Address correspondence and reprint requests to Mark P. Mattson, Laboratory of Neurosciences, National Institute on Aging Gerontology Research Center 4F02, 5600 Nathan Shock Drive, Baltimore, Maryland 21224, USA. E-mail: email@example.com
The development of the nervous system is regulated by trophic signals that control cell proliferation, differentiation, and survival. Numb is an evolutionarily conserved protein identified by its ability to control cell fate in the nervous system of Drosophila. Mammals express four isoforms of Numb that differ in the length of a phosphotyrosine-binding (PTB) domain and a proline-rich region (PRR). Using PC12 cells stably expressing each of the human isoforms, we show that Numb regulates sensitivity of the cells to neurotrophic factor-induced differentiation and neurotrophic factor withdrawal-induced death in an isoform-specific manner. Numb isoforms containing a short PTB domain enhance the differentiation response to NGF and enhance apoptosis upon NGF withdrawal; Numb isoforms containing a long PTB domain exhibit the same sensitivity to NGF as vector-transfected cells. These effects of Numb were found to be independent of the length of the PRR. In undifferentiated conditions, the levels of full-length TrkA and of phosphorylated p44/p42 mitogen-activated protein kinase (MAPK) are increased in cells expressing Numb isoforms with a short PTB domain, indicating an up-regulation of NGF signaling pathways. Furthermore, we provide evidence that the mechanism whereby short PTB domain Numb isoforms sensitize cells to trophic factor deprivation-induced apoptosis involves elevations in intracellular calcium concentrations. Our results suggest that Numb sensitizes cells to neurotrophin responses in an isoform-specific manner, an effect that may play an important role in the development and plasticity of the nervous system.
The formation and plasticity of neuronal circuits is regulated by environmental signals that control the differentiation, synaptogenesis, and survival of neurons. Prominent among such signals are neurotrophic factors such as nerve growth factor (NGF), brain-derived neurotrophic factor, and basic fibroblast growth factor (Mattson et al. 1989; Vogel 1993; Patapoutian and Reichardt 2001). NGF stimulates neurite outgrowth and neurotransmitter synthesis in sympathetic and sensory neurons, as well as in certain populations of neurons in the brain (Davies 1994). During differentiation, sympathetic neurons become dependent upon NGF for their survival, such that reduced availability of NGF triggers programmed cell death or apoptosis. Cell death induced by trophic factor deprivation involves mitochondrial alterations (Deshmukh et al. 2000; Kirkland and Franklin 2001), increased oxyradical production (Greenlund et al. 1995), and destablization of cellular calcium homeostasis (Cheng and Mattson 1991; Cheng et al. 1993; Guo et al. 1997). The effects of NGF on neuronal differentiation and survival are mediated by the membrane receptor tyrosine kinase TrkA and signaling cascades that regulate the expression of calcium ion channels (Lewis et al. 1993), genes encoding antiapoptotic proteins such as nuclear factor kappa B (NF-κB) and Bcl-X (Bui et al. 2001), and genes encoding pro-apoptotic proteins including Bax and Bim (Brunet et al. 2001). Components of NGF signaling through TrkA include Ras and the mitogen-activated protein (MAP) kinases ERK1/2, which regulate transcription via cAMP response element binding protein (CREB) and/or the serum response factor (SRF)–Elk complex, and PI3 kinase, which activates Akt thereby promoting the formation of Bax:Bcl-XL heterodimers and cell survival (Patapoutian and Reichardt 2001; Sofroniew et al. 2001).
