A molecular basis of survival from neuronal injury is essential for the development of therapeutic strategy to remedy neurodegenerative disorders. In this study, we demonstrate that an EF-hand Ca2+-binding protein neuronal Ca2+ sensor-1 (NCS-1), one of the key proteins for various neuronal functions, also acts as an important survival factor. Overexpression of NCS-1 rendered cultured neurons more tolerant to cell death caused by several kinds of stressors, whereas the dominant-negative mutant (E120Q) accelerated it. In addition, NCS-1 proteins increased upon treatment with glial cell line–derived neurotrophic factor (GDNF) and mediated GDNF survival signal in an Akt (but not MAPK)-dependent manner. Furthermore, NCS-1 is significantly up-regulated in response to axotomy-induced injury in the dorsal motor nucleus of the vagus neurons of adult rats in vivo, and adenoviral overexpression of E120Q resulted in a significant loss of surviving neurons, suggesting that NCS-1 is involved in an antiapoptotic mechanism in adult motor neurons. We propose that NCS-1 is a novel survival-promoting factor up-regulated in injured neurons that mediates the GDNF survival signal via the phosphatidylinositol 3-kinase–Akt pathway.
Neuronal apoptosis is induced by numerous stressors and underlies many human neurodegenerative disorders, such as Alzheimer's and Parkinson's disease. Under such apoptotic conditions, several neurotrophic factors such as glial cell line–derived neurotrophic factor (GDNF) and brain-derived neurotrophic factor (BDNF) can activate the antiapoptotic process to rescue neurons from death. However, the signaling pathway leading to cell survival is not yet completely understood. GDNF was reported to exert a potent survival-promoting activity in neurons (Henderson et al., 1994; Oppenheim et al., 1995; Yan et al., 1995) and to reduce neuronal death induced by various toxic challenges both in vitro (Nicole et al., 2001) and in vivo (Wang et al., 2002; Kirik et al., 2004). Recent evidence suggests that a part of molecular mechanisms for GDNF-induced cell survival relates to an increase in intracellular Ca2+ concentration, and it subsequently activates some survival pathways such as the phosphatidylinositol 3-kinase (PI3-K)–Akt pathway (Perez-Garcia et al., 2004).
Ca2+ is the most versatile and important intracellular messenger in neurons, regulating a variety of neuronal processes such as neurotransmission and signal transductions. The various actions of Ca2+ are mediated by a large family of EF-hand Ca2+-binding proteins, which may act as Ca2+ sensors or Ca2+ buffers. One of them, neuronal Ca2+ sensor-1 (NCS-1; mammalian homologue of frequenin), was originally identified in Drosophila melanogaster in a screen for neuronal hyperexcitability mutants (Mallart et al., 1991). Overexpression of NCS-1 has been shown to enhance evoked neurotransmitter release and exocytosis (Pongs et al., 1993; Olafsson et al., 1995). NCS-1 directly interacts with phosphatidylinositol 4-hydroxykinase (PI4-K; Hendricks et al., 1999; Weisz et al., 2000) and enhances neuronal secretion by modulating vesicular trafficking steps in a phosphoinositide-dependent manner (Koizumi et al., 2002). We have previously demonstrated that NCS-1 modulates the voltage-gated K+ channel Kv4 (Nakamura et al., 2001). Subsequently, certain voltage-gated Ca2+ channels have also been reported to be regulated by NCS-1 (Weiss et al., 2000; Wang et al., 2001; Tsujimoto et al., 2002). Furthermore, NCS-1 enhances the number of functional synapses (Chen et al., 2001), potentiates paired pulse facilitation (Sippy et al., 2003), and may be involved in associative learning and memory in Caenorhabditis elegans (Gomez et al., 2001). Despite the participation of NCS-1 in a wide range of biological functions, however, the role of NCS-1 in neuronal survival under pathophysiological conditions or the involvement of NCS-1 in neurotrophic factor–mediated neuroprotection are unknown.
Because we found that the expression levels of NCS-1 is significantly higher in immature brain (Nakamura et al., 2003) and a remarkable similarity exists between immature and injured neurons during the development and regeneration process, respectively (Nabekura et al., 2002b), these findings prompted us to study the expression level and the functional roles of NCS-1 in damaged neurons.
In this study, we found that NCS-1 is a survival-promoting factor, which increases the resistance of neurons to several kinds of stressors. In addition, NCS-1 is up-regulated in response to axonal injury in adult motor neurons, and this protects cells from apoptosis. Furthermore, NCS-1 mediates GDNF-induced neuroprotection via activation of Akt pathways. This is the first study demonstrating a novel role of NCS-1 on neuronal survival.
