Survival of adult neurons lacking cholesterol synthesis in vivo
© Fünfschilling et al; licensee BioMed Central Ltd. 2007
Received: 21 August 2006
Accepted: 02 January 2007
Published: 02 January 2007
Cholesterol, an essential component of all mammalian plasma membranes, is highly enriched in the brain. Both during development and in the adult, brain cholesterol is derived from local cholesterol synthesis and not taken up from the circulation. However, the contribution of neurons and glial cells to total brain cholesterol metabolism is unknown.
Using conditional gene inactivation in the mouse, we disrupted the squalene synthase gene (fdft1), which is critical for cholesterol synthesis, in cerebellar granule cells and some precerebellar nuclei. Mutant mice showed no histological signs of neuronal degeneration, displayed ultrastructurally normal synapses, and exhibited normal motor coordination. This revealed that these adult neurons do not require cell-autonomous cholesterol synthesis for survival or function.
We conclude that at least some adult neurons no longer require endogenous cholesterol synthesis and can fully meet their cholesterol needs by uptake from their surrounding. Glia are a likely source of cholesterol in the central nervous system.
Cholesterol is an essential, integral membrane component of mammalian plasma membranes. Genetic or pharmacological interference with cholesterol synthesis during development has dramatic effects on mice, ranging from embryonic lethality to severe brain abnormalities including holoprosencephaly [1–3]. Defects of cholesterol homeostasis in the adult brain are linked to neurodegenerative diseases like Niemann-Pick type C disease or Alzheimer's disease [4–6] or even to general synaptic degeneration .
Several studies investigated total cholesterol production in and excretion by the brain [8, 9], but very little is known about the contribution of individual cell types. Oligodendrocytes, which require large amounts of cholesterol for the production of myelin, produce most of the cholesterol cell-autonomously [10, 11]. Astrocytes are assumed to be net cholesterol producers, as they express ApoE in vivo and can secrete cholesterol in ApoE-containing particles in vitro [reviewed in ]. Cultured neurons from different sources can grow neurites in the absence of added cholesterol, but in cultured retinal ganglion cells, cholesterol was shown to be rate-limiting for synaptogenesis . Such cell culture experiments led to the hypothesis that some developing neurons produce their own cholesterol, but switch to astroglial cholesterol supply at later stages. However, in vivo evidence for this hypothesis is missing . We therefore set forth to study the role of neuronal cholesterol synthesis in the adult brain.
We recently generated a mouse line with a conditional null allele of the squalene synthase gene (fdft1flox/flox) . Squalene synthase catalyzes the first committed step in sterol biosynthesis, and the complete knockout displays an early embryonic lethal phenotype . In the conditional mouse line, an essential exon of the squalene synthase gene is flanked by loxP site. When crossed to a mouse line that expresses cre recombinase in a cell-type specific manner, this leads to the cell-type specific deletion of squalene synthase and thus cholesterol production. When squalene synthase is eliminated in myelinating glia cells, mutant animals show severe hypomyelination and one third of the animals die before weaning . When squalene synthase is deleted embryonically in nestin-positive neuronal and glial precursors , no viable pups are born, highlighting the importance of brain cholesterol production during development (data not shown). To target adult neurons, we chose the mouse line Tg(mα6-cre)B1LFR, which expresses cre recombinase specifically in postmitotic, postmigratory cerebellar granule cells and a few brainstem nuclei . In contrast to the above-mentioned examples, these mutant animals developed normally and were indistinguishable from control littermates up to at least the age of six months. This demonstrates that some adult neurons do not require continuous cell-autonomous cholesterol synthesis and suggests that they survive utilizing horizontal cholesterol flux in vivo.
