Functional and immunocytochemical ... - Wiley Online Library

JOURNAL OF NEUROCHEMISTRY

| 2010 | 115 | 684–693

doi: 10.1111/j.1471-4159.2010.06957.x

*School of Biological Sciences and Centre for Brain Research, University of Auckland, Auckland, New Zealand Department of Anatomy with Radiology and Centre for Brain Research, Faculty of Medical and Health Sciences, University of Auckland, Auckland, New Zealand

Abstract Creatine uptake by neurons requires a specific creatine transporter (CRT). The purpose of the present work was to investigate the activity and localization of the CRT in primary cultures of hippocampal neurons obtained from 18-day rat embryos. Creatine uptake increased as the neurons differentiated in culture. Immunofluorescence microscopy showed most of the CRT was associated with dendrites, although some CRT was present in axons and axon terminals. Neurons contained high levels of Na+-dependent creatine transport activity (Km = 45.5 lM; Vmax, 1719 pmol creatine/min/mg protein) which was inhibited by competitive inhibitors of the CRT. The IC50 for guanidinoacetate, a precursor of creatine,

was 712 lM, 15-fold higher than the Km for creatine. Incubation of neurons with 1 mM creatine resulted in the accumulation of high levels of creatine which affected the Vmax but not the Km for creatine transport. The rate of creatine release from neurons increased in the absence of Na+ showing the importance of the electrochemical gradient for creatine retention. This is the first detailed study of the CRT in neurons and identifies primary cultures of rat hippocampal neurons as a good model for future studies of the CRT in relation to the effects of creatine on neuronal function and viability. Keywords: creatine transporter, creatine uptake, cultured rat hippocampal neurons, immunolocalization. J. Neurochem. (2010) 115, 684–693.

Creatine and the creatine kinase/phosphocreatine system play a critical role in the normal brain and under neuropathological conditions (Wyss and Kaddurah-Daouk 2000; Schlattner et al. 2006; Brosnan and Brosnan 2007; Andres et al. 2008). Genetic defects in either of the two enzymes, L-arginine : glycine amidino transferase (AGAT) or S-adenosy-Lmethionine: N-guanidinoacetate methyltransferase (GAMT), required for creatine synthesis, or the specific creatine transporter (CRT) required for its cellular uptake, result in a similar phenotype; an almost complete lack of creatine in the brain, developmental delays in speech and language, autism and seizures (Stockler et al. 2007). Unlike defects in creatine biosynthesis, symptoms from CRT deficiency are not moderated by creatine supplementation (Rosenberg et al. 2004; Stockler et al. 2007) indicating the important role the CRT plays in mediating the effects of creatine on brain function and development. Rabbit muscle and brain cDNAs were shown to encode a high affinity Na+- and Cl)-dependent creatine transporter (Guimbal and Kilimann 1994). The transporter was inhibited by creatine analogues such as b-guanidinopropionate (GPA) and guanidinoacetate (GA), but not by amino acids or other

compounds lacking an amidino group. These properties have been confirmed for cloned creatine transporters from a variety of species [for review, see (Christie 2007)]. The CRT is abundant in tissues with high energy demands, e.g. skeletal muscle, heart, brain and retina and also absorptive functions such as the kidney and intestine (Guimbal and Kilimann 1993; Gonzalez and Uhl 1994; Schloss et al. 1994; Saltarelli

684

Received June 9, 2010; revised manuscript received July 12, 2010; accepted August 12, 2010. Address correspondence and reprint requests to A/Professor David L Christie, School of Biological Sciences, The University of Auckland, Private Bag 92019, Auckland 1142, New Zealand. E-mail: [email protected] Abbreviations used: AGAT, L-arginine : glycine amidino transferase; CRT, creatine transporter; cycloCr, cyclocreatine; DIV, days of in vitro culture; E18, 18-day embryonic; FBS, fetal bovine serum; GA, guanidinoacetate; GAMT, S-adenosy-L-methionine: N-guanidinoacetate methyltransferase; GBA, guanidinobutyric acid; GPA, b-guanidinopropionate; KRH, Krebs Ringer–HEPES buffer; MAP2, microtubule associated protein 2; NBG, Neurobasal medium supplemented with B-27 and GlutaMAX; NR1, NMDA glutamate receptor R1; PBS, phosphatebuffered saline; PSD-95, post-synaptic density 95 kDa protein; SDS, sodium dodecyl sulfate.

2010 The Authors Journal of Neurochemistry 2010 International Society for Neurochemistry, J. Neurochem. (2010) 115, 684–693

