neurological defects and die during embryogenesis [20]. Recent studies support a role of JNK in the regulation of neurite outgrowth during development [21,22]. JNK has also been implicated in regulating transcriptional events that regulate neurite outgrowth in PC12 cells [23] and axon regeneration in dorsal root ganglion neurons [24,25]. More importantly, Oliva et al., showed that inhibition of JNK activity by pharmacological or molecular approaches block axonogenesis but does not inhibit neurite formation or prevent dendritic differentiation [21].
Here, we describe the effect of several PPARc agonists in neurite and axonal elongation of hippocampal neurons. We found that PPARc activation promotes axon elongation by a mechanism that involved JNK activation. Treatment with TZDs significantly increased axonal growth and the use of PPARc antagonists like GW 9662, abolished axonal elongation induced by TZDs. Neurite outgrowth was not significantly increased by treatment with TZDs, indicating that PPARc-induced effects are particularly strong on axonal growth. Pharmacological inhibitors of JNK pathway prevented TZDs-induced axonal elongation, and more importantly, activation of PPARcsignificantly increased JNK activation on hippocampal neurons.
Altogether, these results suggest a novel role of PPARc participating in axogenesis and neuronal polarity mediating activation of JNK. These observations extend previous studies that showed a protective role of PPARc in neurodegenerative diseases and validate a potential use of PPARc activators against the neuronal damage observed in neurodegenerative diseases.

Ethics statement
Sprague-Dawley rats used in these experiments were housed at the Faculty of Biological Sciences of the Pontificia Universidad Católica de Chile and handled according to guidelines outlined and approved by the Institutional Animal Care and Use Committee at the Faculty of Biological Sciences of the Pontificia Universidad Católica de Chile.

Primary rat hippocampal culture
Hippocampi from Sprague-Dawley rats at embryonic day 18 were dissected, and primary hippocampal cultures were prepared as previously described [26,27]. Pregnant dams (18 days) were anesthetized with CO 2 before obtaining the 18-day rat embryos used for the hippocampal cell cultures. All procedures were performed in agreement with the animal handling and bioethical requirements established by Institutional Animal Care and Wellbeing Committee at the Faculty of Biological Sciences of the Pontificia Universidad Católica de Chile. Hippocampal neurons were seeded in poly-L-lysine-coated wells. Then, cultured hippocampal neurons were treated with PPARc agonists: TGZ (10 mM), RGZ (10 mM), and CGZ (10 mM) for 24, 48, and 72 h.
During treatment, hippocampal neurons were observed and images were taken using video microscopy.

Cell fractionation and Western blot analysis
After indicated treatments, hippocampal neurons were homogenized, and centrifuged at 100,0006 g at 4uC for 1 h. Supernatants were collected and analyzed by 10% SDS-PAGE. Protein bands were transferred to nitrocellulose membranes, and detected with appropriate primary antibodies [28,29].

Morphometric analysis
Hippocampal neurons plated on poly-L-lysine-coated covers (25,000 cells/cover) treated with PPARc agonists were observed from time 0 to 72 h, and neuronal development was followed using a Zeiss Axiovision fluorescence microscope equipped with a culture chamber (37uC and 5% CO 2 ) and video recording system [27]. The following neurite morphology parameters were evaluated: axonal length, length of minor processes (neurites) and neuronal polarity. For the analysis, an axon-like neurite was defined as a process at least twice as long as the other neurites of the same cell, with a minimum length of 50 mm [30]. A total of 200 cells from 3 independent hippocampal cultures were analyzed for each experimental condition and time point. Additionally, using the same protocol described above, we immunolabeled hippocampal neurons exposed to the different experimental conditions with monoclonal anti-tau-1 antibody (1:500), or loaded neurons with Calsein AM dye (Molecular Probes), in order to evaluate morphometric parameters. Neuronal complexity analysis (Scholl analysis) was made according to Codocedo et al. [30]. Scholl analysis is a quantitative measure of the size and shape of the dendritic tree [31]. In our studies, it represents a measure of how axon length is changing in relation of neuronal soma [30]. The total length of axons and neurites were quantified using Image-Pro plus software as previously described [11,26]. Differences among groups were evaluated by the analysis of variance and Student-Newman-Keuls test.

Wnt 5A conditioned Medium
Wnt 5A conditioned medium was generated according to Farias et al [32]. Briefly, human embryonic kidney 293 (HEK-293) cells were transiently transfected by calcium phosphate precipitation with an empty vector pcDNA (control) or a pcDNA containing sequences encoding for Wnt 5A constructs [32]. The presence of Wnt-5A ligands in the conditioned medium was verified by Western blot analysis using an antibody against the hemagglutinin (HA) epitope [32].

