Molecular dissection of TrkA signal transduction pathways mediating differentiation in human neuroblastoma cells


Activation of the neurotrophin receptor TrkA by its ligand nerve growth factor (NGF) initiates a cascade of signaling events leading to neuronal differentiation in vitro and might play an important role in the differentiation of favorable neuroblastomas (NB) in vivo. To study TrkA signal transduction pathways and their effects on differentiation in NB, we stably expressed wild-type TrkA and five different TrkA mutants in the NGF unresponsive human NB cell line SH-SY5Y. Resulting clones were characterized by TrkA mRNA and protein expression, and by autophosphorylation of the receptor. Introduction of wild-type TrkA restored NGF responsiveness of SH-SY5Y cells, as demonstrated by morphological differentiation, activation of mitogen-activated protein kinases (MAPK) and induction of immediate-early genes. Expression of TrkA in the absence of NGF resulted in growth inhibition of transfectants compared to parental cells, whereas NGF-treatment increased their proliferation rate. Analysis of downstream signal transduction pathways indicated that NGF-induced differentiation was dependent on TrkA kinase activity. Our data suggest that several redundant pathways are present further downstream, but activation of the RAS/MAPK signaling pathway seems to be of major importance for NGF mediated differentiation of NB cells. Our results also show that the signaling effector SH2-B is a substrate of NGF-mediated Trk signaling in NB, whereas it is not activated by NGF in rat pheochromocytoma PC12 cells. This might explain the differences we observed in TrkA signaling between neuroblastoma and PC12 cells. Further insight into TrkA signaling may suggest new options for the treatment of NB.


Neuroblastoma is one of the most common pediatric neoplasms and is derived from the neural crest. Neurotrophic factors and their tyrosine kinase receptors (Trks) play an important role in the pathogenesis, biology and clinical behavior of NB (Brodeur, 1993). Observations from several independent studies suggest that high expression of the neurotrophin receptor TrkA is present in NB with favorable biological features and highly correlated with patient survival (Nakagawara et al., 1993; Kogner et al., 1993). Activation of TrkA by its ligand NGF leads to survival and differentiation of TrkA expressing cells in vitro (Klein et al., 1991) and might play an important role in the regression or differentiation of NB in vivo.

Signal transduction pathways used by TrkA have been studied mainly in the PC12 rat pheochromocytoma cell line. Following NGF binding, TrkA receptors rapidly become phosphorylated on tyrosine residues, and their tyrosine kinase domain is activated (Klein et al., 1991). Phosphorylated tyrosine residues in the TrkA cytoplasmic domain serve as anchors for binding downstream signaling molecules (Schlessinger and Ullrich, 1992). Proteins known to become phosphorylated and activated in response to NGF include phospholipase-Cγ1 (PLC-γ), phosphatidylinositol-3-kinase (PI3K), the adapter protein Shc and the Suc-associated neurotrophic factor-induced tyrosine phosphorylated target (SNT, also called FRS2) (Greene and Kaplan, 1995; Kaplan and Miller, 1997). These proteins couple TrkA to several intracellular signaling pathways like the RAS/MAPK pathway (for a review see Kaplan and Miller, 1997). Activation of MAPK, also known as extracellular signal-regulated kinases (ERKs) (Greene and Kaplan, 1995) is followed by activation of transcription factors and induction of immediate-early genes. NGF-treated PC12 cells cease proliferating, exhibit somatic hypertrophy, acquire neurites, differentiate and show a dependence on NGF for survival in serum-free medium.

Previous work using TrkA mutants defective in association sites for intracellular effector molecules has indicated the existence of several signal transduction pathways used by the neurotrophins to mediate differentiation. However, critical elements of the TrkA signaling pathway may vary in cell lines derived from different tissues. NGF stimulation of TrkA leads to differentiation of neurons, whereas it induces proliferation of fibroblasts and apoptosis of human medulloblastoma cells (Muragaki et al., 1997), further suggesting that the response to TrkA activation by NGF is cell type specific.

As the NGF/TrkA pathway may be playing an important role in the regression or differentiation of NB, we studied TrkA signal transduction and its effects in the context of NB cells. Most established NB cell lines are neither dependent on nor responsive to the presence of NGF in vitro (Matsushima and Bogenmann, 1990; Azar et al., 1990). Introduction of an exogenous TrkA-gene can restore the ability of NB cells to morphologically differentiate in response to NGF, suggesting that one major defect in NB may be that the constitutive TrkA expression level is too low (Matsushima and Bogenmann, 1993; Lavenius et al., 1995). The human SH-SY5Y NB cell line expresses very low levels of the TrkA receptor, but does not respond to NGF by differentiation or by survival in serum-free medium (Lavenius et al., 1995). In this study we expressed exogenous wild-type TrkA (WT-TrkA) and five TrkA mutants defective in different protein association sites in human SH-SY5Y cells to study TrkA signal transduction pathways mediating differentiation in NB.