The expression level of NCS-1 protein increases with neuronal injury
To examine the expression level of NCS-1 in injured neurons, we performed unilateral vagal axotomy (transaction of nerves) on adult rats. 1 d to 2 mo after the in vivo axotomy, brainstems, including the bilateral dorsal motor nucleus of the vagus (DMV) neurons, were isolated. Immunohistochemical staining and computerized image analysis of frozen sections revealed that axotomy significantly (more than threefold) increased the expression level of NCS-1 in the DMV when compared with those on the control side at 1 wk after the surgery (Fig. 1, A–C). NCS-1 immunoreactivity was mainly expressed in cell bodies of neurons, as shown using hematoxylin counterstaining to identify the nuclei (Fig. 1 B, brown staining accompanied with blue staining; depicted by arrows). The increase in NCS-1 level started at 1 d after axotomy, reached a peak at 1 wk, and gradually decreased to control levels over the next 2 mo (Fig. 1 D). We also conducted quantitative immunoblot analysis on tissue samples from DMV neurons 1 d and 1 wk after axotomy, expressing NCS-1 density relative to levels of glyceraldehyde-3-phosphate dehydrogenase (GAPDH). The results confirmed the immunohistochemistry experiments, with levels of NCS-1 protein in ipsilateral DMV being increased significantly (by about threefold) by 1 wk after axotomy (Fig. 1, E and F).
Up-regulation of NCS-1 protein was also observed using a different type of stressor. Continuous treatment of neurons with colchicine for 4 d, which disrupts tubulin polymerization and blocks axonal transport, also increased NCS-1 expression levels (1.40 ± 0.03-fold above control levels; P < 0.05; n = 4), indicating that NCS-1 is up-regulated in vivo in response to two different kinds of stressors—one being mechanical and the other being chemical injury.
Expression of NCS-1 renders PC-12 cells more tolerant to stressors
To study the physiological role of NCS-1 in damaged neurons, we next examined the effect of NCS-1 overexpression on the susceptibility of cells to several kinds of stressors. PC-12 cells stably transfected with either the NCS-1 expression vector (NCS-1/PC-12) or the vector alone (vector/PC-12) were differentiated into neuronlike cells, and the resistance to H2O2 toxicity was compared between these two groups. As shown in the immunofluorescent micrographs and immunoblot in Fig. 2 (A, B, and G), the expression level of NCS-1 was found to be significantly higher in NCS-1/PC-12 cells compared with vector-transfected cells, although these cells also had some endogenous NCS-1.Treatment with a relatively high dose (300 μM) of H2O2 for 3 d in the absence of pyruvate resulted in severe cellular damage in vector/PC-12 control cells; most cells were rounded up and detached from the substratum (Fig. 2 E). In contrast, the same treatment caused only a little damage to cells overexpressing NCS-1 (Fig. 2 F), indicating that the expression of NCS-1 rendered PC-12 cells more tolerant to H2O2 toxicity. The expression of NCS-1 reduced cell death caused by treatment with up to 1,000 μM H2O2 (Fig. 2 H). The aforementioned results were obtained from three cell lines transfected with NCS-1 and with corresponding vector-transfected control. A similar beneficial effect of NCS-1 on cell survival in response to 300 μM H2O2 was seen in two PC-12 cell lines that were not treated with NGF (not depicted).
NCS-1 promotes the long-term survival of primary cultured cortical neurons under stress and normal conditions
To further confirm the involvement of NCS-1 in neuronal survival, we overexpressed NCS-1 or its mutant E120Q in primary cultured embryonic rat cortical neurons that express endogenous NCS-1. The E120Q mutant possesses an amino acid substitution within the third EF-hand Ca2+-binding motif, which impairs Ca2+ binding (Jeromin et al., 2004) but preserves the interaction with target proteins and, thereby, exerts a dominant-negative effect by disrupting the function of endogenous NCS-1 (Weiss et al., 2000). We used an adenoviral transfer system to transiently deliver the cDNA encoding NCS-1 together with EGFP (using an internal ribosome entry site–containing vector) and its E120Q mutant form into neurons cultured for 5 d in neurobasal medium containing B27 trophic supplements. As indicated by cells with EGFP fluorescence and nuclei stained with Hoechst 33258, nearly 70% of neurons were successfully infected with each virus at 3 d after infection (Fig. 3 A). We examined the effects of overexpression of wild-type and dominant-negative NCS-1 on neuronal survival under stress caused by B27 withdrawal, which has been reported to induce neuronal apoptosis (Brewer, 1995; Cheng et al., 2003). As shown in Fig. 3 B, B27 withdrawal promoted cell death in vector-treated control neurons (left; also compare the vector groups with and without B27 in Fig. 3 D). Overexpression of NCS-1, on the other hand, significantly rescued cells from death (Fig. 3 B, middle). In contrast, the expression of E120Q resulted in more severe cell death accompanying bleb formation (Fig. 3 B, right). To quantitatively analyze the time course for the changes in cell viability, the total number of surviving cells from the same field was counted daily by phase-contrast microscopy during 9 d (see Materials and methods). The results show that high cell viability was preserved upon expression of the wild-type NCS-1, whereas cell viability was reduced after the expression of E120Q; i.e., the number of days required to reach 70% cell viability were 5, 8, and 3 d for vector, NCS-1, and E120Q groups, respectively (Fig. 3 C). The expression levels of NCS-1 in each group of neurons before and after adenovirus infection in the absence of B27 trophic supplements are shown in the immunoblot (Fig. 3 E). Essentially the same results were obtained by counting neurons with condensed nuclei using Hoechst staining (Fig. 3, F and G), thus reinforcing the finding that the expression of NCS-1 protects neurons from cell death under apoptotic conditions. Furthermore, when B27 trophic supplement was kept in the culture medium (which is a less stress condition), the dominant-negative effect of E120Q was more clearly observed when compared with the vector control group (Fig. 3 D; also see A, where some blebs were observed in the neurons infected with E120Q mutant). In other preliminary experiments (not depicted), although the time course of the loss in cell viability was variable, overexpression of NCS-1 consistently delayed the loss of cell viability when B27 supplements were omitted, and the expression of E120Q always increased the rate of cell death when B27 was present. These results suggest that endogenous NCS-1 is playing an important role in keeping the long-term survival of cultured neurons under normal conditions in addition to the protective role from stress under apoptotic conditions.
