Apolipoprotein E Receptors Mediate Neurite Outgrowth through Activation of p44/42 Mitogen-activated Protein Kinase in Primary Neurons*

Several ligands of the endocytic low density lipoprotein receptor-related protein (LRP), such as apoE-containing lipoproteins and activated (cid:1) 2-macroglobulin ( (cid:1) 2M*), promote neurite outgrowth, suggesting that LRP may have signaling functions. In this study, we found that the treatment of neurons with (cid:1) 2M* significantly increased the individual length (by 71%) and numbers (by 139%) of neurites of primary mouse cortical neurons. These effects were blocked by the LRP antagonist, the receptor-associ-ated protein. We found similar neurite outgrowth with purified apoE3 and a tandem apoE peptide containing only the receptor-binding domain. To investigate the intracellular pathway of the LRP signaling involved in neurite outgrowth, we tested the effects of (cid:1) 2M* on the phosphorylation of the mitogen-activated protein (MAP) extracellular signal-regulated kinases 1 and 2 (ERK1/2). We found that 1) phospho-MAP kinase levels were altered within 30 min after treatment with (cid:1) 2M*, 2) the MAP kinase inhibitor, PD98059, specifically blocked the (cid:1) 2M*-induced neurite outgrowth, 3) manipulating intracellular calcium by BayK or BAPTA altered the neurite outgrowth and associated changes in the phospho-MAP kinase levels, which were blunted by (cid:1) 2M*, 4) (cid:1) 2M* promoted the phosphorylation of the transcription factor CREB through MAP kinase, and 5) LRP-specific antibodies increased levels of phosphorylated MAP kinase and phosphorylated CREB. The effects of (cid:1) 2M*, apoE3, and apoE peptides increased LRP levels in the cortical neurons, whereas LRP receptor-associated protein reduced dendritic LRP expression. These results demonstrate that p44/42 MAP kinase plays an important role in LRP-medi-ated neurite outgrowth with activation involving the effects on calcium homeostasis and downstream effects involving the activation of gene transcription through CREB. Data Analysis— For calcium studies, each consisted of two or three culture sets of cortical neurons in which 5–15 neuronal somata in each field were measured. For immunohistochemical, calcium, and bio- chemical studies, the data from several cultures were pooled for statistical analyses. Values are expressed as the (cid:8) S.E. Statistical significance was determined by one-way analysis of variance (ANOVA) followed by the Fisher post-hoc test for multiple comparisons. p (cid:5) 0.05 was considered indicative of a statistically significant difference.

LRP 1 is a 600-kDa multifunctional cell surface receptor containing multiple ligand binding sites and a high affinity Ca 2ϩ -binding site, which is important for receptor conformation and ligand recognition (1). LRP directs ligands including apolipoprotein E (apoE) (2) and activated ␣2-macroglobulin (␣2M*) (1) to degradation. LRP is strongly expressed in the central nervous system (3). Among the diverse ligands for LRP, ␣2M* is of particular interest because of its robust association with cytokines (4 -6) and neurodegeneration (7). ␣2M is a large tetrameric protein that has established roles as a multifunctional proteinase inhibitor and in the binding and clearance of a variety of small molecules including cytokines (8), growth factors (9), and endogenous soluble ␤-amyloid peptide (10,11). When ␣2M has been "activated" by protease or chemical modification (␣2M*), it becomes a competent ligand for binding and clearance by LRP. Although ␣2M* can be a neurotrophic factor, little is known regarding the downstream signaling of ␣2M*. To assess the potential functions of ␣2M* in the central nervous system, we examined the short term and long term effects of ␣2M* exposure on an important neuronal function in the cortex of the brain, neurite outgrowth.
Neuronal differentiation and axonal growth are controlled by a variety of factors (12). In a highly simplified model (12), the signals generated within peripheral domain and central domains of the growth cone have distinct functions in controlling growth cone motility. Signals generated in the peripheral domain, such as intracellular calcium, are instructive signals for growth and guidance, whereas those generated in the central domain, such as activation of the mitogen-activated protein (MAP) extracellular signal-regulated kinases 1 and 2 (ERK1/2), are more likely to be permissive signals. However, in neurodegenerative disorders such as Alzheimer's disease, the factors that alter neurite outgrowth involving calcium MAP kinase cascade could promote degenerative processes leading to the neuritic dystrophy observed.
