TrkA as a Life and Death Receptor: Receptor Dose as a Mediator of Function¹ Free

NGF³ is the most well-studied representative of a family of trophic factors that includes brain-derived neurotrophic factor, neurotrophin-3, and neurotrophin 4/5. It has been known to have pleiotropic effects on different cell types (1, 2, 3, 4, 5). Initially, NGF was shown to promote the proliferation, survival, and maturation of neural target cells (6, 7, 8, 9). Cell death is generally thought to be a consequence of NGF withdrawal, rather than NGF exposure (10, 11, 12).

In recent years, however, it has been demonstrated that the biological effects of NGF are mediated by two classes of receptors: (a) p75 glycoprotein, which belongs to the superfamily of tumor necrosis factor receptors; and (b) TrkA glycoprotein, a transmembrane tyrosine kinase of 140 kDa with a cytoplasmic tyrosine kinase domain (13). The binding of NGF to each of these receptors triggers particular cascades of cellular signaling events. The binding of NGF to TrkA, the high-affinity NGF receptor, results in TrkA phosphorylation that will in turn lead to a scaffolding role for TrkA and recruitment of several adapter proteins and enzymes that ultimately propagate the NGF signal (14, 15). Among these proteins, the adapter protein Shc and phospholipase C have been involved in the activation of ERKs (16, 17). In contrast, the binding of NGF to its low-affinity receptor, p75, has variably been found to induce or prevent apoptosis. p75 has been shown to activate programmed cell death through a mechanism involving the stress kinase JNK (18, 19, 20). However, this receptor also activates NF-κB, which is thought to promote survival and counterbalance the proapoptotic signal through up-regulation of TRAF1, TRAF2, and the inhibitor of apoptosis proteins (e.g., c-IAP1 and c-IAP2) and blocking of activation of the caspase pathway (21, 22, 23, 24, 25).

Although the roles of NGF receptors in transducing signals are beginning to be elucidated, because the expression of TrkA and p75 is cell type specific, cell line specific, and even cell cycle specific, the precise effects and signaling mechanisms of ligand activation of NGF receptors are still the subject of considerable investigation and controversy. For example, the role of NGF in the induction and progression or maturation and regression of neural tumors remains speculative. In present study, we used native and trkA-transfected PC12 pheochromocytoma cells to examine the effects on cell survival of TrkA and p75 ligand activation by NGF in cells of varied TrkA content and to explore the signaling mechanisms triggered by binding of NGF to its receptors in these cells.

NGF was obtained from Boehringer Mannheim. Preparation and characterization of the monoclonal antibody mAbNGF30 have been described previously (26). This antibody binds to the p75 binding site of NGF, blocking the binding of NGF to p75 and converting NGF to a TrkA-specific ligand. K252α, a TrkA tyrosine kinase inhibitor, was purchased from Calbiochem (San Diego, CA) and prepared in DMSO as a 100 μm stock solution. Anti-TrkA and anti-phosphotyrosine (clone 4G10) antibodies were obtained from Upstate Biotechnology (Lake Placid, NY). Mouse monoclonal anti-MEK kinase1 IgG, rabbit polyclonal anti-MEK1, rabbit polyclonal anti-ERK1, mouse monoclonal anti-phospho-ERK, rabbit polyclonal anti-MEK4, goat polyclonal anti-p-CREB, rabbit polyclonal anti-MEK3, mouse monoclonal anti-phospho-MEK3/6 IgG, mouse monoclonal anti-p38, rabbit polyclonal anti-NF-κB p65, rabbit polyclonal anti-c-IAP2, and rabbit antimouse IgG-horseradish peroxidase antibodies were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Rabbit polyclonal anti-phospho-p38 MAPK, rabbit polyclonal anti-phospho-MEK1/2, and rabbit polyclonal anti-phospho-SEK1/MKK4 antibodies were purchased from Cell Signaling Technology (Beverly, MA). Cy2-conjugated AffiniPure rabbit antimouse IgG(H+L) and Cy3-conjugated AffiniPure goat antirabbit IgG(H+L) antibodies were obtained from Jackson ImmunoResearch Laboratories, Inc. (West Grove, PA) and used for fluorescent immunostaining of cells. 7-AAD, PI, Hoechst dye 33342, saponin, and LDH assay kits were purchased from Sigma (St. Louis, MO). N2 supplement (100×) was obtained from Invitrogen (Carlsbad, CA).