TrkA expression and autophosphorylation of stable SY5Y transfectants

To generate SH-SY5Y cell clones that stably express wild-type (WT) and mutant TrkA, the cells were transfected with neomycin-selectable retroviral constructs: SY5Yvec (empty vector), SY5Y-WT TrkA (wild-type TrkA), SY5Y-K538N, SY5Y-Y785F, SY5Y-Y490F, SY5Y-ΔKFG and SY5Y-YY785/490FF (Table 1). After G418-selection, single cell clones were characterized with respect to the level of TrkA mRNA and protein expression by RT–PCR and Western blot. TrkA mRNA and protein expression was undetectable in parental SY5Y and SY5Yvec cells. Furthermore, transcripts for TrkB, TrkC, and p75 LNGFR were very low or absent in SY5Y cells (data not shown). However, representative clones of SY5Y cells transfected with wild-type and mutant TrkA demonstrated high expression of exogenous TrkA mRNA (data not shown) and protein (Figure 1a). From each construct we selected three independently derived clones with high levels of TrkA protein expression for further characterizations.

Table 1 Nomenclature of TrkA transfectants
Figure 1

Trk protein expression and autophosphorylation in transfected SH-SY5Y cells. (a) Representative panel of SY5Y single cell clones. Lysates from SY5Y cells, transfected with the indicated TrkA constructs and equalized for total protein, were subjected to SDS–PAGE and Western blotting with C14-anti-pan Trk. SY5Yvec (empty vector) as negative control. The lower band represents unglycosylated TrkA. (b) NGF-induced autophosphorylation of TrkA in representative single cell clones. The indicated cells were serum-starved and treated with (+) or without (−) NGF 100 ng/ml for 9 min. Equal amounts of total protein lysates were immunoprecipitated with C14-anti-pan Trk, resolved by SDS–PAGE and immunoblotted with 4G10 anti-phosphotyrosine

To study TrkA receptor activation induced by NGF treatment, cells were exposed to NGF for 9 min. This resulted in strong autophosphorylation of tyrosine residues in TrkA of all constructs except the kinase-inactive mutant K538N and SY5Yvec (Figure 1b). The blot was stripped and rehybridized with anti-pan Trk to confirm the identity of the phospho-TrkA band and to control for equal loading of immunoprecipitated TrkA (data not shown). The band in untreated Y785F and YY785/490FF cells suggests a low level of autophosphorylation even in the absence of exogenous NGF in these cell types. These findings were consistent in three different single cell clones analysed for each construct. However, there were no morphological alterations in the absence of ligand in Y785F or YY785/490FF cells.

NGF responsiveness of transfectants

To study the morphological response of SY5Y cells expressing TrkA or its mutants to NGF, we treated the transfectants with NGF in medium containing 10% FBS. None of the clones differentiated spontaneously in the absence of NGF. SY5Yvec and SY5Y-K538N cells showed no response to NGF treatment. However, obvious signs of morphological differentiation (such as outgrowth of long stable neurites, formation of cell aggregates and somatic hypertrophy) was seen in all other TrkA transfected clones within 4–6 days of NGF-treatment (Figure 2). The demonstrated differentiation was surprising for SY5Y cells transfected with the KFG mutant or the double mutant YY785/490FF, as cells with these TrkA mutants do not show any signs of differentiation in the PC12 cell system (Stephens et al., 1994; Peng et al, 1995). However, the outgrowth of stable neurites in these two mutants occurred about 3–4 days later and in about 50–70% of the cells, whereas the wild-type as well as the Y490F- and Y785F-single-mutants showed neurites in 80–90% of the cells. None of the differentiated clones showed a uniform morphology, suggesting that only a subpopulation of cells within a cloned cell population are capable of a full morphological response to NGF. Morphological differentiation in response to NGF was stable in all analysed cell clones.

Figure 2

Morphology of wild-type and mutant TrkA transfected SH-SY5Y cells. SY5Y transfectants were cultured in 10% serum medium for 12 days with (+NGF) or without 100 ng/ml NGF. No differentiation could be demonstrated in untreated cells (here SY5Y-WT TrkA as representative example) and in NGF treated SY5Yvec (empty vector) and K538N (kinase-inactive) cells. All other mutants show signs of morphological differentiation after treatment with NGF. Phase contrast microscopy,×320

Effects on proliferation

We compared the growth rate of transfected cells in medium containing 10% serum in the absence of exogenous NGF (Figure 3a). Compared with parental SY5Y cells, TrkA transfected cells showed a severe growth inhibition. This inhibition was evident in wild-type TrkA transfected cells, as well as in cells transfected with the Shc mutant, the KFG mutant and to a certain extent in the double mutant cells, whereas cells transfected with the kinase-inactive mutant and the PLCγ1-mutant showed a growth rate comparable to parental SY5Y cells. Thus, the PLCγ1 binding site seems to be important for the inhibition of proliferation associated with WT TrkA expression. Although morphological differentiation continued after treatment with NGF for more than a week, the cells did not stop proliferating. An MTT-assay revealed no difference in the growth rate of untreated and NGF-treated SY5Yvec cells, whereas a significant increase was demonstrated in the growth rate of NGF-treated SY5Y-WT TrkA cells compared to untreated SY5Y-WT TrkA cells (Figure 3b). The growth curve for SY5Y-K538N was similar to that of SY5Yvec (no difference in proliferation mediated by NGF), whereas the growth curves for SY5Y-Y785F, -Y490F, -ΔKFG and -YY785/490FF were similar to SY5Y-WT TrkA (significant increase in the growth rate in the presence of NGF). We used BrdU-uptake to document that this increase in cell number was due to a higher proliferation rate and not due to a decrease in cell death (data not shown).