Mutant and control animals were obtained at the expected Mendelian ratios and were viable and fertile. Preliminary examination of mutant animals that harbored additionally the R26R-EYFP reporter allele revealed that cerebellar granule cells, the pontine nuclei, and the reticulotegmental nucleus of the pons were present and expressed GFP (Figure 1D–E). The majority of the fibers in the middle cerebellar peduncle, the tract carrying the axons form the pontine nuclei and the reticulotegmental nucleus of the pons to the cerebellar hemispheres, were YFP-positive, indicating that a large fraction of the pontine precerebellar neurons were targeted and had proper projections into the cerebellum (Figure 1E–F). This suggested that the targeted neurons survived despite cre-expression and presumed absence of a functional fdft1 gene. To confirm that the fdft1 locus was indeed recombined, we prepared genomic DNA from two exemplary regions: the cerebellum and the pontine nuclei. Southern blot on DNA from 2 months old animals showed that about 80 % of the total cerebellar DNA was recombined and thus lacked a functional squalene synthase gene (Figure 1G). Granule cells make up approximately 80% of all cells in the cerebellum , and they are the only cre-expressing cells in the cerebellum of Tg(mα6-cre)B1LFR. Therefore virtually all cerebellar granule cells were targeted in our mutants and lost functional squalene synthase gene. A time course using quantitative PCR analysis provided equivalent results with 57 +/- 9 % (mean +/- SEM, n = 3) recombination at postnatal day 16 and 74 +/- 4 % (mean +/- SEM, n = 4) at one month of age. Recombination in the pontine nuclei was also assayed by quantitative PCR and was found to be 9 +/- 1 % (mean +/- SEM, n = 3) at postnatal day 16 and 13 +/- 4 % (mean +/- SEM, n = 4) at one month of age. Considering the relatively crude dissection including some neighboring tissue, this confirmed that a significant fraction of the projection neurons in the pontine nuclei lost a functional squalene synthase gene.
Western blot analysis did not reveal a significant difference in the amount of squalene synthase protein between mutants and controls in extracts from total cerebellum or the pontine nuclei, neither at P16 (Figure 1H), nor at one or two months of age (data not shown). Thus, either the affected neurons only contribute little to the total pool of squalene synthase protein and thus cholesterol synthesizing capacity, or other cells might have upregulated the fdft1 gene to compensate.
Analysis of the brain stem revealed a normal size and location of the pontine nuclei and the reticulotegmental nucleus of the pons (Figure 3A). Staining for GFAP revealed no signs of astrogliosis (Figure 3B). Examination of hematoxylin and eosin stained sections revealed normal cytoarchitecture and fiber tract distribution in the pontine nuclei (Figure 3C and data not shown). The axons from the pontine nuclei and the reticulotegmental nucleus of the pons project to the cerebellar hemispheres as a thick bundle, the middle cerebellar peduncle. The size of the middle cerebellar peduncle was unchanged (Figure 3D), indicating that the axons were projecting normally to their target.
Since the cerebellum is required for motor coordination, cerebellar defects lead to characteristic deficits in motor coordination tasks like the Rotarod . Here, mutant and control animals performed equally well at all speed levels tested, indicating proper cerebellar function (Figure 4B).
Using mouse genetics, we have generated a novel mouse mutant lacking cellular cholesterol synthesis in defined populations of adult neurons, the cerebellar granule cells and neurons in some precerebellar nuclei. These mice were phenotypically indistinguishable from controls and showed no motor deficits. This is particularly remarkable since two "links" of the cerebellar circuit were affected: a subset of mossy fibers (including fibers from the pontine nuclei and from the reticulotegmental nucleus of the pons) as well as all granule cells. Histological and ultrastructural analysis of adult mice failed to reveal any signs of neurodegeneration or inflammation. Thus, some adult neurons do not require cell autonomous cholesterol synthesis.
Genetic or pharmacological perturbation of cholesterol metabolism is known to exert devastating effects on developing neurons both in vivo and in vitro. Depending on the time point, this includes neural tube closure defects, holoprosencephaly, neuronal cell death, reduction of dendritic outgrowth, and the inhibition of dendritic and synaptic maturation [1, 13, 22]. Only by choosing a cre line that recombines fdft1flox/flox after the targeted neurons have reached their final position, and after axons and dendrites have been formed, it was possible to exclude such developmental defects. Synaptogenesis and synaptic maturation, however, continues for several weeks after onset of cre-recombination . Although we cannot exclude subtle defects of these mutants in synaptic function, which can only be detected using electrophysiological recordings or refined motor learning tasks , we have documented a remarkably intact development and maintenance of cerebellar granule cells and neurons in the pontine nuclei in vivo with absence of any degeneration or cell death in the adult cerebellum or brain stem. Since other defects in cholesterol metabolism lead to neurodegeneration (i.e. Niemann-Pick Type C disease, Alzheimer's disease ), the critical difference of our transgenic model is that only a defined subgroup of neurons is affected, whereas other neurons, oligodendrocytes and astrocytes cells maintained cholesterol synthesis.
