HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 275, Issue 26, 20179-20187, June 30, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/26/20179    most recent M001793200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Henderson, L. P. Articles by Maue, R. A. Search for Related Content PubMed PubMed Citation Articles by Henderson, L. P. Articles by Maue, R. A. Social Bookmarking                      What's this?

Embryonic Striatal Neurons from Niemann-Pick Type C Mice Exhibit Defects in Cholesterol Metabolism and Neurotrophin Responsiveness^*

Leslie P. Henderson^§, Li Lin, Anita Prasad, Colleen A. Paul^§, Ta Yuan Chang^§, and Robert A. Maue^§¶

From the Departments of  Physiology and ^§ Biochemistry, Dartmouth Medical School, Hanover, New Hampshire 03755

Received for publication, March 2, 2000, and in revised form, April 12, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Niemann-Pick type C (NP-C) disease is a progressive and fatal neuropathological disorder previously characterized by abnormal cholesterol metabolism in peripheral tissues. Although a defective gene has been identified in both humans and the npc^nih mouse model of NP-C disease, how this leads to abnormal neuronal function is unclear. Here we show that whereas embryonic striatal neurons from npc^nih mice can take up low density lipoprotein-derived cholesterol, its subsequent hydrolysis and esterification are significantly reduced. Given the importance of cholesterol to a variety of signal transduction mechanisms, we assessed the effect of this abnormality on the ability of these neurons to respond to brain-derived neurotrophic factor (BDNF). In contrast to its effects on wild type neurons, BDNF failed to induce autophosphorylation of the TrkB receptor and to increase neurite outgrowth in npc^nih neurons, despite expression of TrkB on the cell surface. The results suggest that abnormal cholesterol metabolism occurs in neurons in the brain during NP-C disease, even at embryonic stages of development prior to the onset of phenotypic symptoms. Moreover, this defect is associated with a lack of TrkB function and BDNF responsiveness, which may contribute to the loss of neuronal function observed in NP-C disease.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Niemann-Pick type C (NP-C)¹ disease is a fatal, autosomal recessive disorder resulting in progressive central nervous system deterioration and premature death. Analysis of this disorder has benefited from feline and murine (npc^nih mutant mice) models that recapitulate much of the human pathology (for review see Refs. 1 and 2). For example, both NP-C patients and npc^nih mutant mice are characteristically asymptomatic at birth. Tremor and ataxia occur later, and in npc^nih mutant mice these symptoms appear after approximately 1 month (3). Recently, a gene mutated in NP-C disease (NPC1) was cloned from humans (4) and mice (5). The NPC1 protein is thought to be important for cholesterol trafficking and metabolism (2, 6, 7), and a diagnostic hallmark of NP-C disease is lysosomal storage of lipids (primarily unesterified cholesterol) in peripheral tissues. Despite the fact that progressive neurological deterioration is ultimately the cause of premature death, neither the levels of cholesterol nor of phospholipids are grossly elevated in npc^nih mutant mouse brain (1 and 2) leaving the basis for this deterioration unknown.

Postmortem studies have characterized widespread anatomical abnormalities in brains of both humans and animals with NP-C disease (1). The severest pathology is often in regions involved with extrapyramidal motor control, including cerebellum, basal ganglia, red nucleus, and spinal cord (3, 8-11). For example, in the basal ganglia, distended neuronal cell somata with displaced nuclei, neurofibrillary tangles, and process degeneration are prevalent (12, 13). Despite the relative plethora of data describing morphological abnormalities, little is known about how neuronal function is compromised, when the biochemical and physiological abnormalities arise during development, or whether the defects originate in glia and/or neurons in NP-C disease.

Previous studies have demonstrated that cultures from embryonic striata are advantageous for assessing neuronal differentiation and the actions of neurotrophic factors. These cells can be maintained in vitro in a defined, serum-free medium (14, 15), and similar to intact striatum (for review see Refs. 16 and 17), more than 90% of the cultured cells are neurons that synthesize -aminobutyric acid (14, 18). Brain-derived neurotrophic factor (BDNF) promotes the differentiation of striatal neurons in vitro as it does in vivo (14, 18), consistent with expression of TrkB receptors in these neurons (19). These neurons also respond to neurotrophin-3, but not to nerve growth factor (14, 18), consistent with the expression of TrkC, but not TrkA, in embryonic rat brain (20, 21).