Figure 3

(a) Growth rate of transfected SH-SY5Y cells in the absence of NGF. The indicated transfectants were cultured in 10% serum medium in 96 well plates. MTT-Assay was performed after 0, 3, 6, 9 and 12 days for each transfectant. Shown are the representative results of one of three different clones tested. Data points are the mean of triplicates. Standard deviations (SDs) were less than 10%. (b) Effect of NGF on growth of transfectants. SY5Y-WT-TrkA and SY5Yvec as representative examples were grown in 10% serum-medium with (+) or without (−) NGF 100 ng/ml. MTT-Assay was performed on day 0, 3, 6, 9 and 12. Data points are the mean of triplicates. Standard deviations (SDs) were less than 10%. Independent experiments with three different clones showed similar results

Activation of downstream signaling elements

To investigate whether deletions in certain protein association sites (and therefore elimination of effector docking) disrupted the TrkA signaling cascade leading to differentiation, we monitored the tyrosine phosphorylation of several downstream effector proteins known to be involved in NGF signaling. Interaction of PLCγ1 with human Trk requires autophosphorylation of the receptor at tyrosine 785 (Loeb et al., 1992; Marshall, 1995). We confirmed this in our system: as shown in Figure 4a, PLCγ1 is not phosphorylated in K538N, Y785F and YY785/490FF mutated cells, but it is phosphorylated and activated in SY5Y-WT-TrkA as well as the KFG- and Y490F mutated cells.

Figure 4

Activation of downstream signaling elements. The indicated serum-starved cells were treated with (+) or without (−) NGF 100 ng/ml for 9 min. (a) Lysates were equalized for total protein and immunoprecipitated with anti-Shc, anti- PLCγ 1 or SNT-p13 Suc agarose. Immunoprecipitates were analysed by Western blot with 4G10 anti-phosphotyrosine. The three arrows pointing to Shc represent the 46, 52 and 66 kD isoforms. (b) Aliquots of cell lysates (equalized for total protein) were subjected to SDS–PAGE and Western blotting with anti-MAPK (upper panel) as control for equal loading, and anti-active MAPK (lower panel)

The Shc protein, which is involved in the activation of Ras (Pellici et al., 1992), undergoes NGF-promoted TrkA association and tyrosine phosphorylation following autophosphorylation of tyrosine 490 (Loeb et al., 1992). Figure 4 shows that tyrosine phosphorylation of Shc proteins is indeed mediated in all Trk mutants except K538N, Y490F and YY785/490FF.

It has been shown that the KFG motif is necessary for NGF mediated activation of the neuronal specific SNT/FRS2 protein in PC12 cells (Loeb and Greene, 1993). Nonetheless, we were able to show SNT phosphorylation in response to NGF in SY5Y cells transfected with the KFG mutant, as well as in all other mutants except for the kinase-inactive K538N mutant. Therefore, SNT/FRS2 tyrosine phosphorylation was not dependent on association with the KFG motif of TrkA in human NB cells.

The MAP kinases or ERKs are downstream elements of the NGF signaling cascade that are activated by a Ras-dependent pathway (Robbins et al., 1992). To assess the capacity of the Trk mutants to regulate this pathway, we compared NGF-mediated activation of MAPK by Western blotting with an anti-active MAPK-antibody. Figure 4b shows that wild-type TrkA and all mutated TrkA receptors (except the kinase-inactive mutant) mediate activation of MAPK 1 and 2, although the latter (p44) appears to be less activated in SY5Y-YY785/490FF cells. This suggests at least partial activation of the Ras pathway by all mutant receptors, even the double mutant, or alternatively that the MAP kinases can be activated by another mechanism.

SH2-B and APS have been identified recently as substrates of Trk receptors in developing cortical neurons that bind to Grb2 (Qian et al., 1998). As shown in Figure 5a, we detected mRNA expression of SH2-B in parental SY5Y cells and in all TrkA transfectants. We also demonstrated NGF-induced activation of SH2-B in wild-type TrkA transfected cells and all mutants except the kinase-inactive K538N mutant (Figure 5b). In contrast, we could not detect any tyrosine-phosphorylation of SH2-B in NGF-treated PC12 cells, although the TrkA receptor was clearly phosphorylated and activated in these cells (Figure 5b). Thus, activation of the signaling element SH2-B may account for some of the observed differences in Trk signaling between neuroblastoma and PC12 cells. We also detected mRNA expression of human APS in SY5Y cells and all TrkA transfectants, but we were unable to demonstrate protein expression or phosphorylation of APS by Western blot analysis (data not shown).