NCS-1 mediates GDNF-induced cell survival
A large body of evidence suggests that neuronal survival is promoted by neurotrophic factors such as BDNF and GDNF (Boyd and Gordon, 2003). Because the long-term application of GDNF has been reported to enhance the expression of NCS-1 in Xenopus laevis motor neurons (Wang et al., 2001), we attempted to clarify the role of NCS-1 as a downstream mechanism of GDNF-induced cell survival in rat cortical neurons. When primary cultured cortical neurons were treated with 10 ng/ml GDNF for 2 d after the withdrawal of B27 supplements, neuronal survival was significantly enhanced when compared with time-matched control (Fig. 4, A [top] and B). Interestingly, the expression of NCS-1 mimicked the survival-promoting effects of GDNF; i.e., NCS-1 exerted a robust survival effect even in the absence of GDNF (Fig. 4, A [middle] and B). Most strikingly, the expression of the dominant-negative NCS-1 mutant E120Q largely prevented cell survival induced by GDNF (Fig. 4, A [bottom] and B). Immunoblot analysis revealed that the application of 10 ng/ml GDNF resulted in a significant increase in the expression level of endogenous NCS-1 within 10 min, which further increased at 40 min in these neurons (Fig. 4 C). The amount of NCS-1 remained elevated through 2 d of exposure to GDNF (not depicted). These results show that the treatment of GDNF increases the expression level of NCS-1, which subsequently promotes neuronal survival, suggesting that GDNF-induced neuroprotection is at least in part mediated by NCS-1.
Activation of the PI3-K–Akt pathway is involved in NCS-1–induced neuronal survival
In neurons, GDNF has been reported to promote cell survival via activation of signaling cascades involving the PI3-K–Akt pathway (Soler et al., 1999; Takahashi, 2001). In accordance with these studies, we also observed that exposure of primary cultured cortical neurons to GDNF resulted in a large increase in phospho-Akt levels (Fig. 5 A, left). Therefore, it was of interest for us to test whether NCS-1 also activates this kinase. We examined the effect of NCS-1 expression or its dominant-negative form on Akt phosphorylation in the presence or absence of B27 trophic supplements. When B27 trophic supplements were absent, the expression of NCS-1 significantly enhanced the phosphorylation of Akt, whereas expression of the dominant-negative mutant E120Q had little effect when compared with control vector-infected neurons (Fig. 5, A [middle] and B). On the other hand, when B27 supplements were present, a relatively high level of phosphorylated Akt was observed in the vector-treated control group (Fig. 5, A [right] and B). Additional expression of exogenous NCS-1 further increased the Akt phosphorylation level, whereas expression of the dominant-negative mutant suppressed phosphorylation (Fig. 5, A [right] and B). Thus, the phosphorylation levels of Akt in each group of neurons were well correlated with their viabilities, as shown in Fig. 3 D.
In addition, pretreatment of cultured cortical neurons with LY294002, an inhibitor of PI3-K, completely abolished both GDNF- and NCS-1–induced neuronal survival, whereas PD98059, an inhibitor of MAPK kinase (MEK), did not (Fig. 5, C and D). These results suggest that the NCS-1–induced survival-promoting effect is mediated via the PI3-K–Akt pathway but not the MAPK pathway in cultured cortical neurons.