Both neuronal subtypes analyzed in this study are glutamatergic neurons derived form the rhombic lip, but they represent two very different size classes of neurons: cerebellar granule cells are among the smallest neurons in the brain and have simple dendrites and unmyelinated axons, whereas the neurons in the pontine nuclei are large, have complex dendrites, and long myelinated axons. Since these two very different populations of neurons show a similar tolerance for loss of the fdft1 gene, this might be a general feature of adult neurons. However, the adult cerebellum has only about half the steady state level of cholesterol production compared to the forebrain, indicating that the cholesterol requirements might differ between different brain regions . A further indication for diversity comes from culture experiments where hippocampal neurons have different cholesterol requirements compared to cortical neurons .
Our results support the hypothesis that adult neurons reduce cholesterol production, and that glial cells take over this task. Our western blot results even indicate that adult granule cells barely contribute to the cerebellar pool of squalene synthase, i.e. cholesterol synthesis. Alternatively, other cell types might have upregulated the steady state level of squalene synthase protein to compensate. In the pontine nuclei, the small fraction of recombined cells (only about 10 %) in the dissected tissue might make a reduction of squalene synthase protein difficult to detect. When squalene synthase was eliminated in myelinating glia, the cholesterol transport protein ApoE and its receptor LRP were upregulated in the spinal cord, suggesting that cholesterol uptake by oligodendrocytes was increased to compensate for the loss of cholesterol synthesizing capacity . We could not detect an upregulation of LRP in the cerebellum (data not shown). This might be a further indication that adult neurons already express sufficient amounts of LRP to import cholesterol from ApoE-containing lipoprotein particles, or they might utilize a different cholesterol uptake system. Oligodendrocytes and astrocytes are the most likely candidates responsible for providing cholesterol in a paracrine fashion. Astrocytes have been shown to secrete cholesterol in the form of ApoE-containing particles . Indeed, cerebrospinal fluid contains about 8 μg/ml cholesterol , a concentration sufficient to sustain full neuronal growth and maturation in vitro .
Our finding that adult neurons can survive and function in the absence of cholesterol synthesis demonstrates the importance of horizontal cholesterol flux in vivo. It also highlights the importance of glial cells on brain cholesterol homeostasis.
Mice and behavior
All animal experiments were carried out in compliance with animal policies of the State of Niedersachsen, Germany. Animals were maintained on a mixed genetic background (C57/Bl6, Sv129). Motor behavior analysis was performed on an accelerating Rotarod (Hugo Basile 7650) at fixed speed levels as described .
20 μg of total cerebellar DNA were digested with HindIII, and a radiolabeled Southernblot was performed using standard procedures. The template was a 343 bp fragment of intron IV of the fdft1 gene, generated by PCR from genomic DNA using the primers GAGATGGACGTCCTCCCAGCCTGAGGC and GCAGCCTACCAACTCCTAGCCTGGC. The fragment of the flox allele migrates at 2.9 kb, the recombined allele at 6.1 kb.
Quantitative PCR was performed using 10 ng of target DNA in a 12.5 μl assay using SYBR Green PCR Master Mix (Applied Biosystems 4367659) on an Applied Biosystems 7500 Fast real time PCR system in duplicates. The standard curve was generated using DNA from heterozygous knockout animals fdft1flox/rec (50 % recombination) and fdft1flox/flox animals (0 % recombination). To achieve a standard curve between 50 % and 100 % recombination, 20 ng of fdft1flox/rec DNA were set as 100 % recombination, and all samples falling in that range were supplemented with 10 ng of "dummy DNA" (fdft1flox/flox) to equalize the amount of DNA. The values were standardized to an independent genomic marker to correct for small differences in the amount of DNA. Primers AAGGCTGGATCCGTCGAGGTG and GTCTTCAAGAATTCCGATCATATTCA amplify a 112 bp band specific for the recombined allele.