The actions of NGF and other trophic factors on post-mitotic neurons have been extensively studied, yet the signaling mechanisms that control the decision of neural progenitor cells to either self-renew or differentiate into neurons are largely unknown. One signaling pathway controlling such cell fate decisions involves a cell surface receptor known as Notch that, when activated, is proteolytically processed resulting in the release of a cytoplasmic fragment that translocates to the nucleus and regulates gene expression (Artavanis-Tsakonas et al. 1999). Disruption of the Notch gene in mice results in severe developmental defects and embryonic lethality, supporting a major role for Notch in regulating cell fate (Conlon et al. 1995). A second protein that regulates neural precursor cell fate is Numb, originally discovered in Drosophila where it appears to play a role in cell fate determination by antagonizing the function of Notch (Guo et al. 1996; Zhong et al. 1996). Mice deficient in Numb die during early development, around embryonic day 11.5, and exhibit impaired neuronal differentiation leading to profound defects in cranial neural tube closure (Zhong et al. 2000; Zilian et al. 2001). Numb contains two protein–protein interaction domains: a phosphotyrosine-binding (PTB) domain and a proline-rich region (PRR), the latter of which functions as an SH3-binding domain (Guo et al. 1996; Verdi et al. 1996, 1999). Whereas only one form of Numb has been identified in Drosophila, mammals produce four isoforms of Numb that differ in the length of their PTB (lacking or containing an 11 amino acid insert) and PRR (lacking or containing a 48 amino acid insert) domains (Dho et al. 1999; Verdi et al. 1999). In the mouse P19 embryonic carcinoma cell line, it was shown that expression of Numb isoforms with short PRR promote differentiation but not proliferation, whereas expression of Numb isoforms with long PRR promote proliferation but not differentiation (Verdi et al. 1999). Based on these observations, we tested the hypothesis that Numb modulates neurotrophic factor-induced differentiation and survival responses of neural cells depending on the composition of the protein–protein interaction domains.
Materials and methods
Cell culture and experimental treatments
PC12 cells were transfected with the expression vector pcDNA3.1 with or without cDNAs for each of the human Numb isoforms (Verdi et al. 1999). Stably overexpressing clones were selected with G418 as described (Guo et al. 1997). Cultures were maintained at 37°C (5% CO2 atmosphere) in RPMI medium supplemented with 10% heat-inactivated horse serum, 5% heat-inactivated fetal bovine serum, antibiotic, and 0.5 mg/mL G418. Murine NGF (2.5S) waspurchased from Gibco-BRL (Gaithersburg, MD, USA) as a 100-µg/mL stock. Differentiation was induced by replacing the culture maintenance medium with RPMI containing 1.5% total serum and 50 ng/mL NGF. To induce apoptosis, the NGF-containing medium was replaced with the same medium lacking NGF. In undifferentiated cells, apoptosis was induced by replacing the culture maintenance medium with serum-free RPMI. Dantrolene (sodium salt) and bradykinin were purchased from Sigma Chemical Co. and stock solutions were prepared in serum-free RPMI; BAPTA-AM and PD 98059 were purchased from Calbiochem and stock solutions were prepared in dimethyl sulphoxide (DMSO).
The methods for immunoblotting analyses were similar to those described previously (Guo et al. 1998; Pedersen et al. 2002). Briefly, aliquots equivalent to 50 µg of total protein lysate were separated by electrophoresis in a 12% sodium dodecyl sulfate (SDS)–polyacrylamide gel, transferred to a nitrocellulose sheet, and incubated for 2 h at room temp with an antibody recognizing all Numb isoforms (mouse monoclonal IgG1 at a final concentration of 0.25 µg/mL; Transduction Laboratories, Lexington, KY, USA), phosphorylated ERK1/2 (mouse monoclonal IgG2a at a final concentration of 200 ng/mL; Santa Cruz Biotechnology, Santa Cruz, CA, USA), or TrkA [the antibody was raised against an extracellular domain of rat TrkA (RTA); used at a dilution of 1 : 5000; Clary et al. 1994]. The nitrocellulose sheet was further processed using horseradish peroxidase-conjugated secondary antibodies (Vector Laboratories, Burlingame, CA, USA) and a chemiluminescence detection method (Amersham, Piscataway, NJ, USA). The membrane probed with the anti-pERK1/2 antibody was incubated in Ponceau S solution (Sigma, St Louis, MO, USA; diluted 1 : 10 with dH20) and reprobed with an antibody recognizing ERK1/2 (rabbit polyclonal IgG at a final concentration of 200 ng/mL in blocking solution; Santa Cruz Biotechnology). The membranes incubated with the anti-Numb or the anti-TrkA antibodies were stained with Ponceau S solution to visualize proteins and confirm equal loading of proteins onto the gels.