Previous studies from our laboratory and others demonstrate that ␣2M* is involved in numerous central nervous system functions including up-regulation of LRP expression (13), calcium signaling (14,15), and neuronal protection (16,17). In this study, we sought to determine the pathways involved in ␣2M* promotion of neurite outgrowth via LRP in primary cortical neurons. We found that the treatment of neurons with ␣2M* significantly increased the neurite outgrowth in a LRP receptor-associated protein (RAP)-blockable manner dependent on the phosphorylation of MAP kinase. In addition, the phosphorylation of MAP kinase depends on changes in calcium homeostasis altered by ␣2M* treatment and promotes the phosphorylation of CREB.

MATERIALS AND METHODS
Cell Culture-Cortical neurons from 16-day-old embryonic Swiss Webster mice were isolated by a standard enzyme treatment protocol (15,18). Cortices were dissociated in calcium-free saline and plated on poly-D-lysine (Sigma) coated tissue culture dishes at the density of 5 ϫ 10 4 cell/ml. The neurons were grown in neurobasal medium (Invitrogen) plus 10% fetal bovine serum with 25 M penicillin-streptomycin. 1 h after plating, the medium with serum was replaced with medium containing the B-27 supplement (Invitrogen).
Treatment with LRP Ligands-Primary neuronal cultures were treated with 500 nM ␣2M*. ␣2M* was added into B-27/neurobasal medium 24 h after plating. A similar treatment protocol was applied to other LRP ligands or drugs such as apoE (100 nM), tandem apoE peptide (100 nM), and RAP (500 nM). Polyclonal LRP antibodies raised against purified human LRP or against the cytoplasmic domain of LRP were a kind gift of Dr. Dudley Strickland (American Red Cross). Control cultures consisted of sister cultures, which were untreated with LRP ligands. The neurons were observed for 48 -72 h in culture. In short term treatments, LRP ligands were added into medium for 30 min. Control cultures (sister cultures) were not treated with LRP ligands. The medium was replaced by physiological saline before calcium measurements were made. The composition of the physiological saline was (in mM): 140 NaCl; 3.5 KCl; 0.4 KH 2 PO 4 ; 0.33 Na 2 HPO 4 ; 2 MgSO 4 ; 2.2 CaCl 2 ; 10 glucose; and 10 HEPES-NaOH, pH 7.3.
Neurite Outgrowth Analysis-Neurite outgrowth was assessed by checking the cell morphology using an MRC 1024 confocal microscope system. Microscopic images from random fields were captured and digitized by Bio-Rad software. ␣2M* and related treatments were added to the neuronal culture at 1 day in vitro. The length and numbers of neurites were measured 48 -72 h after treatment with ␣2M*. The cortical cultures were washed, fixed, and immunostained overnight at 4°C. Anti-MAP2 mouse monoclonal antibody (Upstate Biotechnology, Lake Placid, NY) was used as a neuronal marker. R829, a rabbit polyclonal antibody against LRP holoprotein (a kind gift of Dudley Strickland) was used to investigate LRP expression and distribution. MAP2 and LRP antibodies were detected with fluorescein-linked antimouse Ig and Cy3-linked anti-rabbit Ig, respectively. The length of neurites from all of the neurons within a field was stereologically quantified and expressed as the total length (in mm). The stereological analysis of neurite outgrowth was carried out using the computerassisted microscope allowed for precise well defined movements along the x and y axes. Total neurite length was measured as described by Larsen (19) and Berezovska et al. (20) using a quadratic lattice ran-domly placed on the sampled neuron: b ϭ /2 ϫ d ϫ I (where d is the inverse line density and I is the number of intersections between the neurite and the grid line). Morphologically, the alterations of neurite outgrowth were compared between differentiation stages of primary neurons in the presence and in the absence of drug treatment. For example, the alterations of neurite outgrowth between the round cells and cells with short minor processes were assessed, such as cells with a single neurite or bipolar cells with processes no longer than approximately 4 -5 diameters of the cell body or between cells with well developed ramified processes. The average length or number of neurites was obtained by dividing the total length by the number of cells within a given field or by dividing the total number of neurites by the number of cells within a given field.