10 μg of extract were separated by SDS PAGE and blotted onto nitrocellulose. The blot was probed with mouse anti squalene synthase (BD Transduction Laboratories 611808, 1:1000), and later reprobed with mouse anti glyceraldeyde-3-phosphate dehydrogenase (GAPDH; Advanced Immunochemicals #RGM2, 1:5000). Bands were detected using a fluorescent secondary antibody and scanned on an Odyssey infrared scanner (LI-COR Biosciences).
Histology and immunohistochemistry
For all histological analyses, three mutants and three control littermates were analyzed. Mice were anesthetized with Halothane and perfused intracardially with 15 ml of Hanks' balanced salt solution (HBSS) (PAA Laboratories H15-010), followed by 50 ml of 4 % formaldehyde in 100 mM phosphate buffer (PB). Brains were postfixed in 4 % formaldehyde (in PB) overnight and then either embedded in paraffin or stored in PBS. For sectioning on the vibratome (Leica VT 1000 S), sagittal halves of brains were embedded in 2 % low-melt agarose and cut at 50 μm. Vibratome sections were stained free floating. Paraffin embedded tissue was sectioned at 5 μm on a sliding microtome (Microm HM400). Sections were stained with hematoxylin/eosin, for Nissl or immunohistochemically and photographed on a Zeiss Axiophot or a Leica DMRXA. The cerebellar overview bright field pictures were manually cleared of background and brightness/contrast on all pictures was optimized using the software Deneba Canvas. Confocal pictures were taken on a Zeiss LSM 510 Meta microscope. Antibodies used: rabbit anti S100β (SWANT code 37, 1:1000), mouse anti GFP (Molecular Probes A11120, 1:500), mouse anti GFAP (Novocastra NCL-GFAP-GA5, 1:200), rabbit anti GABAAα6 (Chemicon AB5610, 1:500), rat anti Mac-3 (Pharmingen 01781, 1:5000), rabbit anti calbindin (Swant CB-38a 1:5000). Golgi impregnation was performed using a Rapid Golgi Stain Kit (FD Neuro Technologies PK401) on entire halves of freshly dissected cerebellum.
Animals were anesthetized with Halothane and perfused intracardially with 15 ml of HBSS, followed by 50 ml of fixative (2.5 % glutaraldehyde, 4 % paraformaldehyde in phosphate buffered saline ), and the whole cerebellum was postfixed overnight in fixative. Midsagittal 1 mm slices of whole cerebellum were embedded in Epon, blocks were trimmed to contain lobules 6/7. Ultrathin sections were contrasted with 1 % uranyl acetate and lead citrate . Parallel fiber to Purkinje cell synapses were identified by their morphology . Pictures were taken on a LEO 912AB electron microscope (Zeiss, Oberkochen, Germany) with an on-axis 2048 × 2048 CCD-camera (Proscan, Scheuring, Germany) at 10 000x, and digitally enhanced using the program ImageJ .
We thank Annette Fahrenholz, Torben Ruhwedel, and Miriam Teroerde for technical assistance. This work was supported by a Marie Curie Fellowship (509506) to UF and the Deutsche Forschungsgemeinschaft (SFB 523) for KAN.