Cultures were fixed for 30 min with 4% paraformaldehyde in phosphate-buffered saline (PBS). For immunostaining, cells were incubated for 5 min in a solution of 0.2% Triton X-100 in PBS, for 1 h in blocking solution (0.2% Triton X-100, 5% normal horse or goat serum in PBS), and then incubated overnight at 4°C in the presence of either the anti-Numb antibody (final concentration of 0.25 µg/mL) or the anti-TrkA antibody (RTA; 1 : 5000 dilution) in blocking solution. Following washes with PBS, the cells were incubated for 1 h in PBS containing biotinylated secondary antibody, further washed with PBS, and incubated for 30 min in the presence of fluorescein-labeled avidin (Vector Laboratories). Images were acquired using a Zeiss CLSM510 confocal laser scanning microscope with a 60× oil immersion objective (488 nm excitation and 510 nm emission). To quantify the images of Numb immunoreactivity, average pixel intensity per cell was determined using software supplied by the manufacturer.
Assessments of neurite outgrowth and cell death
Parameters of neurite outgrowth were quantified from phase-contrast images of cells acquired using a 40× microscope objective lens andaHamamatsu camera. The length of individual neurites was measured using National Institutes of Health (NIH) Image software. Cell death was quantified in cultures stained with the fluorescent DNA-binding dye Hoechst 33342 as described previously (Pedersen et al. 2002). Hoechst-stained cells were visualized under epifluorescence illumination (340 nm excitation and 510 nm barrier filter) using a 40× oil immersion objective and a Nikon Eclipse TE300 inverted microscope.
Measurements of intracellular calcium and oxyradical levels
Intracellular free calcium levels ([Ca2+]i) were quantified by fluorescence imaging of the calcium indicator dye fura-2 as described in our previous studies (Guo et al. 1997, 1998; Chan et al. 2002). Briefly, cells were incubated for 10 min in the presence of 5 µm of the acetoxymethylester form of fura-2 (Molecular Probes, Eugene, OR, USA), washed three times with serum-free RPMI, and imaged 20–40 min later using a Zeiss Axiovert microscope (40× oil immersion objective) coupled to an Attofluor imaging system. The average [Ca2+]i in individual cells was determined from the ratio of the fluorescence emissions obtained using excitation wavelengths of 340 and 380 nm. The system was calibrated using solutions containing either no Ca2+ or a saturating level of Ca2+ (1 mm) using the formula: [Ca2+]I = Kd[(R-Rmin)/(Rmax-R)](Fo/Fs). The levels of mitochondrial reactive oxygen species were measured by confocal imaging of the fluorescent probe dihydrorhodamine-123 (DHR; Molecular Probes) using methods described previously (Guo et al. 1998).
Numb modulates neural cell differentiation in an isoform-specific manner
Rat PC12 cells differentiate and acquire properties of sympathetic neurons in response to NGF, and are thereafter dependent upon NGF for their survival (Greene 1978; Fujita et al. 1989).
To determine if there are differential effects of Numb isoforms on neuronal differentiation and survival, we generated clonal lines of PC12 cells stably overexpressing each of the human Numb isoforms as follows: Numb1, long PTB–long PRR (LPTB/LPRR); Numb2, long PTB–short PRR (LPTB/SPRR); Numb3, short PTB–long PRR (SPTB/LPRR), and Numb4, short PTB–short PRR (SPTB/SPRR) (Fig. 1a). Multiple clones of cells transfected with each of the Numb isoforms were selected, and cells transfected with empty vector were used as the control in all studies. Overexpression of Numb was determined by quantification of immunostaining (Fig. 1b) and immunoblotting (Fig. 1c) data; clonal lines with approximately threefold higher levels of Numb than in vector-transfected cells were chosen for experiments. Immunostaining revealed that the Numb protein was localized in cytoplasmic compartments where it was concentrated in punctate regions throughout the cell body and in neurites, but was absent from the nucleus (Fig. 1b). All of the forthcoming results were obtained with the set of clones shown in Fig. 1(b), but each result was confirmed with at least one additional set of clonal cell lines.
Previous studies have shown that activation of the cyclic AMP second messenger pathway can induce neurite outgrowth in PC12 cells (Heidemann et al. 1985). Exposure of vector-transfected cells and cells expressing Numb isoforms with a long PTB domain to forskolin, an agent that activates adenylate cyclase and increases cyclic AMP production (Hansen et al. 2000), caused an increase in the length of short neurites during a 48-h exposure period (Fig. 2a). In contrast, cells expressing Numb isoforms with a short PTB domain responded to forskolin by extending long neurites, i.e. lengths at least twice the diameter of the cell body, with most cells having neurite lengths in the range of 80–120 µm (Fig. 2a). We next determined the effects of Numb on cell differentiation induced by NGF. During a 3-day exposure to NGF, there was a threefold increase in neurite length in cells expressing a Numb isoform containing a short PTB domain, while neurite length was increased by only 20% in vector-transfected cells and by 80% in cells expressing a Numb isoform containing a long PTB domain (Fig. 2b). Although the major effect of Numb was to enhance neurite elongation, the isoform containing a short PTB domain also increased thenumber of neurites/cell by 30% upon NGF treatment (Fig. 2c). Cells expressing Numb isoforms containing a short PTB domain spontaneously extended neurites under basal culture conditions, whereas vector-transfected cells and cells expressing Numb isoforms with a long PTB domain exhibited undifferentiated morphologies (Fig. 2d).
Isoform-specific effects of Numb on trophic factor withdrawal-induced apoptosis
We quantified cell death after NGF withdrawal by determining the percentage of cells with condensed or fragmented nuclei, an indicator of apoptosis. Untransfected PC12 cells degenerate over a period of many days after NGF withdrawal, and we therefore chose to examine an earlier time point (24 h) prior to death of the majority of cells. Whereas very few cells died during this short period of NGF deprivation in cultures of vector-transfected cells and cells expressing Numb with a long PTB domain, there was a large and highly significant increase in the extent of apoptosis in cultures of cells expressing the Numb isoform with a short PTB domain (Fig. 3a). To further examine theeffectsofNumb on trophic factor-dependent cell survival, we withdrew serum from undifferentiated clones and quantified apoptotic cell death. Whereas 15–20% of the cells exhibited apoptotic nuclei in vector-transfected cells and in cellsexpressing Numb isoforms with a long PTB domain, 45–50% of the cells expressing Numb isoforms with a short PTB domain exhibited apoptotic nuclei (Fig. 3b). These results demonstrate that expression of Numb isoforms with a short PTB domain render PC12 cells more dependent on trophic factors for survival, even under basal culture conditions.
Involvement of calcium release and oxidative stress in the cell death-enhancing effects of short PTB Numb isoforms
Calcium release from the endoplasmic reticulum (ER) is increasingly recognized for its involvement in regulating adaptive responses of neurons to environmental signals, and for its contribution to neuronal death in both physiological and pathological settings (Berridge 1998; Mattson et al. 2000). We previously showed that calcium release from the ER, resulting in sustained elevation of cytoplasmic calcium levels, plays a pivotal role in trophic factor deprivation-induced death of PC12 cells (Guo et al. 1997). Pre-treatment of PC12 cells with dantrolene (an agent that blocks calcium release from ryanodine-sensitive ER stores) or with the intracellular calcium chelator BAPTA-AM abolished the cell death-enhancing effects of short PTB domain Numb isoforms (Fig. 4a). We next determined the intracellular free calcium levels in cells expressing short and long PTB domain Numb isoforms, in comparison to control cells, following trophic factor deprivation. Basal calcium levels were similar among Vector, SPTB/SPRR, and LPTB/SPRR cells (90–130 nm). However, a statistically significant, twofold increase in the levels of intracellular calcium occurred within 10 h of trophic factor deprivation in cells expressing the short PTB domain Numb isoform, in contrast to vector-transfected and LPTB/SPRR cells that showed only 20–40% increases in calcium levels (Figs 4b and c). When the cells were challenged with bradykinin, an agonist that causes calcium release from the ER (Neuhaus et al. 1991), the highest levels of calcium were reached in cells expressing the Numb isoform with a short PTB domain (Figs 4b and c).
Cell death triggered by trophic factor deprivation has been shown to involve oxidative stress with increases in superoxide anion radical, hydrogen peroxide, and peroxynitrite being detected within several hours of withdrawal (Greenlund et al. 1995; Mattson et al. 1995; Estevez et al. 1998; Castagne et al. 1999). We therefore assessed the levels of oxidative stress in control and Numb-expressing cells under basal culture conditions and after trophic factor deprivation. The fluorescent probe dihydrorhodamine (DHR) was used to measure relative levels of hydroxyl radical and peroxynitrite generated by the mitochondria (Keller et al. 1998). No difference in basal levels of DHR fluorescence was observed between the clones. In vector-transfected cells and in cells expressing Numb isoforms with a long PTB domain, the levels of DHR fluorescence increased by approximately 30% and 50%, respectively, during a 10-h period after withdrawal of trophic support (Fig. 5). However, in cells expressing Numb isoforms containing a short PTB domain, there was a larger increase in mitochondrial oxyradical production, with levels approximately twofold greater than the basal level 10 h after withdrawal of trophic support. Our results suggest that Numb isoforms containing short PTB domains enhance processes involved in oxyradical production.
Increased levels of TrkA and activated p44/p42 MAPK in cells expressing short PTB Numb isoforms
We hypothesized that the enhanced sensitivity of cells expressing short PTB Numb isoforms to NGF-induced differentiation and trophic factor withdrawal-induced apoptosis results from up-regulation of the NGF signaling pathway. We tested this possibility by determining the levels of TrkA and the activation of downstream kinases known to be involved in NGF–TrkA signaling. Vector-transfected cells and cells expressing each Numb isoform were left untreated or were differentiated with NGF. The cells were then immunostained with an antibody that selectively recognizes TrkA (Clary et al. 1994; Averill et al. 1995). TrkA immunoreactivity was highest in both undifferentiated and differentiated cells expressing Numb isoforms with a short PTB domain compared to cells expressing long PTB domain Numb or cells transfected with empty vector (Fig. 6a). Previous studies have shown that the TrkA antibody used here recognizes two bands on immunoblots, with the higher molecular weight band (140 kDa) corresponding to full-length, functional TrkA and the lower molecular weight band (110 kDa) being an immature, non-functional form of TrkA (Clary et al. 1994). In lysates prepared from undifferentiated cells, the intensity of a band corresponding to full-length TrkA was markedly increased in cells expressing Numb isoforms with a short PTB domain compared to vector-transfected cells and cells expressing Numb isoforms with a long PTB domain (Fig. 6b).
Among the kinases activated in response to neurotrophins, p44/p42 MAPK may play a particularly important role in regulating cell survival (Hetman and Xia 2000). Undifferentiated cells expressing Numb isoforms with a short PTB domain had higher levels of phosphorylated p44/p42 MAPK (pERK1/2) compared to vector-transfected cells and cells expressing long PTB domain Numb isoforms (Fig. 7a). The overall levels of ERK1/2 varied among the clones, but this could not account for the differences observed for phospo-p44/p42. To determine if activation of p44/p42 MAPK plays a role in the differential effects of Numb isoforms on cell survival, we pre-treated cells with PD 98059, a selective inhibitor of MAP kinase kinase (MEK) shown to inhibit neurotrophin responses (Bonni et al. 1999), and then subjected the cells to trophic factor deprivation. Pre-treatment with the MEK inhibitor significantly attenuated apoptosis in cells expressing Numb isoforms with a short PTB domain (Fig. 7b). These results indicate that the effects of Numb isoforms with a short PTB domain on neuronal cell survival involve an up-regulation of NGF signaling through TrkA and p44/p42 MAPK.