Western Blot Analyses-For analysis of active MAP kinase and CREB, cell lysates were prepared from the primary cortical neuronal cultures. The neuronal cultures were lysed in 50 mM Tris-HCl, pH 8.0, containing 0.5 M NaCl, 4 M leupeptin, 2 M pepstatin, 1.5 M aprotinin, 400 M phenylmethylsulfonyl fluoride, and 0.5% Triton X-100. The total cell lysates were centrifuged at 2000 rpm for 30 s, and the supernatant was analyzed by the immunoblot. Proteins were denatured, reduced, and separated by 4 -12% Tris-glycine SDS-polyacrylamide gel electrophoresis (Novex, San Diego, CA). Proteins were transferred to nitrocellulose at 380 mA for 60 min, and the blotted membrane was blocked with 5% milk in Tris-buffered saline containing 0.05% Tween 20 for 30 min at room temperature. The blots were incubated with polyclonal antibodies against phospho-MAP kinase (Promega, Madison, WI) or phospho-CREB-1 (Santa Cruz Biotechnology, Santa Cruz, CA) in phosphate-buffered saline with Tween 20 containing 5% milk overnight at 4°C. The phospho-MAP kinase detects the active form of MAP kinase dually phosphorylated at Thr 183 and Tyr 185 . The phospho-CREB antibody recognizes the active form of CREB phosphorylated at Ser 133 . From the same blots, actin (AC-40, Sigma) was detected by specific monoclonal antibody to ensure that equal protein was present in each lane. Immunoreactivity was detected using horseradish peroxidaselinked anti-rabbit IgG developed with a chemiluminescent reagent and exposed to film. Analyses with a Bio-Rad GS-700 imaging densitometer were recorded as the percentages of the sister cultures.
Intracellular Calcium Measurement-Intracellular calcium was determined for individual cells using standard microscopic Fluo-3 digital imaging (15). Cortical neurons were loaded with 1 M Fluo-3/AM, and live video imagines of selected microscopic fields were recorded with a photomultiplier (Hamamatsu Photonics, Hamamatsu City, Japan) and digitized by computer with a Bio-Rad imaging time course software (Imaging Research Inc.). The somata of ϳ5-10 cells in each microscopic field were individually measured. Intracellular calcium levels were estimated by converting fluorescent intensity to intracellular calcium concentration using the following formula: OR) and in vivo under saturating calcium concentrations facilitated by introducing extracellular calcium into cells with the calcium ionophore A23187 (Molecular Probes). The calcium calibration in vivo was consistent with calcium calibration in vitro, and the in vitro calibrations were applied in the current study. All of the experiments were performed at room temperature (ϳ23°C).
Transfection of Neurons with LRP-Primary mouse cortical neurons were cultured in neurobasal medium with B-27 supplement. Transient transfection of the cells was performed using a calcium phosphate method (21). Cells were plated into 4-well chambers 1 day before the transfection with an expression vector of LRP fused to EGFP, which was generated from ligating human LRP cDNA digested with restriction enzymes of XhoI and Bsu36I and synthetic oligomers of 5Ј-TGAG-GACGAGATAGGGGACCCCTTGGCAA-3Ј and 5Ј-AGCTTTGCCAAGG-GGTCCCCTATCTCGTCC-3Ј containing the last part of LRP without a stop codon and HindIII site. These then were inserted into the XhoI and HindIII sites of expression vector pEGFP-N1 (Clontech). A mixture of 8 g of plasmid DNA, 9 l of 2.5 M CaCl 2 , and 100 l of water were made in 100 l of HEPES buffered saline and left for 15-30 min at room temperature. 25 l of this mixture then was added to the cells cultured in Dulbecco's modified Eagle's medium in each well. Cells were washed with Dulbecco's modified Eagle's medium after 20 min and maintained in the conditioned medium collected before transfection. LRP-EGFP expression in primary neurons was not detectable until 18 h after transfection. It was measured 48 h after transfection by confocal microscopy. Microscopic images of LRP-EGFP from random fields were captured and digitized by Bio-Rad software. Control cells transfected with vector alone demonstrated no fluorescence.