Native [∼40,000 p75 receptors/cell and ∼400 TrkA receptors/cell (p75+++ TrkA+)] and trkA-transfected [∼40,000 of each receptor/cell (p75+++ TrkA+++)] PC12 cells were the kind gift of Dr. H. Uri Saragovi (26). All cells used in these studies were demonstrated to be Mycoplasma free using a MycoTect Kit (Life Technologies, Inc.). Both native and trkA-transfected PC12 cells were maintained with regular medium [DMEM supplemented with 10% horse serum, 5% fetal bovine serum (Atlanta Biologicals, Norcross, GA), and 1.1% penicillin/streptomycin (Invitrogen)] for at least 2 days, followed by replacement of regular medium with serum-free medium [DMEM supplemented with 1% 100× N2 supplement solution (insulin, 500 μg/ml; human transferrin, 10,000 μg/ml; progesterone, 0.63 μg/ml; putrescine, 1,611 μg/ml; selenite, 0.52 μg/ml)]. After 2 days of incubation with serum-free medium, the cells were treated and examined as indicated below. G418 [5 mg/ml; 0.4% (v/v)] was used as selection antibiotic for trkA-transfected PC12 cells throughout the experimental period. The cells were fed twice weekly and periodically examined by immunofluorescence staining to ensure maintenance of the high level of expression of the TrkA receptor in trkA-transfected PC12 cells.

Cells were grown in 6-well tissue culture plates (BioFlex coated with collagen I; Flexcell International Corp., Hillsborough, NC). After reaching 80% confluence, the cells were washed three times with PBS, fixed in 2% paraformaldehyde for 20 min, and washed three more times with PBS. The cells were then permeabilized by incubation in 5% nonfat dry milk containing 1% Triton X-100 for 30 min, followed by three PBS washes. After 2 h of incubation with rabbit polyclonal anti-TrkA or mouse monoclonal anti-p75 antibody at 37°C, the cells were washed another three times with PBS and incubated for 2 h with a 1:500 dilution of Cy3-conjugated AffiniPure goat antirabbit IgG or Cy2-conjugated AffiniPure rabbit antimouse IgG(H+L) antibodies, followed by three additional PBS washes. The membranes were removed from the wells and placed onto glass microscope slides. A Zeiss light microscope equipped for epifluorescence illumination was used to examine and photograph these preparations.

Cell injury was estimated morphologically by phase-contrast light microscopy and quantitated by adherent cell count and spectrophotometric measurement of LDH release. Adherent cell number was determined in vehicle- and NGF (2 nm)-treated cell cultures, as we have described previously for neuroblastoma cells (7, 27, 28). Briefly, adherent cells were manually counted in each of four high-power fields from each cell culture well. Results are expressed as the mean ± SE of the four determinations. Measurement of LDH activity was performed using a LDH assay kit (Sigma) according to the manufacturer’s instructions. Values were normalized to LDH measurements on vehicle-treated sister cultures. Each experiment involving determination of LDH was performed on sister cultures from single plating. Results represent the findings from four such independent experiments.

Apoptotic cells were quantified by flow cytometry considering 7-AAD staining intensity to be proportional to the DNA content (29). In short, after harvesting, the cells were washed once in PBS and once in PBS/0.05% saponin, followed by the addition of 4 μg of 7-AAD in 1 ml PBS/saponin to the samples. The cells were incubated at room temperature in the dark for 30–60 min, and DNA histograms were obtained using a CellQuest apparatus and CellQuest software (Becton Dickinson). Data on 10⁴ cells were collected. Electronic gates were set for viable and apoptotic cells with 2–4 n DNA and subnormal DNA contents, respectively, and for exclusion of debris. The percentage of apoptosis was calculated as (number of apoptotic cells/number of total cells) × 100.