Recent evidence indicates that localization of sphingolipids and cholesterol in membrane domains, or "lipid rafts," is essential for assembly and activity of specific transmembrane signaling complexes (for review see Refs. 22 and 23) and implies that defects in cholesterol metabolism could interfere with neurotrophin signaling and neuronal differentiation. Here, we demonstrate that abnormalities in the metabolism of exogenously supplied LDL-derived cholesterol are evident in embryonic striatal neurons from npc^nih mutant mice and that these deficits are accompanied by loss of BDNF-mediated neurite outgrowth and TrkB receptor activation. Consistent with this, TrkB activation also fails to occur in wild type striatal neurons depleted of cholesterol. The results are important for understanding neurotrophin-mediated signal transduction as well as the mechanisms underlying NP-C disease.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Primary Cultures of Embryonic Striatal Neurons from npc^nih Mice with Specific Genotypes-- Mice that were wild type (+/+), heterozygous (+/), or homozygous (/) with respect to the mutation in the NPC1 gene (5) were obtained from an established breeding colony of npc^nih mice at Dartmouth Medical School (original breeding pairs were provided courtesy of Dr. P. Pentchev, National Institutes of Health). Mice were maintained under temperature-controlled conditions on a 12:12 h light/dark cycle with food and water ad libitum. The genotype of individual animals, including embryos, was determined from genomic DNA isolated from a small piece of the tail, using minor modifications of the polymerase chain reaction (PCR)-based method and oligonucleotide primers described in previous work (5). Specifically, after 0.5-cm pieces of tail tissue were incubated overnight at 55 °C in 500 µl of lysis buffer (100 mM Tris, 5 mM EDTA, 0.2% SDS, 20 mM NaCl) containing 10 µl of 10 mg/ml proteinase K (Sigma), 5-µl aliquots of the supernatants were incubated with ~3.5 pmol of forward (5'-GGTGCTGGACAGCCA AGTA-3') and reverse (5'-GATGGTCTGTTCTCCCATG-3') primers (Operon Technologies Inc., Alameda, CA) for 5 min at 94 °C and for 30 PCR cycles (94 °C for 1 min, 55 °C for 2 min, 72 °C for 3 min). Agarose gel electrophoresis was then used to determine the size of the PCR products, which were indicative of the genotype (1048 bp for +/+, 1209 bp for /, and products of both sizes for +/ npc^nih mice). In the mice genotyped at the time of weaning (n = 517), the percentage of / mice (18%) was lower than predicted from simple Mendelian inheritance, similar to previous reports regarding postnatal animals (3). However, in the E16 embryos that were used for neuronal cultures (n = 388), 26% were of the / genotype.

A separate culture of striatal neurons was made from each embryo, and a tail snip was collected from the embryo for subsequent PCR analysis, so that all of the striatal neurons in a given culture dish had the same identified genotype. Previously reported techniques for culturing embryonic striatal neurons (14) were modified in order to routinely make cultures from the striatal tissue of an individual embryo. Briefly, striatal tissue was incubated at 37 °C for 15 min in serum-free Dulbecco's modified Eagle's medium containing 0.25% trypsin (Worthington) and then at 37 °C for 10 min in 0.2 mg/ml DNase (Sigma). The tissue pieces were then gently triturated in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum before the cell suspension was diluted with a chemically defined, serum-free culture medium (24). Neurons were then plated into tissue culture dishes coated with polylysine (Sigma) and merosin (Chemicon, Temecula, CA). For the analyses of morphology, TrkB surface expression, or TrkB activation, cells from a single embryo were plated into two grid-etched 35-mm culture dishes (Nunc, Naperville, IL) at a density of 3 × 10⁵ cells per dish. For analysis of lipid metabolism, striatal neurons from an individual embryo were plated into a single well of a 12-well tissue culture plate (Costar, Cambridge, MA). For the morphometric analyses, 4 h after the cells were plated, BDNF (Upstate Biotechnology, Inc., Lake Placid, NY) was added to one dish of each pair at a final concentration of 50 ng/ml and was present throughout the 7-day culture period. For analysis of TrkB activation, BDNF was added to one dish of each pair 3 days after plating, and the neurons were harvested 5 min later.

Analysis of GAD Immunocytochemistry-- The expression of the enzyme glutamic acid decarboxylase (GAD) was detected according to Mizuno et al. (18). Briefly, striatal cultures were rinsed with PBS 3 times for 5 min each before they were fixed in 4% paraformaldehyde in 0.12 M sucrose for 30 min at room temperature. After 3 more rinses with PBS for 10 min each, the neurons were permeabilized for 15 min in 0.2% Triton X-100 in PBS. The neurons were then incubated overnight at 4 °C with an anti-GAD mouse monoclonal antibody (catalog number 1522 825; Roche Molecular Biochemicals) at 1 µg/ml. Cultures were rinsed 3 times with PBS for 5 min each before they were incubated with a biotinylated goat anti-mouse IgG secondary antibody and then an avidin-biotin peroxidase complex (ABC Kit, Vector Laboratories, Burlingame, CA). Antibody labeling was visualized using 3,3'-diaminobenzidine tetrahydrochloride as the chromogen.

Analysis of Cholesterol and Phospholipid-- Estimates of total and free cholesterol were made according to modifications of procedures described in Cadigan et al. (25). Cells were lysed in 0.2 M NaOH, the NaOH samples neutralized by addition of HCl and phosphate buffer, and the lipids Folch-extracted according to Chang et al. (26). After drying under N₂, samples were resuspended in isopropyl alcohol, and cholesterol determinations were made using a fluorometric procedure (27), either with or without cholesterol esterase preincubation.

Total phospholipid content was determined using assays of inorganic phosphorus after acid hydrolysis, according to modifications of procedures described by Kagawa and Racker (28). Cultured neurons were lysed in 0.2 M NaOH and extracted with chloroform/methanol (2:1) and then dried under N₂. Following drying, a solution of 1% of ammonium molybdate and a 9:1 mixture of H₂SO₄/HClO₄ were added to the samples before they were vortexed and incubated at 180 °C for 1 h. Samples were cooled, diluted with distilled H₂O, and incubated in boiling water for 7-10 min in a solution consisting of Fiske and Subbarow Reducer (Sigma) and 5% ammonium molybdate. Samples were cooled and assayed by optical density.