Statistical analysis
Results were expressed as the mean 6 standard error (S.E.M.). Differences among groups were evaluated by analysis of variance and Student-Newman-Keuls test. Students t test was used for analyzing data for Western blot and image analysis. P,0.05 was regarded as statistically significant.

PPARc activation promotes axonal elongation on hippocampal neurons
PPARcactivation with TGZ prevents neuronal cell death and calcium stress induced by Ab peptide [11]. In that study, PPARc activation by agonists induced an increase of axonal caliber and neurite length on hippocampal neurons [11]. Previous evidence suggests that PPARc activation promotes neurite extension in PC12 cells exposed to soluble Nerve Growth Factor (NGF) [16]. Treatment with the PPARc agonist TGZ (10 mM) for 24 h accelerated axonal development on hippocampal neurons (Fig. 1). Similar results were obtained with other PPARc activators including RGZ (10 mM) and CGZ (10 mM) ( Fig. 2A, C). Neuronal development was evaluated measuring axonal growth (Fig. 1B), neuronal polarity (Fig. 1C), and neurite outgrowth (Fig. 2B). Treatment with TGZ induced a two-fold increase in the axonal length compared with untreated neurons (Fig. 1A, B). Additionally, TGZ induced a substantial increase in the percentage of hippocampal neurons showing neuronal polarization (Fig. 1C). We also observed that in hippocampal cultures exposed to TGZ for 72 h, around 98% of the neurons showed a polarized phenotype, which means that they developed a distinguishable axonal process with minor secondary processes (Fig. 1C) [30]. These results suggest that activation of PPARcby TZDs drugs promotes axonal growth and neuronal polarity in rat hippocampal neurons.

Blockage of PPARc activation prevented the increase in axonal growth in hippocampal neurons treated with TZDs
To corroborate the effects observed with TGZ, we tested other PPARc activators belonging to the TZDs family, like RGZ and CGZ, and the specific PPARc antagonist GW 4662 (GW) [15]. TZDs drugs have been used for the treatment of diabetes mellitus type 2 [15], and their use have recently been associated with a significant recovery of memory impairment in Alzheimer's disease patients [33]. GW is an antagonist of the PPARc receptor. In ours hands, it was capable of preventing neuronal cell death protection induced by TGZ in Ab-treated neurons [11]. Figure 2 shows the effect of PPARc agonists in neurite and axonal outgrowth in presence and absence of 5 mM GW. Measurement of total neurite length in hippocampal cultures treated with TZDs plus GW did not show significant differences compared with untreated neurons (Fig. 2B). Further studies in neurons treated with TZDs plus GW showed a significant reduction in axonal length (Fig. 2C). These indications suggest that TZDs-mediated effect were PPARcdependent and were mainly observed in the axon. In addition, RGZ and CGZ increased the percentage of polarized neurons, similar to the effect observed after TGZ treatment showed in Figure 1. This effect was also abolished by incubation with GW (data not shown).