Sister cultures of PC12 cells were treated with vehicle, NGF (2 nm) alone or a preincubated (15 min) mixture of mAbNGF30 and NGF (mAbNGF30:NGF = 2:1; final NGF = 2 nm) for 48–72 h. For the experiments with K252α, the cells were pretreated with 100 nm K252α for 1 h before the addition of NGF. After the treatment, the cells were observed under the microscope, and representative photomicrographs were taken to demonstrate cell morphology. Then, the cells were harvested, washed once in PBS and once in PBS/0.05% saponin, and incubated with 7-AAD (2 mg/ml) for 30–60 min in the dark. Flow cytometric analysis was performed as described above.

For the TUNEL test, DNA ends were labeled using an In Situ Cell Death Detection Kit (Roche). Briefly, monolayers of native and trkA-transfected PC12 cells grown on 6-well BioFlex tissue culture plates (Flexcell International Corp.) were treated with 2 nm NGF for 0, 6, 12, 18, and 24 h, respectively, followed by two washes with PBS. Then the cells were fixed with 4% paraformaldehyde in PBS at 15–25°C for 1 h and incubated with permeabilization solution (0.1% Triton X-100 in 0.1% sodium citrate) for 2 min on ice. The cells were washed twice with PBS, incubated in 50 μl of TUNEL reaction mixture for 1 h at 37°C, and washed twice with PBS. DNA fragmentation was observed under a Zeiss light microscope equipped with epifluorescence illumination.

Intracellular soluble DNA fragmentation analysis was performed as follows. Cells plated in 75-cm² flasks were washed twice with PBS and lysed in cell lysis solution [2% Triton X-100, 20 mm EDTA (pH 8.0), 50 mm Tris-HCl (pH 7.5), 5 × 10⁶ cells/50 μl] for 10 min. After a 2-min centrifugation (6,000 rpm, 4°C), the supernatant was incubated with 1% SDS and 5 μg/ml RNase A for 2 h at 37°C. Then, 2.5 μg/ml (final concentration) proteinase K was added, and incubation was continued for another 2 h. The preparation was precipitated overnight with 0.5 volume of 10 m ammonium acetate and 2.5 volumes of ethanol at −20°C and centrifuged at 12,000 × g (4°C) for 30 min. The pellet was air dried, washed in 70% ethanol, and resuspended in 20 μl of TE buffer (10 mm Tris and 1 mm EDTA). Aliquots of the suspension containing equal cell number equivalents were subjected to agarose gel electrophoresis.

The cells grown on 6-well BioFlex tissue culture plates were treated with either 2 nm NGF, 2 nm NGF + 4 nm mAbNGF30, or 2 nm NGF +100 nm K252α as described above. Forty-eight h later, the cells were harvested, washed once with PBS, and stained with PI (20 ng/ml) for 5 min at room temperature. After washing with PBS, the cells stained with PI were subsequently stained with bisbenzimide (Hoechst dye 33342; 25 μg/ml) for 5 min, fixed with 2% paraformaldehyde for 10 min, and washed with PBS three times. A Zeiss light microscope equipped with epifluorescence illumination was used for all observations.

At the indicated time points after incubation with NGF (2 nm), native and trkA-transfected PC12 cells were lysed in radioimmunoprecipitation assay buffer [10 mm Tris (pH 8), 150 mm NaCl, 0.1% NP40, 0.5% sodium deoxycholate, 0.1% SDS, 1 mm phenylmethylsulfonyl fluoride, 4 μg/ml aprotinin, and 1 mm sodium orthovanadate]. Subsequently, the protein concentrations of the lysates were estimated using the Bio-Rad Protein Assay (Bio-Rad Laboratories, Richmond, CA) with BSA as a standard. An aliquot of each lysate containing 150 mg of protein was loaded onto each lane and electrophoresed on a 10% SDS-polyacrylamide gel, followed by blotting onto a nitrocellulose membrane (Bio-Rad Laboratories). After blotting, nonspecific binding was blocked with 5% nonfat dry milk in PBS, and the membrane was incubated with the appropriate primary antibody, diluted in 5% nonfat dry milk in PBS at 20°C for 2 h, and then washed and incubated with horseradish peroxidase-conjugated secondary antibodies for 1 h. The membrane was finally washed and developed with Western Blotting Chemiluminescence Luminol Reagent (Santa Cruz Biotechnology) following the manufacturer’s instructions. In all cases, the blots stained for TrkA, MEK1, ERK, MEK3/6, p38, and MEK4 were stripped using standard methods and restained for their corresponding phosphorylated proteins, respectively. For experiments involving NGF + K252α, 100 nm K252α was administered to the cells 1 h before NGF treatment. Nonimmune serum was used as negative control, and optical scanning of the films was performed using a Bio-Rad Optiscan optical scanner.

