Par-4 Inhibits Choline Uptake by Interacting with CHT1 and Reducing Its Incorporation on the Plasma Membrane*

CHT1 is a Na+- and Cl–-dependent, hemicholinium-3 (HC-3)-sensitive, high affinity choline transporter. Par-4 (prostate apoptosis response-4) is a leucine zipper protein involved in neuronal degeneration and cholinergic signaling in Alzheimer's disease. We now report that Par-4 is a negative regulator of CHT1 choline uptake activity. Transfection of neural IMR-32 cells with human CHT1 conferred Na+-dependent, HC-3-sensitive choline uptake that was effectively inhibited by cotransfection of Par-4. Mapping studies indicated that the C-terminal half of Par-4 was physically involved in interacting with CHT1, and the absence of Par-4·CHT1 complex formation precluded the loss of CHT1-mediated choline uptake induced by Par-4, indicating that Par-4·CHT1 complex formation is essential. Kinetic and cell-surface biotinylation assays showed that Par-4 inhibited CHT1-mediated choline uptake by reducing CHT1 expression in the plasma membrane without significantly altering the affinity of CHT1 for choline or HC-3. These results suggest that Par-4 is directly involved in regulating choline uptake by interacting with CHT1 and by reducing its incorporation on the cell surface.

Cholinergic neurotransmission plays a major role in the regulation of many physiological functions (1)(2)(3). For example, basal forebrain cholinergic neurons are involved in learning and memory processes, and hypoactivity of the cholinergic system may be directly responsible for the cognitive deficit in Alzheimer's disease (AD) 1 (4 -8). Reductions in choline acetyltransferase (ChAT) activity and a profound loss of cholinergic neurons in the nucleus basalis magnocellularis have been observed in AD brain (5)(6)(7)(8)(9)(10)(11)(12). High affinity choline uptake is generally believed to be the rate-limiting step in acetylcholine synthesis in cholinergic neurons and is essential for cholinergic transmission in the central nervous system (13)(14)(15)(16). Reduced choline uptake has been implicated in degenerating neurons in AD (15,17), although spared cholinergic nerve terminals in AD may show a compensatory increase in choline uptake (18,19).
CHT1 is a Na ϩ -and Cl Ϫ -dependent, hemicholinium-3 (HC-3)sensitive, high affinity choline transporter, and it is located specifically in cholinergic neurons (13). The human CHT1 (hCHT1) gene is ϳ25 kb in length, contains nine exons, and has been mapped to chromosome II at IIq11-12 (13,20,21). The hCHT1 protein is composed of 580 amino acids with a predicted molecular mass of ϳ63 kDa. Sequence analysis of the hCHT1 protein predicts 13 transmembrane domains (13,20,21). Par-4 (prostate apoptosis response-4) is a leucine zipper protein that plays an important role in neuronal dysfunction and cell death in neurodegenerative disorders (22)(23)(24)(25)(26)(27)(28). Levels of Par-4 expression are significantly increased in cortical, hippocampal, and cholinergic neurons in in vivo and in vitro models of AD (24), and blockade of Par-4 expression prevents neuronal death (23,24,27,29). In addition, Par-4 contributes to apoptosis and aberrant production of amyloid ␤-peptide 1-42 in neural cells expressing presenilin-1 mutations (24,30). Of importance, an aberrant increase in Par-4 activity results in a significant reduction of ChAT activity under both apoptotic and non-apoptotic conditions (31). These results indicate that Par-4 may directly and/or indirectly regulate cholinergic signaling. We now report that Par-4 colocalizes and interacts with the high affinity choline transporter CHT1 in neuronal cells and that Par-4⅐CHT1 complex formation inhibits the choline uptake activity of CHT1 by reducing the incorporation of CHT1 on the cell surface.