To further understand the upstream mechanism of the NCS-1–induced activation of Akt, we next examined the effect of overexpression of NCS-1 and E120Q on the subcellular localization of Akt/PKB in living cells. Akt/PKB is known to be translocated to the plasma membrane when it is fully activated upon phosphorylation and bound with its substrates PtdIns(3,4)P2 and PtdIns(3,4,5)P3 (Alessi et al., 1996). We constructed the GFP-tagged pleckstrin homology (PH) domain of Akt/PKBα (EGFP-Akt/PKB-PH) and transiently cotransfected it into CCL39 cells, which express a small amount of endogenous NCS-1, together with NCS-1, E120Q, or empty vector. 2 d later, the subcellular localization of EGFP-tagged Akt/PKB-PH was assessed on a confocal microscope. Akt/PKB-PH was diffusely localized in the cytosol of vector-transfected control cells (Fig. 5 E). Interestingly, Akt/PKB-PH became localized in the peripheral region of cells when NCS-1 was coexpressed, but this peripheral localization was abolished when E120Q was coexpressed (Fig. 5 E). Qualitatively similar results were also obtained when primary cultured cortical neurons were treated with the same vectors; i.e., Akt/PKB-PH was localized in the peripheral regions of neurons when NCS-1 was overexpressed, but a more diffuse localization pattern was observed when E120Q was overexpressed (Fig. 5 F). The distribution pattern of Akt/PKB-PH in vector-transfected neurons was similar to that of NCS-1–overexpressing cells (not depicted). These results strongly demonstrate that NCS-1 increases the levels of plasma membrane PtdIns(3,4)P2 and PtdIns(3,4,5)P3 and, thus, activates Akt/PKB in living cells.
Dominant-negative NCS-1 accelerates the in vivo axotomy-induced loss of neurons
We examined the effects of the overexpression of NCS-1 and its dominant-negative mutant on the survival of these neurons to clarify the physiological role of NCS-1 in injured motor neurons in vivo. One side of vagus nerves of adult rats were axotomized as previously described (Fig. 1) and infected with adenoviral vectors encoding NCS-1, E120Q, or EGFP vector alone, and neuronal degeneration was evaluated by histological analysis. 1 wk after axotomy, nearly 30% of nerve cells were found to be EGFP positive in the injured side (Fig. 6 A). There were clear differences in the staining pattern between control and injured sides for all groups, probably because the regeneration process, such as activation of the surrounding glial cells, was ongoing on injured sides. However, the number of surviving motor neurons stained with hematoxylin were not significantly decreased at the injured side for vector-treated DMV sections (Fig. 6, B and D; examples of counted neurons are indicated by black arrows in C). This would probably be the result of natural antiapoptotic mechanisms induced by injury, which exist in mature neurons as previously reported (Benn and Woolf, 2004). Because the expression level of NCS-1 was significantly increased in response to in vivo axotomy (Fig. 1), we hypothesized that NCS-1 may be involved in this antiapoptotic mechanism. If so, blocking of endogenous NCS-1 would reduce this beneficial effect. As expected, the dominant-negative E120Q mutant resulted in a significant loss of neurons in the injured side (Fig. 6, B–D), and some TUNEL-positive nuclei were also detected only in this group (Fig. 6 E, arrows; and its magnified image in F). Considering that the infection efficiency was only ∼30% in these experiments, a large majority of neurons successfully infected with E120Q appear to have undergone apoptosis. Infection of neurons with the functional NCS-1 adenovirus only had a modest effect on neuronal survival. This probably results from both the low infection efficiency and the high levels of endogenous NCS-1 expression in axotomized neurons (Fig. 1) because the NCS-1 effects are already close to maximum. Thus, the dominant-negative mutant E120Q inhibited the survival of adult DMV neurons from axotomy-induced injury, strongly suggesting that NCS-1 is one of the important factors mediating neuronal survival after in vivo axotomy.
Numerous stressors, including physical or chemical injury and genetic abnormalities, lead to neuronal degeneration by programmed cell death along an apoptotic pathway. Under these conditions, some intrinsic and extrinsic factors, including neurotrophic factors, are known to activate the antiapoptotic process to rescue neurons from death. However, the signaling pathway leading to cell survival is not yet completely understood.
In this study, we identified a novel function for the Ca2+-binding protein NCS-1, which (1) promotes the long-term survival of cultured neurons via PI3-K–Akt signaling pathways; (2) mediates, at least in part, GDNF-induced neuroprotection; and (3) is up-regulated in response to axonal injury and plays an important role in the antiapoptotic mechanism in injured motor neurons.
NCS-1 is a novel survival-promoting factor in neuronal cells
We observed that the overexpression of NCS-1 rendered PC-12 cells and primary cultured cortical neurons more tolerant to several kinds of stressors, such as oxidative stress or trophic supplement withdrawal (Figs. 2 and 3), demonstrating that the expression of NCS-1 protects neurons from cell death under apoptotic conditions. In addition, overexpression of an EF-hand dominant-negative mutant E120Q significantly accelerated apoptosis when B27 trophic supplements were kept in the culture medium (Fig. 3 D), suggesting that endogenous NCS-1 is important for keeping the long-term survival of cultured neurons under normal (or less apoptotic) conditions. The latter finding also indicates that Ca2+ binding is required for NCS-1–mediated cell survival. On the basis of these findings, we propose that NCS-1 is a novel member of survival-promoting factors in cultured neurons.