Chemicals-Recombinant human ␣2M (Sigma) was activated with methylamine and stored at Ϫ20°C for no more than 2-3 weeks. Recombinant human RAP was prepared from a glutathione S-transferase fusion protein as described previously (22). BayK 8644 (Research Biochemicals, Inc.) at 10 M was used to increase the intracellular calcium by opening the calcium channels on the cell membrane. Data Analysis-For calcium studies, each protocol consisted of two or three culture sets of cortical neurons in which 5-15 neuronal somata in each field were measured. For immunohistochemical, calcium, and biochemical studies, the data from several cultures were pooled for statistical analyses. Values are expressed as the mean Ϯ S.E. Statistical significance was determined by one-way analysis of variance (ANOVA) followed by the Fisher post-hoc test for multiple comparisons. p Ͻ 0.05 was considered indicative of a statistically significant difference.

Effects of ␣2M* on Neurite Outgrowth and LRP Distribution in Cultured Cortical
Neurons-Cortical neurons obtained from 16-day-old embryonic mice show distinctive developmental changes during the culture period, thus providing an accessible developmental model for analyzing neurite outgrowth. Neurons were treated with 500 nM ␣2M* for 48 -72 h, and total neurite length and numbers were assessed. Fig. 1A shows the micrographs of ␣2M*-induced morphological changes. Microtubules in the primary neuronal cultures were detected by antibody MAP2 (in red). The expression of LRP in the primary neuronal cultures was detected by antibody R829 (in green). The individual length and numbers of neurites were substantially increased in ␣2M*-treated neurons. In addition, the expression of LRP was also increased and diffusely distributed on the neurites. The averaged data in Fig. 1B illustrate that the treatment of primary neuronal cultures with ␣2M* significantly increased neurite length (by 72%) and neurite numbers (by 139%). To test whether the increase in neurite outgrowth was because of ␣2M* binding to LRP, we co-incubated cultures with RAP, which blocks the interactions between LRP and its ligands. The ␣2M*-induced increase in neurite outgrowth was eliminated by co-incubation with RAP in primary neuronal culture (Fig. 1, A and B). In addition, the expression of LRP was restricted to the soma of neurons by RAP treatment. These data suggest that ␣2M* exerts its effect via binding to LRP, because ␣2M*-induced effects on neurite outgrowth could be eliminated by RAP.
We hypothesized that LRP ligands might alter the turnover of LRP, contributing to their effects on neurite outgrowth. To investigate whether the distribution of LRP correlated with neurite outgrowth, we transfected primary cortical neurons with full-length LRP tagged with GFP and exposed neurons to LRP ligands for 48 h (Fig. 2). LRP expression in transfected primary neurons was substantially increased by ␣2M* and apoE3, which also promoted neurite outgrowth. Similar effects were obtained from a tandem apoE peptide (100 nM) containing only the receptor-binding region. Both purified apoE (24,25) and this tandem apoE peptide (23) interact with LRP. RAP treatment alone dramatically altered LRP expression, redistributing LRP from the neurites to the soma. This redistribution of LRP may contribute to the inhibitory effects of RAP in our sets of experiments.