MATERIALS AND METHODS
Primary Neuronal Cultures, Analysis of Cholinergic Neuronal Phenotype, and Trophic Factor Withdrawal-Dissociated hippocampal, cortical, cerebellar, and basal forebrain neuronal cultures were prepared from postnatal day 1 mouse pups using methods similar to those described previously (27). Briefly, hippocampal, cortical, and cerebellar tissues and the pillow-like basal forebrains were removed and incubated for 15 min in Ca 2ϩ -and Mg 2ϩ -free Hanks' balanced saline solution (Invitrogen) containing 0.2% papain. Cells were dissociated by trituration and plated onto polyethyleneimine-coated plastic or glassbottom culture dishes containing minimal essential medium with Earle's salts supplemented with 10% heat-inactivated fetal bovine serum, 2 mM L-glutamine, 1 mM pyruvate, 20 mM KCl, 10 mM sodium bicarbonate, and 1 mM HEPES (pH 7.2). Following cell attachment (3-6 h post-plating), the culture medium was replaced with Neurobasal medium with B-27 supplements (Invitrogen). Experiments were performed in 7-day-old cultures. Expression of the cholinergic neuronal phenotype in basal forebrain cultures was determined by measurement of the specific activity of ChAT in the cultures as described in our previous study (31). ChAT activity is expressed as picomoles of acetylcholine synthesized per mg of total protein in a 30-min period. In addition, as a measure of physiological responsiveness of the basal forebrain cholinergic neurons, nerve growth factor (NGF)-mediated changes in ChAT activity were monitored, as it has been shown that NGF enhances the specific activity of this cholinergic phenotypic marker (31). Primary stocks of mouse 2.5 S NGF were made at 100 g/ml in 0.02% acetic acid. Working solutions of NGF were prepared at 10 g/ml in phosphatebuffered saline (PBS) containing 1 mg/ml bovine serum albumin. Trophic factor withdrawal (TFW) was initiated by washing cultures four times with Locke's buffer (154 mM NaCl, 5.6 mM KCl, 2.3 mM CaCl 2 , 1.0 mM MgCl 2 , 3.6 mM NaHCO 3 , 5 mM glucose, and 5 mM HEPES (pH 7.2)) with subsequent incubation in Locke's buffer. An antisense Par-4 oligonucleotide (5Ј-ATAGCCGCCGGTCGCCATGTT-3Ј) was prepared as 1 mM stocks in sterile water and diluted directly into the culture medium.
Transfection of Neuroblastoma IMR-32 Cells-For transfection of  IMR-32 cells with hCHT1 cDNA, specific PCR primers (forward primer  with added NotI cloning site, 5Ј-gcggccgccatggctttccatgtggaag-3Ј; and  reverse primer with added XhoI cloning site, 5Ј-ctcgagtcactgtaaattatcttcagtcccag-3Ј; designed based on the nucleotide sequence of hCHT1 cDNA, GenBank TM /EBI Data Bank accession number AF276871) were used to amplify the hCHT1 cDNA from human spinal cord cDNA template (Clontech). The PCR-amplified hCHT1 cDNA was subcloned into the NotI and XhoI sites of expression vector pcDNA3.1(ϩ) (Invitrogen) and subsequently confirmed by sequencing. The expression construct was designated pCMV-hCHT1. A full-length Par-4 cDNA was subcloned into the expression vector pREP4 (Invitrogen), yielding the recombinant construct pREP4-Par-4, which encodes a 1.2-kb RNA species and the full-length 38-kDa Par-4 protein (24). Another cDNA fragment representing the C-terminal half deletion mutant of Par-4 (Par-4⌬CTH, in which nucleotides 541-1267 in Par-4 cDNA were deleted) was similarly subcloned into the expression vector pREP4, yielding the recombinant construct pREP4-Par-4⌬CTH, which encodes a 700-bp RNA species and the N-terminal half (i.e. 163 amino acids, ϳ18 kDa) of the Par-4 protein, as described in our previous studies (23,24). Human neuroblastoma IMR-32 cells (American Type Culture Collection) were maintained at 37°C in an atmosphere of 95% air and 5% CO 2 in Eagle's minimal essential medium supplemented with nonessential amino acids and 10% heat-inactivated fetal bovine serum. Human IMR-32 cells were transfected with pCMV-hCHT1 or pREP4-Par-4 using Lipo-fectAMINE (Invitrogen) according to the manufacturer's instructions. Additional IMR-32 cells were cotransfected with pREP4-Par-4 and pCMV-hCHT1. For mapping studies, IMR-32 cells were cotransfected with pREP4-Par-4⌬CTH and pCMV-hCHT1. For control purposes, parallel cultures of IMR-32 cells were transfected with pcDNA3.1 and/or pREP4 vectors alone. Transfected cells were selected with G418 (400 g/ml; for cells transfected with pCMV-hCHT1) and/or hygromycin (400 g/ml; for cells transfected with pREP4-Par-4) for 4 weeks. Surviving clones were selected for this study based on their similar levels of expression of hCHT1 and/or Par-4 as assessed by Western blotting.