The metabolism of LDL-derived cholesterol linoleate by cultures of embryonic striatal neurons was analyzed 72 h after plating. Human LDL (d = 1.019-1.063 g/ml) was prepared from plasma by sequential flotation in the presence of protease inhibitors as described previously (25). For measuring uptake, hydrolysis, and re-esterification of LDL-derived cholesteryl linoleate, LDL was labeled with [1,2,6,7-³H]cholesteryl linoleate according to Roberts et al. (29), using the lipoprotein-free fraction of fresh human serum as the source of cholesterol ester exchange protein. Preliminary experiments indicated that only minimal levels of hydrolysis and esterification were measurable after incubation with [³H]cholesterol linoleate-LDL for 3-12 h but that significant levels were observed after 24 h (data not shown). Therefore, for all of the experiments described, a 24-h incubation period was used. Neurons were incubated in medium containing 40-50 µg/ml [³H]cholesteryl linoleate-LDL, the cells harvested in 0.2 M NaOH, the lysate neutralized with HCl, and the lipids extracted as described previously (26). Extracted lipids were separated by TLC, using silver nitrate-impregnated Silica G TLC plates (Analtech Inc., Newark, DE). The R_F values were as follows: cholesterol ~0.1; cholesteryl linoleate ~0.3; and cholesteryl oleate ~0.6, respectively. Uptake, percent hydrolysis, and percent esterification were analyzed as described previously (30). Control experiments showed that there was an average recovery of 60% for cholesterol, cholesteryl oleate, and cholesteryl linoleate after lipid extraction and TLC analysis. Total cellular proteins were determined using a BCA Protein Assay Reagent Kit (Pierce).

Morphometric Analysis-- Cultures of embryonic striatal neurons maintained in the absence or presence of BDNF (Upstate Biotechnology, Inc.) for 7 days (see above) were fixed in 4% paraformaldehyde (Fisher). Minor modifications of previously published procedures (14, 31, 32) were then used to analyze the cells. Briefly, 20 images were captured from each dish using a Zeiss Axiovert microscope with a × 20 Achroplan objective modified for Hoffman modulation contrast optics (Carl Zeiss Inc., Thornwood, NY), a Dage-MTI 300-RC CCD camera (Michigan City, IN), and a Power Macintosh 8600/300. Beginning at a randomly chosen grid in the upper corner of the culture dish, every 6th grid was selected for image capture so that the extent of the culture was sampled without bias for specific cells. Each image was analyzed for the number of cell bodies, somal area, neurite length per cell body, the number of branch points per cell body, and the number of primary neurites per cell body using NIH Image software augmented with a subroutine for collecting and tabulating the results of the analysis (courtesy of C. Daghlian, Dartmouth Medical School). Data for each experimental condition and each genotype were collected from 3 to 5 separate platings and 3-5 animals.

Surface Biotinylation-- Cultures from a single embryo (see above) were washed 3 times with cold PBS before 1 dish of each pair was incubated in 1 ml of PBS containing 0.5 mg/ml sulfo-NHS-biotin (EZ Link; Pierce) for 30 min at room temperature. All of the cultures were then incubated in 2 ml of PBS containing 0.5% bovine serum albumin for 10 min before the cultures were washed 3 times with PBS and then lysed for 20 min at 4 °C in 0.1 ml of lysis buffer (500 mM NaCl, 0.1% SDS, 1% Triton X-100, 0.5% sodium deoxycholate, 0.02% sodim azide, and 10 mM sodium vanadate, to which 10 µM aprotinin, 1 µM leupeptin, and 1 mM phenylmethylsulfonyl fluoride were added just prior to use). After total protein concentration in the samples was determined using a BCA Protein Assay Reagent kit (Pierce), equivalent amounts of protein from biotinylated and non-biotinylated cells were then incubated with streptavidin-agarose (Pierce) for 2 h at 4 °C to precipitate biotinylated proteins. The precipitated proteins and aliquots of the remaining lysates were then separated by SDS-PAGE using 10% gels prior to Western blot analysis (see below).

Western Blot Analysis of TrkB Receptor Activation and Surface Expression-- To assess TrkB activation, cultures of embryonic striatal neurons that were untreated or treated with BDNF (see above) were lysed in 0.1 ml of lysis buffer (see above), and total protein concentration in the samples was determined using a BCA Protein Assay Reagent kit (Pierce). Equivalent amounts of protein from untreated and BDNF-treated cells were then separated by SDS-PAGE using 5% gels. For both the surface biotinylation assays and activation assays, proteins separated by SDS-PAGE were electrophoretically transferred to Immobilon-P polyvinylidene difluoride membranes (Millipore Corp., Bedford MA). After washing with TBS (20 mM Tris-Cl, 150 mM NaCl, pH 7.4) and then TBST (20 mM Tris-Cl, 150 mM NaCl, 0.05% Tween 20), membranes were blocked in TBST for 4 h at room temperature for anti-phosphotyrosine blotting or overnight at 4 °C in TBS for anti-TrkB, anti-Trk, or anti-tubulin blotting. Membranes were then incubated with a 1:1000 dilution of an anti-phosphotyrosine antibody (4G10; Upstate Biotechnology, Inc.) at room temperature overnight or incubated at room temperature for 1 h with either a 1:200 dilution of an anti-TrkB antibody (H121; Santa Cruz Biotechnology, Santa Cruz, CA), a 1:10,000 dilution of anti-tubulin antibody (DM1A; Sigma), or, in some cases, a 1:500 dilution of an affinity-purified pan-specific Trk polyclonal antibody (C14; Santa Cruz Biotechnology). The membranes were washed 3-4 times for 5 min each in TBST at room temperature before they were incubated with a secondary antibody (1:1000; goat anti-mouse IgG-POD/anti-rabbit IgG-POD, Roche Molecular Biochemicals) for 45 min. After 3-4 washes for 5 min each in TBST at room temperature, antibody binding was detected using an enzyme-linked chemiluminescence detection kit (Supersignal, Pierce) and visualized on autoradiographic film (Eastman Kodak Co.).