Time-dependent Effects of ␣2M* on the Levels of Phospho-MAP Kinase-To examine the role of MAP kinase in neurite outgrowth in response to ␣2M* stimulus, phospho-MAP kinase (Thr 183 /Tyr 185 ) was measured in primary cortical neurons treated with ␣2M* with an antibody against this active form of MAP kinase. In both control-and ␣2M*-treated neurons, phospho-MAP kinase (p42/44) was detectable at all time points in cell lysates. ␣2M* increased the phospho-MAP kinase levels at 30 min, and this increase disappeared by 48 h (Fig. 3A). To test whether the LRP-mediated pathway was involved in the effects FIG. 6. Neurite outgrowth mediated by interaction of ␣2M* and intracellular calcium. Neurite length and numbers were determined as in Fig. 1, and cells were treated as in Fig. 4. The representative micrographs are shown in panel A, and the average data of neurite outgrowth are shown in panel B. Results are pooled from two to three sets of cultures, and each culture includes five fields containing 3-5 cells/field. *, p Ͻ 0.05 versus control; ϩ, p Ͻ 0.05 versus ␣2M*-treated group. Neurite outgrowth was significantly inhibited by the calcium chelator BAPTA and promoted by the calcium channel agonist BayK. Incubation of cells with BAPTA prevented the significant ␣2M* promotion of neurite outgrowth. PMA, phorbol 12-myristate 13-acetate. of ␣2M* on the levels of phospho-MAP kinase, we co-incubated cultures with RAP (500 nM) to block the interactions between LRP and ␣2M*. Surprisingly, RAP alone also increased the levels of phospho-MAP kinase at 5, 15, and 30 min and 48 h (Fig. 3B). However, there was no change in the phospho-MAP kinase level when the neurons were co-cultured with ␣2M* and RAP. These results suggest that the effect of ␣2M* on phospho-MAP kinase is a time-dependent process. To address whether these effects were because of binding to LRP or another member of the LDL receptor family, we treated cells with an antibody raised against holo-LRP. This antibody, but not the one directed against the LRP intracellular domain, increased the levels of phospho-MAP kinase (Fig. 3C). Thus, the activation of LRP alone can increase the levels of phospho-MAP kinase. The effect of PD98059, an inhibitor of MAP kinase activation, was used to determine whether the enhancement in the levels of active MAP kinase was necessary for the ␣2M*-induced increase in neurite outgrowth. Fig. 4 shows the effects of PD98059 on the ␣2M*-induced increases in phospho-MAP kinase levels (panel A) and neurite outgrowth (Fig. 4, B and C). When the neurons were co-incubated with PD98059, ␣2M* did not induce a visible increase either in phospho-MAP kinase levels (Fig. 4A) or in neurite outgrowth (Fig. 4, B and C). These results indicate that the changes in neurite outgrowth produced by ␣2M* required activation of MAP kinase in primary cortical neurons.
Interaction of ␣2M*, Intracellular Calcium, and Phosphorylation of MAP Kinase-We have previously reported that ␣2M* altered calcium influx via the NMDA receptor (15), affecting both endogenous calcium oscillations and calcium influx after NMDA stimulation. Here we observed that calcium oscillations in the cortical cultures often were synchronized among several neurons in a microscopic field, suggesting that the network synaptic activity played a role in the generation of the oscillations. Bath application of tetrodotoxin, a treatment that blocks synaptic transmission in the neuronal cultures, blocked the intracellular Ca 2ϩ oscillations (data not shown), consistent with a dependence of the oscillations on network synaptic activity. The synchronized calcium oscillations across neurons in a microscopic field depend on NMDA receptor activation (15) and on calcium release from intracellular calcium stores. ␣2M* dramatically reduced the amplitude of spontaneous intracellular calcium oscillations in cortical neurons (Fig. 5A). This alteration of synchronized calcium oscillations by ␣2M* remained when neurons were co-cultured with RAP (Fig. 5B).
Calcium transients encode information across a range of frequencies and direct neurite extension (26 -29). Growth cone calcium waves regulate the rate of axon extension, which is inversely proportional to their frequency (30 -33). We tested whether intracellular calcium alterations affected neurite outgrowth in our culture system. Cells were treated with BAPTA to chelate intracellular calcium or BayK to promote calcium influx through surface calcium channels. Reduction of intracellular calcium by BAPTA reduced neurite outgrowth, while increased intracellular calcium by BayK promoted outgrowth (Fig. 6). Treatment with ␣2M* showed the expected increase in neurite outgrowth, but this increase was significantly inhibited by co-incubation with BAPTA (Fig. 6).
We tested whether the effects of calcium levels on neurite outgrowth were mediated through the effects on activation of MAP kinase. The levels of phospho-MAP kinase were increased by BayK after 30 min of treatment but not after 48 h of treatment (Fig. 5C). There was no further alteration in the phospho-MAP kinase levels when cells were also treated with ␣2M*. In contrast, BAPTA substantially reduced the levels of active MAP kinase after 30 min and this reduction by BAPTA was partially blunted by ␣2M* pretreatment. However, after 48 h of BAPTA treatment, the levels of phospho-MAP kinase were substantially reduced, even in the combined treatment (Fig. 5C). Thus, the effects of BayK and BAPTA on neurite outgrowth correlate with their effects on phospho-MAP kinase at early (30 min) time points.