Figure 5

Expression and activation of SH2-B. (a) Semiquantitative RT–PCR of the indicated cell lines with biotinylated SH2-B specific primers and biotinylated GAPD primers as internal control. High expression of SH2-B mRNA is demonstrated in all cell lines. (b) The indicated transfectants were treated with (+) or without (−) NGF 100 ng/ml for 9 min. Aliquots of cell lysates (equalized for total protein) were subjected to immunoprecipitation with anti-phosphotyrosine, followed by Western blot with anti-SH2-B. This antibody was recommended for Western blot by the manufacturer, but not for immunoprecipitation. The antibody is specific for SH2-B of mouse, rat or human origin. The blot was stripped and reprobed with anti-pan Trk (lower panel) to control for equal loading of immunoprecipitated TrkA and to confirm activation of the Trk receptor in PC12 cells

Induction of immediate early genes

A functional NGF response in neuroblastoma cells is defined as neurite outgrowth and the induction of immediate-early genes like FOS and the NGF inducible (NGFI)-A, -B and -C genes (Matsushima and Bogenmann, 1993; Sheng and Greenberg, 1990). At least two different single cell clones of each TrkA construct were screened by Northern blot for their ability to induce FOS and NGFI-A, -B and -C gene transcripts in response to NGF. Cell clones of SY5Yvec and SY5Y-K538N did not show any induction of immediate early genes in response to NGF. However, clear induction of the immediate-early gene NGFI-A could be demonstrated for all other constructs (Figure 6). In comparison, there was no induction of NGFI-B in the PLCγ1 mutant and the Shc/PLCγ1 double mutant, and just modest induction of NGFI-C in the double mutant. Also FOS was not induced in the double mutant and only moderately in the PLCγ1 mutant. Thus, the PLCγ1 binding site seems to be important for immediate-early gene induction.

Figure 6

Induction of immediate-early gene expression by NGF. Indicated cell lines were serum-starved and treated with (+) or without (−) 100 ng/ml NGF for 40 min prior to isolation of total RNA. Northern blot analysis was carried out using 20 μg RNA per lane. Equal loading was confirmed by ethidium bromide staining of RNA in the agarose gel. Blots were sequentially hybridized with probes for NGFI-A, -B, -C and FOS

Effect of PD 98059 and LY 294002 on differentiation of TrkA transfectants

To test whether NGF mediated differentiation can be disrupted by inhibition of the Ras/MAPK pathway or if other Trk signaling pathways like the PI3K survival pathway are also involved in differentiation, we treated our different SY5Y transfectants with PD 98059 or with LY 294002 along with NGF. PD 98059 is a specific MEK-inhibitor that disrupts the Ras/MAPK signaling pathway (Alessi et al., 1995), whereas LY 294002 selectively inhibits survival pathways mediated via PI3K signaling (Vlahos et al., 1994). The vast majority of PD 98059 treated wild-type and mutant SY5Y-TrkA transfectants failed to show neurites, growing in 10% serum medium containing 100 ng/ml NGF, and no activation of MAPK could be detected in the presence of NGF+PD 98059 (data not shown). However, the number of neurite bearing cells in the same transfectants treated with NGF+LY 294002 or with NGF+DMSO (solvent control) was not reduced in comparison to cells treated with NGF alone (data not shown). Thus, the Ras/MAPK signaling pathway seems to be of major importance for mediating differentiation of neuroblastoma cells.


Transfection of SH-SY5Y cells with TrkA restores NGF responsiveness

We have investigated the role of TrkA and its signaling pathway in cell growth and differentiation of NB cells. Our results demonstrate that transfection of the NB cell line SH-SY5Y with exogenous WT-TrkA restores NGF responsiveness in terms of TrkA autophosphorylation, phosphorylation of downstream signaling molecules, induction of immediate-early genes and outgrowth of stable neurites. These data are in agreement with those reported previously (Lavenius et al., 1995; Poluha et al., 1995), and suggest that the failure of parental SY5Y cells to respond to NGF-treatment is a consequence of low TrkA receptor expression, whereas the TrkA signaling pathway itself seems to be intact in these cells.

Introduction of TrkA does not lead to growth arrest of NGF treated SY5Y cells

Our results demonstrate that introduction of TrkA into SH-SY5Y cells leads to a marked growth inhibition in the absence of NGF, whereas NGF treatment results in increased proliferation. These data are consistent with the previous observation that TrkA-transfected SH-SY5Y cells demonstrated an increased uptake of thymidine after NGF-stimulation, indicating increased mitotic activity (Lavenius et al., 1995). IGF-I- and basic FGF-differentiated SH-SY5Y cells do also continue to proliferate (Lavenius et al., 1994). However, our data are in contrast to two reports about TrkA-transfected NB cells, in which NGF treatment led to dramatic reduction of cell growth in the MYCN amplified NB cell line HTLA230 (Matsushima and Bogenmann, 1993) and in SH-SY5Y cells (Poluha et al., 1995). The presence of MYCN-amplification in Matsushima's study might lead to additional effects on neurotrophin signal transduction. During sympathetic neuronal development in vitro, TrkA expression is preceded by growth arrest, suggesting that differentiation and growth arrest are two distinctly regulated phenomena in nontransformed neuroblasts (Anderson, 1993). The signaling pathway leading to growth arrest of NB cells might also be different and/or independent from the pathway leading to differentiation. Thus it might require additional neuronal cell-specific components that distinguish differentiation from a mitotic response. Moreover, in most cell systems Ras/MAPK signaling is typically associated with survival and proliferation (Marshall, 1995). For instance, TrkA transfected NIH3T3 cells also proliferate in response to NGF. Some effects of TrkA signaling on the proliferation behavior of SH-SY5Y cells might also be due to overexpression of TrkA, as the levels of TrkA expression reached by retroviral transfection of a cell line might exceed biological levels of TrkA expression in NB tumors in vivo. During the propagation of TrkA single cell clones we observed a progressive loss of TrkA expression after several passages. This was accompanied by a gradual decrease in the NGF mediated proliferation advantage and a longer time of NGF treatment (up to 12 days) before morphological signs of differentiation occurred. Thus, the passage number and level of TrkA expression seem to be of importance for proliferation and differentiation studies.