NCS-1 mediates GDNF-induced cell survival via activation of the PI3-K–Akt survival pathway
We found that treatment of cultured cortical neurons with a neurotrophic factor GDNF increased the expression level of NCS-1 (Fig. 4 C) and enhanced neuronal survival (Fig. 4, A and B), which is consistent with a previous study reporting that GDNF enhanced the expression of frequenin/NCS-1 in Xenopus motor neurons (Wang et al., 2001). GDNF-induced increase in the NCS-1 level appeared to be caused by the synthesis of protein and/or mRNA but not by the prevention of NCS-1 degradation because GDNF did not raise the expression level of NCS-1 in the presence of the inhibitor of protein synthesis cycloheximide (10 μg/ml for 20 h; not depicted). In contrast to the vector-treated control neurons, GDNF did not further enhance the survival effect in neurons overexpressing NCS-1 (Fig. 4, A and B), suggesting that cell viability was already sufficiently high under this condition. Strikingly, the survival-promoting effect of GDNF was largely prevented by overexpression of the dominant-negative mutant E120Q (Fig. 4, A and B), suggesting that NCS-1 mediates the GDNF survival signal.
GDNF activates at least two intracellular pathways in neurons: one involving the PI3-K–Akt pathway and another involving the MAPK (p42 and p44, also called ERK1 and ERK2) pathway. However, PI3-K but not the MAPK pathway has been reported to be responsible for GDNF-mediated neuronal survival in motor neurons (Soler et al., 1999). In accordance with this study, we observed that exposure of primary cultured cortical neurons to GDNF resulted in a large increase in the phospho-Akt level (Fig. 5 A). In the same way, the overexpression of NCS-1 also dramatically enhanced the phosphorylation levels of Akt both in the presence and absence of B27 trophic supplements, whereas the overexpression of dominant-negative mutant E120Q did not (Fig. 5 A). In addition, the NCS-1–nduced survival-promoting effect was largely inhibited by the PI3-K inhibitor LY294002 but not the MEK inhibitor PD98059 (Fig. 5, C and D). Furthermore, NCS-1 increased the plasma membrane PtdIns(3,4)P2 and PtdIns(3,4,5)P3 levels, which indicates the activation of Akt in intact cells (Fig. 5, E and F). These results strongly suggest that NCS-1 is a novel downstream target that mediates GDNF survival signal through activation of the PI3-K–Akt pathway.
Possible mechanisms of the action of NCS-1
Several possible mechanisms may underlie the survival action of NCS-1. We observed that Akt/PKB-PH was recruited to the plasma membrane when NCS-1 was coexpressed, suggesting that NCS-1 acts upstream of the Akt pathway. NCS-1 was previously reported to activate PI4-K (Hendricks et al., 1999), which increases the level of plasma membrane PtdIns(4)P, the substrate of PI3-K, as well as PI5-K. Therefore, upon activation of these kinases, other phosphoinositides would be produced. Indeed, it has been reported that the overexpression of NCS-1 significantly increased both PtdIns(4)P and PtdIns(4,5)P2 levels in PC-12 cells (Koizumi et al., 2002). Furthermore, PtdIns(3,4)P2 and PtdIns(3,4,5)P3, the substrates of Akt/PKB, would also be produced, which, in turn, would activate the Akt pathway (Cantley, 2002). As expected, the overexpression of NCS-1 increased the plasma membrane PtdIns(3,4)P2 and PtdIns(3,4,5)P3 levels both in CCL39 cells and neuronal cells (Fig. 5, E and F), enhanced the phosphorylation of Akt (Fig. 5 A), and promoted neuronal survival (Fig. 3). Therefore, we propose that activation of such a phosphatidylinositol pathway is a mechanism for the survival action of NCS-1. In addition, we observed that in contrast to GDNF, our preliminary data show that BDNF did not increase the expression level of NCS-1 (unpublished data) despite the reported survival-promoting effect of BDNF in cultured cortical neurons (Cheng et al., 2003). Interestingly, BDNF-induced survival signaling has been reported to be mediated by CaM, another Ca2+-binding protein in cortical neurons (Cheng et al., 2003), and CaM has been reported to directly activate PI3-K (Perez-Garcia et al., 2004). Therefore, NCS-1 mediates the GDNF signal by activating PI4-K, whereas CaM mediates the BDNF signal by activating PI3-K. These two signals would lead the survival signal to the Akt pathway.