PKC is an upstream regulator of MAP kinase activation (34,35), which is not only associated with intracellular calcium (36 -38) but also required for activation and nuclear translocation of active MAP kinase (39,40). Fig. 5C shows the involvement of PKC in the short term effects of ␣2M* on the levels of phospho-MAP kinase. 30 min after treatment of primary cortical neurons with phorbol 12-myristate 13-acetate or ␣2M*, the levels of phospho-MAP kinase were substantially increased and the effects were much greater with the combined treatment. The increased levels of phospho-MAP kinase returned to base-line levels by 48 h (Fig. 5C). Activation of PKC by ␣2M* is consistent with its role in the activation of MAP kinase.
␣2M* Binding to LRP Induces CREB Activation via MAP Kinase-The transient effects of ␣2M* and intracellular calcium alterations on kinase activation could have long term effects on processes such as neurite outgrowth if these transient changes alter gene transcription. Thus, we sought to determine whether ␣2M*-induced phosphorylation of MAP kinase led to the phosphorylation of CREB. Immunoblots with an antibody against phospho-CREB showed that phospho-CREB was significantly increased in primary neurons in response to both short term (383% control, p Ͻ 0.0006) and long term (205% control, p Ͻ 0.002) ␣2M* treatment (Fig. 7A). There were no visible changes in the levels of control protein ␤-actin (bottom panel).
To determine whether MAP kinase was required for CREB activation in response to ␣2M*, we treated cells with the MAP kinase inhibitor, PD98059, for 30 min or 48 h (Fig. 7B). PD98059 treatment blocked CREB phosphorylation (110 Ϯ 10.31% control after 30 min treatment and 73 Ϯ 12.41% control after 48 h treatment; p Ͻ 0.01 compared with ␣2M*-treated cells without PD98059) (Fig. 7B). These data indicate that the activation of CREB by ␣2M* is mediated by the activation of MAP kinase.
Similar experiments were performed to determine whether LRP activation could promote CREB phosphorylation. Co-incubation of cultures with RAP abolished the ␣2M*-induced increase in phospho-CREB, similar to the results with PD98059 (Fig. 8, A and B). RAP alone did not alter the levels of phospho-CREB. Similar to ␣2M*, an anti-LRP antibody also caused an increase in the levels of phospho-CREB (Fig. 8, C and D). Together, these data show that ␣2M* promotes neurite outgrowth, p44/42 MAP kinase activation, and activation of CREB via binding to LRP. DISCUSSION One of the earliest indications that apoE receptors might have signaling functions in the central nervous system was that several ligands of LRP promoted neurite outgrowth (41,42). In recent years, these signaling functions have been dem- onstrated in vitro and in vivo with apoE receptors altering neuronal calcium homeostasis (14,15,43), kinase activation (43,44), and neuronal migration (45,46). Our current study demonstrates that neurite outgrowth involves a signaling pathway that can be promoted through the activation of apoE receptors as supported by several experiments. First, ␣2M* treatment significantly increased the individual length and numbers of neurites in primary mouse cortical neurons in a RAP-blockable fashion. Second, the MAP kinase inhibitor, PD98059, diminished the ␣2M*-induced increase in the levels of phospho-MAP kinase and blocked the ␣2M*-induced neurite outgrowth. Third, the effects of ␣2M* on neurite outgrowth were inhibited by chelating intracellular calcium levels with BAPTA. Fourth, ␣2M* and an LRP-specific antibody stimulated the phosphorylation of CREB through interactions with LRP and activation of MAP kinase. Taken together, our data demonstrate that LRP mediates neurite outgrowth through the effects on intracellular calcium homeostasis and p44/42 MAP kinase activation, leading to the effects on CREB transcription regulation.