Creatine transporter in rat hippocampal neurons | 685

et al. 1996; Peral et al. 2002). The CRT is member A8 of the solute carrier 6 family of transporters (SLC6A8) that includes transporters for neurotransmitters such as GABA, dopamine and serotonin (Christie 2007; Dodd and Christie 2007). Creatine deficiency because of mutations in the CRT/SLC6A8 gene is considered to be the next most common form of X-linked mental retardation to Fragile X (Rosenberg et al. 2004; Stockler et al. 2007). Creatine is a highly effective neuroprotective agent in animal models of neurological diseases and brain injury and is currently being trialed in humans as a potential therapeutic for several neurodegenerative diseases where there are known associated bioenergetic deficits (Wyss and Schulze 2002; Andres et al. 2008). These include: Huntington’s disease, Parkinson’s disease, Alzheimer’s disease, and amyotrophic lateral sclerosis. The National Institute of Neurological Disorders and Stroke is testing creatine in one of the largest clinical trials ever for Parkinson’s disease (Couzin 2007). To further understanding of how creatine provides neuroprotection and affects brain function, we recently carried out a detailed study on the immunolocalization of the CRT in rat brain (Mak et al. 2009). The CRT protein was predominantly found in neurons, principally those present in regions involved in motor and sensory processing and in learning, memory and limbic functions. CRT was high in the olfactory bulb, granule cells of the dentate gyrus of the hippocampus, pyramidal neurons of the cerebral cortex, Purkinje cells of the cerebellum, motor and sensory cranial nerve nuclei in the brainstem and the dorsal and ventral horns of the spinal cord (Mak et al. 2009). The CRT was associated with both cell bodies and dendrites of pyramidal cells in the cerebral cortex. CRTpositive apical dendrites were observed to project from layer V to layer II, raising interesting new questions about CRT trafficking in neurons, and suggesting it may play a role in supporting energy metabolism in neuronal subcompartments. In the present work, we have characterized the activity and localization of the endogenous CRT in cultured hippocampal neurons form embryonic (E18) rats (Kaech and Banker 2006). Creatine has been shown to protect these neurons from toxicity from glutamate and b-amyloid (Brewer and Wallimann 2000). The current study is the first to our knowledge to characterize the activity and localization of the CRT in cultures of primary neurons.

Materials and methods Materials All cell culture media were purchased from Invitrogen (Carlsbad, CA, USA). Papain was purchased from Worthington Biochemical Corporation (Lakewood, NJ, USA). [14C]creatine was purchased from American Radiolabelled Chemicals (St Louis, MO, USA). All other chemicals are from Sigma (St Louis, MO, USA). Plasticware for cell culture was purchased from BD Biosciences (Franklin

Lakes, NJ, USA). Coverslips were from Menzel Gla¨ser (Braunschweig, Germany). CRT antibody was produced in rabbits and affinity-purified as described previously (Mak et al. 2009). Antibodies used for immunofluorescence were specific for microtubule associated protein 2 (MAP2; clone HM-2, Sigma M4403), Tau-1 (clone PC1C6, Chemicon, Temecula, CA, USA, MAB3420), post synaptic density 95 kDa protein (PSD-95; Affinity Bioreagents, Golden, CO, USA, MA1-045), synaptophysin (clone SY38, Pierce, Rockford, IL, USA) and NMDA glutamate receptor R1 (clone 54.1, BD Biosciences). AlexaFluor secondary antibodies were from Molecular Probes (Eugene, OR, USA). Secondary and tertiary reagents for immunohistochemistry were from Sigma. The actin antibody used for western blotting was received from Sigma (clone AC-15, A5441). Secondary antibodies were from Bio-Rad (Hercules, CA, USA). Chemiluminescent substrate was from Pierce. Preparation of hippocampal neurons Procedures for use of animals were approved by the University of Auckland Animal Ethics Committee. Primary cultures of hippocampal neurons were prepared using a modified method based on that originally described (Kaech and Banker 2006). Briefly, E18 embryos from time-mated Wistar rats were removed from the uterus after killing of the dam, into a solution of ice-cold HEPES-buffered Hank’s balanced salts solution containing penicillin and streptomycin. The heads were removed and the hippocampi dissected out with the aid of a stereo microscope. Isolated hippocampi were digested in 0.8 lg/mL papain in HEPES-buffered Hank’s balanced salts solution for 10 min at 37C, then washed in warm modified Eagle’s medium/ 10% fetal bovine serum (FBS)/20 mM glucose for 2 min. The modified Eagle’s medium solution was replaced with NeurobasalTM medium supplemented with B-27 and GlutaMAXTM (referred to as NBG) and the hippocampal cells dissociated by trituration using a fire-polished Pasteur pipette. The resultant cell suspension was diluted and plated onto polylysine-coated dishes at 150 000 cells per well (40 000 cells/cm2) in 12-well dishes (Falcon 356470) for functional assays. Cells for immunofluorescence were plated out at 8000 cells/cm2 on 18 mm round #1.5 coverslips that were treated with 50% nitric acid, washed with water, sterilized by heating to 180C for 2 h, and coated overnight with 10 lg/mL polylysine. Cultures were maintained at 37C/5% CO2 in a humidified incubator for up to 21 days in vitro (DIV). Half the media was exchanged for fresh NBG every 7 days. Uptake of radiolabelled creatine by cultured neurons Creatine transport activity of cultured neurons was measured with [14C]creatine. Cultures were washed once with Krebs Ringer– HEPES buffer (KRH; 120 mM NaCl, 4.7 mM KCl, 2.2 mM CaCl2, 1.2 mM MgSO4, 1.2 mM KH2PO4, 25 mM HEPES, 10 mM glucose, pH 7.4), then incubated for 15 min with 20 lM [14C]creatine (specific activity 11 mCi/mmol) in KRH buffer at 37C. The incubation medium was removed, followed by three washes of icecold KRH and cells solubilised in 0.5% Triton X-100. Samples of solubilised cells were taken for liquid scintillation counting and protein determination. For uptake experiments without sodium, NaCl was replaced with N-methylglucamine. Curve fitting of kinetic data was by nonlinear, one-site binding (hyperbola) analysis using GraphPad Prism version 4.0 (GraphPad Software, San Diego, CA, USA). In some experiments, creatine transport assays were carried


686 | J. R. Dodd et al.