Cholesterol Depletion of Cultured Embryonic Striatal Neurons-- Cholesterol depletion of wild type embryonic striatal neurons was achieved using previously established procedures (33, 34). Specifically, after 3 days in culture, lovastatin (kindly provided by Merck) and mevalonate (Sigma) were added to the culture medium at final concentrations of 4 µM and 0.25 mM, respectively. Two days later, methyl--cyclodextrin (Sigma) was added to the culture medium at a final concentration of 5 mM for 15 min. BDNF (50 ng/ml) was then added to specific culture dishes, and after 5 min the cells were harvested. Control cells (untreated and BDNF-treated) were incubated in medium containing 0.01% Me₂SO, as Me₂SO was used in preparing the stock solution of lovastatin.

Statistical Analyses-- All values in the text represent the means ± S.E. Data were analyzed using a one-way analysis of variance followed by post hoc comparisons using a Student's two-tailed t test.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Abnormal Cholesterol Metabolism in Cultures of Embryonic Striatal Neurons from npc^nih Mice-- In peripheral tissues, the hallmark of the NP-C phenotype is abnormal lipid metabolism, which leads to the accumulation of unesterified cholesterol (1). However, in the central nervous system it was unclear from previous work whether abnormalities in cholesterol metabolism occur in neurons, particularly during critical periods of embryonic development. Whereas abnormal cholesterol accumulation has not been detected by biochemical assays of whole brain in NP-C disease (1, 2, 35), impaired cholesterol esterification has been reported for mixed neuronal/glial cultures from npc^nih mutant mice (36), and there are preliminary indications of cholesterol accumulation, as assessed by histological techniques, in individual neurons during NP-C disease.² Therefore, lipid metabolism was analyzed in cultures of embryonic striatal neurons from npc^nihmice that were wild type (+/+), heterozygous (+/), or homozygous (/) for the mutation in the NPC1 gene (5). These cultures are essentially glial-free and can be maintained in the absence of serum, which is advantageous for the analysis of lipid metabolism in neurons. However, mice homozygous for the NPC1 mutation do not breed successfully (1), and it was necessary to mate heterozygous mice in order to obtain embryos. Because these embryos could have different genotypes, previously established techniques for culturing striatal neurons (14) were modified in order to make cultures from an individual embryo, and a previously reported PCR-based strategy (5) was used to determine the genotype of each embryo at the time the cultures were made (see "Experimental Procedures").

Striatal cultures prepared from +/+, +/, and / npc^nih embryos appeared similar to each other and to cultures of rat embryonic striatal neurons described previously (14, 18). Cells exhibiting neuronal morphology and obvious neurite outgrowth were abundant in all of the cultures (Fig. 1). As previously reported for the rat cultures (14), greater than 95% of these cells expressed neuron-specific enolase, and there were no significant differences in the average number of immunopositive neurons in cultures prepared from +/+ and / npc^nih mice (data not shown). Moreover, in cultures analyzed for the expression of GAD, the synthetic enzyme for the neurotransmitter -aminobutyric acid, 96.9% of the +/+ neurons examined (n = 326) and 97.2% of the / neurons examined (n = 178) were immunopositive for this enzyme (Fig. 1, A and C), consistent with results obtained from rat striatal cultures (14, 18). Taken together, the results suggest that, like the neurons from +/+ npc^nih mice, the neurons from / npc^nih mice can survive and differentiate in the culture conditions used and that the NPC1 mutation does not appear to lead to the selective survival of a specific subclass of neurons.

View larger version (51K): [in this window] [in a new window]   Fig. 1.   GAD expression in embryonic striatal neurons from wild type and mutant npcnih mice. Striatal neurons from +/+ (A and B) or / (C) npcnih mouse embryos were isolated and maintained in vitro for 7 days in a serum-free, defined culture medium before they were fixed and processed for immunocytochemistry. GAD immunoreactivity was visualized using a biotinylated secondary antibody, an avidin-biotin-peroxidase complex reagent, and 3,3'-diaminobenzidine tetrahydrochloride as the chromogen. When incubated with an anti-GAD antibody, essentially all of the neurons were immunopositive in cultures from both +/+ (A) and / (C) npcnih mice. When the primary antibody was omitted (B), no staining was observed. Scale bar, 20 µm.