On the other hand, it is also possible that NCS-1 promotes neuronal survival by some other mechanisms in addition to activation of the Akt pathway. For example, the survival-promoting effect of NCS-1 appears to be analogous to that of the recently characterized antiapoptotic protein family called inhibitors of apoptosis, which suppress apoptosis through the direct inhibition of caspases (Liston et al., 2003). Some of these proteins, such as neuronal apoptosis inhibitory protein and X-linked inhibitors of apoptosis protein, have been reported to be essential for GDNF-mediated neuroprotective effects in injured motor neurons in vivo (Perrelet et al., 2002). Furthermore, recent evidence demonstrates that neuronal apoptosis inhibitory protein interacts with hippocalcin, another closely related Ca2+-binding protein that affects caspase-12 activity (Korhonen et al., 2005) and protects neurons against Ca2+-induced cell death (Mercer et al., 2000). Therefore, we do not exclude the possibility that like these proteins, NCS-1 also exerts a more direct effect on some caspases. We are currently investigating the possible interaction of these proteins.
As NCS-1 is known to interact with voltage-gated K+ channels (Kv4; Nakamura et al., 2001), it might also increase the resistance of neurons to excitotoxic apoptosis through the activation of K+ channels. Increased outward K+ current would prevent neurons from reaching firing threshold and, thereby, prevent cells from Ca2+ overload leading to cell death.
NCS-1 is a novel survival-promoting factor up-regulated in injured neurons
In this study, we found that the expression level of NCS-1 was significantly increased in response to axonal injuries (transection of the vagus nerve as well as treatment of nerves with colchicine) in the DMV neurons of adult rats (Fig. 1). The behavior of NCS-1 appears to be analogous to that of the recently identified protein damage-induced neuronal endopeptidase, which is expressed in response to neuronal damages induced by nerve transection and colchicine treatment in both the central and peripheral nervous systems (Kiryu-Seo et al., 2000). Because antiapoptotic mechanisms are activated in mature neurons in response to stress to protect against accidental apoptotic cell death, it has been described that peripheral axotomy in adult neurons does not result in extensive cell death (Benn and Woolf, 2004). In accordance with this, we also observed that little loss of motor neurons was evident by in vivo axotomy in vector-treated control neurons (Fig. 6, B and D). The expression of exogenous NCS-1 did not exert the further beneficial effect (Fig. 6, B and D). This marginal effect of exogenous NCS-1 (compared with the vector control group) would be the result of the increased expression level of endogenous NCS-1 in axotomized neurons, which occurs for all groups. In contrast, overexpression of dominant-negative E120Q significantly decreased the number of surviving neurons (Fig. 6, B–D) and produced TUNEL-positive apoptotic neurons at the injured side (Fig. 6, E and F), indicating that disruption of NCS-1 function increased the vulnerability of DMV neurons to axotomy.
Overexpression of NCS-1 rendered PC-12 cells resistant to H2O2 toxicity even in the absence of GDNF (Fig. 2), suggesting that NCS-1 itself is enough to promote cell survival. This is consistent with our view that NCS-1 is the downstream target for GDNF. Because growing evidence indicates that nerve injury leads to the up-regulation of multiple antiapoptotic molecules, including GDNF (Liberatore et al., 1997; Yamamoto et al., 1998; Wang et al., 2002), it is possible that neuronal damages induced by in vivo axotomy enhance the synthesis and/or secretion of GDNF, which, in turn, up-regulates NCS-1 expression and promotes neuronal survival in injured neurons. Although an underlying mechanism would be different, the up-regulation of NCS-1 has also been reported in the cortex of schizophrenic and bipolar patients, demonstrating the involvement of NCS-1 in neurological disease (Koh et al., 2003).
In conclusion, we characterized a novel function of NCS-1 mediating a GDNF-induced neuroprotective effect via activations of Akt kinase. Furthermore, we found that NCS-1 is up-regulated in response to nerve injury and plays an important role in the antiapoptotic mechanism in adult motor neurons. Our present findings would provide new and basic insights into the mechanism of neuronal regeneration.
Materials And Methods
Plasmids and viral vectors
E120Q NCS-1 point mutant was generated with a conventional PCR protocol using the wild-type rat NCS-1 (GenBank/EMBL/DDBJ accession no. L27421) as a template and was sequenced to confirm the mutation.
Akt/PKBα cDNA was cloned from the human kidney cDNA library (CLONTECH Laboratories, Inc.), and NH2-terminally tagged fluorescent protein EGFP-Akt/PKB-PH was constructed incorporating a fragment of 750 bp, encoding the first 250 amino acids of PKBα (containing the PH domain) into EGFP-vector as described previously (Currie et al., 1999).