out in the presence of 1 mM guanidinopropionic acid (GPA), guanidinobutyric acid (GBA), cyclocreatine (cycloCr), GA, GABA, and arginine. Calculation of the IC50 value for GA used one-site competition analysis (GraphPad Prism). Release of pre-loaded [14C]creatine from cultured neurons Cells were grown for 10 days in vitro. The growth medium was then removed, and half of the conditioned medium combined with an equal volume of fresh NBG media containing [14C]creatine to give a final concentration of 1 mM (0.54 mCi/mmol). The cells were incubated with [14C]creatine for 24 h, after which the media was removed, and the cells washed with KRH buffer before the addition of either: fresh NBG buffer ±1 mM unlabeled creatine; or KRH buffer made with either NaCl or N-methyl glucamine. Samples of the media were taken at various time points for liquid scintillation counting; and the cells were washed, solubilised and intracellular [14C]creatine determined as described for the creatine uptake assays. Immunoblotting of CRT and actin in neuron lysates Sodium dodecyl sulfate (SDS)–polyacrylamide gel electrophoresis and immunoblotting was carried out as described previously (Dodd and Christie 2001). Briefly, neurons were cultured in 12-well dishes for 7–17 days, when the medium was removed and the cells washed in phosphate-buffered saline (PBS). The cells were incubated in lysis buffer [137 mM NaCl, 20 mM Tris, 1 mM EDTA, 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS, protease inhibitors (CompleteTM Mini protease inhibitor mixture; Roche Diagnostics GmbH, Mannheim, Germany) adjusted to pH 7.5 with HCl] for 30 min at 4C with gentle shaking. The lysate samples were transferred to 1.5-mL microcentrifuge tubes, centrifuged (20 000 g for 20 min at 4C) and the supernatants removed, and stored frozen. Aliquots of lysates were thawed, diluted with an equal volume of 2· SDS reducing buffer (125 mM Tris, 4% SDS, 20% glycerol, and 10% b-mercaptoethanol) and incubated for 30 min at room temperature, followed by centrifugation at 10 000 g for 5 min. Samples were run on 10% acrylamide gels and transferred to nitrocellulose membranes. CRT was detected by western blotting using rabbit CRT antibody at 1 : 1000 (Dodd and Christie 2001). Mouse antibody against actin was used at 1 : 1000 dilution. Goat anti-rabbit and anti-mouse IgG-horseradish peroxidase conjugated secondary antibodies with chemiluminescent substrates were used to visualize immunoreactive bands. Images were captured using a CCD camera and LAS-4000 system (Fujifilm, Tokyo, Japan). Immunohistochemistry on whole brain sections The immunohistochemical procedure has been described previously (Mak et al. 2009). Whole brains from E18 rats were dissected as described above and soaked in 4% paraformaldehyde/0.1 M PBS overnight, then transferred into 25% sucrose/0.1 M PBS for 24 h before freezing. 50-lm sagittal sections were cut using a freezing microtome. These were incubated as free-floating sections with the CRT antibody at 1 : 1000 dilution; then goat anti-rabbit biotin at 1 : 500; then ExtrAvidin-peroxide at 1 : 1000, with washes after each step. Colour was developed with exposure to 3,3¢-diaminobenzidene tetrahydrochloride/0.01% hydrogen peroxide for 10 min. Sections were dehydrated and mounted with DPX medium (Merck KGaA, Darmstadt, Germany).

Immunofluorescent labeling of cultured neurons Neurons grown on glass coverslips were fixed with 4% paraformaldehyde in PBS for 20 min, followed by a 15-min wash in 15 mM glycine in PBS. Cells were permeabilised in 0.1% Triton X100/10% FBS/PBS for 10 min. Primary antibodies were incubated with cells for 1 h at the following dilutions: rabbit CRT (1 : 200); mouse MAP2 (1 : 800); mouse tau (1 : 500); mouse synaptophysin (1 : 200); mouse PSD-95 (1 : 200); mouse NMDA glutamate receptor R1 (NR1; 1 : 500) in 10% FBS/PBS. Cells were washed four times in PBS before probing for 1 h with secondary antibodies AlexaFluor488 goat anti-rabbit IgG and AlexaFluor568 goat antimouse IgG at 1 : 500, again in 10% FBS/PBS. Cells were washed four times in PBS before mounting on slides with CitiFluor AF1 mounting medium (CitiFluor Ltd, Leicester, UK). Image collection Samples were visualised on a Leica DMR upright microscope (Leica Microsystems GmbH, Wetzlar, Germany) under bright field and epifluorescent conditions. Leica EN GFP and N2.1 filter cubes were used for epifluorescence. Single channel images were collected with a Leica DC500 CCD digital camera, then optimized for contrast and merged using Adobe Photoshop (Adobe Systems Inc, San Jose, CA, USA).

Results CRT immunoreactivity in E18 rat brain and hippocampus Saggital sections of E18 rat brain showed high levels of immunoperoxidase staining following incubation with affinity-purified CRT antibodies compared with the control (Fig. 1a and b). Strong CRT staining was present in the septal area, hypothalamus, thalamus, midbrain, pontine area and medulla. Discrete cellular staining was seen in regions of the hippocampus, cortex and olfactory bulb (Fig. 1c, d and e). The specific CRT staining of neuroepithelium of the hippocampal formation (regions corresponding to Ammon’s horn, the subicular region and the dentate gyrus) is of interest for the present work (Fig. 1c). Creatine uptake activity and immunolocalization of the CRT in E18 hippocampal neurons during in vitro culture Initial experiments showed the rate of [14C]creatine by cultured hippocampal neurons to remain linear for at least 30 min (data not shown). [14C]creatine uptake activity of hippocampal neurons cultured from 4 to 16 days in vitro is shown in Fig. 2(a). Creatine uptake was strongly Na+dependent and found to increase with the time in culture. This increase coincided with increases in the protein content of neurons associated with the development of processes in vitro. Analysis of extracts from neurons by western blotting showed an increase in CRT protein from 7 to 17 DIV (Fig. 2b). This increase was similar to that seen for actin, indicating that the CRT was not specifically up-regulated. We investigated the localization of the CRT in the neurons as