To determine whether lipid metabolism is abnormal in the embryonic striatal neurons from / npc^nih mice, the levels of total cholesterol, free cholesterol, and total phospholipids were analyzed. Liver tissue, which exhibits elevated levels of both cholesterol and phospholipids in NP-C disease (1, 2), was analyzed in parallel as a positive control. As shown in Fig. 2, A and B, significantly (p < 0.01) higher levels of total and free cholesterol were detected in liver from adult / npc^nih mice than in liver from +/+ or +/ npc^nih mice, consistent with previous studies (35, 37-39). Similarly, the total phospholipid content in adult liver from / npc^nih mice was significantly (p < 0.05) elevated in comparison to the levels in +/+ and +/ npc^nih mice (Fig. 2C), as previously reported (3). To determine if these differences were evident during embryonic development, liver samples from embryonic day 15 (E15) mice were also analyzed. Both total and free cholesterol levels were significantly (p < 0.05) elevated in the liver of / npc^nih mice even at this embryonic age, although the difference was much less pronounced than for adult animals (Fig. 1, A and B). In contrast to embryonic liver, no significant differences were detected in the levels of total (Fig. 2A) or free (Fig. 2B) cholesterol in cultures of embryonic striatal neurons prepared from +/+, +/, or / npc^nih mice. The total phospholipid content of the embryonic neurons was higher than for liver, consistent with previous analysis of brain and liver (3). However, as with total and free cholesterol, there were no significant differences in the total phospholipid content of embryonic striatal neurons from +/+, +/, or / npc^nih mice (Fig. 2C).

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Cholesterol and phospholipid content of liver and embryonic striatal neurons from wild type, heterozygous, and mutant npcnih mice. Levels of total cholesterol (A), free cholesterol (B), and total phospholipids (C), expressed as pmol/µg total cellular protein, were measured in striatal neurons and in liver isolated from +/+, +/, or / npcnih mice. In liver from / adult mice, there were significantly higher levels of total cholesterol (p < 0.01), free cholesterol (p < 0.01), and total phospholipids (p < 0.05) than in liver from +/+ and +/ animals (n = 3 for assays of total and free cholesterol; n = 4 for phospholipid assays). The levels of free and total cholesterol, but not total phospholipids, were also significantly elevated (p < 0.05) in embryonic liver from / animals (n = 3 for assays of total and free cholesterol; n = 15 for phospholipid assays). In contrast, there were no significant differences in the levels of total cholesterol, free cholesterol, or total phospholipids in the striatal neurons from +/+, +/, or / embryos (n = 4 for all assays). Asterisks indicate values significantly different from those for the +/+ genotype.

Although fluorometric assays did not reveal a marked accumulation of cholesterol, there were significant differences in the metabolism of cholesterol in the embryonic striatal neurons from / npc^nih mice. Specifically, taking advantage of the access to the neurons afforded by the in vitro cultures, [³H]cholesterol linoleate-LDL was used as a means to deliver radiolabeled cholesterol to the neurons. Despite a significantly (p < 0.005) higher level of uptake of the radiolabeled cholesterol linoleate (Fig. 3A), striatal neurons from / embryos exhibited significantly lower levels of cholesterol linoleate hydrolysis (p < 0.0005) and re-esterification (p < 0.003) than cultures derived from +/+ embryos (Fig. 3, B and C). Whereas the hydrolysis and re-esterification of free cholesterol in neurons from +/ embryos was intermediate compared with neurons from +/+ and / neurons (Fig. 3, B and C), the values were not significantly different from those observed for neurons from +/+ npc^nih mice. Overall, the results demonstrate for the first time that clear deficits in cholesterol metabolism exist in neurons of the central nervous system of / npc^nih mice. In addition, these deficits are apparent at embryonic stages of development, well before the outward manifestations of the disease.

View larger version (9K): [in this window] [in a new window]   Fig. 3.   Metabolism of [3H]LDL-derived cholesterol linoleate in cultured embryonic striatal neurons. Cultures of striatal neurons isolated from +/+, +/, or / npcnih mouse embryos were incubated with LDL labeled with [3H]cholesterol linoleate, and cell lysates were analyzed for radiolabeled metabolites by thin layer chromatography. As shown in A, striatal neurons from / embryos (n = 7) exhibited significantly higher levels of 3H-LDL-derived cholesterol linoleate uptake (p < 0.005) when compared with striatal neurons from either +/+ (n = 3) or +/ (n = 6) embryos. Despite the increased uptake, striatal neurons from / animals (n = 8) exhibited significantly lower levels of hydrolysis (p < 0.0001) of [3H]cholesterol linoleate (B) and esterification (p < 0.03) of [3H]cholesterol released from [3H]cholesterol linoleate (C) when compared with neurons from +/+ (n = 6) or +/ (n = 9) embryos. Asterisks indicate values significantly different from those for the +/+ genotype.

BDNF Fails to Elicit Morphological Responses in Embryonic Striatal Neurons from npc^nih Mice-- Cholesterol is thought to be important for the growth of neurons (40, 41) and has been postulated to play a critical role in the localization and function of signaling molecules (22, 23, 42). To determine if the deficits in cholesterol metabolism observed in striatal neurons from / npc^nih mice were associated with deficits in neurite outgrowth and other morphological responses to neurotrophins, standard measurements (14, 18, 31, 32) were made of striatal neurons that had been isolated from +/+ and / embryos and maintained in the absence or presence of BDNF for 7 days. Initial comparison of untreated neurons from +/+ and / npc^nih mice revealed differences in the extent of neurite outgrowth. Specifically, the number of primary neurites per cell, the number of neurite branch points per cell, and the total neurite length per cell in neurons from / embryos were all significantly (p < 0.001) less than observed for neurons from +/+ embryos (Fig. 4A). When untreated and BDNF-treated cultures of the same genotype were compared, the morphological differences between neurons from +/+ and / animals were accentuated. Consistent with previous analyses of neurons isolated from embryonic rat striata (14, 18), BDNF-treated striatal neurons from +/+ mice exhibited significant increases in somal area and all three of the indices of neurite outgrowth when compared with untreated striatal neurons from the same embryo (Fig. 4B). In stark contrast, BDNF failed to elicit significant changes in any of the measured parameters in striatal neurons from / animals (Fig. 4C). Thus, in striatal neurons, the deficits associated with the mutation in the NPC1 gene include a marked reduction in morphological differentiation.