Adenovirus containing wild-type NCS-1 and the E120Q mutant inserts were generated by cotransfecting either of these plasmids and pBHG11 (Microbix Biosystems, Inc.) into HEK 293 cells. Viral DNA was isolated from the supernatant in the wells displaying the cytopathic effect. Replication-incompetent virus containing DNA inserts were plaque-purified twice and grown on HEK 293 cells to produce large amounts of adenovirus. Tissue culture supernatant containing adenovirus was concentrated by centrifugation over cesium chloride. The titers of viral stocks were 2.2 × 1010 pfu/ml for EGFP–NCS-1, 1.1 × 1010 pfu/ml for EGFP-E120Q, and 2.2 × 1010 pfu/ml for EGFP-vector.
PC-12 cells stably transfected with vector alone or vector containing cDNA coding for the wild-type NCS-1 (several clones) were grown onto collagen-coated (500 μg/ml of type I; Sigma-Aldrich) culture dishes in growth medium (DME containing 10% horse serum, 5% FBS, and 400 μg/ml geneticin and gentamicin) as described previously (Koizumi et al., 2002). When cells became 80% confluent, they were switched to the differentiation medium (growth medium with half serum) supplemented with 100 ng/ml NGF-7S (Invitrogen).
Primary culture of cortical neurons was performed using the cortex from Sprague-Dawley rats at embryonic day 18. In brief, cortical tissues were isolated from whole brain, minced into small pieces, and digested for 10 min at 37°C in a 20-U/ml papain solution containing 0.002% DNase I (Worthington Biochemical Corp.). After titration of the enzymatic activity, cells were mechanically dissociated by several passages through pipette tips. After centrifugation, cells were resuspended in neurobasal medium supplemented with B27 trophic factors (both from Invitrogen), whose compositions were reported previously (Brewer et al., 1993). They were then plated onto culture dishes coated with 0.1% polyethylenimine at a density of 2.5–5 × 104 cells/cm2 for cell survival assay and 105 cells/cm2 for immunoblot analysis.
CCL39 cells and primary cultured rat cortical neurons were plated onto collagen-coated glass coverslips and cultivated for 1 d. They were then transiently transfected with the EGFP-Akt/PKB-PH construct together with either NCS-1, E120Q, or pCDNA3 (1:3 ratio) using LipofectAMINE 2000 (Invitrogen) and were subjected to immunocytochemistry. In brief, cells were fixed with 4% PFA, permeabilized with 0.2% Triton X-100, and blocked with 5% BSA. They were then incubated for 1 h with anti–NCS-1 antibody (1:200) followed by incubation with secondary antibodies (FITC- or rhodamine-conjugated goat anti–rabbit IgG; 1:200; Jackson ImmunoResearch Laboratories). After extensive wash with PBS, cell images were scanned on a laser confocal microscope (MRC-1024K; Bio-Rad Laboratories) or obtained with conventional epifluorescence illumination (BX50WI; Olympus) with a cooled CCD camera (CoolSNAP; Photometrics) using a 0.9-W 60× water immersion objective lens. Immunocytochemistry for PC-12 cells were also performed in the same way.
Evaluation of neuronal survival
The primary cultured neurons were infected with viruses at a multiplicity of infection of 100 pfu/cell at 5 d after plating. Under this condition, we found that nearly 70% of the neurons were infected by monitoring EGFP fluorescence (Fig. 3 A). The number of living neurons was counted within the fixed area of images taken by a digital camera (Coolpix 4500; Nikon). To count the number of cells always within the same area, a grid seal with numbering (Asahi Techno glass) was stuck on the bottom of each culture dish. The number of living neurons remaining at each day was expressed as a percentage of the initial number. Neurons showing the degenerating stage characterized by nuclear condensation, membrane blebbing, or extensive neurite fragmentation were excluded. Four different regions were selected from one dish, and six separate experiments were performed for each condition.
To identify and quantify apoptotic neurons, cells were fixed with 4% PFA and were stained with Hoechst 33258. Coverslips were mounted onto glass slides, and cells were observed under epifluorescence illumination on an inverted microscope (1X71; Olympus) using a 40× NA 1.35 oil immersion objective lens (Olympus). Cells were considered apoptotic if their nuclear chromatin was condensed or fragmented, whereas cells were considered viable if their chromatin was diffusely and evenly distributed throughout the nucleus (Fig. 3 F).
DMV tissue samples were obtained by scratching the DMV neurons from several frozen sections of brainstem (described in the next section) using pulled glass capillary under the light microscope. These tissue samples or cultured cells (PC-12 cells and cortical neurons) were then solubilized in SDS-PAGE sample buffer containing protease and phosphatase inhibitors and subjected to immunoblot analysis using image density software (Scion Image; Scion Corp.) as previously described (Nakamura et al., 2001). Primary antibodies used were anti–NCS-1 antibody (1:1,000), which was previously described (Jeromin et al., 1999), and publicly available antibodies: monoclonal anti-GAPDH antibody (1:1,000) obtained from Chemicon as well as antiphospho-Akt antibodies (detectable for the phosphorylation of Thr308 and Ser473; 1:1,000) and anti-Akt antibody (1:1,000; both from Cell Signaling Technology). Secondary antibodies used were HRP-conjugated anti–rabbit and anti–mouse antibodies or a combination of biotinylated anti–rabbit (or mouse) antibodies (Zymed Laboratories) and HRP-conjugated streptavidin (Zymed Laboratories).