Mutation of certain effector protein association sites still allows differentiation of NB cells in response to NGF

To investigate the role of Trk associating proteins and their intracellular targets in Trk signal transduction pathways, we expressed TrkA receptors encoding mutations in different protein association sites in SY5Y cells.

SY5Y-K538N mutant

 As expected our studies with this kinase-inactive TrkA mutant did not show any phosphorylation of the receptor or other tyrosines.

SY5Y-Y785F, SY5Y-Y490F and SY5Y-double mutants

 The Trk receptor contains two autophosphorylated tyrosines in the cytoplasmic region at positions 490 and 785 of human TrkA. Tyrosine 785 has been shown to be the site for PLC-γ1 interactions (Obermeier et al., 1993; Loeb et al., 1994). PLC-γ1 promotes activation of protein kinase C and the Ras/MAPK cascade (Loeb et al., 1994; Rhee and Choi, 1992). Tyrosine 490 has been identified as the Shc binding site (Stephens et al., 1994). Shc associates with the adapter protein Grb2, which couples receptors to SOS, a regulator of p21 Ras activity (Rozakis-Adcock et al., 1992). It has been suggested that Trk receptors can activate MAPK via either Shc- or PLC-γ1-dependent signaling pathways. Our results indeed indicate that mutation in either the Shc- or PLC-γ1 association site does not prevent NGF-induced differentiation in NB cells. However, a difference in proliferation behavior was demonstrated, as the Y785F mutant did not show any growth inhibition in the absence of NGF in contrast to WT-TrkA and Y490F. Therefore, the PLC-γ1 association site might have an additional function in signaling leading to growth arrest. This might also be reflected by the different pattern of immediate-early gene induction demonstrated in the PLC-γ1 mutant. With respect to the behavior of the double mutant YY785/490FF, our observations are different from those demonstrated in PC12 cells, where receptors containing mutations in both of these sites failed to exhibit stable NGF-mediated neurite outgrowth and MAPK activation (Stephens et al., 1994). In contrast to this, YY785/490FF expressed in SH-SY5Y extended long, stable neurites in response to NGF treatment, and we showed MAPK activation and induction of immediate-early genes NGFI-A and NGFI-C after NGF stimulation. However, the percentage of differentiated cells was markedly less compared to WT-TrkA and the single mutants Y490F and Y785F. The different pattern of immediate-early gene induction might also reflect some restrictions in differentiation signaling in the simultaneous absence of Shc- and PLC-γ1 activity. Thus, the Ras/MAPK pathway certainly plays a major role in TrkA mediated differentiation signaling in NB cells, but additional Ras-dependent and/or -independent pathways leading to differentiation may exist. The discrepancy in results of functional analysis of mutated Trk receptors in PC12 compared to NB cells might also be due to a different level of Trk receptor expression. However, it is likely that other differences in the TrkA signaling pathways of human NB and rat PC12 pheochromocytoma cells (like activation of SH2-B) account for the discrepancies in the results studying this mutant.

SY5Y-ΔKFG mutant

 The three-residue juxtamembrane sequence KFG has been shown to mediate phosphorylation of SNT/FRS2 in PC12 cells (Peng et al., 1995). In PC12 cells, mutation of KFG leads to prevention of neuritogenesis, but promotes survival of the cells and activation of MAPKs. Surprisingly, NGF-treatment of the ΔKFG mutant introduced into SH-SY5Y cells leads to SNT-phosphorylation, indicating that KFG cannot be the only binding site for SNT in NB cells. Corresponding to SNT activation, we demonstrated neuritogenesis and induction of immediate-early genes in KFG-mutated cells, which is also different than the published behavior of this mutant in PC12 cells. Our data are in agreement with the most recently published study of Meakin et al. (1999) who demonstrated that the Shc binding site of human TrkA (Y490F) is at least one additional binding site for human SNT/FRS-2. The recently published characterization of SNT/FRS2 reveals that SNT binds to the adapter protein Grb2/SOS, which links receptor tyrosine kinases with the Ras pathway (Kouhara et al., 1997). Thus SNT might be an additional pathway leading to Ras activation.