View larger version (20K): [in this window] [in a new window]   Fig. 4.   Morphometric comparison of embryonic striatal neurons from wild type and npcnih mutant mice. Cumulative histograms from morphometric analyses of embryonic striatal neurons from +/+ and / npcnihmice. A, in cultures maintained in defined medium in the absence of identified neurotrophic factors for 7 days, measurements were made from 344 neurons from +/+ embryos and 715 neurons from / npcnih embryos. The values obtained for neurons from / embryos are plotted as a percent of the values for neurons from +/+ embryos. The average number of primary neurites per cell, number of neurite branch points per cell, and total neurite length per cell in neurons from / animals were all significantly less than observed for neurons from +/+ embryos (p < 0.001). Asterisks indicate values significantly different from those for the +/+ genotype. B, in cultures from +/+ embryos, measurements were made from neurons maintained in defined medium for 7 days in the absence (control; n = 154) or presence (+BDNF; n = 154) of 50 ng/ml BDNF. Values for BDNF-treated neurons are plotted as a percent of the values for the untreated neurons from the same embryo (see "Experimental Procedures"). BDNF treatment elicited significant increases in somal area (p < 0.01), number of primary neurites per cell (p < 0.01), number of neurite branch points per cell (p < 0.001), and total neurite length per cell (p < 0.001). Asterisks indicate values for the BDNF-treated neurons that are significantly different from those for untreated neurons. C, in cultures from / embryos, measurements were made from neurons maintained in defined medium for 7 days in the absence (control; n = 378) or presence (+BDNF; n = 335) of 50 ng/ml BDNF. Data were analyzed and compared as described in B. In contrast to its effects on neurons from +/+ embryos, BDNF did not elicit any significant changes in neurons from / embryos.

BDNF Fails to Activate TrkB Receptors in Embryonic Striatal Neurons from npc^nih Mice-- Among the transmembrane molecules thought to be concentrated in cholesterol-enriched membrane domains or lipid rafts is the TrkB receptor tyrosine kinase activated by BDNF (23, 42, 43). To determine if the diminished response to BDNF by striatal neurons from / npc^nih mice could be attributed to a lack of TrkB receptor signaling, Western blot analysis was used to assess TrkB receptor expression and the initial autophosphorylation of the receptor that occurs in response to BDNF (for review see Ref. 44). Cultures of embryonic striatal neurons from +/+, +/, and / npc^nih embryos were untreated or treated with 50 ng/ml BDNF for 5 min before they were harvested, and the lysates were analyzed using previously developed procedures (45, 46). This concentration of BDNF does not activate TrkC (47, 48), and these neurons do not appear to express TrkA (see Introduction), so this treatment selectively activates TrkB in these cells. As shown in Figs. 5 and 6, when the BDNF-induced autophosphorylation of TrkB receptors was assessed using an anti-phosphotyrosine antibody, there were striking differences in the results obtained. Consistent with previous demonstrations of BDNF-dependent activation of TrkB in other populations of primary neurons from wild type rodents (48, 49), there were obvious increases in receptor autophosphorylation in striatal neurons from +/+ and +/ mice (Figs. 5A and 6). In contrast, there was no detectable increase in TrkB autophosphorylation in neurons from / npc^nih mice (Fig. 5A). This difference was consistently observed in multiple experiments (n = 4) and could not be attributed to a lack of TrkB protein in the neurons from / mice (Fig. 5A). It was also not due to elevated levels of the truncated, kinase-inactive form of TrkB in the / neurons inhibiting the activation of the full-length TrkB, as the smaller (~95 kDa) truncated TrkB was not detected in any of the embryonic striatal cultures (Fig. 5A), consistent with previous analysis of neurons from the embryonic central nervous system (20, 50). Finally, although the levels of TrkB in the lysates appeared to be comparable (Fig. 5A), a possible explanation for the difference in levels of TrkB autophosphorylation could have been a lack of surface expression of TrkB in the neurons from the / npc^nih mice. Therefore, we used an approach previously described for analysis of TrkB surface expression in primary neurons (51), in which surface proteins were biotinylated, precipitated from cell lysates using streptavidin-agarose, and then analyzed by SDS-PAGE and Western blotting. The detection of TrkB in the precipitates appeared to be a specific indication of surface expression, as TrkB receptors did not precipitate from cultures that were not exposed to the biotinylating reagent nor was the intracellular protein tubulin biotinylated or precipitated under any condition (Fig. 5B). More important, in the two separate experiments in which it was determined, there were comparable levels of biotinylated TrkB receptors in neurons from / and /+ mice (Fig. 5B). Thus, the inability of BDNF to activate TrkB in the striatal neurons from / mice cannot be attributed to a lack of surface expression of TrkB proteins, and instead the results strongly suggest that the deficits in cholesterol metabolism in neurons from / npc^nih mutant mice interfere with the ability of BDNF to activate TrkB receptors.