In vivo axotomy and colchicine treatment
The method of vagus axotomy was described previously (Nabekura et al., 2002a). In brief, 4–6-wk-old Sprague-Dawley rats were deeply anesthetized with 50 mg/kg pentobarbital, and axotomy of the vagus motor neurons was performed with fine scissors at the unilateral vagus nerve at the neck. Injured neurons were confirmed by detecting the fluorescence of Di-I in the DMV, which had been placed at the proximal cut site of the nerve bundle (Nabekura et al., 2002a).
To test the effects of colchicine, an implantable polymer containing 10% (wt/vol) colchicine was made by mixing colchicine with ethylene-vinyl acetate copolymer (Elvax) followed by drying as described previously (Kakizawa et al., 2000). Solid slices (∼1 mm2) were placed around the unilateral vagal nerve to allow the continuous release of colchicine from slices. The skin incision was closed, and rats were returned to the cage after awaking from the anesthetic.
1 d to 2 mo (usually 1 wk) after receiving ipsilateral vagal axotomy, brainstems were quickly removed, and 8–10-μm-thick frozen sections were cut. Immunohistochemistry was performed using the labeled biotin–streptavidin method. In brief, after fixation and blocking, the sections were incubated at 4°C overnight with a rabbit polyclonal antibody against NCS-1 at a dilution of 1:15,000 and were sequentially incubated with a biotinylated anti–rabbit secondary antibody and a HRP-conjugated streptavidin–biotin complex (GE Healthcare). The colored reaction product was developed with DAB solution. The sections were lightly counterstained with hematoxylin to visualize nuclei. Images were acquired using a digital camera (FX380; Olympus) equipped with an image filing software (FLVFS-LS; Flovel).
Comparison of the expression level of NCS-1 between injured and control sides were performed using computerized image analysis (Win Roof; Mitani Corp.). In brief, the DMV region from the injured side was at first selected, and the image was converted to binary images by thresholding so that only the area highly stained with anti–NCS-1 antibody could be detected. The same threshold level was used for both the control and injured DMV in each tissue section. The highly stained area was summated and represented as normalized values.
Neuronal degeneration was evaluated by counting surviving neurons as described previously (Rothstein et al., 2005) as well as by TUNEL staining using the apoptag peroxidase in situ Apoptosis Detection Kit (Chemicon). In brief, in vivo axotomy was performed as described above, and, at the same time, adenoviral vectors carrying EGFP only, EGFP plus NCS-1, or E120Q (109 pfu each) was injected into the stump of the nerve using a 34-gauge needle. 1 wk after axotomy, paraffin-embedded serial sections (3–4 μm) were made from the brainstem. After they were deparaffinized, sections were directly stained with hematoxylin/eosin to visualize the structure of the DMV region. TUNEL staining was performed in accordance with the manufacturer's method. The sections were lightly counterstained with methyl green. Control sections were treated similarly but incubated in the absence of TdT enzyme. To confirm whether the adenoviral vectors were transferred to the DMV neurons, another set of animals were treated in the same way. 8-μm frozen sections were cut 1 wk after operation, and EGFP signals were viewed under a fluorescence microscope (1X71; Olympus). All image acquisitions were performed at room temperature, and images were subsequently processed using Adobe Photoshop (version 7) and Adobe Illustrator (version 10) software. All experiments conformed to the Guiding Principles for the Care and Use of Animals approved by the Council of the Physiological Society of Japan. All efforts were made to minimize the number of animals used and their suffering.
Comparisons between two groups were performed using the paired or unpaired t test. Values of P < 0.05 were considered statistically significant. All summarized data are expressed as means ± SEM.
We thank Dr. Andrew Moorhouse (The University of New South Wales, Sydney, Australia) for critical reading of this manuscript. We also thank Dr. Takeharu Nishimoto and Dr. Hideo Nishitani (Kyushu University, Fukuoka, Japan) for useful discussions of this study.
This work was supported, in part, by a Grant-in-Aid for Scientific Research (17590196) and Priority Areas (13142210) from the Ministry of Education, Culture, Sports, Science and Technology of Japan and the Japan Heart Foundation and by a grant from the National Institutes of Health/National Institute of Neurological Disorder and Stroke (NS38126).
Abbreviations used in this paper: BDNF, brain-derived neurotrophic factor; DMV, dorsal motor nucleus of the vagus; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; GDNF, glial cell line–derived neurotrophic factor; MEK, MAPK kinase; NCS-1, neuronal Ca2+ sensor-1; PH, pleckstrin homology; PI3-K, phosphatidylinositol 3-kinase; PI4-K, phosphatidylinositol 4-hydroxykinase.