For immunohistochemical staining, PC12 cells were grown in N2 supplement-containing serum-free medium on 6-well tissue culture plates for at least 2 days. The cells were incubated with 2 nm NGF for 0, 5, or 15 min, respectively; washed three times with PBS; fixed in 95% ethanol for 20 min; and washed three more times with PBS. The cells were then permeabilized by incubation in 2% Triton X-100 for 30 min followed by three PBS washes. After blocking of nonspecific staining (0.5% BSA + 2% horse serum in PBS; 10 min), the cells were incubated with anti-p65 antibody for 2 h at 37°C and then incubated with a 1:100 dilution of goat antirabbit antibody-Cy3 immunoconjugate followed by three additional washes (10 min each) with PBS. The membranes were removed from the wells and placed onto glass microscope slides to examine the translocation of the NF-κB p65 component from the cytoplasm to the nucleus using a Zeiss light microscope equipped for epifluorescence illumination. Nonspecific immunostaining (negative control) was assessed by subjecting sister cultures to the same procedure minus treatment with the primary antibody.

For studies involving the comparison of multiple samples, statistical significance was assessed by one-way ANOVA followed by Fisher’s protected least significant difference test. For studies involving the comparison of paired samples, statistical significance was assessed by Student’s (paired) t test. In all cases, P < 0.05 was considered to be indicative of statistical significance.

The present studies use native (p75:TrkA = 100:1) and trkA-transfected (p75:TrkA = 100:100) PC12 cells, respectively, to examine the response of these cells to NGF alone. Fig. 1 confirms the differential expression level of TrkA protein (Fig. 1, A and C) and similar expression level of p75 protein (Fig. 1, B and D) in native (Fig. 1, A and B) and trkA-transfected (Fig. 1, C and D) PC12 cells.

When maintained in N2-supplemented serum-free medium, native PC12 cells are adherent and stellate (Fig. 2,A, a). In contrast, trkA-transfected PC12 cells are adherent and triangular (Fig. 2,A, c). After 48 h of treatment with NGF (2 nm), the morphology of native PC12 cells remains unchanged (Fig. 2,A, b), but trkA-transfected PC12 cells start to shrink, round up, and detach from the culture surface (see arrows in Fig. 2 A, d).

To quantify cell injury, the effects of NGF on cell membrane integrity and cell adhesion were examined for native and trkA-transfected PC12 cells after 48 h of NGF treatment. As shown in Fig. 2, B and C, NGF induced a decrease in the number of adherent cells and a concomitant LDH release from trkA-transfected PC12 cells. Neither of these was observed in native PC12 cells.

PC12 cells undergo apoptosis in response to a number of diverse insults (30). We have demonstrated a correlation between loss of cell adhesion and apoptosis induction in neural crest cells (31). We therefore determined whether or not the NGF-induced death of trkA-transfected PC12 cells is accompanied by the classical features of apoptosis. As demonstrated in Fig. 3 A, a time-dependent increase in the fraction of the cells with sub-2 n DNA content was found after NGF treatment. However, the caspase 3 inhibitor AcDEVD-CHO did not inhibit NGF-induced cell death.

DNA laddering, another hallmark of apoptosis, is thought to result from the activation of an endonuclease capable of cleaving nuclear DNA at internucleosomal sites (32). As displayed in Fig. 3,B, no DNA ladder pattern was seen at any time point after treatment in either vehicle- or NGF-treated trkA-transfected PC12 cells. This result is completely in agreement with our observations using the TUNEL assay (Fig. 3 C). Cells treated with vehicle or 2 nm NGF for varying periods demonstrate no NGF- or time-dependent change in TUNEL staining. For each of the five experimental conditions depicted, only one or two nuclei were stained per ×200 field.

To further characterize cell death in this model, we performed nuclear double staining with both PI and Hoechst dye 33342 (Fig. 3 D). Typically, necrotic cells stain with PI; normal and apoptotic cells do not. All nuclei stain with Hoechst dye, and each gives a characteristic nuclear morphology upon staining.

NGF treatment (6, 12, 18, and 24 h) did not significantly increase the number of PI-stained nuclei relative to vehicle treatment (P > 0.05), suggesting that neither necrotic cell death nor oncosis accounts for the cell death induced by NGF treatment. No nuclear fragmentation was seen after NGF treatment. However, a statistically significant, time-dependent increase was seen in the number of abnormally small nuclei stained with Hoechst dye alone. These findings suggest that trkA-transfected PC12 cells undergo atypical apoptosis (i.e., programmed cell death without all of the typical morphological and biochemical features of apoptosis) after NGF treatment.

We next hypothesized that ligand activation of the TrkA receptor is responsible for induction of cell death in trkA-transfected PC12 cells. To test this hypothesis, we used K252α to determine whether the NGF-TrkA pathway is involved in the mechanism by which NGF induces cell death. K252α has been reported to selectively inhibit TrkA tyrosine phosphorylation (33). Sister-cultured cells were pretreated with vehicle or K252α (100 nm) for 1 h before NGF treatment. K252α reduced the fraction of dead trkA-transfected PC12 cells after NGF treatment (NGF, 57.4%; NGF + K252α, 23.2%). No change was seen in native PC12 cells after NGF treatment (NGF, 11.5%; NGF + K252α, 11.4%; Fig. 4, A and B). These data suggest that ligand activation of the TrkA receptor is involved in NGF-induced death of trkA-transfected PC12 cells.

Similarly, we used a competitive inhibitor of NGF p75 binding (26), mAbNGF30, to determine whether the NGF-p75 pathway is involved in the pathway by which NGF induces cell death. As shown in Fig. 4, A and B, mAbNGF30 does not alter the effect of NGF on trkA-transfected PC12 cells (NGF, 57.4%; NGF + mAbNGF30, 56.2%), implying that ligand activation of the p75 receptor is not involved in NGF-induced death of trkA-transfected PC12 cells.

Several signaling pathways are triggered by the binding of NGF to TrkA. These pathways include NGF/TrkA-MEK1-ERK1/2-p-CREB, NGF/TrkA-MEK3/6-p38 MAP, and NGF/TrkA-MEK4-JNK-caspase 3. Among these pathways, NGF/TrkA-MEK1-ERK1/2-p-CREB has been recognized as a pathway that facilitates cell survival and growth (34, 35, 36). Here, we compared activation of the TrkA-MEK1-ERK1/2-p-CREB signaling pathway by NGF in the two PC12 cell lines. Although phosphorylation of the TrkA receptor was detectable in both native (Fig. 5,A, a) and trkA-transfected (Fig. 5,A, b) PC12 cells after NGF treatment, the time course of generation of the phosphorylated product in response to NGF treatment is quite different between these lines (Fig. 5 A, c). In native PC12 cells, the increase in phosphorylation of TrkA was detected at 15 min after NGF treatment; in contrast and not surprisingly, in trkA-transfected PC12 cells, the percentage of phosphorylation of TrkA was much stronger and earlier (5 min after NGF treatment).

MEK1 is activated downstream of TrkA phosphorylation and upstream of the MAPKs (e.g., ERK1/2) in this pathway. Fig. 5,B shows a similar expression of MEK1 in native and trkA-transfected PC12 cells. However, whereas the phosphorylation of MEK1 is unchanged in response to NGF treatment of native PC12 cells (Fig. 5,B, a), this phosphorylation decreased markedly in trkA-transfected PC12 cells (Fig. 5,B, b), implying that TrkA phosphorylation inhibits the phosphorylation or induces dephosphorylation of MEK1 in trkA-transfected PC12 cells. The ratio of absorbance of phospho-MEK1 to MEK1 demonstrates semiquantitatively this difference between these cell lines (Fig. 5 B, c).

ERKs are involved downstream of MEK1 phosphorylation in this pathway (37). Fig. 5,C demonstrates the differential expression of ERK after NGF treatment between native and trkA-transfected PC12 cells. NGF treatment markedly increased ERK1/2 phosphorylation in native PC12 cells (Fig. 5,C, a). In contrast, ERK1/2 phosphorylation increased only slightly in trkA-transfected PC12 cells (Fig. 5,C, b). The normalized ratio of phospho-ERK to ERK further illustrates this differential phosphorylation response (Fig. 5,C, c). Fig. 5 E demonstrates that preincubation with K252α (100 nm) abrogated NGF-induced phosphorylation of ERK protein in both cell lines, suggesting that the phosphorylation of ERK in response to NGF treatment is indeed induced by ligand activation of TrkA receptor.

Considering that p-CREB is downstream of ERK phosphorylation, we also evaluated the phosphorylation of CREB in the two cell lines. The results demonstrate that NGF treatment induced an increase in p-CREB content in native PC12 cells, but not in trkA-transfected (Fig. 5 D, a–c) PC12 cells.

These data suggest that NGF induces activation of the TrkA-MEK1-ERK1/2-p-CREB signaling pathway in native PC12 cells, but not in trkA-transfected PC12 cells.

We next determined the effects of NGF on signaling via the NGF/TrkA-MEK3/6-p38 MAP pathway. Western blotting shows that no phosphorylation of MEK3/6 was found in native PC12 cells (Fig. 6,A, a and c); in contrast, MEK3/6 phosphorylation was observed 5 min after NGF treatment in trkA-transfected PC12 cells (Fig. 6,A, b and c). Additional studies on the phosphorylation of p38 MAPK after NGF treatment demonstrate that p38 MAPK was markedly phosphorylated after NGF treatment in trkA-transfected cells, but not in native PC12 cells (Fig. 6 B). Preincubation with K252α (100 nm) abrogated NGF-induced phosphorylation of p38 MAPK in trkA-transfected cells, implying that the phosphorylation of p38 MAP in response to NGF treatment is also induced by ligand activation of the TrkA receptor.

The NGF/TrkA-MEK4-JNK-caspase 3 pathway has been recognized as a cell death signaling pathway (38, 39, 40). Interestingly, there was no difference in activation of this pathway between the two PC12 cell lines (data not shown).

It has been reported that the NGF/p75-NF-κB-c-IAP2 signal transduction pathway plays an important role in programmed cell death regulation in PC12 cells (41). We therefore examined the effects of NGF on translocation of the NF-κB p65 component using fluorescent immunohistochemical staining. As demonstrated in Fig. 7, A (native PC12 cells) and B (trkA-transfected PC12 cells), under control conditions (i.e., without NGF treatment), p65 protein primarily distributed in the cytoplasm and the nuclear membrane is clearly visible (as indicated by the arrowheads in Fig. 7,A, a and Fig. 7,B, d). In sharp contrast, similar translocation of p65 protein from the cytoplasm to the nucleus was observed after NGF (2 nm) treatment for 5 min (as indicated by the arrows in Fig. 7,A, b and Fig. 7,B, e) or 15 min (Fig. 7,A, c and Fig. 7 B, f) in the two cell lines.

It has been reported that NF-κB activates a group of gene products that function cooperatively at the earliest checkpoint to suppress p75-mediated apoptosis. c-IAP2, a member of the mammalian IAP family, is one of these gene products and is identified as a target of NF-κB transcriptional activity. In agreement with the results shown in Fig. 7, A and B, up-regulation of c-IAP2 protein was detected in both native and trkA-transfected PC12 cells (Fig. 7 C) after NGF treatment. These findings suggest that although NGF induces activation of the NGF/p75-NF-κB-c-IAP2 signal transduction pathway in both cell lines, there is no difference in this activation between native and trkA-transfected cells.

The NGF-TrkA ligand-receptor pair is thought to have a critical role in the regulation of cell survival, proliferation, and differentiation in neural crest-derived cells, including neuroblastoma cells. Ironically in view of the trophic and neuroprotective role of TrkA in most systems, TrkA overexpression in neuroblastomas increases the likelihood of apoptotic spontaneous regression or chemotherapeutic responsiveness of neuroblastoma (42). In the present study, native and trkA-transfected PC12 cells (Fig. 1) were used to examine the effects of modulation of the cellular TrkA content on the outcome of NGF ligand-mediated activation of the TrkA receptor. The expression level of TrkA in native and trkA-transfected PC12 cells was similar to that reported for good and poor prognosis neuroblastomas, respectively (42). Surprisingly from a neurobiological standpoint, but in concert with clinical observations, NGF induced time-dependent cell death, instead of cell survival, of trkA-transfected, but not native, PC12 cells (Fig. 2). This is in direct contrast to the mitogenic and survival effects of NGF described in a number of different cell types (26, 41, 43, 44).