- Incardona JP, Roelink H: The role of cholesterol in Shh signaling and teratogen-induced holoprosencephaly. Cell Mol Life Sci. 2000, 57: 1709-1719. 10.1007/PL00000653.View ArticleGoogle Scholar
- Tozawa R, Ishibashi S, Osuga J, Yagyu H, Oka T, Chen Z, Ohashi K, Perrey S, Shionoiri F, Yahagi N, Harada K, Gotoda T, Yazaki Y, Yamada N: Embryonic lethality and defective neural tube closure in mice lacking squalene synthase. J Biol Chem. 1999, 274: 30843-30848. 10.1074/jbc.274.43.30843.View ArticleGoogle Scholar
- Ohashi K, Osuga J, Tozawa R, Kitamine T, Yagyu H, Sekiya M, Tomita S, Okazaki H, Tamura Y, Yahagi N, Iizuka Y, Harada K, Gotoda T, Shimano H, Yamada N, Ishibashi S: Early embryonic lethality caused by targeted disruption of the 3-hydroxy-3-methylglutaryl-CoA reductase gene. J Biol Chem. 2003, 278: 42936-42941. 10.1074/jbc.M307228200.View ArticleGoogle Scholar
- Moebius FF, Fitzky BU, Glossmann H: Genetic defects in postsqualene cholesterol biosynthesis. Trends Endocrinol Metab. 2000, 11: 106-114. 10.1016/S1043-2760(00)00235-6.View ArticleGoogle Scholar
- Vance JE, Hayashi H, Karten B: Cholesterol homeostasis in neurons and glial cells. Semin Cell Dev Biol. 2005, 16: 193-212. 10.1016/j.semcdb.2005.01.005.View ArticleGoogle Scholar
- Yanagisawa K: Cholesterol and pathological processes in Alzheimer's disease. J Neurosci Res. 2002, 70: 361-366. 10.1002/jnr.10348.View ArticleGoogle Scholar
- Koudinov AR, Koudinova NV: Cholesterol homeostasis failure as a unifying cause of synaptic degeneration. J Neurol Sci. 2005, 229-230: 233-240. 10.1016/j.jns.2004.11.036.View ArticleGoogle Scholar
- Dietschy JM, Turley SD: Thematic review series: brain Lipids. Cholesterol metabolism in the central nervous system during early development and in the mature animal. J Lipid Res. 2004, 45: 1375-1397. 10.1194/jlr.R400004-JLR200.View ArticleGoogle Scholar
- Lund EG, Xie C, Kotti T, Turley SD, Dietschy JM, Russell DW: Knockout of the cholesterol 24-hydroxylase gene in mice reveals a brain-specific mechanism of cholesterol turnover. J Biol Chem. 2003, 278: 22980-22988. 10.1074/jbc.M303415200.View ArticleGoogle Scholar
- Jurevics H, Morell P: Cholesterol for synthesis of myelin is made locally, not imported into brain. J Neurochem. 1995, 64: 895-901.View ArticleGoogle Scholar
- Saher G, Brugger B, Lappe-Siefke C, Möbius W, Tozawa R, Wehr MC, Wieland F, Ishibashi S, Nave KA: High cholesterol level is essential for myelin membrane growth. Nat Neurosci. 2005, 8: 468-475.Google Scholar
- Fagan AM, Holtzman DM: Astrocyte lipoproteins, effects of apoE on neuronal function, and role of apoE in amyloid-beta deposition in vivo. Microsc Res Tech. 2000, 50: 297-304. 10.1002/1097-0029(20000815)50:4<297::AID-JEMT9>3.0.CO;2-C.View ArticleGoogle Scholar
- Mauch DH, Nägler K, Schumacher S, Göritz C, Müller EC, Otto A, Pfrieger FW: CNS synaptogenesis promoted by glia-derived cholesterol. Science. 2001, 294: 1354-1357. 10.1126/science.294.5545.1354.View ArticleGoogle Scholar
- Pfrieger FW: Outsourcing in the brain: do neurons depend on cholesterol delivery by astrocytes?. Bioessays. 2003, 25: 72-78. 10.1002/bies.10195.View ArticleGoogle Scholar
- Tronche F, Kellendonk C, Kretz O, Gass P, Anlag K, Orban PC, Bock R, Klein R, Schütz G: Disruption of the glucocorticoid receptor gene in the nervous system results in reduced anxiety. Nat Genet. 1999, 23: 99-103. 10.1038/12703.View ArticleGoogle Scholar