View larger version (39K): [in this window] [in a new window]   Fig. 5.   BDNF fails to activate TrkB receptors expressed on the surface of striatal neurons from npcnih mutant mice. Cultures of striatal neurons were isolated from +/ or / npcnih embryos and maintained in culture for 3 days before neurons of a given genotype were either untreated () or treated (+) with 50 ng/ml of BDNF for 5 min. Western blot analysis was carried out using an anti-TrkB antibody to assess levels of TrkB protein expression (TrkB) and an anti-phosphotyrosine antibody to detect autophosphorylated (activated) TrkB proteins (Trk-P). As shown in A, BDNF elicited a marked increase in the levels of activated TrkB in striatal neurons from +/ embryos but no detectable increase in neurons from / embryos. B, striatal neurons from +/ or / mice were either untreated () or treated with a biotinylating reagent (+) before the biotinylated proteins on the cell surface were precipitated, and the precipitated proteins (ppt) and aliquots of the remaining lysates (lys) were subjected to Western blot analysis using an anti-TrkB antibody (TrkB) or an anti-tubulin antibody (Tubulin). When the anti-TrkB antibody was used to analyze the precipitated material (TrkB/ppt), TrkB receptors on the surface of +/ and / neurons were detected in samples from cells treated with the biotinylating reagent (+) but not in samples from untreated cells (), despite the presence of TrkB in the lysates from the untreated cells (TrkB/lys). In contrast, the intracellular protein tubulin was not biotinylated or precipitated under any conditions (Tubulin/ppt), despite the fact it could be detected in all of the lysates (Tubulin/lys).

View larger version (68K): [in this window] [in a new window]   Fig. 6.   BDNF fails to activate TrkB receptors in wild type striatal neurons acutely depleted of cholesterol. Western blot analysis was carried out using an anti-TrkB antibody to assess levels of TrkB protein expression (TrkB) and an anti-phosphotyrosine antibody to detect autophosphorylated (activated) TrkB proteins (Trk-P) in striatal neurons isolated from +/+ embryos. As shown, there was a clear increase in TrkB activation in response to BDNF when the neurons were maintained in control conditions (left). However, there was no detectable increase in TrkB activation in striatal neurons that had been previously depleted of cholesterol (right) via prior treatment with lovastatin, mevalonate, and methyl- -cyclodextrin.

BDNF Fails to Activate TrkB Receptors in Cholesterol-depleted Striatal Neurons from Wild Type Mice-- To determine further if alterations in cellular cholesterol interfere with TrkB receptor function, embryonic striatal neurons from +/+ npc^nih mice were subjected to a previously established procedure for depleting cholesterol in primary neurons (30, 31), and the response of the neurons to BDNF was assayed by Western blot analysis of TrkB receptors, as described above. The neurons were incubated in culture medium containing lovastatin and mevalonate for several days prior to a brief (~20 min) exposure to methyl--cyclodextrin, a treatment that has been shown to deplete ~70% of the cholesterol in primary neurons (34). Attempts to treat the cultures with cyclodextrin for longer periods (i.e. 12 h) dramatically reduced cell viability (data not shown), as observed for other populations of primary neurons (33). Control cultures were incubated in 0.01% Me₂SO, which was used as the delivery vehicle for the lovastatin. As shown in Fig. 6, while a typical increase in TrkB autophosphorylation occurred in the control neurons, there was no discernible increase in autophosphorylation in the cholesterol-depleted neurons in response to BDNF. This result was observed in several independent experiments (n = 3) and is consistent with the data obtained with neurons from / npc^nih mice. Together the results indicate that abnormalities in cellular cholesterol inhibit TrkB receptor function and suggest a cellular mechanism that may contribute to the neurodegenerative changes in NP-C disease.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Niemann-Pick type C disease is a progressive and fatal neuropathological disorder. Although this disease is associated with abnormal cholesterol metabolism in peripheral tissues, the basis for the progressive neurodegeneration and abnormal brain function is unclear. Here we show that although embryonic neurons from the brain of / npc^nih mice can survive in vitro and are not grossly abnormal in appearance, they exhibit significant deficits in cholesterol metabolism, neurite outgrowth, and neurotrophin responsiveness, all of which may contribute to the progression of NP-C disease.

Accumulation of unesterified cholesterol in peripheral fibroblasts is the diagnostic hallmark of NP-C disease, yet elevated cholesterol has not been reported in previous analyses of the brains of NP-C patients or npc^nih mutant mice (1, 2). Similarly, we did not detect an abnormal accumulation of cholesterol in striatal neurons isolated from / npc^nih mice at an embryonic stage of development. Studies in vivo suggest that minimal amounts of LDL-derived cholesterol are taken up by the brain, especially during embryonic and early postnatal development (39, 52). However, although it appears that neurons rely heavily on de novo synthesis of cholesterol during initial development (52-54), it has been suggested that exogenous uptake, as part of cholesterol recycling in the brain, may be important for subsequent neuronal remodeling (Refs. 55 and 56 and for discussion see Ref. 39). Consistent with this idea, we show that embryonic striatal neurons take up exogenously supplied cholesterol linoleate. More important, we find that subsequent to its uptake, the neurons from / npc^nihmice exhibit significant deficits in their ability to hydrolyze and esterify this cholesterol. These results extend the findings of a previous study showing an impaired cholesterol metabolism in mixed cultures of glia and neurons from newborn / npc^nih mice (36) by providing direct evidence this abnormality occurs in neurons. Taken together, the findings suggest that in NP-C disease errors in neuronal cholesterol metabolism occur early in brain development and may have an accumulating influence as the remodeling of neuronal connections continues in the brain.