(a)

(a)

1 mm

(b) (b)

1 mm

(c)

(d)

(e)

Fig. 2 Expression of creatine uptake activity and the CRT protein during in vitro culture of neurons. (a) Creatine uptake activity was assessed in neurons cultured for 4–16 days in vitro (DIV). Cells were incubated for 15 min with 20 lM [14C]creatine in KRH buffer with (filled squares) and without Na+ (empty squares). The protein content of neurons from each time point is also shown (filled triangles). All points indicate the mean ± SEM of three separate experiments, each in triplicate. (b) Detection of CRT and actin by immunoblotting of lysate samples of neurons cultured for 7, 10, 14 and 17 DIV. The positions of molecular mass marker proteins and bands corresponding to the CRT (70 kDa) and actin (42 kDa) are indicated.

Fig. 1 Immunohistochemistry of CRT in sagittal sections from E18 rat embryo brains. (a) Brain section incubated without CRT antibody; (b) adjacent section exposed to CRT antibody and secondary antibodies. (c, d and e) Expanded views of regions of the hippocampus, cortex and olfactory bulb (boxed in b).

these markers, the pattern of staining was most similar to that of NR1 (Fig. 4).

they differentiated in vitro by immunofluorescence microscopy (Fig. 3). CRT immunofluorescence was associated with both the soma and extensive processes of neurons (7–17 DIV). Double labeling with MAP2 antibodies indicated that that much of the CRT immunoreactivity was associated with dendrites (Fig. 3). Some CRT positive processes were however not labeled by MAP2. Double labeling of the CRT, in combination with tau or synaptophysin antibodies, indicated that the CRT was also associated with some axons and axon terminals (Fig. 4). The CRT did not appear to be present in all axons. The association of CRT with dendritic processes was further established through use of specific markers for postsynaptic regions of dendrites (PSD-95 and NR1). The CRT was present within regions of dendrites containing these markers. While CRT immunostaining did not overlap with

Properties of creatine transporter present in cultured rat hippocampal neurons The saturation kinetics of creatine uptake of hippocampal neuron cultures is shown in Fig. 5(a). Na+-dependent creatine uptake was saturable (Km = 45.5 ± 5 lM; Vmax = 1719 ± 69 pmol creatine/min/mg protein; n = 4). A low level of creatine uptake occurred in the absence of Na+. This increased in a linear fashion with increasing creatine concentration suggesting it was because of diffusion. The uptake of [14C]creatine by hippocampal neurons was inhibited by creatine, and all known competitive inhibitors of the CRT (GPA, GBA, cycloCr and GA) but not by GABA or arginine (Fig. 5b). GA is a precursor in the synthesis of creatine and may be a physiological substrate for the creatine transporter. It was therefore of interest to investigate the ability of GA to compete with creatine for uptake into



Fig. 3 Localization of CRT in rat hippocampal neurons during in vitro culture by double immunofluorescent labeling. Comparison of immunofluorescent labeling of neurons with CRT antibodies with a dendritic marker (MAP2 antibody) at four time points during development (4, 10, 14 and 17 DIV). Single channel images (left and middle) are merged for the panel on right, with CRT in green and MAP2 in red. Arrows show points of co-localization while arrowheads indicate single labeling only. Scale bar, 10 lm.

Fig. 4 Comparison of CRT immunofluorescence with axonal, pre- and post-synaptic markers. Neurons at 14 DIV labeled with antibodies against CRT (green) and either tau, synaptophysin (sy38), PSD-95, or NR1 (red). Single channel images (left and middle) are merged for the panel on right. Scale bar, 10 lm.

neurons. The IC50 for GA inhibition of creatine uptake was 712 ± 140 lM (Fig. 5c).

and creatine-supplemented cells were very similar (31.2 ± 3.7 and 26.4 ± 2.5 lM, respectively).

Effect of creatine accumulation on the kinetics of creatine uptake in cultured neurons To investigate whether accumulation of creatine limits creatine uptake, cultured neurons were incubated with and without 1 mM creatine for 24 h, and washed prior to incubation with [14C]creatine. The saturation kinetics of neurons exposed to creatine were compared to those of cells cultured in medium without creatine (Fig. 6). The inset shows the Eadie-Hofstee plot of the data. The Vmax for creatine treated cells was 2.4-fold less than for untreated cells (379 ± 23 and 903 ± 72 pmol creatine/min/mg protein, respectively). The Km values for creatine uptake in control

The retention of creatine accumulated by neurons requires extracellular Na+ Neurons were incubated with 1 mM [14C]creatine for 16 h (sufficient time for creatine accumulation to reach a steady state). To determine whether cellular creatine was released through an exchange mechanism, cells were washed and incubated with unlabelled creatine (1 mM) in NBG culture medium. The amount of [14C]creatine, retained by the cells or released into the culture medium, is shown in Fig. 7(a and b). There was a significant increase in [14C]creatine release during incubation. A small, but significant, additional increase in [14C]creatine release occurred when



(a)

(b) Fig. 6 Effect of exposure of neurons to extracellular creatine on the saturation kinetics of creatine uptake. Cultured neurons were tested for creatine accumulation after 10 DIV. Prior to creatine uptake determination, neurons were incubated for 24 h in fresh NBG media with (triangles) or without (squares) 1 mM creatine. Cells were then incubated with increasing amounts of [14C]creatine for 15 min in KRH buffer. Results are also displayed in an Eadie-Hofstee plot (inset). Points indicate the mean and SEM of three experiments, each in triplicate.

(c)

Fig. 5 The kinetics and specificity of creatine transport activity of cultured neurons. Experiments used cells cultured for 10 DIV. (a) Cells incubated with increasing amounts of [14C]creatine for 15 min in KRH buffer with (filled square) and without (empty square) Na+. Sodiumdependent uptake was calculated at a Vmax of 1719 ± 69 pmol/min/mg protein and a Km of 45.5 ± 5.0 lM. Sodium-independent uptake was linear. Points indicate the mean and SEM of four experiments, each in triplicate. (b) Neurons were incubated with 20 lM [14C]creatine in KRH buffer for 15 min in the presence of creatine analogues, GABA and arginine, all at 1 mM. Values indicate the mean and SEM of three experiments, each in triplicate. Paired t-tests were performed for each analogue using GraphPad Prism, comparing each data set to control (untreated) cells. ***p < 0.001; **p < 0.01; and ns indicates no significant difference. (c) Neurons were incubated with 20 lM [14C]creatine in KRH buffer for 15 min along with increasing amounts of GA. Points indicate the mean and SEM of four experiments, each in triplicate. The IC50 for GA inhibition of creatine uptake was calculated as 712 ± 140 lM by one-site competition analysis (GraphPad Prism).

creatine was present in the incubation medium. This indicates that creatine release from neurons does not occur primarily through an exchange mechanism. We further investigated whether the retention of accumulated creatine was dependent on extracellular Na+. Neurons pre-loaded with [14C]creatine (16-h incubation with 1 mM creatine) were incubated in KRH buffer with or without Na+. [14C]creatine, retained by the cells or released into the culture medium, is shown in Fig. 7(c and d). The rate of [14C]creatine release from the cells was significantly greater in the absence of Na+.

Discussion Creatine, an energy metabolite and a well-known dietary supplement, is receiving increasing attention as a neuroprotective agent with major clinical trials underway for Parkinson’s and Huntington’s disease. Remarkably, potential therapeutic applications of creatine have overtaken our understanding of how creatine is taken up and distributed within the brain. It is known that the CRT is expressed in brain endothelial cells (Ohtsuki et al. 2002) which enables transport across the blood brain barrier. The CRT is predominantly located in neurons present in specific subregions of rat brain (Mak et al. 2009). The present work is the first detailed study of the activity and localization of the CRT in rat neurons. High levels of CRT were present in brain tissue from rat E18 embryos as shown by immunoperoxidase staining. The



(a)

(b)

(c)

(d)

Fig. 7 Effect of extracellular creatine, and Na+, on the release of creatine accumulated by neurons. After 10 DIV, cultured neurons were incubated with 1 mM [14C]creatine in NBG for 24 h. This medium was then replaced, with either NBG ± creatine or KRH ± Na+, and [14C]creatine retained by the cells or released into the medium determined at time intervals of up to 4 h. (a and b) Medium was replaced with NBG with (triangles) or without (squares) 1 mM creatine. (c and d) Medium was replaced with KRH with (filled squares) or

without Na+ (empty squares). The data represent the mean ± SEM of at least three experiments, each performed in triplicate. The data from each treatment in (a), (b), (c), and (d) were analysed by one-way ANOVA (***p < 0.001; **p < 0.01; *p < 0.05; ns, not significant); data from matching time-points, from (a) and (c), and (b) and (d), were analysed using a paired t-test (***p < 0.001; **p < 0.01; *p < 0.05; ns, not significant).

overall level of CRT expression in E18 rat brain was higher than that seen by us previously in adult rat brain (Mak et al. 2009). These results are consistent with early expression of the CRT observed in embryonic brain and other tissues both from the rat (Braissant et al. 2005) and spiny mouse (Ireland et al. 2009). Interestingly, each of these studies found CRT

expression to precede expression of one of the creatine biosynthetic enzymes, GAMT, suggesting uptake may be more important than biosynthesis for supplying creatine to the developing brain. Discrete cellular staining was apparent in the hippocampus as well as regions of the developing cortex and olfactory bulb.