PPARc agonists induced PPARc expression and its axonal localization in hippocampal neurons
We evaluated by immunofluorescence protein expression and localization of PPARcreceptor in hippocampal neurons in response to TZDs. Figure 3 shows representative immunofluorescence images and analysis of the levels and distribution of PPARc in neurons exposed to 10 mM TZDs for 72 h. TZDs induced a robust increase in PPARc levels, in comparison with untreated neurons (Fig. 3A). Additionally, we observed a significant axonal localization of PPARc in neurons treated with PPARc agonists (Fig. 3A). Immunofluorescence studies evidenced a robust and close localization between anti-tau-1 and anti-PPARc antibody in TZDs-treated neurons. PPARc staining of untreated neurons predominated in the nucleus with not apparent co-localization between tau-1 and PPARc in axons (Fig. 3). Interestingly, in hippocampal cultures co-treated with TZDs and 10 mM GW, PPARc levels were significantly decreased, indicating that the effect of TZDs were mediated by specific activation of PPARc (Fig. 3B). Quantitated data from representative images of neurons treated with TDZs and immunolabeled for tau-1 and PPARcindicated that PPARc activation by TZDs significantly increased protein PPARc levels in hippocampal neurons (Fig. 3C). The immunofluorescence data presented above was corroborated by western blot studies made in hippocampal neurons treated with increasing concentrations of CGZ, and in the presence of GW (Fig. 3D). Treatment with CGZ increased PPARc protein levels, effect that was prevented by GW (Fig. 3D). These results suggest that PPARc activation by TZDs increased PPARc protein levels, and also promoted localization of PPARcin the axon of hippocampal neurons. This effect could facilitate the accelerated axonal growth observed in the TZDs-treated neurons. Previous evidence suggests that neurite elongation induced by PPARc agonists in PC12 cells is produced by activation of MAPK, p38, and JNK kinase [16]. Additionally, studies in knock out mice for JNK showed a delay in neuronal development with evident signs of neurodegeneration [34]. To study the possible role of JNK in TZDs-induced axonal elongation, we studied hippocampal neurons treated with PPARc agonists (10 mM) in the presence of the specific JNK inhibitor SP 600125 (SP; 100 nM) [17,32]. Figure 4A shows representative confocal images of neurons exposed to the indicated conditions for 72 h. Inhibition of JNK prevented axonal elongation induced by TZDs (Fig. 4A). The effect was significant only for average axonal length (Fig. 4C). In contrast, quantification of independent experiments did not show statistical differences for neurite total length in neurons treated with PPARc agonists in presence of SP (Fig. 4B). Additional quantification analysis indicated that TZDs-induced axonal growth was dependent on JNK activation (Figs. 5 and Fig S2). A time course of hippocampal neurons exposed to 10 mM CGZ in the presence or absence of 100 nM SP and labeled with anti-tau 1 antibody to specifically detect the axon, indicated that the increased axonal growth was totally prevented by the JNK inhibitor SP (Fig. S1). Additional analysis of neuronal complexity (Scholl analysis) supports the role of JNK in axonal elongation induced by TZDs (Fig. 5) [30,31]. Scholl analysis indicated that TZDs treatments clearly induced axon elongation and pretreatment with SP totally prevented this effect (Fig. 5A, B, C). These results suggest that PPARc activation promotes axonal elongation by the activation of JNK in hippocampal neurons. Figure 6 shows representative confocal images from neurons double labeled with anti tau-1 and anti-phosphorylated JNK (p-JNK) antibodies after being treated with TGZ, RGZ and SP for 72 h. Anti-p-JNK shows the activation of the JNK pathway [29]. There was a strong increase in p-JNK levels in TZDs-treated neurons (Fig. 6A). p-JNK was mainly localized in the axon, suggesting that activation of JNK may participate in axonal elongation induced by TZDs (Fig. 6A). Additionally, immunofluorescence analysis of TZDs-treated neurons showed a conspicuous co-localization of p-JNK and anti-tau 1 labeling (Fig. 6A). As was expected, SP reduced p-JNK levels, and reorganized p-JNK localization towards a cytoplasmic pattern (Fig. 6B). In addition, dose response studies showed that CGZ induced a significant increase in p-JNK expression evaluated by western blot (Fig 7).

PPARc agonists induce JNK activation in primary hippocampal neurons
Interestingly, increased levels of p-JNK were not observed when hippocampal cultures were cultured in the presence of 5 mM GW, suggesting a specific role for PPARc on the control of JNK activation.

Axonal elongation induced by TZDs is not mediated by external signal response kinase (ERK) activation
In this paper, we show that activation of PPARc receptors by TZDs enhances axon growth through JNK activation. However, it was previously suggested that PPARc activators induced neurite outgrowth of PC12 cells [16] and differentiation of embryonic midbrain cells [35] by participation of JNK, p38, and ERK   [16,35]. To study the possible role of ERK in the increase of axon growth produced by TZDs, we treated hippocampal neurons with PPARc activators in the presence and absence of 5 mM PD 98059 (PD), which is a well-know inhibitor of ERK [32]. Figure 8A shows representative confocal images of hippocampal neurons untreated and treated with 10 mM CGZ and CGZ+PD during 72 h, and immunostained against tau-1 (Fig. 8A). These studies revealed that inhibition of ERK has not apparent effect on the axonal elongation induced by CGZ (Fig. 8A, B). In addition, we evaluated the activation levels of ERK in hippocampal neurons treated with increasing concentrations of CGZ in the presence of GW (Fig. 8C). Western blot studies indicated that treatment with 10 mM CGZ significantly increased p-ERK levels compared with untreated neurons (Fig. 8C). However, inhibition of PPARc activation by GW was not able to prevent p-ERK levels increased by CGZ (Fig. 8C). These studies suggest that ERK is not participating in the increased axonal growth produced by TZDs in hippocampal neurons.