Previous studies suggest that cholesterol is important for neurite outgrowth. In particular, inhibition of cholesterol synthesis impairs the neurite outgrowth from sympathetic neurons (41), whereas exogenous cholesterol enhances neurite outgrowth from dorsal root ganglion neurons (40). Furthermore, interfering with LDL receptor-independent means by which cholesterol uptake may occur results in simplified dendritic arborizations from neurons of the central nervous system, including decreases in the number and average length of neurites (57, 58). Here we show that abnormal metabolism of cholesterol in neurons from / mice is accompanied by deficits in neurite outgrowth that mirror those reported in these previous studies, including a reduced number of primary neurites, fewer branch points, and diminished neurite length. Thus, the abnormal cholesterol metabolism associated with mutation of the NPC1 gene may ultimately compromise both the complexity and the maintenance of neuronal arbors and lead to eventual loss of neuronal function in NP-C disease.

In addition to the observed deficiencies in cholesterol metabolism and neurite outgrowth, we show for the first time that striatal neurons from / npc^nih mice do not respond to BDNF with an increase in neurite outgrowth. Errors in cholesterol metabolism in NP-C disease could potentially interfere with BDNF signaling in a variety of ways, as cholesterol and other lipid metabolites have important roles in multiple aspects of neuronal signal transduction. For example, sphingomyelin plays an important role in the signaling from the p75 neurotrophin receptor (59). In addition, cholesterol is necessary for the covalent modification and function of molecules such as ras that are involved in neurotrophin signaling pathways (60, 61). Finally, gangliosides, which are elevated in NP-C disease (1, 2), have been shown to alter receptor tyrosine kinase signaling (62-65). However, because these aspects of signaling may be affected in NP-C disease, perhaps most consistent with our data from both wild type and mutant mice is the idea that the errors in cholesterol homeostasis that arise as a result of the NPC1 mutation interfere with TrkB signaling by altering lipid rafts. Specifically, recent studies have demonstrated that localization of sphingolipids and cholesterol in lipid rafts is essential for the assembly and activity of transmembrane signaling complexes in these specialized membrane domains (22, 23, 42). Relevant to these studies, a previous report suggests that mutation of the NPC1 gene is associated with alterations in the cholesterol composition of the rafts isolated from peripheral tissues (66). Chemical depletion of cholesterol from the cell surface disassembles raft-enriched membrane invaginations referred to as caveolae and interferes with the function of receptor tyrosine kinases (67, 68). TrkA activation is also disrupted upon altered expression of caveolin-1 (46), a cholesterol-binding protein that plays a key role in controlling the level of cholesterol in the plasma membrane (69, 70). We find that although the levels of TrkB protein are comparable in striatal neurons from wild type and mutant mice, the BDNF-induced autophosphorylation of TrkB was essentially eliminated in neurons from the / npc^nih mice. Furthermore, TrkB activation was also abolished in wild type striatal neurons acutely depleted of cholesterol, consistent with the fact that the TrkB receptor tyrosine kinase is found in lipid rafts (43). When taken together, the results support the assertion that the abnormal cholesterol metabolism associated with the NPC1 mutation interferes with BDNF responsiveness by altering lipid rafts and interfering with TrkB activation.

Evidence suggests that BDNF and the activation of TrkB play prominent roles in the maintenance and modulation of neuronal function. For example, BDNF modulates axonal and dendritic arborization, the expression of peptides and calcium-binding proteins, and synaptic transmission (71-74). TrkB signaling has also been shown to be important in the mature nervous system, as indicated by recent experiments in which expression of this receptor was conditionally knocked out in forebrain neurons of postnatal mice. Although the knockout did not result in gross abnormalities in brain structure, these animals exhibited selective deficits in both synaptic and cognitive functions (75). Thus, our finding that neurons from / npc^nih mice exhibit deficits in BDNF-mediated TrkB signaling may not only provide a basis for understanding the aberrant neuronal morphology described in these animals but may also provide insight into the cognitive deficits that characterize the postnatal progression of NP-C disease. Finally, by identifying specific cellular and molecular deficits in neuronal function in NP-C disease, our results provide insight not only into the neuropathology of this disorder but also into neurotrophin signaling and the functional significance of cholesterol in the brain.

	ACKNOWLEDGEMENTS

We thank D. Ginty (The Johns Hopkins University School of Medicine) for Trk antibodies and useful suggestions; D. Compton (Dartmouth Medical School) for tubulin antibodies; and C. Chang and J. Cruz (Dartmouth Medical School) for helpful discussions.

	FOOTNOTES

^* This work was supported by grants from the Ara Parseghian Medical Research Foundation (to R. A. M., L. P. H., and T. Y. C.) and the National Niemann-Pick Disease Foundation (to L. P. H.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^¶ To whom correspondence should be addressed: Dept. of Physiology, Dartmouth Medical School, Hanover, NH 03755. Tel.: 603-650-1311; Fax: 603-650-1128; E-mail: robert.maue@dartmouth.edu.

Published, JBC Papers in Press, April 17, 2000, DOI 10.1074/jbc.M001793200

² K. Suzuki, personal communication.

	ABBREVIATIONS

The abbreviations used are: NP-C, Niemann-Pick type C; BDNF, brain-derived neurotrophic factor; LDL, low density lipoprotein; PCR, polymerase chain reaction; PBS, phosphate-buffered saline; GAD, glutamic acid decarboxylase.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES