LAMTOR2-Mediated Modulation of NGF/MAPK Activation Kinetics during Differentiation of PC12 Cells

LAMTOR2 (p14), a part of the larger LAMTOR/Ragulator complex, plays a crucial role in EGF-dependent activation of p42/44 mitogen-activated protein kinases (MAPK, ERK1/2). In this study, we investigated the role of LAMTOR2 in nerve growth factor (NGF)-mediated neuronal differentiation. Stimulation of PC12 (rat adrenal pheochromocytoma) cells with NGF is known to activate the MAPK. Pharmacological inhibition of MEK1 as well as siRNA–mediated knockdown of both p42 and p44 MAPK resulted in inhibition of neurite outgrowth. Contrary to expectations, siRNA–mediated knockdown of LAMTOR2 effectively augmented neurite formation and neurite length of PC12 cells. Ectopic expression of a siRNA-resistant LAMTOR2 ortholog reversed this phenotype back to wildtype levels, ruling out nonspecific off-target effects of this LAMTOR2 siRNA approach. Mechanistically, LAMTOR2 siRNA treatment significantly enhanced NGF-dependent MAPK activity, and this effect again was reversed upon expression of the siRNA-resistant LAMTOR2 ortholog. Studies of intracellular trafficking of the NGF receptor TrkA revealed a rapid colocalization with early endosomes, which was modulated by LAMTOR2 siRNA. Inhibition of LAMTOR2 and concomitant destabilization of the remaining members of the LAMTOR complex apparently leads to a faster release of the TrkA/MAPK signaling module and nuclear increase of activated MAPK. These results suggest a modulatory role of the MEK1 adapter protein LAMTOR2 in NGF-mediated MAPK activation required for induction of neurite outgrowth in PC12 cells.


Introduction
Signaling pathways in eukaryotic cells are often controlled by the formation of specific signaling complexes which are coordinated by scaffold and adaptor proteins. A well-studied signaling pathway is the mitogen-activated protein kinase (MAPK/ERK) cascade, which underlies the regulation of many cellular processes [1]. The discrete dynamics of MAPK activation are believed to be the underlying cause of differences in cellular response [2,3,4,5]. To date, several scaffold proteins have been identified that facilitate MAPK activation in mammalian cells, such as the kinase suppressor of Ras 1 (KSR1) and the MEK1 partner (LAMTOR3/ MP1), which is recruited to late endosomes by the adapter protein LAMTOR2/p14 [1,5,6,7]. Besides its role as a scaffold for MAPK signaling, the LAMTOR2/LAMTOR3 complex has been shown to be important for endosomal biogenesis and routing of receptors such as the epidermal growth factor receptor (EGFR) [7,8].
Early observations in our laboratory indicated that LAMTOR2 [15] is an essential modulator of NGF-mediated differentiation. In view of the critical role of LAMTOR2 for the stability of the entire LAMTOR complex and the controversial role of mTOR1 signaling in neuronal differentiation [16,17,18,19], in this study, we focused on the role of LAMTOR2 in NGF/MAPK-mediated differentiation of PC12 cells. This cell line has been extensively used as a model for investigating NGF-induced signal transduction events because it can mimic NGF-induced survival or differentiation observed in neuronal cells [20].
The aim of the present study was to investigate the role of LAMTOR2/MAPK module in neuronal signaling. We were able to show that LAMTOR2 is a negative regulator for NGFmediated neurite formation in PC12 cells.

Cell Culture
PC12 cells were grown in RPMI 1640 medium supplemented with 1% Pen/Strep, 1% L-glutamine, 10% horse serum and 5% fetal calf serum at 37uC with 5% CO 2 on collagen-S type I-coated culture dishes. Cells were subcultured at a density of about 80% or 2 days before starting an experiment. The effects of NGF are mediated by two different cell surface receptors (reviewed in [21]: the receptor tyrosine kinase TrkA [22,23] and p75NTR (also called p75) [24,25], both of which are present on PC12 cells [26].

Neurite Outgrowth Assay
Wild type or transfected cells were incubated for various time periods (6-72 h) with NGF or EGF (concentrations according to data in the literature [4] and our own preliminary experiments). Thereafter, pictures were taken of 3-5 independent microscopic fields under blind trial conditions, and neurite-bearing cells were counted. Cells with neurites longer than twice their cell diameter were defined as cells with neurites. Results were calculated and presented either as percentages of neurite-bearing cells [neuritebearing cells 6 100/total cells] or fold of control [% neuritebearing treated cells/% neurite-bearing control cells]. Neurite length was measured using ImageJ, linking neurite length to the amount of pixels measured.

Phalloidin Staining
PC12 cells were cultured on chamber slides and treated once with NGF (5 ng/mL) or alternatively pretreated with MEK1inhibitor PD098059 (50 mM). For studies, cells were generally stained with phalloidin 16 h after incubation. Briefly, cells were washed with 16 phosphate buffered saline (PBS) and fixed for 10 min at room temperature (RT) in 4% PFA. After extensive washing in 16PBS, cells were permeabilized with 0.5% Triton X-100 for 15 min at RT. Washes followed and cells were stained with 0.2 mM tetramethylrhodamine isothiocyanate-labeled phalloidin (Phalloidin-TRITC) and Hoechst 33342 (10 mg/mL) for 40 min at RT in the dark. To remove unbound phalloidin conjugate, cells were washed several times with 16PBS. Chamber slides were cover slipped with MOWIOL 4-88. Cells of 3-5 independent microscopic fields were visualized under blind trial conditions under a fluorescence microscope (Zeiss Axioplan2, Austria) equipped with a spot camera (RT-slider 2.3.1 Visitron Systems, Germany) using Hoechst and TRITC filters.

GAP-43 Staining
Transfected PC12 cells were grown on coated chamber slides or glass coverslips and stimulated with or without NGF (25 ng/ml) for 24 h. After washing the cells in prewarmed 16 PBS, cells were fixed with 4% PFA supplemented with 5% sucrose for 10 min at RT, washed again, permeabilized in ice cold methanol for 2 min and blocked for 30 min with 3% BSA. For labeling, cultures were incubated with monoclonal primary antibody (anti-pan GAP-43 clone GAP-7B10) diluted 1:1000 in 1% BSA overnight at 4uC. After several washing steps in 16 PBS, incubation for 30 min at RT with the secondary antibody solution (Alexa 555-conjugated goat anti-mouse IgG of F(ab) 2 fragments, 1:1000 in 1% BSA) followed. Cells were covered with MOWIOL 4-88, and pictures were taken under blind trial conditions using a fluorescence microscope (Zeiss Axioplan2, Austria) equipped with a spot camera (RT-slider 2.3.1 Visitron Systems, Germany).

Colocalization Experiments and Image Analysis
For colocalization analysis, PC12 cells were transfected with control or specific siRNA for LAMTOR2. One day after transfection, cells were put on ice for about 5 min and thereafter left untreated (control) or treated with NGF (200 ng/ml) for various time points (0, 5 and 15 min). After stimulation, cells were washed once in 16 PBS, fixed in 4% PFA for 10 min, permeabilized in 0.2% Triton-X-100 for 2 min and blocked in blocking solution (0.5 g gelatin in 10 ml H 2 O, 2.5 ml 500 mM NH 4 Cl, BSA, 12.5 ml 26 CB (20 mM Pipes [pH 6.8], 300 mM NaCl, 10 mM EGTA, 10 mM glucose, 10 mM MgCl 2 )) for 30 min at RT. Anti-EEA1 (1:200 in blocking solution) and antipan-TrkA (1:200 in blocking solution) were applied for 2 h at RT in the dark. After extensive washing, secondary antibodies (goat anti-mouse Alexa 568 1:2500 and goat anti-rabbit Alexa 488 1:1000 in blocking solution) were added for 40 min in the dark followed by 3-4 washing steps with 16 PBS. In the last step, Hoechst 33342 (8 mM) was added as nuclear marker. Cells were mounted in MOWIOL 4-88 and documented in one section per cell (at the level of the nucleus) using the Leica TCS SP5 confocal microscope followed by analysis with Huygens Professional 3.7 software (Scientific Volume Imaging, Laapersveld, The Netherlands). Images were deconvolved applying the classical maximum likelihood estimation (CMLE) with the number of iterations set to 100, the quality change threshold to 0.1% and the signal-to-noise ratio varying between 7 and 15 in each channel. Colocalization analysis was performed with the ''Colocalization Threshold'' plugin of ImageJ. The threshold is determined in two stages and hence the percentage of voxels, which have both channel intensities above threshold, is expressed as percentage of the total number of pixels in the image.

Bio-Plex Phosphoprotein Detection Assay
1610 4 untransfected and transfected (control, LAMTOR2/p14 siRNA 6 human LAMTOR2/p14 ortholog) cells were lysed by adding lysis buffer (Biorad Factor I 1:250, Factor II 1:500, PMSF 2 mM) directly to the cells. After 20 min on ice, a centrifugation step (4500 g, 20 min, 4uC) followed. The supernatant was diluted in one volume assay buffer and analyzed using the Bio-Plex phosphoprotein detection assay. Briefly, nonmagnetic beads coupled to antibodies against the target protein were added to the filter plate. After washing steps, lysates were added and incubated overnight (gentle shaking at 4uC). This was followed by several washing steps. Biotin-labeled detection antibodies specific for a secondary epitope on the target were added, and after a 30 min incubation step at RT, streptavidin-PE was used for detection with the Bio-Plex system.

Quantitative Real-time RT-PCR
Cells-to-CT technology enables reverse transcription of lysates from cells without preceding RNA purification. Briefly, transfected (LAMTOR2/p14 siRNA and control siRNA) cells were washed in cold 16PBS. Thereafter, lysis solution (1:100) was added and after 5 min incubation at RT, cell lysis was stopped by adding stop solution. Reverse transcription and real-time PCR analysis was carried out directly thereafter. Expression analysis was done in duplicates on an ABI PRIM 7000 Sequence Detection System with TaqMan gene-expression assays. To determine the relative quantification of a target gene (LAMTOR2) in comparison to a reference gene (GAPDH), the mathematical model presented by PE Applied Biosystems (Perkin Elmer) and described by Pfaffl [31] was used: 2 2''ct method.

Statistical Analysis
Results are presented as means 6 SEM of the indicated number of independent experiments. The SPSS 19.0 statistics program was applied for analysis of experiments. The unpaired two-tailed t-test was used to compare two independent groups. P-values ,0.05 were considered statistically significant (*P,0.05, **P,0.01, ***P,0.001). When more than two groups were compared, data were analyzed using one-way ANOVA followed by Dunnett's multiple comparison test. P-values ,0.05 were considered statistically significant (*P,0.05, **P,0.01, ***P,0.001).

Results and Discussion
When PC12 cells were incubated with the two growth factors EGF or NGF, only the latter efficiently supported the development of neurites, which was monitored over a time period of three days in PC12 cells (7.9461.22% percentage of neurite-bearing cells) ( Fig. 1A and B). This result stands in line with previous observations in PC12 cells representing a model of choice to compare the signaling of growth factors such as NGF and purine nucleosides related to differentiation processes and of mitogenic growth factors such as EGF [27,32,33].
Based on our earlier observation indicating that the adapter protein LAMTOR2 [15] is an essential player in NGF-mediated differentiation, we further focused on the role of the MAPK/ LAMTOR module. Transfection of PC12 cells with synthetic specific siRNA for LAMTOR2 resulted in efficient knockdown (e.g. 87.7867.86% LAMTOR2 mRNA and 60.25613.54% LAMTOR protein expression at 48 hours), as shown by analysis of LAMTOR2 mRNA (Fig. 1C) and protein expression levels ( Fig. 1D and F) in PC12 cells. This also resulted in reduced protein expression of the other members of the LAMTOR complex, LAMTOR1 (p18; 7.6361.50% reduction), LAMTOR3 (MP1; 55.91614.68% reduction), LAMTOR4 (C7orf59; 51.85612.52% reduction) and LAMTOR5 (HBXIP; 41.06610.33% reduction) at 48 hours ( Fig. 1E and F). Knockdown of LAMTOR2 also decreased protein stability of the other four LAMTOR components. This result is in agreement with previous reports on destabilization of the entire LAMTOR/Ragulator complex following inhibition of LAMTOR2 [9,11].
In a next step, we investigated the effect of LAMTOR2 knockdown on neurite formation of PC12 cells. To our surprise, knockdown of LAMTOR2 did not inhibit neurite formation but significantly augmented both NGF-mediated neurite outgrowth ( Fig. 2A and B), and neurite length (Fig. 2C). In parallel, the expression of the plasticity protein GAP-43 was analyzed in control and LAMTOR2 knockdown PC12 cells that were stimulated with and without NGF in the presence and absence of the TrkA inhibitor K-252a (Fig. 2D). GAP-43 expression was enhanced in cells transfected with LAMTOR2 siRNA as compared to cells transfected with control siRNA, both in unstimulated ( Fig. 2Di and ii) and in NGF-stimulated PC12 cells (Fig. 2D iii and iv). Addition of K-252a resulted in a drastic downregulation of GAP-43 expression levels in control-and LAMTOR2-siRNA transfected NGF-stimulated PC12 cells (Fig. 2D v and vi).
Ectopic expression of a siRNA-resistant LAMTOR2 ortholog (expression was checked by western blotting, B. T, data not shown) reversed this phenotype (Fig. 3A and B), also for neurite length (data not shown), ruling out nonspecific off-target effects of the LAMTOR2 siRNA approach.
Due to the tight connection of LAMTOR2 and p42/ 44 MAPK, we wanted to analyze the role of MAPK activation in NGF-mediated differentiation. First we approached this question by using the pharmacological inhibitor of MEK1 (PD098059) in our cell system. We observed that PD098059 significantly blocked NGF-mediated differentiation (Fig. 4A). Activity of p42/44 MAPK was significantly augmented following stimulation with NGF, as shown by Bio-Plex phosphoprotein detection (Fig. 4B) and western blot analysis (Fig. 4C). In agreement with this, knockdown of MAPK to the level of about 30% protein expression [27] induced inhibition of NGF-mediated neurite formation (Fig. 4D and E). The effects of siRNA against p44 and p42, however, were not additive. There are several possible explanations for this phenomenon. The results may reflect a non-redundant and isoform-selective role of both p42 and p44 MAPK in NGF-induced neurite formation. Of note, the question whether p42 and p44 MAPK are functionally redundant is still a matter of debate. Human p42 and p44 MAPK are 84% identical in sequence, are regulated similarly, contribute to intracellular signaling by phosphorylating a largely common subset of substrates and share many, if not all, functions [34,35]. However, Yao et al. reported that p42 and p44 MAPK are not functionally redundant [36]. Alternatively, the observed result may be explained by the haplosufficiency of either p42 and p44 MAPK following siRNA-mediated knockdown, as the remaining leftover of about 30% protein kinase allow, in part, an almost normal cellular function.
We were therefore interested to find out whether LAMTOR2 may directly regulate NGF-mediated MAPK activity. PC12 cells were transfected with siRNA to specifically knockdown LAM-TOR2 protein expression. Lysates of PC12 cells were then tested for the activation of MAPK. Mechanistically, LAMTOR2 siRNA treatment significantly enhanced NGF-dependent MAPK activities (Fig. 5A). This result was confirmed by western blot analysis of phospho MAPK as compared to pan MAPK/pan Akt (Fig. 5B). These effects were again reversed upon expression of the siRNAresistant LAMTOR2 ortholog (Fig. 5C).
In addition, the distribution of active MAPK as compared to pan MAPK, LAMTOR2 and pan Akt in the cytoplasmic and nuclear fraction were analyzed after NGF stimulation. Interestingly, we observed that LAMTOR2 knockdown decreased MAPK activity in the cytoplasm by 28.87610.85%, whereas active MAPK in the nucleus was increased by 23.5767.94% (Fig. 6A). In contrast to the even distribution of Akt, LAMTOR2 protein expression was predominant in the cytoplasmic fraction (Fig. 6B).
Next, we raised the question whether LAMTOR2 was involved in endosomal trafficking of the NGF receptor TrkA. PC12 cells were transfected with control or siRNA specific for LAMTOR2. Cells were then stimulated with NGF, and colocalization of endosomes and TrkA receptor analyzed. Studies of intracellular trafficking of the NGF receptor TrkA revealed a fast (5 min) colocalization with early endosomes, which was enhanced by 19.3966.97% by LAMTOR2 siRNA. Yet, after 15 minutes, a significant decrease (44.45615.79%) of TrkA receptor colocalization with early endosomes was observed ( Fig. 7A and B). Concomitant with this result, LAMTOR2 knockdown decreased colocalization of TrkA with late endosomes (B.T. unpublished observation). Due to the abnormal, peripheral distribution of late endosomes in LAMTOR2 knockdown cells, we did not pursue this line of research. Inhibition of LAMTOR2 and concomitant destabilization of the remaining LAMTOR complex apparently leads to a faster release of the TrkA/MAPK signaling module and nuclear increase of activated MAPK.
Our results are in agreement with previous findings [37] that, in fibroblasts obtained from patients suffering from immunodeficiency syndrome caused by deficiency of LAMTOR2, the late endosomes were no longer concentrated in the perinuclear area, but redistributed to the cell periphery, suggesting a role of LAMTOR2 in the control of late endosomal compartment configuration. In LAMTOR2 knockout mouse embryonic fibroblasts (MEFs), late endosomes, multivesicular bodies (MVBs) and lysosomes, but not early endosomes were displaced to the cell periphery [7,8]. Interestingly, MVBs and lysosomes were also displaced to the cell periphery in MEK2/2 MEFs suggesting that the proper positioning of late endosomes require LAMTOR2-LAMTOR3-MEK1 signaling [8]. These data also indicate an exciting link between MAPK signal transduction, organelle biogenesis and intranuclear transport. In a future study, it will be interesting to identify the interplay of NGF receptors and the LAMTOR2-MAPK-MEK1 module. Data presented in this study are in line with previous findings [38], providing evidence that TrkA predominantly recycles back to the cell surface after ligand treatment, whereas TrkB is predominantly sorted to the degradative pathway. A further goal will be the development of a primary nerve cell model. While PC12 cells may be maintained without nerve growth factor (NGF) in culture and the role of NGF in PC12 cell differentiation is clear and reproducible [3,20,33], 'developing sympathetic neurons' largely depend on NGF for survival and die by apoptosis after NGF withdrawal [39]. For this reason, the TrkA receptor was also called a ''dependence receptor,'' (reviewed in [40]). Biochemical analysis of the actions of NGF upon primary peripheral neurons has often been hampered by the lack of a variety of neuronal cell models responsive to NGF, but do not require it for survival and also because it is difficult to obtain large numbers of sympathetic neurons for in vitro studies [39]. Furthermore, in a primary neuron model the development of sympathetic neurons might be critically regulated by two neurotrophins NT3 and NGF, acting through a common receptor TrkA, as reported earlier [41]. For these reasons, whether the results of the present study using PC12 cells can be extrapolated to primary neurons remains to be investigated.

Conclusion
Taken together, our data clearly identify a modulatory role of the MEK1 adapter protein LAMTOR2 in NGF-mediated MAPK activation kinetics and neurite outgrowth induction of PC12 cells. Of note, stimulation with EGF significantly enhanced neither neurite formation nor MAPK activity in our system. The MEK1/ MAPK pathway constitutes an essential element in NGF-mediated differentiation, shown by pharmacological inhibition of MEK1 with PD098059 and siRNA-mediated knockdown of MAPK. Knockdown of LAMTOR2 unexpectedly led to a positive feedback coupling through the upregulation of MAPK activation and to an intensified differentiation process. Based on our own data and data from other studies [9,11], we speculate that knockdown of LAMTOR2 unleashes the NGF receptor signaling pathway from late-endosomal degradation, likely via downregulation of the LAMTOR/Ragulator complex thereby allowing a prolongation of the MAPK signal and increased nuclear entry of active MAPK, which favors the differentiation signal. Figure 7. LAMTOR2 is involved in endosomal trafficking. (A, B) PC12 cells were transfected with control or specific siRNA for LAMTOR2. After 24 h, cells were stimulated with NGF (200 ng/ml) for various time points (0 = control, or 5 and 15 min), fixed and stained for EEA1 (early endosomes), pan-TrkA and Hoechst (nucleus). EEA 1 is shown in red and pan-TrkA in green. Colocalization of EEA1 and pan-TrkA is shown in yellow (arrows). Scale bar = 10 mm. Colocalization tests were performed as described in material and methods section. Colocalization in LAMTOR2 knockdown cells is expressed as fold of control (f.o.c.; control values in control siRNA transfected cells: 0 min 20.9968.55%, 5 min 22.7467.94%, 15 min 24.9469.04%). Values represents the means 6 SEM, n = 3-5. Differences were analyzed using unpaired two-tailed t-test *p,0.05. doi:10.1371/journal.pone.0095863.g007