Because of the role of NGF in mediating (45) or preventing (31, 46) apoptosis in neural crest cells, we tested the hypothesis that NGF-induced cell death in trkA-transfected PC12 cells is apoptotic in nature. Apoptosis can be distinguished from necrosis by its involvement of endonuclease activation leading to internucleosomal double-stranded chromatin (DNA) breaks and by the absence of the early cell membrane damage characteristic of necrosis (47, 48). We therefore used nuclear fluorescence double staining and DNA laddering techniques to test this hypothesis. To our surprise, we found that cell death induced by ligand activation of the TrkA receptor is not accompanied by the typical characteristics of apoptosis. In addition, AcDEVD-CHO, an inhibitor of caspase 3, does not inhibit NGF-TrkA-induced cell death (Fig. 3). These data indicate that the cell death induced by NGF ligand activation of the TrkA receptor is distinct from apoptosis by morphological, biochemical, and physiological criteria. However, this cell death is also not simple necrosis/oncosis because most of the damaged cell nuclei are stained with Hoechst dye 33342 alone, and not with PI, implying that there is no damage to the cell membrane at a time point when there is significant alteration in nuclear chromatin morphology. The process by which these cells die is therefore an atypical form of apoptosis (i.e., a form of programmed cell death that is distinct from classical apoptosis).

Although apoptosis and necrosis have been defined as two distinct models of cell death, the two may occur simultaneously within a cell population and may involve common signaling and execution mechanisms. To determine the relative incidence of necrosis and atypical apoptosis, respectively, in the cell death induced by NGF, we counted the nuclei stained with either PI alone or PI + Hoechst dye 33342 (which identifies necrotic cells) and the nuclei with abnormal morphology (lobulated or fragmented nuclei) stained with Hoechst dye 33342 alone (which identifies apoptotic cells), respectively. Cell death induced by NGF treatment (48 h) of trkA-transfected PC12 cells was primarily programmed cell death (apoptotic cells: 70.35 ± 6.37% in the NGF-treated group versus 10.17 ± 4.29% in the vehicle-treated group). Necrosis accounted for only a small fraction of the dead cells seen (necrotic cells: 30.32 ± 2.09% in the NGF-treated group versus 8.26 ± 2.75% in the vehicle-treated group). Additionally, the incidence of necrosis in these NGF-treated cultures is not related to the duration of exposure to NGF, suggesting again that programmed cell death, rather than necrosis, is the primary form of cell death in response to NGF treatment.

NGF signal transduction occurs through its high- and low-affinity receptors, TrkA and p75, respectively. The presence of two different receptors that act both independently and cooperatively is thought to account, in part, for the diverse effects of NGF. Binding to p75 has previously been shown to either induce (45, 49, 50, 51) or prevent (31, 46) apoptosis, depending upon the cell involved. Binding to TrkA is thought primarily to induce protection from apoptosis, cell proliferation, and neurite outgrowth in neural cells (44, 52, 53, 54, 55). In the present study, however, we demonstrate that in PC12 cells with up-regulated trkA, cell death induced by NGF is driven by TrkA-mediated, rather than p75-mediated, signal transduction. Accordingly, K252α, a tyrosine kinase inhibitor reported to be relatively selective for TrkA (33), abrogates the cell death induced by NGF in trkA-transfected PC12 cells. Conversely, mAbNGF30, a monoclonal antibody that occupies and inactivates the p75 binding site of NGF (26), has no impact on the effects of NGF on trkA-transfected cells.

It is not known how NGF stimulates growth through the TrkA receptor in some tumors but inhibits growth through the same receptor in others. Distinct patterns of NGF receptor expression and/or localization may confer such growth-stimulatory or -inhibitory effects. Although NGF inhibits apoptosis of PC12 cells under normal growth conditions (56, 57, 58), it could increase cell death under abnormal growth conditions. The present study demonstrates the dose dependence of the function of the TrkA receptor and the signaling pathway activated by NGF binding to it. MAPK activities (ERK and p38 MAP) are activated differentially as a function of the expression level of the TrkA receptor in PC12 cells. In native PC12 cells with native expression of the TrkA receptor, the ERK, but not the p38 MAP, pathway was activated by NGF binding to TrkA. In contrast, in trkA-transfected PC12 cells, the p38 MAP, but not the ERK, pathway was activated by NGF binding to TrkA (Fig. 5).