Treatment with ligand Wnt 5A and TGZ increased axon growth through the JNK pathway
Wnt proteins are morphogens that play important roles during embryogenesis [36]. Wnt proteins signal through at least two different pathways: canonical and non-canonical [32,36]. In the canonical pathway, Wnt signals through Dishevelled (Dvl) to increase cytoplasmicb-catenin levels, and then b-catenin enters the nucleus, where it co-activates transcription of Wnt target genes [36]. Non-canonical Wnt signaling pathways mediate several cellular processes through different molecular intermediates, including Rho-GTPases, intracellular calcium levels and JNK activation [32]. Recently, it has been shown that the ligand Wnt 5A, an activator of non-canonical Wnt pathway, could play a role in the process of axonal growth and guidance [37,38]. Treatment with Wnt 5A increased axon outgrowth and enhances the vesicle transport to growth cones in cortical neurons [38]. In addition, we previously reported that treatment with Wnt 5A rapidly induced activation of JNK pathway [32]. However, the mechanism for the participation of Wnt 5A in axon elongation is not completely elucidated. Therefore, we treated hippocampal neurons with conditioned medium containing Wnt 5A during 72 h, and then neurons were fixed and double staining with anti-tau1 and anti-p-JNK antibodies, and axon length was analyzed (Fig. 9). Representative confocal images showed that treatment with Wnt 5A significantly increased axonal elongation compared with untreated neurons (Fig. 9A, B). Interestingly, axonal growth increase by Wnt 5A was abolished in the presence of JNK inhibitor SP, suggesting that JNK could be involved in this process (Fig. 9A, B). As we previously observed in this paper, treatment with TZDs induced axonal elongation through JNK pathway (Figs. 4, 5). Therefore, we evaluated axon length in hippocampal neurons treated for 72 h with both Wnt 5A and TGZ. Treatment with Wnt 5A+TGZ induced a significant increase in axonal growth. However, this increase was not significant compared with neurons treated with Wnt 5A or TGZ per separate (Fig. 9B). In addition, p-JNK levels were evaluated in neurons treated with Wnt 5A or Wnt 5A+TGZ, in the presence of SP. Immunofluorescence analysis indicated that Wnt 5A+TGZ treatment for 72 h increased p-JNK levels and this increment was prevented using JNK inhibitor SP. These observations suggest that Wnt 5A and TGZ stimulates axonal growth using a common pathway, in this case, JNK pathway.
Altogether, these observations suggest that JNK kinase plays an important role for axonal elongation induced by PPARc activators (TZDs) in hippocampal neurons. Both pathways can contribute to neuronal development by promoting the extension of the neuronal processes, and represent a novel therapeutic strategy to promote neuronal protection in neurodegenerative diseases.

Discussion
Neurite network loss and axonal degeneration has been observed in a wide range of neurodegenerative disorders [5,6]. These features are common in neurodegenerative diseases, producing anomalous synaptic function, and neuronal cell death [5,39,40]. Ab peptide induces a severe neurite network loss and axonal degeneration in different neuronal cell types [41]. Therefore, it is important to understand how these neurodegenerative changes evolve in order to design new strategies to repair the loss of connections. Here, we showed that PPARc activation promoted axonal growth in rat hippocampal neurons, effect that was mediated by the activation of JNK kinase induced by activation of PPARc. Previous studies indicate that PPARc activation is involved in differentiation of adipocytes and oligodendrocytes [10,14]. Our findings are in agreement with increased evidence that suggest that PPARc has a role in neuronal repair [42,43]. TZDs drugs (e.g., TGZ, RGZ, and CGZ) are PPARc agonists that increase peripheral insulin sensitivity and stimulate mitochondrial biogenesis and function [15,44]. Recently, clinical trials showed that pioglitazone improved memory and cognition in a subset of AD patients [33] as well as reduced learning and memory deficits in a mouse model for AD [45]. In addition, other studies describe that PPARc activation protects from neuronal ischemia, glutamate toxicity, and long terminal potential (LTP) impairment in an AD mice model overexpressing APP protein [46]. Moreover, we showed that PPARc activation prevents Ab neurotoxicity effects [11,12], and RGZ treatment protected from mitochondrial failure induced by mutant hunting- tin expression [44]. PPARc activation and the induction of peroxisomes prevented neuritic network loss and axonal damage induced by Ab [28]. In fact, the peroxisome proliferation effect induced by Wy (a peroxisome proliferator) is associated with the activation of the PPARaresponse [28]. PGC1-a, a transcriptional factor involved in mitochondrial biogenesis, is involved in this process [47,48]. Additionally, evidence indicates that PGC1-a could be playing a role in the pathogenesis of Huntington Disease (HD), evidence that support the importance of PPARc receptor in the neuropathological mechanisms of various neuronal disorders [48,49].
These events are in agreement with our findings that led us to propose a role for PPARc activation on the promotion of neuronal development, especially on axonal elongation. TZDs treatment promoted axonal growth and this effect was totally prevented by GW 4622, a specific PPARc antagonist. In addition, co-treatment with the JNK inhibitor SP600125 prevented axonal elongation induced by TZDs, further supporting the participation of PPARc pathway. Previous evidence suggests that PPARcis involved in PC12 differentiation induced by nerve growth factor (NGF) through activation of MAPK and JNK [9,14,16]. Interestingly, Brodbeck et al. showed that treatment with RGZ significantly increased dendritic spine density in a dose-dependent manner in primary cortical rat neuron cultures [42]. This effect was abolished by GW9662, suggesting that RGZ exerts its effect by activating the PPARc pathway [42]. Our observations are in agreement with these studies and confirm the potential role of PPARc promoting neuronal development and synaptic regeneration, by increasing axonal length and dendritic spine density in hippocampal neurons. Our results suggest that PPARc promoted axonal elongation by the activation of JNK kinase. There are interesting observations that associate the JNK pathway with neuronal polarity [4,21]. JNK activity is maintained at an extremely high level in the embryonic brain compared with other MAP kinase-related enzymes [4]. Previous studies show severe impairments on dendritic structure in the cerebellum and motor cortex of c-Jun N-terminal kinase 1 (JNK1)-deficient mice [34]. JNKs may influence cytoskeletal reorganization via the phosphorylation of proteins regulating microtubule stability, including doublecortin (DCX), stathmin family protein (SCG10), and microtubuleassociated proteins, MAP2 and MAP1B [34,50].
Interestingly, it has been shown that activated JNK is required for axonogenesis but not for the formation of minor processes or development of dendrites in hippocampal neurons [21]. Pharmacological blockage of JNK pathway inhibited axonal elongation resulting in a phenotype that may lack a defined axon [21]. In our studies, inhibition of JNK significantly prevented axonal elongation induced by TZDs and the phenotype showed by hippocampal neurons resembled that described by Oliva et al. [21]. Therefore, activation of JNK pathway appears to mediate induction of axonal growth by PPARc.
It has been shown that ATF-2 is required for maximal and accurate PPARc transcription [51,53]. ATF-2 directly binds to the PPARc promoter and activates their transcription to regulate adipocyte differentiation [51]. Therefore, activation of ATF-2 through JNK pathway could be involved in the axonal elongation increase induced by PPARc agonists in hippocampal neurons. Further studies are required to evaluate ATF-2 involvement in TZDs-induced axonal elongation in hippocampal neurons.
Finally, our work presents evidence that support the role of PPARc activation through JNK pathway in neuronal development. Combined activation of these two pathways could be beneficial for the promotion of neuroprotective effects in various neurodegenerative disorders.

Conclusions
Our results suggest that PPARc stimulation by TZDs induces axonal growth in hippocampal neurons. Treatment with different PPARc activators significantly increased axonal elongation without effects over other neuronal properties. The use of GW9662, a specific PPARc antagonist, and SP 600125, an inhibitor of JNK, prevented these changes. Interestingly, other reports show an important role of JNK controlling the neuronal polarity. Our studies showed that JNK activity could be modulated by PPARc activators, suggesting that the increase in axonal elongation induced by PPARc agonists is mediated by JNK. Altogether, our results suggest that PPARc stimulation could contribute to the development and maintenance of a proper neuronal connectivity. Figure S1 Troglitazone increases axonal elongation in hippocampal neurons. Hippocampal neurons recently plated were treated with 10 mM troglitazone (TGZ) and axonal development was observed by video microscopy. Neurons were mounted in a culture chamber controlling temperature, CO 2 , and humidity. Images were taken every hour using a cool CCD fluorescence camera (Zeiss, Germany). (TIFF)  Figure S2 PPARc activation increase of axonal elongation is mediated by JNK activation. Hippocampal neurons treated with CGZ, SP, and CGZ+SP were fixed at the indicated times and immunofluorescence against tau-1 was done. Axonal length was evaluated using Image Pro software. Data are the mean 6 S.E.M. of 4 independent experiments, *p,0.05 and **p,0.01. (TIFF)

Author Contributions
Conceived and designed the experiments: RAQ JAG NCI. Performed the experiments: RAQ JAG IA DC. Analyzed the data: RAQ RvB MB NCI. Contributed reagents/materials/analysis tools: RAQ NCI. Wrote the paper: RAQ RvB MB NCI.