Our results demonstrate that Numb isoforms differentially modulate two major responses of neural cells toneurotrophic factors, namely, stimulation of neurite outgrowth and cell survival dependency. These findings provide a novel view of Numb as a regulator of neurotrophic factor signaling, and suggest this action of Numb as a potential mechanism for modulating neural cell development. The identification of Numb and the initial supporting evidence that it promotes neurogenesis came from studies of Drosophila, where Numb was shown to be required for neuronal differentiation of external sensory organ precursor cells (Lin and Schagat 1997). There is now considerable evidence that Numb specifies the neuronal fate in a variety of lineages of the vertebrate CNS (Wakamatsu et al. 1999, 2000; Zilian et al. 2001), where it is believed to promote the progenitor cell fate during asymmetric division (Zhong et al. 2000). In the mouse, it has been shown that Numb is expressed not only in progenitor cells, but also in differentiating neurons during development and in adult neurons throughout the nervous system (Zhong et al. 1997), indicating a role for Numb once the neuronal fate has been specified. Again, based primarily upon data obtained in Drosophila, it is proposed that Numb exerts its effects by antagonizing Notch signaling (Frise et al. 1996; Guo et al. 1996; Wai et al. 1999). However, evidence suggests that Numb may have functions independent of Notch (Spana et al. 1995; Zilian et al. 2001), supported by the finding that there is a differential distribution of Numb and Notch1 during mouse cortical neurogenesis (Zhong et al. 1997). Thus, although activation of Notch has been shown to inhibit neurite outgrowth (Berezovska et al. 1999; Sestan et al. 1999), neuronal differentiation of PC12 cells promoted by short PTB domain Numb isoforms is not necessarily due to antagonism of Notch signaling.
Numb isoforms with a short PTB domain, in contrast to isoforms with a long PTB domain, enhanced the neurite outgrowth response to NGF and enhanced apoptosis resulting from trophic factor deprivation. These effects were found to be independent of the length of the PRR domain. The later result is particularly interesting given that in two cell culture models of mammalian neurogenesis, the P19 embryonic carcinoma cell line and the multipotent neural crest stem cell line MONC-1, overexpression of Numb isoforms containing short PRR domains biased the cells toward a neuronal fate (Verdi et al. 1996, 1999). Specifically, isoforms with a short PRR domain promoted neuronal differentiation, whereas isoforms with a long PRR domain promoted cell proliferation (Verdi et al. 1996, 1999). It was further demonstrated in these studies that expression of human SPRR Numb isoforms in Drosophila directed neuronal cell fate choice during sensory organ precursor development, an effect that was abrogated by expression of human LPRR Numb isoforms. The effects of Numb on neuronal cell fate choice in Drosophila were found to be independent of the length of the PTB domain and, in fact, overexpression of the mouse Numb PTB domain in MONC-1 cells biases differentiation away from neuronal fates (Verdi et al. 1996). Although speculative at present, the possibility exists that the two domains of Numb differentially regulate intrinsic neurogenic programs during development (PRR) and responses to neurotrophic factors once the neuronal fate has been specified (PTB).