The role of the MAP kinase cascade in neurite outgrowth has been extensively studied in the neuronal cell lines such as PC12 cells. The MAP kinase cascade is required for growth factor-induced differentiation of naive PC12 cells (47,48), although it is not sufficient for neurite outgrowth (47,49). The activation of the MAP kinase cascade appears to be a permissive signal involved in making cells competent to extend neurites in response to a growth factor stimulus and calcium signaling (12,50). The role of the MAP kinase cascade in neurite outgrowth in primary neurons was demonstrated using MAP kinase inhibitors in our study of ␣2M*-induced neurite outgrowth (Fig. 3) and in studies of growth factor-induced neurite outgrowth (51,52). Our results also demonstrate that MAP kinase activation in primary cortical neuron is only a permissive signal, which depends also on regulation of calcium homeostasis to promote the neurite outgrowth. Manipulating intracellular calcium alone altered the activation of MAP kinase (Fig. 5) and altered the associated neurite outgrowth (Fig. 6). RAP treatment alone altered MAP kinase ERK1/2 activation (Fig. 3) but did not change the intracellular calcium homeosta-sis ( Fig. 5B) (16). RAP treatment alone also did not promote neurite extension (data not shown and Fig. 2). We hypothesize that ␣2M* affects calcium homeostasis and MAP kinase activation and that both pathways are necessary for neurite outgrowth.
Calcium is a key second messenger within growth cones that can increase the rate of growth cone extension (53,54), turn growth cones, and induce growth cone collapse (55). The differential effects of calcium on growth cone motility (e.g. growth and retraction) can be explained in part by the "set-point messenger" hypothesis (56). In this scheme, there is an optimal or set point level of calcium that promotes maximal growth. Increasing calcium levels toward the set point will increases motility, whereas increasing it above the set point decreases motility. Both spatial and temporal changes in calcium concentration are likely to play determining roles in growth cone behavior (57). ␣2M* affects calcium influx through NMDA channels (14,15,43), and here we show that it diminishes calcium oscillations (Fig. 5). Manipulating intracellular calcium affects the activation of MAP kinase, and such effects were steadied in the presence of ␣2M* (Fig. 5), implying that the effects of ␣2M* on activation of MAP kinase are subsequent to its effects on the modulation of intracellular calcium. We suggest that the stabilization of intracellular calcium by ␣2M* is involved in the promotion of neurite outgrowth in the calcium set-point messenger model. Activation of MAP kinase causes it to translocate to nuclei of neurons where it can then activate CREB (35,58,59). Thus, MAP kinase represents a point of convergence for cell surface signals regulating cell growth, division, differentiation, and protection (60). Two distinct mechanisms may mediate the early and late phases of CREB activation. The initial CREB activation may be dependent on calcium homeostasis and MAP kinase activation (Fig. 7, A and B), whereas the delayed CREB activation may be attributable to calcium signaling to NMDA, which is responsible for the dephosphorylation of CREB (61). CREB, a multipurpose transcription factor, is involved in the synaptic plasticity, cell division, proliferation, and protection of neurons from neurodegeneration (62,63). Activation of CREB through LRP-Ca 2ϩ -MAP kinase signaling pathway in primary cortical neurons fits well with the numerous effects of ␣2M* on neurons such as up-regulation of LRP expression (13), neuronal protection (16,17), and promotion of neurite outgrowth.
LRP ligands induced both neurite outgrowth and dendritic localization of LRP in primary neurons (Figs. 1 and 2). These effects were blocked by co-incubation with RAP, the inhibitor of the low density lipoprotein receptor family. RAP did not promote neurite outgrowth and decreased LRP expression in neuronal processes, causing a largely somatic distribution after RAP treatment. Our earlier work (13) shows that RAP treatment did not affect the overall levels of LRP in neurons (64). The redistribution of LRP by RAP could also play a role in the ability of RAP to prevent ␣2M* and other ligands from promoting neurite outgrowth via LRP. Fig. 9 illustrates a model of ␣2M* regulation of neurite outgrowth. In the early stages, LRP activation alters intracellular calcium homeostasis. Such calcium signaling initiates the downstream signaling by activating the PKC-MAP kinase cascade. Although activation of these kinases may be transient, their effects on the activation of the transcription factor CREB allow the effects to alter gene expression over long periods of time, leading to neurite outgrowth and LRP up-regulation (13). This pathway is apparent for different ligands of LRP including ␣2M* and apoE but not RAP, which does not cause changes in calcium homeostasis (14,15,43) and leads to a dramatic redistribution of LRP away from the cell surface (Fig. 8). Thus, the activation of LRP by ligands up-regulated temporarily during acute phase response (but present on amyloid deposits chronically) leads to a complex cascade of calcium and kinase signaling that promotes neurite outgrowth.