Activation of SH-2B may account for differences in Trk signaling between NB and PC12 cells

Most recently two novel substrates of Trk receptors, rAPS and SH2-B, have been identified and characterized in developing cortical neurons (Qian et al., 1998). Both are closely related Src homolog 2 (SH2) and pleckstrin-homology (PH) domain-containing signaling molecules that bind to Grb2. They are sufficient to mediate NGF induced activation of Ras/MAPK, morphological differentiation and survival, when they are transfected into NGF-unresponsive PC12nnr5 cells together with the Trk mutant F8 (defective in differentiation signaling) (Qian et al., 1998). In our study, we demonstrated that NGF-mediated activation of TrkA leads to tyrosine-phosphorylation of SH2-B in wild-type and all our mutant TrkA transfected SY5Y cells, whereas no activation of SH2-B could be detected in NGF treated PC12 cells. Our data are in agreement with Qian and colleagues (Qian et al., 1998), who also did not detect tyrosine phosphorylated rAPS or SH2-B in extracts from NGF treated PC12 cells, suggesting that PC12 cells do not express rAPS or SH2-B. This might explain the differences in TrkA signaling observed between PC12 and NB cells. As SH2-B was also demonstrated to be a substrate of TrkA in sympathetic neurons, NB cells might more closely resemble signaling pathways of sympathetic neurons than PC12 cells. We failed to detect tyrosine-phosphorylation or protein expression of human APS in our transfectants, although we could detect APS expression on mRNA levels by RT–PCR (data not shown). This might be due to the antibody we used. On the other hand, the data of Qian et al. (1998) indicates that rAPS and SH2-B are substrates of TrkB and TrkC in embryonic cortical neurons, but only SH2-B was demonstrated to be a substrate of TrkA in neonatal sympathetic neurons. Thus, APS might not be a good substrate of TrkA signaling in NB.

The Ras/MAPK pathway is the main signaling pathway leading to differentiation of neuroblastoma cells

The use of the specific MEK-inhibitor PD 98059 in our study demonstrates that the Ras/MAPK signaling pathway is of major importance to mediate NGF-induced differentiation, whereas inhibition of the PI3K survival pathway has no influence on differentiation signaling. However, a small part of each cell population (WT-TrkA and all mutants) treated with NGF+PD 98059 still formed cell aggregates and about 3–7% of the cells demonstrated outgrowth of stable neurites independently of MAPK activation. This might be due to incomplete inhibition of the Ras signaling pathway by PD 98059, as it has been described before (Alessi et al., 1995). Alternatively, a small percentage of cells might be able to use additional, MAPK independent pathways leading to differentiation.

Taken together our results indicate that the promotion of neurite outgrowth and proliferation in NB cells is a complex, cooperative process that results from the stimulation of several different signaling pathways that can at least in part be compensated by each other. The pathways in human NB cells might be different from Trk signaling pathways in rat PC12 cells, due to either cell type or species differences. Activation of SH2-B is one mechanism that may account for these differences. Thus, we provided evidence that different signaling elements might be used in both systems. The Ras/MAPK pathway seems to be the major signaling pathway for mediating NGF-induced differentiation of NB cells. With the NGF-responsive wild-type and mutant TrkA-transfected clones presented here, we have a model that allows the detailed study of NGF-induced signal transduction mechanisms in a human NB cell system. Furthermore, the detailed comparison of TrkA- and TrkB-mediated signaling appears to be especially important in understanding the molecular events responsible for the different biological behavior of favorable and unfavorable human NB. A more detailed insight into the mechanisms regulating differentiation and proliferation might suggest new options for the treatment of NB.

Material and methods

Cell culture and transfection

SH-SY5Y is a neuronal clone from the NB cell line SK-N-SH and has been described previously (Biedler et al., 1973). WT-TrkA and the TrkA-mutants (Table 1) were a generous gift from David Kaplan and were cloned into the retroviral expression vector pLNCX (Miller and Rosman, 1989; Stephens et al., 1994). The cells were grown under standard conditions (Nakagawara et al., 1993). For differentiation studies NGF-treated (every 2 days 100 ng/ml murine NGF, 2.5S, Promega, Madison, WI, USA) and untreated cultures were scored after 12 days by determining the presence of neurites three times the length of the cell body on cells chosen at random (n=300). Reported values represent the mean values of triplicate experiments conducted with each clonal line, and 2–3 clonal lines of the wild-type and each mutant were studied. Representative fields of cells were photographed under phase contrast microscopy. All pLNCX-TrkA constructs were transfected into the packaging cell line Bing by electroporation as described (Greene et al., 1997). Virus-containing supernatant (10 ml) from these cells added to LipoTaxi (Stratagene, La Jolla, CA, USA) was used to infect 70% confluent SH-SY5Y cell cultures (Greene et al., 1997). Transfected cells were selected with 500 μg/ml G418 (Sigma, St Louis, MO, USA) and subcloned by limited dilution to obtain clonal lines derived from single cells. As negative controls, SY5Y cells were also infected with a retrovirus bearing the pLNCX empty retroviral vector (SY5Yvec). The identity of all mutants was confirmed by sequencing after transfection.