In accordance with CRT immunohistochemical staining, primary cultures of rat E18 hippocampal neurons expressed very high levels of creatine transport activity. Saturation kinetics of neurons (10 DIV) indicated a Vmax of 1.7 nmol creatine/min/mg protein. Higher maximal rates of transport are only seen following heterologous expression of the CRT; a Vmax of 24.9 nmol/min/mg protein was seen in a HEK293 cell line stably expressing very high levels of the CRT (Dodd et al. 1999; West et al. 2005). The Km of the CRT in primary neurons (45.5 ± 5 lM) is similar to the range of reported values for the rat transporter (Christie 2007). Rat neuronal CRT exhibited a very strong Na+-dependence, and in common with the known properties of creatine transporters (Christie 2007) creatine uptake was effectively inhibited by known competitive inhibitors of the CRT (GPA, GBA, cycloCr and GA) but not by arginine or GABA. Together, these data indicate that creatine uptake in cultured rat hippocampal neurons occurs primarily through the CRT. The rate of creatine uptake in cultured neurons from rat E18 hippocampal tissue is approximately 3-, and 150-times higher than that seen previously for astroglia-rich mouse brain cell cultures and neuron-rich cultures of embryonic rat brain, respectively (Moller and Hamprecht 1989). Astrogliarich mouse brain cell cultures contain a variety of astroglial precursor cells but neurons are absent (Hamprecht and Loffler 1985). It is known that mature astrocytes do not express the creatine transporter (Braissant et al. 2001) although creatine transport activity may be associated with astrocyte or glial precursor cells. A disadvantage of the astroglia-rich cultures was that higher concentrations of creatine resulted in significant uptake through diffusion (Moller and Hamprecht 1989). In view of the results of the present study, we were surprised that previous work with neuron-rich cultures (Moller and Hamprecht 1989) did not demonstrate higher levels of creatine transport activity. We consider that the high level of Na+-dependent creatine transport activity seen in rat E18 hippocampal cultures is likely to be a combination of the use of dissected hippocampal tissue, a region containing high CRT, and the use of Neurobasal medium with a B-27 supplement, which reduces the growth of glial cells to give a nearly pure neuronal population (Brewer et al. 1993). The widespread expression of creatine biosynthetic enzymes AGAT and GAMT in cells within the brain indicate that some of the creatine in the brain comes through synthesis from precursor amino acids, arginine and glycine (Braissant et al. 2001, 2007). Recent work, however, suggests that relatively few cells (12%) are able to synthesize creatine, i.e. express both AGAT and GAMT. A similar portion of cells express both CRT and GAMT leading to the suggestion that GA, produced and secreted by cells expressing AGAT, may be taken up by the CRT and converted to creatine through the action of GAMT (Braissant et al. 2007). We found that GA is a relatively poor competitive inhibitor of creatine uptake by rat

neurons in vitro; the IC50 for GA was 712 ± 140 lM compared with the Km for creatine of 45.5 ± 5 lM. Km values for GA transport by the CRT expressed in oocytes and HEK293 cells were reported as 269 ± 54 lM and 412 ± 50 lM, respectively (Tachikawa et al. 2008). These results appear to contradict recent work with organotypic cultures of brain cells showing GA uptake to be inhibited by a 5-fold excess of creatine, as effectively as creatine uptake was inhibited by a 5-fold excess of GA (Braissant et al. 2010). One explanation is that the latter study investigated steady-state accumulation of GA and creatine (incubation with 200 lM creatine or GA over a 12 h) while we determined the effect of GA on the initial rate of creatine uptake (incubation with 20 lM creatine for 15 min). Regardless, there remains the possibility that GA uptake by neurons in vivo may occur through the CRT, in particular for the endogenous synthesis of creatine. While GA in CSF is about 100 times lower than creatine, GA may only be three to five times lower than creatine in whole brain [references cited in (Braissant and Henry 2008)]. Furthermore, where a neuron expressing the CRT is adjacent to a cell synthesizing and releasing GA, it may be possible for GA to pass directly from one cell to the other. Interestingly, the level of creatine transport activity increased in rat neurons as they were cultured in vitro. The cells do not divide in culture, so it appeared likely that the CRT may be associated with the extensive processes that are produced as the cultures differentiate. This was supported by the association of increased creatine transport activity with increased protein content, and confirmed by western blotting that showed CRT to increase in parallel with actin used as a standard. Immunofluorescence microscopy identified CRT staining of processes in neurons (7–17 DIV). Most of the CRT was associated with dendrites, as considerable overlap was observed with MAP2 immunostaining. The results from cultured neurons are consistent with immunostaining of sections from adult rat brain where CRT was found to be associated with apical dendrites of pyramidal cells in the cerebral cortex (Mak et al. 2009). CRT was also present in some axons as shown by CRT staining of processes unstained by MAP2 antibodies, but containing the axonal marker tau. Occasionally, co-staining was observed with the terminal marker synaptophysin. The presence of CRT in axons and axon terminals is consistent with the recent identification of CRT activity and protein in rat brain synaptosomes (Peral et al. 2010). Double labelling experiments with both PSD-95 and the NMDA receptor R1 subunit indicated that the CRT was present in processes containing these dendritic post-synaptic markers. In view of this, creatine has been shown to increase the activity of dendritic mitochondria and synaptogenesis in hippocampal neurons (Li et al. 2004). Furthermore the presence of CRT near NMDA receptors may account for the neuroprotective effects of creatine against glutamate excitotoxicity (Brewer and Wallimann



2000). Recently it has been shown that mitochondria accumulate near synapses following glutamate stimulation (Cai and Sheng 2009; Macaskill et al. 2009). If the CRT is located near NMDA receptors, creatine may be delivered directly to mitochondria, increasing their activity and ability to handle Ca2+ influx. Thus, expression of the CRT in dendritic compartments of neurons is consistent with a role for creatine in promoting mitochondrial function, synaptogenesis and neuronal activity. Incubation of hippocampal neurons with 1 mM creatine for 24 h resulted in the accumulation of very high levels of intracellular creatine. This concentration of creatine is higher than present in human CSF (17–90 lM), but lower than estimates of the total creatine concentration in normal human brain (5.1–6.3 mM) [references cited in (Braissant and Henry 2008)]. We chose 1 mM creatine to ensure the neuronal CRT was fully saturated; 1 mM creatine has also been found to protect cultured hippocampal neurons from glutamate excitotoxicity (Brewer and Wallimann 2000). Calculations based on the data in Fig. 7(a and c) taking into account the protein content estimated/well, and a value of 4.4 lL/mg protein for the neuronal cell volume (Olson and Li 2000) gives an average intracellular creatine concentration of 20 mM. Similar experiments with HEK293-CRT cells gave an estimate for the intracellular concentration of creatine of 30 mM (Dodd et al. 1999). These high levels of intracellular creatine also affected the saturation kinetics of the CRT resulting in a decrease in the Vmax of 60%. This is most likely because of the high intracellular concentration of creatine ( 20 mM) affecting the dynamic equilibrium of the CRT (Richerson and Wu 2003). In previous work, HEK293-CRT cells were able to maintain high levels of accumulated creatine, even when incubated for long periods (up to 7 h) in media without creatine (Dodd et al. 1999). In contrast, 30–40% of the creatine accumulated by neurons was released during incubation for 1–2 h. The rate of release of accumulated [14C]creatine was increased only slightly by the presence of extracellular creatine, indicating that neurons release intracellular creatine by mechanisms other that the exchange of substrate. The rate of release was however significantly increased in the absence of Na+. It is well-established that creatine uptake through the CRT is Na+- and Cl-- dependent (Christie 2007) so it is possible that the apparent increase in release is because of the absence of reuptake. In this case, however, we would have expected to see loss of [14C]creatine in exchange for non-radioactive creatine in the presence of Na+. Neurons, unlike HEK293 cells, are excitable cells so it is also possible that release is because of reversal of the CRT because of high intracellular [creatine] and the altered electrochemical gradient for Na+ ([Na+]IN > [Na+]OUT). It has been established that the GABA transporter, GAT-1, which is structurally and functionally closely related to the CRT (Dodd and Christie 2007) has a reversal potential close

to the resting potential of neurons (Wu et al. 2007). It has been suggested that the calculated driving force for a transporter should be viewed in a similar way to an ion channel; when the membrane potential is equal to the reversal potential there will be no net flux, and when it is not a driving force will be generated to move the solute in the direction needed to return the membrane potential back to the reversal potential (Richerson and Wu 2003). Thus, it is entirely possible that activation of ion channels and neural activity may lead to the release of creatine from neurons as has been observed in brain slices (Almeida et al. 2006). In summary, primary cultures of hippocampal neurons provide a good model system to study the neuronal CRT. Future studies of CRT function and trafficking in these cells may help unravel the relationship between sites of creatine uptake and the role of creatine in stimulating mitochondrial and synaptic function.

Acknowledgements The authors declare they have no conflicts of interest. The research was supported by a grant to the authors from the Auckland Medical Research Foundation. We thank Toby Lowe for help with the sectioning of rat embryo brains.

References Almeida L. S., Salomons G. S., Hogenboom F., Jakobs C. and Schoffelmeer A. N. (2006) Exocytotic release of creatine in rat brain. Synapse 60, 118–123. Andres R. H., Ducray A. D., Schlattner U., Wallimann T. and Widmer H. R. (2008) Functions and effects of creatine in the central nervous system. Brain Res. Bull. 76, 329–343. Braissant O. and Henry H. (2008) AGAT, GAMT and SLC6A8 distribution in the central nervous system, in relation to creatine deficiency syndromes: a review. J. Inherit. Metab. Dis. 31, 230–239. Braissant O., Henry H., Loup M., Eilers B. and Bachmann C. (2001) Endogenous synthesis and transport of creatine in the rat brain: an in situ hybridization study. Brain Res. Mol. Brain Res. 86, 193–201. Braissant O., Henry H., Villard A. M., Speer O., Wallimann T. and Bachmann C. (2005) Creatine synthesis and transport during rat embryogenesis: spatiotemporal expression of AGAT, GAMT and CT1. BMC Dev. Biol. 5, 9. Braissant O., Bachmann C. and Henry H. (2007) Expression and function of AGAT, GAMT and CT1 in the mammalian brain. Subcell. Biochem. 46, 67–81. Braissant O., Beard E., Torrent C. and Henry H. (2010) Dissociation of AGAT, GAMT and SLC6A8 in CNS: relevance to creatine deficiency syndromes. Neurobiol. Dis. 37, 423–433. Brewer G. J. and Wallimann T. W. (2000) Protective effect of the energy precursor creatine against toxicity of glutamate and beta-amyloid in rat hippocampal neurons. J. Neurochem. 74, 1968–1978. Brewer G. J., Torricelli J. R., Evege E. K. and Price P. J. (1993) Optimized survival of hippocampal neurons in B27-supplemented Neurobasal, a new serum-free medium combination. J. Neurosci. Res. 35, 567–576. Brosnan J. T. and Brosnan M. E. (2007) Creatine: endogenous metabolite, dietary, and therapeutic supplement. Annu. Rev. Nutr. 27, 241–261.



Cai Q. and Sheng Z. H. (2009) Moving or stopping mitochondria: Miro as a traffic cop by sensing calcium. Neuron 61, 493–496. Christie D. L. (2007) Functional insights into the creatine transporter. Subcell. Biochem. 46, 99–118. Couzin J. (2007) Clinical research. Testing a novel strategy against Parkinson’s disease. Science 315, 1778. Dodd J. R. and Christie D. L. (2001) Cysteine 144 in the third transmembrane domain of the creatine transporter is located close to a substrate-binding site. J. Biol. Chem. 276, 46983–46988. Dodd J. R. and Christie D. L. (2007) Selective amino acid substitutions convert the creatine transporter to a gamma-aminobutyric acid transporter. J. Biol. Chem. 282, 15528–15533. Dodd J. R., Zheng T. and Christie D. L. (1999) Creatine accumulation and exchange by HEK293 cells stably expressing high levels of a creatine transporter. Biochim. Biophys. Acta 1472, 128–136. Gonzalez A. M. and Uhl G. R. (1994) ‘Choline/orphan V8-2-1/creatine transporter’ mRNA is expressed in nervous, renal and gastrointestinal systems. Brain Res. Mol. Brain Res. 23, 266–270. Guimbal C. and Kilimann M. W. (1993) A Na(+)-dependent creatine transporter in rabbit brain, muscle, heart, and kidney. cDNA cloning and functional expression. J. Biol. Chem. 268, 8418–8421. Guimbal C. and Kilimann M. W. (1994) A creatine transporter cDNA from Torpedo illustrates structure/function relationships in the GABA/ noradrenaline transporter family. J. Mol. Biol. 241, 317–324. Hamprecht B. and Loffler F. (1985) Primary glial cultures as a model for studying hormone action. Methods Enzymol. 109, 341–345. Ireland Z., Russell A. P., Wallimann T., Walker D. W. and Snow R. (2009) Developmental changes in the expression of creatine synthesizing enzymes and creatine transporter in a precocial rodent, the spiny mouse. BMC Dev. Biol. 9, 39. Kaech S. and Banker G. (2006) Culturing hippocampal neurons. Nat. Protoc. 1, 2406–2415. Li Z., Okamoto K., Hayashi Y. and Sheng M. (2004) The importance of dendritic mitochondria in the morphogenesis and plasticity of spines and synapses. Cell 119, 873–887. Macaskill A. F., Rinholm J. E., Twelvetrees A. E., Arancibia-Carcamo I. L., Muir J., Fransson A., Aspenstrom P., Attwell D. and Kittler J. T. (2009) Miro1 is a calcium sensor for glutamate receptordependent localization of mitochondria at synapses. Neuron 61, 541–555. Mak C. S., Waldvogel H. J., Dodd J. R., Gilbert R. T., Lowe M. T., Birch N. P., Faull R. L. and Christie D. L. (2009) Immunohistochemical localisation of the creatine transporter in the rat brain. Neuroscience 163, 571–585. Moller A. and Hamprecht B. (1989) Creatine transport in cultured cells of rat and mouse brain. J. Neurochem. 52, 544–550. Ohtsuki S., Tachikawa M., Takanaga H., Shimizu H., Watanabe M., Hosoya K. and Terasaki T. (2002) The blood-brain barrier creatine

transporter is a major pathway for supplying creatine to the brain. J. Cereb. Blood Flow Metab. 22, 1327–1335. Olson J. E. and Li G. Z. (2000) Osmotic sensitivity of taurine release from hippocampal neuronal and glial cells. Adv. Exp. Med. Biol. 483, 213–218. Peral M. J., Garcia-Delgado M., Calonge M. L., Duran J. M., De Horra La M. C., Wallimann T., Speer O. and Ilundain A. (2002) Human, rat and chicken small intestinal Na+ – Cl) – creatine transporter: functional, molecular characterization and localization. J. Physiol. 545, 133–144. Peral M. J., Vazquez-Carretero M. D. and Ilundain A. A. (2010) Na(+)/ Cl())/creatine transporter activity and expression in rat brain synaptosomes. Neuroscience 165, 53–60. Richerson G. B. and Wu Y. (2003) Dynamic equilibrium of neurotransmitter transporters: not just for reuptake anymore. J. Neurophysiol. 90, 1363–1374. Rosenberg E. H., Almeida L. S., Kleefstra T. et al. (2004) High prevalence of SLC6A8 deficiency in X-linked mental retardation. Am. J. Hum. Genet. 75, 97–105. Saltarelli M. D., Bauman A. L., Moore K. R., Bradley C. C. and Blakely R. D. (1996) Expression of the rat brain creatine transporter in situ and in transfected HeLa cells. Dev. Neurosci. 18, 524–534. Schlattner U., Tokarska-Schlattner M. and Wallimann T. (2006) Mitochondrial creatine kinase in human health and disease. Biochim. Biophys. Acta 1762, 164–180. Schloss P., Mayser W. and Betz H. (1994) The putative rat choline transporter CHOT1 transports creatine and is highly expressed in neural and muscle-rich tissues. Biochem. Biophys. Res. Commun. 198, 637–645. Stockler S., Schutz P. W. and Salomons G. S. (2007) Cerebral creatine deficiency syndromes: clinical aspects, treatment and pathophysiology. Subcell. Biochem. 46, 149–166. Tachikawa M., Fujinawa J., Takahashi M. et al. (2008) Expression and possible role of creatine transporter in the brain and at the bloodcerebrospinal fluid barrier as a transporting protein of guanidinoacetate, an endogenous convulsant. J. Neurochem. 107, 768– 778. West M., Park D., Dodd J., Kistler J. and Christie D. (2005) Purification and characterization of the creatine transporter expressed at high levels in HEK293 cells. Protein Expr. Purif. 41, 393–401. Wu Y., Wang W., Diez-Sampedro A. and Richerson G. B. (2007) Nonvesicular inhibitory neurotransmission via reversal of the GABA transporter GAT-1. Neuron 56, 851–865. Wyss M. and Kaddurah-Daouk R. (2000) Creatine and creatinine metabolism. Physiol. Rev. 80, 1107–1213. Wyss M. and Schulze A. (2002) Health implications of creatine: can oral creatine supplementation protect against neurological and atherosclerotic disease? Neuroscience 112, 243–260.