- Fünfschilling U, Reichardt LF: Cre-mediated recombination in rhombic lip derivatives. Genesis. 2002, 33: 160-169. 10.1002/gene.10104.PubMed CentralView ArticleGoogle Scholar
- Srinivas S, Watanabe T, Lin CS, William CM, Tanabe Y, Jessell TM, Costantini F: Cre reporter strains produced by targeted insertion of EYFP and ECFP into the ROSA26 locus. BMC Dev Biol. 2001, 1: 4-10.1186/1471-213X-1-4.PubMed CentralView ArticleGoogle Scholar
- Altman J, Bayer SA: Development of the cerebellar system: in relation to its evolution, structure, and functions. 1997, , CRC Press, 1Google Scholar
- Sotelo C: Cerebellar synaptogenesis: what we can learn from mutant mice. J Exp Biol. 1990, 153: 225-249.Google Scholar
- Rhyu IJ, Abbott LC, Walker DB, Sotelo C: An ultrastructural study of granule cell/Purkinje cell synapses in tottering (tg/tg), leaner (tg(la)/tg(la)) and compound heterozygous tottering/leaner (tg/tg(la)) mice. Neuroscience. 1999, 90: 717-728. 10.1016/S0306-4522(98)00518-1.View ArticleGoogle Scholar
- Goddyn H, Leo S, Meert T, D'Hooge R: Differences in behavioural test battery performance between mice with hippocampal and cerebellar lesions. Behav Brain Res. 2006Google Scholar
- Fan QW, Yu W, Gong JS, Zou K, Sawamura N, Senda T, Yanagisawa K, Michikawa M: Cholesterol-dependent modulation of dendrite outgrowth and microtubule stability in cultured neurons. J Neurochem. 2002, 80: 178-190. 10.1046/j.0022-3042.2001.00686.x.View ArticleGoogle Scholar
- Shibuki K, Gomi H, Chen L, Bao S, Kim JJ, Wakatsuki H, Fujisaki T, Fujimoto K, Katoh A, Ikeda T, Chen C, Thompson RF, Itohara S: Deficient cerebellar long-term depression, impaired eyeblink conditioning, and normal motor coordination in GFAP mutant mice. Neuron. 1996, 16: 587-599. 10.1016/S0896-6273(00)80078-1.View ArticleGoogle Scholar
- Keller RK, Small M, Fliesler SJ: Enzyme blockade: a nonradioactive method to determine the absolute rate of cholesterol synthesis in the brain. J Lipid Res. 2004, 45: 1952-1957. 10.1194/jlr.D400007-JLR200.View ArticleGoogle Scholar
- Ko M, Zou K, Minagawa H, Yu W, Gong JS, Yanagisawa K, Michikawa M: Cholesterol-mediated neurite outgrowth is differently regulated between cortical and hippocampal neurons. J Biol Chem. 2005, 280: 42759-42765. 10.1074/jbc.M509164200.View ArticleGoogle Scholar
- Wahrle SE, Jiang H, Parsadanian M, Legleiter J, Han X, Fryer JD, Kowalewski T, Holtzman DM: ABCA1 is required for normal central nervous system ApoE levels and for lipidation of astrocyte-secreted apoE. J Biol Chem. 2004, 279: 40987-40993. 10.1074/jbc.M407963200.View ArticleGoogle Scholar
- Carter RJ, Morton AJ, Dunnett SB: Current Protocols in Neuroscience, Chapter 8.12 "Motor Coordination and Balance in Rodents". Current Protocols in Neuroscience. Edited by: Taylor G. 2001, New York, John Wiley & Sons, IncGoogle Scholar
- Karlsson U, Schultz RL: Fixation of the Central Nervous System from Electron Microscopy by Aldehyde Perfusion. I. Preservation with Aldehyde Perfusates Versus Direct Perfusion with Osmium Tetroxide with Special Reference to Membranes and the Extracellular Space. J Ultrastruct Res. 1965, 12: 160-186. 10.1016/S0022-5320(65)80014-4.View ArticleGoogle Scholar
- Reynolds ES: The use of lead citrate at high pH as an electron-opaque stain in electron microscopy. J Cell Biol. 1963, 17: 208-212. 10.1083/jcb.17.1.208.PubMed CentralView ArticleGoogle Scholar
- Palay SP, Chan-Palay V: Cerebellar Cortex. 1974, Berlin, Springer-VerlagView ArticleGoogle Scholar
- ImageJ. [http://rsb.info.nih.gov/ij/]
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.