Immunocytochemistry and Confocal Laser Scanning Microscopy-The affinity-purified rabbit anti-hCHT1 polyclonal antibody was a generous gift from Dr. Takashi Okuda (University College London, London, United Kingdom). This antibody has been reported to be able to detect the CHT1 protein on Western blots as two major ϳ45and 80-kDa bands that represent oligomers and/or heterogeneous glycosylation of the CHT1 protein (32,33). In CHT1-transfected IMR-32 cells, the antibody recognized three major bands of ϳ80, 68, and 45kDa. The 68-kDa band corresponds closely to the molecular mass predicted from hCHT1 primary structure (see "Results"). IMR-32 cells expressed little or no endogenous CHT1 (see "Results"). The anti-Par-4 antibody was a mouse monoclonal antibody raised against full-length rat Par-4 (Santa Cruz Biotechnology). This antibody recognizes both human and rodent Par-4 proteins at ϳ38 kDa. For immunocytochemical analysis of hCHT1 and Par-4 expression, the cultured cells were fixed for 30 min in 4% paraformaldehyde and PBS, and membranes were permeabilized by incubation in 0.2% Triton X-100 and PBS. Cells were incubated for 1 h in blocking serum (5% normal goat serum in PBS). Cells were then exposed to primary antibodies (1:1000 dilution of rabbit anti-hCHT1 polyclonal antibody and 1:100 dilution of mouse anti-Par-4 monoclonal antibody) overnight at 4°C, followed by incubation for 1 h with a mixture of Texas Red-labeled anti-rabbit and fluorescein-labeled antimouse secondary antibodies (Vector Labs, Inc., Burlingame, CA). Images of Par-4 and/or CHT1 immunofluorescence were acquired using a confocal laser scanning microscope (dual wavelength scan) with a ϫ60 oil immersion objective. All images were acquired using the same laser intensity and photodector gain to allow quantitative comparisons of relative levels of fluorescence in the cells. The average pixel intensity/ cell and sites of colocalization of immunoreactivities were determined using Fluoview Version 2.0 software.
[ 3 H]Choline Uptake Assay-[ 3 H]Choline uptake assays were performed essentially as described previously (33,34 In these experiments, experimental data were normalized by protein content, and specific [ 3 H]choline uptake is presented as percent uptake above background (background choline uptake was determined in cells transfected with the empty vectors alone). The Na ϩ dependence of [ 3 H]choline uptake was assessed in buffer containing 140 mM NaCl, 2 mM KCl, 1 mM CaCl 2 , 1 mM MgCl 2 , 10 mM HEPES, 5 mM Tris, and 10 mM glucose (pH 7.4) using isotonic replacement of NaCl with LiCl as described previously (33,34). To analyze the effect of Par-4 on the choline uptake activity of hCHT1, IMR-32 cells were cotransfected with CHT1 and Par-4; the [ 3 H]choline uptake activity of hCHT1 was analyzed at increasing concentrations of Immunoprecipitation/Western Blot Analysis-The methods used have been described previously (35). In brief, to examine Par-4⅐CHT1 complex formation, aliquots of cell lysates containing 200 g of protein were incubated for 1 h at 4°C with appropriate dilutions of mouse anti-Par-4 or rabbit anti-hCHT1 antibodies in immunoprecipitation buffer (150 mM NaCl, 2 mM EDTA, 1% Nonidet P-40, 5 g/ml leupeptin, 5 g/ml aprotinin, 2 g/ml pepstatin A, 0.25 mM phenylmethylsulfonyl fluoride, and 50 mM Tris (pH 7.6)). Protein A-Sepharose resin (30 l/sample; Amersham Biosciences) was added to the extracts for collection of the immunocomplexes, which were then washed three times with immunoprecipitation buffer and solubilized by heating in Laemmli buffer containing 2-mercaptoethanol at 100°C for 4 min. 50 g of solubilized proteins were separated by electrophoresis on a 4 -12% gradient SDS-polyacrylamide gel and then transferred to a nitrocellulose sheet. For Western blot analysis, the nitrocellulose sheet was blocked with 5% milk, followed by 1-h incubation in the presence of anti-hCHT1 or anti-Par-4 primary antibody. The membrane was further processed using horseradish peroxidase-conjugated secondary antibody, and immunoblotted proteins were detected by chemiluminescence using the ECL system (Amersham Biosciences). To further confirm that Par-4/CHT1 interaction was specific, similar immunoprecipitation/Western blotting methods were employed to examine whether Par-4 interacts with a similar transporter protein, the sodium/ glucose cotransporter SGLT1, using a rabbit anti-SGLT1 polyclonal antibody (Chemicon International, Inc.). Equal loading was verified by probing the blots with anti-␤-actin antibody. Western blot images were acquired and quantified using Kodak Image Station 2000R and Kodak Digital Science 1D Version 3.6 software.
[ 3 H]Hemicholinium-3 Binding Assay-[ 3 H]HC-3 (128 Ci/mmol; PerkinElmer Life Sciences) binding experiments were performed using intact cells as described previously (33). In brief, transfected cells were washed with Krebs-Ringer HEPES buffer and preincubated for 1-2 h at 37°C for equilibrium. Binding assays were performed at 4°C for 1-2 h and terminated by three washes with ice-cold Krebs-Ringer HEPES buffer. Saturation binding was determined using 0.1-10 nM [ 3 H]HC-3 with or without 1 M unlabeled HC-3. Specific binding was determined by subtracting binding in the presence of 1 M unlabeled HC-3 from total binding. Because HC-3 is hydrophilic and cell membrane-impermeable, the HC-3 binding activity represents the expression level of the transporter in the plasma membrane (33). K d for HC-3 was calculated by nonlinear least-square fits using the Hill equation (SigmaPlot Version 8.0).
Cell-surface Biotinylation Assay-These assays were used to analyze the cell-surface expression of CHT1 because biotinylated proteins are considered to be mostly expressed on the cell surface since they are greatly depleted of the intracellular marker actin (33). The methods were essentially the same as described previously (33). In brief, cells were treated with sulfosuccinimidyl 2-(biotinamido)ethyl-1,3Ј-dithiopropionate (1 mg/ml; Pierce) in PBS/calcium/magnesium buffer (138 mM NaCl, 2.7 mM KCl, 1.5 mM KH 2 PO 4 , 9.6 mM Na 2 HPO 4 , 0.1 mM CaCl 2 , and 1 mM MgCl 2 (pH 7.4)) at 4°C for 30 min, followed by washing with 100 mM glycine in PBS/calcium/magnesium buffer. Cells were solubilized with assay buffer (10 mM Tris (pH 7.4), 150 mM NaCl, 1% Triton-X-100, 1% sodium deoxycholate, and 0.1% SDS) and then centrifuged at 17,000 ϫ g for 30 min at 4°C. Biotinylated proteins were separated from non-biotinylated proteins by incubation with UltraLink TM Plus immobilized streptavidin gel (Pierce) for 1 h at room temperature.
Beads were washed five times with 1 ml of assay buffer, and adsorbed proteins were eluted with SDS sample buffer (50 mM Tris-HCl (pH 6.8), 10% glycerol, 2% SDS, 6% ␤-mercaptoethanol, and 0.001% bromphenol blue) for 30 min at room temperature. 500 g of biotinylated proteins were separated by electrophoresis on a 4 -12% gradient SDS-polyacrylamide gel and then transferred to a nitrocellulose sheet. Western blot analysis with anti-CHT1 or anti-SGLT1 antibody was then performed as described above. Blots were stripped and reprobed with anti-actin antibody (Sigma), which was used as a marker for intracellular proteins. Immunoblot images were acquired and quantified using Kodak Image Station 2000R and Kodak Digital Science 1D 3.6 software.

Par-4 Is Present in Basal Forebrain Cholinergic Neurons and
Colocalizes in the Same Neuron Positive for the Cholinergic Marker CHT1-Primary neuronal cultures of different brain regions (cortical, hippocampal, cerebellar, and basal forebrain) were established, and expression of the cholinergic neuronal phenotype in basal forebrain cultures was confirmed by measurement of the specific activity of ChAT in the cultures as described in our previous study (31). As shown in Fig. 1, primary basal forebrain neuronal cultures showed the most significant amount of basal ChAT activity (an average of ϳ3832 pmol/mg of protein/30 min) compared with cortical, hippocampal, and cerebellar neuronal cultures. In addition, NGF-mediated changes in ChAT activity were also significantly greater in basal forebrain neuronal cultures than in cortical, hippocampal, and cerebellar cultures (Fig. 1b). These results are consistent with the overall cholinergic phenotype of the basal forebrain neuronal cultures. Western blot analysis showed that basal levels of Par-4 were relatively high in the basal forebrain, hippocampus, and cerebral cortex and low in the cerebellum (Fig. 1, c and d). We next examined whether Par-4 colocalizes in the same basal forebrain neuron positive for the cholinergic marker CHT1. Immunocytochemical analysis of cultured basal forebrain cholinergic neurons revealed that both Par-4 and CHT1 were predominantly located in cytoplasmic compartments (Fig. 1e). This observation is consistent with previous reports showing that, although CHT1 is constantly cycled through the plasma membrane and intracellular compartments, CHT1 is predominantly intracellular (33,34). Double labeling immunocytochemistry using confocal laser scanning microscope further showed significant colocalization between Par-4 and hCHT1 in the perikaryon and neurites (Fig. 1e).

CHT1 Confers Na ϩ -dependent, HC-3-sensitive Choline Uptake in IMR-32 Cells: Inhibition by Par-4 in Transfected IMR-32 Cells and in Basal Forebrain Neurons-Since
Par-4 is present and colocalizes with CHT1 in cholinergic neurons, we next tested the hypothesis that Par-4 regulates the specific choline uptake activity of CHT1 in transfected human neuroblastoma IMR-32 cells. The IMR-32 cell line exhibits many cholinergic and adrenergic properties and has been used a model for studying cholinergic and adrenergic neurotransmission (36). The results from representative Western blot analysis showing increased levels of expression of Par-4 and/or CHT1 in transfected IMR-32 cells are given in Fig. 2 (a and b). Note that anti-CHT1 antibody detected the CHT1 protein on Western blots as three major bands of ϳ80, 68, and 45kDa. The 68-kDa band corresponds closely to the molecular mass predicted from hCHT1 primary structure, whereas various other bands may represent oligomers and/or heterogeneous glycosylation of the CHT1 protein. In examining the functional properties of CHT1, we found that transfection of CHT1 resulted in an ϳ2.3-fold increase in specific [ 3 H]choline uptake compared with vector-transfected control cells (Fig. 2c). The increased [ 3 H]choline uptake was completely abolished in the presence of 1 M HC-3. Equimolar replacement of Na ϩ with Li ϩ also abol-ished the choline transport activity of the CHT1-transfected cells (Fig. 2c). The CHT1-mediated [ 3 H]choline uptake was concentration-dependent and saturable (Fig. 2d). Since IMR-32 cells express little or no endogenous CHT1, the background levels of choline uptake observed in vector-transfected control cells and in Par-4-transfected cells were not significantly affected by HC-3 or replacement of Na ϩ with Li ϩ . These results demonstrate that transfection of CHT1 confers Na ϩ -dependent, HC-3-sensitive choline transport activity in IMR-32 cells. Coexpression of Par-4 resulted in a significant (ϳ41%) reduction of Na ϩ -dependent, HC-3-sensitive choline uptake mediated by CHT1 (Fig. 2c). A dose-response analysis with increasing concentrations of [ 3 H]choline also showed that specific choline uptake was significantly reduced in cells coexpressing Par-4 and CHT1 compared with cells transfected with CHT1 alone (Fig. 2d). Kinetic analysis indicated that cells coexpressing Par-4 and CHT1 showed a significant reduction in V max , with no apparent alteration in CHT1 affinity for choline ( Fig.  2d and Table I). To further confirm that the inhibitory effect of Par-4 on CHT1 transporter activity was not an artifact of transfection and overexpression proteins, we examined whether induction of Par-4 would decrease CHT1-mediated choline uptake in primary basal forebrain neurons. We previously reported that apoptotic insults, such as TFW, induce aberrant Par-4 expression in primary neurons (24,28,31,35). As shown in Fig. 2 e and f), a significantly decreased CHT1 transporter activity was observed in primary cholinergic neurons 8 h following TFW, the time point of maximal induction of Par-4 expression after TFW. The decrease in CHT1 activity was largely blocked by pretreatment of neurons with the antisense Par-4 oligonucleotide. These results show that induction of Par-4 was largely responsible for the observed decrease in CHT1 activity following TFW in basal forebrain neurons. The data suggest that inhibition of CHT1 activity by Par-4 is a physiologically relevant and pathologically meaningful event.
Par-4 Interacts with CHT1 in Transfected IMR-32 Cells and in Primary Neurons-The observations that Par-4 colocalizes with CHT1 in cholinergic neurons and that coexpression of Par-4 in IMR-32 cells results in a significant decrease in CHT1mediated, Na ϩ -dependent, HC-3-sensitive choline uptake indicate that Par-4 might directly modulate CHT1 transporter activity. Indeed, this hypothesis was supported by data from our co-immunoprecipitation studies showing that Par-4 physically interacted with CHT1 in transfected IMR-32 cells and in primary neurons. In transfected IMR-32 cells coexpressing Par-4 and CHT1, when immunoprecipitation was performed using the specific antibody against CHT1, a 38-kDa Par-4 band was clearly detected on the immunoblots (Fig. 3a). Reverseorder immunoprecipitation/Western blot analysis of the same transfected cell lines showed similar Par-4⅐CHT1 complex formation (Fig. 3b). Similar Par-4⅐CHT1 complex formation was confirmed in primary basal forebrain neurons at physiological concentrations of these proteins (Fig. 3, c and d). These results demonstrate that Par-4⅐CHT1 complex formation is physiologically relevant. To confirm that Par-4/CHT1 interaction is specific, we examined whether Par-4 would interact with the sodium/glucose cotransporter SGLT1. SGLT1 is a member of the Na ϩ /glucose cotransporter family, shares a relatively significant amino acid homology with CHT1, and is reportedly expressed in the kidney, intestine, and brain (37). Immunoprecipitation/Western blot experiments using tissue extract from the hippocampus and basal forebrain showed that SGLT1 was detected in proteins from total lysate, but not from those immunoprecipitated with anti-Par-4 or anti-CHT1 antibody, indicating that neither Par-4 nor CHT1 interacted with SGLT1 (Fig. 3, e and f).   (23,24). When immunoprecipitation was performed using the specific antibody against CHT1, a 38-kDa Par-4 band was detected in cells cotransfected with full-length Par-4 and CHT1, but not in cells cotransfected with Par-4⌬CTH and CHT1, indicating that it is the C terminus (not the N terminus) of Par-4 that is involved in interacting with CHT1 (Fig. 4a). This finding was further confirmed by reverse-order immunoprecipitation/Western blot analysis of the same transfected cell lines, showing lack of Par-4⅐CHT1 complex formation in cells coexpressing Par-4⌬CTH and CHT1 (Fig. 4b). These results indicate that one or more domains located within the C terminus of Par-4 (amino acids 164 -332) are responsible for interacting with CHT1. We next examined whether Par-4⅐CHT1 complex formation is required for the inhibitory effect of Par-4 on the specific [ 3 H]choline uptake activity of CHT1. As shown in Fig. 4c, although cells coexpressing full-length Par-4 and CHT1 (in which Par-4 forms a complex with CHT1) showed a significantly reduced choline uptake activity of CHT1, coexpression of CHT1 with the C-terminal half deletion mutant of Par-4 (Par-4⌬CTH, which abolishes the interaction between Par-4 and CHT1) precluded the loss of the choline uptake function of CHT1 induced by Par-4.

Par-4⅐CHT1 Complex Formation Alters CHT1 Levels in the Plasma Membrane without a Significant Effect on the Affinity of CHT1 for Choline or HC-3-We next examined whether
Par-4 affects the delivery of CHT1 to the cell membrane using [ 3 H]HC-3 binding assays. Because HC-3 is cell membraneimpermeable, the HC-3 binding activity represents the expression levels of CHT1 in the plasma membrane (33). As shown in Fig. 5a, HC-3 binding was significantly reduced in cells coexpressing Par-4 and CHT1 compared with cells transfected with CHT1 alone. On the other hand, abolishing Par-4/CHT1 interaction by coexpression of CHT1 with Par-4⌬CTH precluded the decrease in HC-3 binding induced by Par-4. Further kinetic analysis indicated that cells coexpressing Par-4 and CHT1 showed an average of 40.6% reduction in V max in choline uptake assays and 31.9% reduction in B max in [ 3 H]HC-3 binding studies, with no apparent alterations in K m for choline or in K d for HC-3 binding (Figs. 2d and Fig. 5a and Table I). The reduced levels of CHT1 at the cell surface in cells transfected with Par-4 was further confirmed by cell-surface biotinylation assays. Biotinylated proteins are considered to be mostly expressed on the cell surface because they are greatly depleted of the intracellular marker actin. As shown in Fig. 5b, immunoblotting of biotinylated protein fractions using anti-CHT1 antibody demonstrated a significant decrease in cellsurface CHT1 immunoreactivity in cells cotransfected with Par-4 and CHT1. Note that little actin immunoreactivity was detected in biotinylated (cell-surface) fractions, indicating very limited contamination of biotinylated fractions with intracellular proteins. To further confirm that the inhibitory effect of Par-4 on cell-surface expression of CHT1 is specific, we examined whether Par-4 would alter cell-surface expression of SGLT1 using cell-surface biotinylation assays. As shown in Fig. 5c, immunoblotting of biotinylated proteins using anti-SGLT1 antibody demonstrated no significant change in cell-surface SGLT1 immunoreactivity in all cell lines examined, regardless of whether Par-4 was transfected or not. These results provide strong and direct experimental evidence that Par-4 overexpression leads to a specific decrease in the amount of CHT1 on the plasma membrane. Taken together, the data suggest that Par-4 reduces the choline uptake by forming a complex with CHT1 via domain(s) located in the C terminus and by inhibiting cellsurface expression of CHT1 without significantly altering the affinity of CHT1 for choline or HC-3.

DISCUSSION
Cholinergic neurotransmission is involved in the regulation of many complex motor, autonomic, behavioral, and cognitive functions (1)(2)(3). Profound loss of cholinergic function has been implicated in several neurodegenerative diseases, such as AD (4 -8, 38, 39). Because cholinergic neurons lack the capacity to synthesize choline de novo, their function depends upon choline uptake. More important, Na ϩ -dependent, HC-3-sensitive, high affinity choline uptake is unique to cholinergic neurons and is generally believed to be the rate-limiting step in acetylcholine synthesis (13). These observations suggest that studies of the regulatory machinery of choline uptake should have significant implications regarding our understanding of normal cholinergic neurotransmission and pathogenesis of cholinergic hypoactivity in AD. The cloning and functional characterization of

alters functional properties of CHT1 in transfected IMR-32 cells
Kinetic analysis showed that coexpression of Par-4 with CHT1 resulted in significant reductions in V max in choline uptake assays and in B max in ͓ 3 H͔HC-3 binding studies, with no apparent alterations in K m for choline or in K d for HC-3 binding. These results indicate that Par-4 reduced the choline uptake by inhibiting cell-surface expression of CHT1 without significantly altering the affinity of CHT1 for choline or HC-3. elegans, rat, human, mouse, Torpedo, and Limulus) mark a significant breakthrough in this effort (13, 20, 21, 33, 34, 40 -48). The data from this study demonstrate that Par-4, a leucine zipper protein previously found to play a significant role in neuronal degeneration and cholinergic signaling, is a novel and effective regulator of CHT1-mediated, Na ϩ -dependent, HC-3sensitive choline uptake. Although Par-4 was initially identified as playing an important role in hippocampal and cortical neuronal cell death in AD (24), an increase in Par-4 expression was subsequently shown to alter elements of cholinergic signaling even under nonapoptotic conditions. For example, overexpression of Par-4 results in a significant non-apoptotic reduction of ChAT activity in transfected neural cells (31), suggesting that Par-4 may be involved in the regulation of cholinergic function. The results from the present study indicate the following. 1) Par-4 was present in primary basal forebrain cholinergic neurons. 2) Transfection of IMR-32 cells with CHT1 conferred Na ϩ -dependent, HC-3-sensitive choline uptake that was effectively inhibited by cotransfection of Par-4. This inhibitory effect of Par-4 was physiologically relevant because it was also observed in primary basal forebrain neurons. 3) Par-4 colocalized with CHT1 in basal forebrain neurons, and co-immunoprecipitation studies showed that Par-4 interacted with CHT1 in transfected IMR-32 cells and in basal forebrain neurons. 4) Neither Par-4 nor CHT1 interacted with SGLT1, a member of the Na ϩ / glucose cotransporter family that shares a significant amino acid homology with CHT1, indicating that Par-4/CHT1 interaction is specific. 5) The CHT1 interaction domain was mapped to the C-terminal half of Par-4 using a C-terminal half deletion mutant of Par-4 (Par-4⌬CTH). 6) Coexpression of CHT1 with Par-4⌬CTH, which abolishes the interaction between Par-4 and CHT1, precluded the loss of the choline uptake function of CHT1 induced by Par-4, indicating that Par-4⅐CHT1 complex formation is necessary for the inhibitory effect of Par-4 on CHT1 activity. 7) Par-4 reduced the choline uptake by inhibiting CHT1 expression on the cell surface without significantly altering the affinity of CHT1 for choline or HC-3. These results suggest that Par-4 plays a direct role in regulating the choline uptake activity of CHT1 in neural cells.
The precise mechanisms by which Par-4⅐CHT1 complex formation alters choline uptake need to be carefully examined. It is noteworthy that, although aberrant induction of Par-4 expression sensitizes neural cells to a variety of apoptotic insults, overexpression of Par-4 alone does not significantly alter cell  4 -6). Preimmune rabbit serum was used as a control (lanes 7-9). The Par-4⅐CHT1 complex was clearly observed in all three clones of transfected cells (lanes 4 -6). The Par-4 protein in the transfected cells was detected on Western blots as a major band of ϳ38 kDa. In b, reverse-order immunoprecipitation/Western blot analysis of the same transfected IMR-32 cells showed similar Par-4⅐CHT1 complex formation (lanes 4 -6). c and d, interaction between endogenous CHT1 and Par-4 in basal forebrain neurons. In c, proteins from total lysates of primary basal forebrain neurons were immunoprecipitated with rabbit anti-hCHT1 antibody (lane 2), followed by Western blotting with anti-Par-4 antibody. Preimmune rabbit serum was used as a control (lane 3). In d, reverse-order immunoprecipitation/Western blot analysis of the same basal forebrain neurons showed similar endogenous Par-4⅐CHT1 complex formation. e and f, neither Par-4 nor CHT1 interacts with the sodium/glucose cotransporter SGLT1. In e, proteins from total lysates of the hippocampal and basal forebrain tissues were immunoprecipitated with rabbit anti-Par-4 antibody (lanes 3 and 4), followed by Western blotting with anti-SGLT1 antibody. Similar data were obtained by reverse-order immunoprecipitation/Western blot analysis of the same transfected cells (data not shown). In f, proteins from total lysates of the hippocampal and basal forebrain tissues were immunoprecipitated with anti-CHT1 antibody (lanes 3 and 4), followed by Western blotting with anti-SGLT1 antibody. Similar data were obtained by reverse-order immunoprecipitation/Western blot analysis of the same transfected cells (data not shown). All Input lanes show 10% of the total protein used in immunoprecipitation experiments. growth rate and is not sufficient to induce spontaneous apoptosis in transfected IMR-32 cells. Therefore, it is unlikely that the reduction in choline uptake induced by overexpression of Par-4 in IMR-32 cells was due to an alteration in growth rate or induction of the cell death process. Although CHT1 is predominantly intracellular (34), CHT1 shuttles constitutively between the plasma membrane and endocytic compartments, and synaptic vesicles may function as a reservoir to deliver CHT1 to the plasma membrane following neuronal excitation (34). Since, under physiological conditions, Par-4 is located primarily in cytoplasmic compartments and is enriched in synaptic terminals (49), one possibility is that Par-4⅐CHT1 complex formation results in an increased retention of CHT1 in intracellular compartments, thereby reducing the incorporation of CHT1 on the plasma membrane. Our data clearly support this hypothesis. The fact that Par-4 interacts with CHT1 in transfected IMR-32 cells as well as in primary neurons demonstrates the physiological relevance of Par-4⅐CHT1 complex formation. The observation that neither Par-4 nor CHT1 interacted with the CHT1 homolog protein SGLT1 indicates that Par-4/CHT1 interaction is specific. Indeed, mapping studies further showed that the C-terminal half of Par-4 was directly involved in interacting with CHT1. Kinetic analyses and data from our HC-3 binding assays clearly showed that Par-4/CHT1 interaction in cotransfected IMR-32 cells resulted in significant reductions in V max (in choline uptake assays) and in B max (in [ 3 H]HC-3 binding studies). These results indicate that Par-4⅐CHT1 complex formation results in an increased retention of CHT1 in intracellular compartments, thereby reducing the incorporation of CHT1 on the plasma membrane. Another possibility is that Par-4 binds to a critical functional domain of CHT1, resulting in an altered CHT1 conformation and therefore reduced binding affinity of CHT1 for choline and/or the substrate transport activity of CHT1. However, kinetic analyses found no apparent alterations in K m for choline or in K d for HC-3 binding, indicating that Par-4⅐CHT1 complex formation does not significantly affect the affinity of CHT1 for choline or HC-3.
The identification of Par-4 as a novel and effective regulator of CHT1 activity is of great significance. Profound loss of cholinergic function has been linked to memory deficit in AD, and accordingly, overall high affinity choline transport is reduced in AD (1, 5-8, 15, 17). Although spared neurons in AD brain may show compensatory up-regulation of the high affinity choline transporter (18,19), the data of the present study suggest that CHT1-mediated, Na ϩ -dependent, HC-3sensitive, high affinity choline uptake is most likely to be significantly reduced in degenerating neurons wherein Par-4 is aberrantly overexpressed. Since aberrant induction of Par-4 expression has been observed in many degenerating neurons in AD patients (24) and since Par-4 expression is induced during the early phase of apoptotic neuronal death (upstream of mitochondrial dysfunction and caspase activa- tion) (24), our data may reconcile the cholinergic and apoptotic hypotheses of AD and suggest that specific inhibition of Par-4 expression in vulnerable neurons in AD brain will not only block engagement of early apoptotic machinery, but also enhance CHT1-mediated, high affinity choline uptake in cholinergic neurons, thereby preserving an effective cholinergic neurotransmission in a stressed environment.
Developmental apoptosis is accomplished relatively quickly (within hours), whereas Alzheimer's pathology may require many years to form. This raises some concern about the significance of Par-4⅐CHT1 complex formation in cholinergic hypofunction in AD. However, the time course required for overall disease development does not necessarily match precisely the time required for neuronal cell death in AD for several reasons. 1) Since only a small number of neurons die at any given time in AD brain, the observation that severe clinical symptoms of AD may take years to develop cannot undermine the importance of acute apoptotic cell death at any given moment in the pathogenesis of AD. 2) Loss of neurons in certain vulnerable regions of AD brain may occur before any signs of dementia can be detected clinically, indicating that, although overall Alzheimer's pathology may require many years to form, the death of a particular neuron may occur relatively early and quickly. 3) It has been noted that not all apoptotic cell death takes the form of acute cell death. Some forms of apoptotic cell death in the central nervous system process may take much longer (days, months, or even years) to accomplish. For example, unlike developmental apoptosis, in which cell death ensues rapidly after initial caspase activation, it has been reported that partial activation of caspase-1 in N2a cells expressing the superoxide dismutase-1 G85R mutation does not provoke rapid cell death, indicating that traditional concepts of the mechanisms and timing of apoptosis may require revision in the context of a complex neurodegenerative death process evolving over months or years (50). In addition, in a complex in vivo environment, anti-apoptotic (cytoprotective) mechanisms are often activated in response to apoptotic insults. In cells in which apoptotic cascades have been initiated, but have not yet reached the point of no return, effective activation of cytoprotective mechanisms may delay or even reverse apoptotic cell death process in vivo. Thus, Par-4⅐CHT1 complex formation may play a significant role in cholinergic dysfunction in neurons undergoing either acute or prolonged cell death processes. Blots were stripped and reprobed with anti-actin antibody, which was used as a marker for intracellular proteins. Note that little actin immunoreactivity was detected in biotinylated (cell-surface) fractions, indicating very limited contamination of biotinylated fractions with intracellular proteins. Similar data were obtained in at least three separate clones of each of the transfected cell lines. c, representative immunoblot showing that transfection of Par-4 and/or CHT1 does not alter surface expression of SGLT1. Immunoblots of 500 g of biotinylated proteins using anti-SGLT1 antibody demonstrated no significant change in cell-surface SGLT1 immunoreactivity in all cell lines examined. Similar data were obtained in at least three separate clones of each of the transfected cell lines.