The binding of NGF to TrkA receptors on the cell surface leads to activation of MEK1 and ERK1/2. The activated ERKs move to the nucleus and mediate phosphorylation of transcription factors, including CREB, a known regulator of genes required for neuron survival and gene expression (59). It has been shown that the NGF-TrkA-ERK-CREB signal transduction pathway mediates cell survival (34, 35, 60). In the present study, we found that there is marked activation of the TrkA-CREB pathway in response to NGF treatment in native PC12 cells, but not in trkA-transfected PC12 cells, despite the 100-fold excess of TrkA receptors in trkA-transfected relative to native PC12 cells (Fig. 5).

Unlike the NGF-TrkA-ERK-CREB pathway, the NGF-TrkA-MEK3/6-p38 MAPK pathway has primarily been reported to be associated with cell death (61, 62, 63, 64). p38 MAPK is the mammalian homologue of the yeast HOG kinase and participates in a signaling cascade that controls cellular responses to cytokines and stress (65). It can be activated by a variety of cellular stresses and growth factors (66, 67, 68). Here we found that, upon the binding of NGF to the TrkA receptor, the phosphorylation of both MEK3/6, an upstream regulator of p38 MAPK, and p38 MAPK was increased in trkA-transfected PC12 cells, but not in native PC12 cells. Furthermore, preincubation with K252α abrogated the increase in phosphorylation of p38 MAP in trkA-transfected PC12 cells. These observations imply that activation of the MEK3/6-p38 MAPK pathway is indeed associated with NGF ligand activation of the TrkA receptor in trkA-transfected PC12 cells.

The NF-κB pathway has been proposed to be involved in TrkA-independent signaling of NGF through the p75 receptor and has been suggested to be associated with cell survival in some systems (25, 41) and apoptosis in others (50, 69). The signal transduction pathways responsible for this association are only recently elucidated. NF-κB is composed of two subunits (p65 and p50) and exists in a complex with an inhibitory protein, termed IκB, in resting cells (22). It has been suggested that ligand-induced trimerization of tumor necrosis factor results in the recruitment of the death domain adaptor protein TRADDM, which in turn recruits and interacts with receptor interaction protein (23). Overexpression of receptor interaction protein activates NF-κB-inducing kinase, which in turn activates IκB kinase complex. Upon phosphorylation by IκB kinase complex, IκB becomes ubiquitinated and degraded by proteosome complexes (24). Once free of this inhibitor, NF-κB dimerizes and becomes an active transcription factor that translocates to the nucleus, activates transcription of NF-κB-responsive genes (23), and up-regulates the expression of the IAP (70, 71).

Recent observations by Lee et al. (72) indicate that unprocessed NGF exhibits an equilibrium binding constant for p75 of 10⁻¹⁰ m, some five times stronger than that for processed NGF. In the present study, 2 nm (i.e., 2 × 10⁻⁹ m) NGF, a concentration 20 times higher than the equilibrium binding constant for p75 of unprocessed NGF, was used to examine the activation of the p75-NF-κB-c-IAP2 pathway by NGF in native and trkA-transfected PC12 cells as a measure of signaling through the p75 receptor. Translocation of p65 protein from the cytoplasm to the nucleus and up-regulation of c-IAP2 protein were observed in both cell lines in response to NGF treatment (Fig. 7). Although the p75-NF-κB-c-IAP2 pathway is activated upon NGF treatment, this pathway cannot account for the difference between the two PC12 lines or for cell death induced by NGF in trkA-transfected PC12 cells.

In summary, these results demonstrate that NGF induces apoptosis of trkA-transfected PC12 cells. Diminished signaling through the NGF-TrkA-MEK1-ERK1/2-CREB pathway and enhanced signaling through the NGF-TrkA-MEK3/6-p38 MAP pathway relative to native cells could contribute to NGF-induced cell death.

We thank Dr. H. Uri Saragovi for providing us with native and trkA-transfected PC12 cell lines.