A number of different proteins with PTB domains have been identified, yet there are only a few for which specific functions are known; prominent among these are Shc and insulin receptor substrate-1 (IRS-1; van der Geer and Pawson 1995; Margolis et al. 1999). Shc is an adaptor protein that contains both an SH2-binding domain and a PTB domain, and plays a critical role in coupling receptor tyrosine kinases to the Ras signaling pathway (Cattaneo and Pelicci 1998). IRS-1 interacts with the insulin receptor and provides a link to the activation of downstream kinases in the PI3 kinase–Akt pathway (Yoshimura et al. 1997). Homodimerized TrkA is coupled, via Shc, to activation of Ras and MAP kinases (Patapoutian and Reichardt 2001; Sofroniew et al. 2001). Because proteins containing PTB domains can interact with receptor tyrosine kinases, and associated signaling proteins, it is reasonable to consider that Numb may modulate NGF responses of developing neurons by interacting with TrkA or other proteins in the NGF–TrkA signaling cascade. We found increased levels of full-length TrkA and of activated p44/p42 MAPK in undifferentiated PC12 cells expressing Numb isoforms with a short PTB domain. Although not tested here, it is possible that short PTB domain Numb isoforms also up-regulate PI3 kinase–Akt signaling, which may explain the partial effect of the MEK inhibitor in preventing cell death caused by these isoforms. While it is not known how Numb up-regulates NGF–TrkA signaling, the enhanced NGF responsiveness of cells expressing Numb isoforms with a short PTB domain appears to be the cause of increased sensitivity of the cells to trophic factor deprivation-induced apoptosis. Indeed, cells cultured in the presence of high concentrations of growth factors have increased susceptibility to death upon growth factor withdrawal (Vander Heiden et al. 2001), an effect that appears to be mimicked by Numb isoforms containing short PTB domains.
Calcium plays a fundamental role in regulating neurite outgrowth and cell survival during development of the nervous system (Mattson 1996), and in the adult nervous system in response to physiological stimuli and pathological processes (Chittajallu et al. 1998; Mattson et al. 2001). In a previous study, we provided evidence that expression of a short PTB domain Numb isoform renders PC12 cells more vulnerable to cell death induced by amyloid β-peptide by disrupting calcium homeostasis (Chan et al. 2002). Pertinent to the present findings are studies showing that NGF and other neurotrophic factors can promote maintenance of intracellular calcium concentrations within a range optimal for neurite outgrowth and cell survival (Mattson et al. 1989; Johnson et al. 1992). Our present findings provide evidence that Numb, in an isoform-specific manner, regulates neurotrophin-mediated differentiation and survival signals by modifying neural cell calcium homeostasis. The enhancement of trophic factor deprivation-induced apoptosis by Numb isoforms with a short PTB domain was abolished by treating cells with dantrolene or BAPTA-AM. Both the short and long PTB domain isoforms of Numb enhanced calcium responses to bradykinin in cells deprived of trophic support, but the peak levels of calcium were highest in cells expressing the short PTB domain isoform. These findings indicate that Numb influences calcium release from the ER, and that the short PTB domain isoforms facilitate the increase in calcium to levels optimal for neuronal differentiation. The regulation of calcium release is a complex process involving IP3 receptors and ryanodine receptors that are controlled by growth factor and neurotransmitter signaling pathways coupled to inositol phospholipid hydrolysis, and by calcium itself (Berridge 1998; Mattson et al. 2000). As such, there are several possible sites where Numb might influence ER calcium regulation.
Several key steps in the cell death program triggered by NGF deprivation have been established in studies of sympathetic neurons developing in culture (Deshmukh et al. 2000).
An early event, occurring within 2–6 h of NGF deprivation, is an increase in levels of reactive oxygen species, which is followed by increases in the levels of c-jun and Bax, caspase activation, and cell death. We found that the early increase in cellular oxidative stress was greatly enhanced in PC12 cells overexpressing Numb isoforms with a short PTB domain, suggesting that Numb modifies a very early step in the cell death pathway triggered by trophic factor deprivation. This result is consistent with the possibility that short PTB domain Numb isoforms enhance the signal (or lack of signal) that results from NGF receptor vacancy. Collectively, our findings reveal novel roles for Numb isoforms in regulating: (i) neuronal differentiation in response to NGF, and (ii) the dependency of neural cells on NGF for survival. Given that many neurotrophic factors employ receptor tyrosine kinases, modulation of these signaling pathways may be a widely used mechanism whereby Numb controls the formation of neuronal circuitry during development.
This work was supported by the National Institute on Aging. SLC is a recipient of a fellowship from The John Douglas French Alzheimer's Foundation. We thank LF Reichardt (Howard Hughes Medical Institute, UCSF) for the kind gift of TrkA antibody.