Immunoprecipitation and protein biochemistry

Cell cultures were treated with NGF (100 ng/ml) after starving for 3 h in serum-free medium. After indicated treatment times, cells were rapidly lysed in 800 μl/dish Nonidet P-40 (NP 40) lysis buffer (1% NP40, 20 mM Tris pH 8.0, 137 mM NaCl, 0.5 mM EDTA, 10% glycerol, 1 mM phenylmethylsulphonyl fluoride, 0.15 U/ml aprotinin, 20 μM leupeptin, 1 mM sodium vanadate). Samples normalized for total protein content were immunoprecipitated with the indicated antibodies overnight at 4°C and precipitates were collected with Protein-A- or Protein-G-Sepharose (Gibco–BRL). The samples were separated by SDS–PAGE, electroblotted onto nitrocellulose and immunostained. Anti-pan Trk C14, Anti-MAPK, Anti-SH2-B, Anti-APS (all from Santa Cruz Biotechnology, CA, USA). Anti-phosphotyrosine, Anti-PLCγ1 and Anti-Shc (all from UBI, Lake Placid, NY, USA) were used as described (Loeb et al., 1992; Peng et al., 1995). Detection of immunocomplexes was conducted using an ECL chemiluminescence system (Amersham Corp., Arlington Heights, IL, USA). For SNT precipitations, 20 μl of p13sucl-agarose (UBI) was added to 400 μg total protein lysate of cells and incubated at 4°C as described (Stephens et al., 1994).

Northern blotting

After starving in serum-free medium for 6 h, total RNA of untreated cells (control) or cells treated with NGF for 40 min at 37°C was extracted according to the method of Chomczynski (Chomczynski and Sacchi, 1987). The purified RNA (20 μg) was separated by electrophoresis through 1% agarose gels and transferred to nylon membranes (Hybond N+, Amersham). Blots were UV cross-linked, hybridized with 32P-dCTP-random-primer-labeled probes to FOS, NGFI-A, -B and -C and exposed to BiomaxMS film (Kodak).


Expression of TrkA, SH2-B and APS was analysed using specific biotinylated primers (sequences available upon request). Total RNA was reverse transcribed and amplified for 20 cycles on a PTC-100 Programmable Thermal Controller (MJ Research, Inc., Watertown, MA, USA) using the Superscript amplification system (Gibco). The PCR products were run on a 6% polyacrylamide gel and transferred to a nylon membrane (Hybond N+, Amersham, IL, USA). Biotinylated signals were detected using the Southern Light Detection system (Tropix, MA, USA) and exposed to X-ray film. Target gene expression (TrkA, SH2-B or APS) was normalized to the coamplified housekeeping geneGAPD (Eggert et al., 2000).

BrdU incorporation

Transfected cells were plated on Lab-Tek chamber slides (Fisher) and grown in 10% serum containing RPMI medium. Untreated and NGF-treated cells were cultured for 4, 8 and 12 days and treated with 10 μM bromodeoxyuridine (BrdU, Boehringer Mannheim) for 5 h to assess mitotic activity. Cells were counterstained with DAPI and the percentage of BrdU positive cells was counted in a total cell number of 400 cells/slide. Each experiment was done in triplicate with three different independently derived clones.

Cell proliferation assay

Transfected cell clones were seeded into 96 well plates at a density of 5×103 cells per well and cultured in 10% serum containing RPMI medium in the presence or absence of NGF (100 ng/ml). Each condition was performed in triplicate. Medium was replaced every 2 days and cultures were maintained for 3, 6, 9 and 12 days. A colorimetric 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) assay was then performed as described (Mosmann, 1983). A multiwell scanner was used to measure the absorbance at 570 nm.

Inhibition assay with PD 98059 and LY 294002

Cells were grown in 10% serum medium and treated every day with the specific MEK-Inhibitor PD 98059 (100 μM, New England Biolabs) or the specific PI3K inhibitor LY 294002 (100 μM, Sigma) 1 h prior to addition of NGF 100 ng/ml as described (Alessi et al., 1995; Vlahos et al., 1994). Control cells were treated with NGF alone or NGF+DMSO to exclude an inhibitory effect of the inhibitor solvent DMSO. The number of neurite bearing cells was counted as described.





tyrosine kinase (receptor)


nerve growth factor


mitogen-activated protein kinase


phospholipase γ1




Suc-associated neurotrophic factor-induced tyrosine phosphorylated target


extracellular signal-regulated kinase






reverse transcriptase


3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide




low affinity nerve growth factor receptor


Growth factor receptor-bound protein 2


protein encoded by the son-of-sevenless gene


enhanced chemiluminescence


  1. Alessi DR, Cuenda A, Cohen P, Dudley DT and Saltiel AR . 1995 J Biol Chem 270: 27489–27494

  2. Anderson DJ . 1993 Annu Rev Neurosci 16: 129–158

  3. Azar CG, Scavarda NJ, Reynolds CP and Brodeur GM . 1990 Cell Growth Differ 1: 421–428

  4. Biedler JL, Helson L and Spengler BA . 1973 Cancer Res 33: 2643–2649

  5. Brodeur GM . 1993 J Natl Cancer Inst 85: 344–345

  6. Chomczynski P and Sacchi N . 1987 Anal Biochem 156–159

  7. Eggert A, Brodeur GM and Ikegaki N . 2000 Biotechniques 28: In press

  8. Greene LA, Cunningham ME, Farinelli SE and Park DS . 1997 Culturing Nerve Cells Banker G and Goslin K eds MIT Press Cambridge, MA

  9. Greene LA and Kaplan DR . 1995 Curr Opin Neurobiol 5: 579–587

  10. Kaplan DR and Miller FD . 1997 Curr Opin Cell Biol 9: 213–221

  11. Kaplan DR and Stephens RM . 1994 J Neurobiol 25: 1404–1417

  12. Klein R, Jing S, Nanduir V, O'Rourke E and Barbacid M . 1991 Cell 65: 189–197

  13. Kogner P, Barbany G, Dominici C, Castello M, Raschella G and Persson H . 1993 Cancer Res 53: 2044–2050

  14. Kouhara H, Hadari YR, Spivak-Kroizman T, Schilling J, Bar-Sagi D, Lax I and Schlessinger J . 1997 Cell 89: 693–702

  15. Lavenius E, Gestblom C, Johansson I, Nanberg E and Pahlmann S . 1995 Cell Growth Differ 6: 727–736

  16. Lavenius E, Parrow V, Nanberg E and Pahlman S . 1994 Growth Factors 10: 29–39

  17. Loeb DM and Greene LA . 1993 J Neurosci 13: 2919–2929

  18. Loeb DM, Stephens RM, Copeland T, Kaplan DR and Greene LA . 1994 J Biol Chem 269: 8901–8910

  19. Loeb DM, Tsao H, Lobb MH and Greene LA . 1992 Neuron 9: 1053–1065

  20. Marshall CJ . 1995 Cell 80: 179–185

  21. Matsushima H and Bogenmann E . 1990 Mol Cell Biol 10: 5015–5020

  22. Matsushima H and Bogenmann E . 1993 Mol Cell Biol 13: 7447–7456

  23. Meakin SO, MacDonald JI, Gryz EA, Kubu CJ and Verdi JM . 1999 J Biol Chem 274: 9861–9870

  24. Miller AD and Rosman GJ . 1989 Biotechniques 7: 980–982

  25. Mosmann T . 1983 J Immunol Methods 65: 55–63

  26. Muragaki Y, Chou TT, Kaplan DR, Trojanowski JQ and Lee VM-Y . 1997 J Neurosci 17: 530–542

  27. Nakagawara A, Arima-Nakagawara M, Scavarda NJ, Azar CG, Cantor AB and Brodeur GM . 1993 N Engl J Med 328: 847–854

  28. Obermeier A, Halfter H, Wiesmueller KH, Jung J, Schlessinger J and Ullrich A . 1993 EMBO J 12: 933–941

  29. Pellici G, Lanfrancone L, Grignani F, McGlade J, Carallo F, Forni G, Nicoletti I, Grignani F, Pawson T and Pelicci PG . 1992 Cell 70: 93–104

  30. Peng X, Greene LA, Kaplan DR and Stephens RM . 1995 Neuron 5: 395–406

  31. Poluha W, Poluha DK and Ross AH . 1995 Oncogene 10: 185–189

  32. Qian X, Riccio A, Zhang Y and Ginty DD . 1998 Neuron 21: 1017–1029

  33. Rhee SG and Choi KD . 1992 J Biol Chem 267: 12393–12396

  34. Robbins DJ, Cheng M, Zhen E, Vanderbilt CA, Feig LA and Cobb MH . 1992 Proc Natl Acad Sci USA 89: 6924–6928

  35. Rozakis-Adcock M, McGlade J, Mbamalu G, Pelicci G, Daly R, Li W, Batzer A, Thomas S, Brugge J, Pelicci PG et al. 1992 Nature 360: 689–692

  36. Schlessinger J and Ullrich A . 1992 Neuron 9: 383–391

  37. Sheng M and Greenberg ME . 1990 Neuron 4: 477–485

  38. Stephens RM, Loeb DM, Copeland TD, Pawson T, Greene LA and Kaplan DR . 1994 Neuron 12: 1–20

  39. Vlahos CJ, Matter WF, Hui KY and Brown RF . 1994 J Biol Chem 269: 5241–5248

Download references


We thank Dr David Kaplan for the TrkA mutants and Dr John Maris for discussing the manuscript. This work was supported by Grants from the Deutsche Krebshilfe/Dr Mildred Scheel Stiftung (A Eggert), the National Institutes of Health Grant NS 34514 (GM Brodeur), and the Audrey E Evans Endowed Chair (GM Brodeur).

Author information



Corresponding author

Correspondence to Garrett M Brodeur.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Eggert, A., Ikegaki, N., Liu, X. et al. Molecular dissection of TrkA signal transduction pathways mediating differentiation in human neuroblastoma cells. Oncogene 19, 2043–2051 (2000).

Download citation


  • neuroblastoma
  • tyrosine kinase receptors
  • neurotrophins
  • signal transduction
  • differentiation

Further reading