• brain;
  • ketone bodies;
  • neuroenergetics;
  • neuron-specific expression;
  • sodium-coupled monocarboxylate transporter;
  • sodium-coupled transport


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References

SMCT1 is a sodium-coupled (Na+-coupled) transporter for l-lactate and short-chain fatty acids. Here, we show that the ketone bodies, β-d-hydroxybutyrate and acetoacetate, and the branched-chain ketoacid, α-ketoisocaproate, are also substrates for the transporter. The transport of these compounds via human SMCT1 is Na+-coupled and electrogenic. The Michaelis constant is 1.4 ± 0.1 mm for β-d-hydroxybutyrate, 0.21 ± 0.04 mm for acetoacetate and 0.21 ± 0.03 mm for α-ketoisocaproate. The Na+ : substrate stoichiometry is 2 : 1. As l-lactate and ketone bodies constitute primary energy substrates for neurons, we investigated the expression pattern of this transporter in the brain. In situ hybridization studies demonstrate widespread expression of SMCT1 mRNA in mouse brain. Immunofluorescence analysis shows that SMCT1 protein is expressed exclusively in neurons. SMCT1 protein co-localizes with MCT2, a neuron-specific Na+-independent monocarboxylate transporter. In contrast, there was no overlap of signals for SMCT1 and MCT1, the latter being expressed only in non-neuronal cells. We also demonstrate the neuron-specific expression of SMCT1 in mixed cultures of rat cortical neurons and astrocytes. This represents the first report of an Na+-coupled transport system for a major group of energy substrates in neurons. These findings suggest that SMCT1 may play a critical role in the entry of l-lactate and ketone bodies into neurons by a process driven by an electrochemical Na+ gradient and hence, contribute to the maintenance of the energy status and function of neurons.

Abbreviations used



glial fibrillary acidic protein


human retinal pigment epithelial


monocarboxylate transporter




phosphate-buffered saline


sodium-coupled monocarboxylate transporter

Glucose is the major energy substrate used by the brain (Pellerin and Magistretti 2004). The metabolism of glucose in the brain is compartmentalized. Even though the metabolic utilization of glucose in the brain is increased during intense neuronal activity, uptake of glucose in astrocytes, rather than in neurons, accounts for most of this activity-associated glucose utilization (Pellerin and Magistretti 2004). Astrocytes convert glucose into l-lactate, a monocarboxylate, and release it into the extracellular medium to be taken up subsequently by neurons. Thus, the energy needs of the active neurons are not met directly by glucose oxidation but rather, by oxidation of l-lactate supplied by astrocytes. l-Lactate is therefore the primary metabolic fuel for neurons under normal physiological conditions. Normal levels of l-lactate in blood are quite significant (1–1.2 mm) and this l-lactate in the circulation is also available for neurons, in addition to the l-lactate generated by astrocytes.

Under circumstances when glucose availability is limited, such as during the suckling period in mammals (Hawkins et al. 1971; Cremer 1982), prolonged starvation and uncontrolled diabetes (Gjedde and Crone 1975; Hawkins et al. 1986), neurons must rely on substrates other than l-lactate to fulfil their high energy demands and sustain normal function. To preserve normal brain function, the body begins to mobilize its fat stores. Fatty acids released from the adipose tissue then undergo β-oxidation in the liver, which produces the ketone bodies β-d-hydroxybutyrate (β-D-HB) and acetoacetate for subsequent use as metabolic substrates by the neurons (Mitchell et al. 1995; Pellerin and Magistretti 2004). The concentration of ketone bodies in the circulation is usually very low (about 0.3 mm) and, under normal circumstances, these substrates are of little physiological relevance as a source of energy in the brain. However, during conditions such as pregnancy, starvation and uncontrolled diabetes, the concentration of these substrates in the blood increases substantially (approximately 10 mm) (Laffel 1999). Recent evidence suggests that astrocytes may also serve as a significant source of ketone bodies for the support of brain function (Guzman and Blazquez 2004). Similar to the interaction between the astrocytes and neurons in l-lactate production and utilization, ketone bodies are also shuttled between astrocytes and neurons (Guzman and Blazquez 2004).

l-Lactate and ketone bodies enter the brain across the endothelial cell layer of the blood–brain barrier via monocarboxylate transporters (MCTs) (Pellerin et al. 2005; Pierre and Pellerin 2005). MCTs belong to the solute-linked carrier gene family, SLC16 (Halestrap and Meredith 2004). Three MCT isoforms (MCT1, MCT2 and MCT4) are expressed in the brain. MCT1 and MCT4 represent low-affinity transporters and are expressed in astrocytes whereas MCT2, a high-affinity transporter, is expressed primarily in neurons (Pellerin et al. 2005; Pierre and Pellerin 2005). MCT1 is also expressed in the blood–brain barrier, where it plays a role in the transfer of blood-borne monocarboxylates into the brain, and its expression is subject to regulation under certain physiological and pathological conditions (Gerhart et al. 1997; Leino et al. 2001). Thus, the existing view of the metabolic utilization of l-lactate and ketone bodies by the brain is as follows. l-Lactate and ketone bodies present in the circulation are transported into the brain across the blood–brain barrier via MCT1, and neurons then take up these metabolites via MCT2. Astrocytes take up glucose and fatty acids, metabolize them to l-lactate and ketone bodies, respectively, and release these metabolites into the extracellular medium via MCT1 and MCT4; neurons take up this astrocyte-derived l-lactate and ketone bodies via MCT2. Therefore, the current belief is that MCT2 is the principal transporter responsible for meeting the energy needs of the neurons by providing l-lactate and ketone bodies under various physiological and pathological conditions.

Recently, we identified an Na+-coupled transporter for the transport of several monocarboxylates, including l-lactate (Ganapathy et al. 2005; Gupta et al. 2006). This transporter, known as SMCT (SLC5A8) (Sodium-coupled MonoCarboxylate Transporter), belongs to the solute-linked carrier gene family SLC5. In the light of the most recent studies in our laboratory that have identified a second isoform of SMCT, SLC5A12 (Srinivas et al. 2005), we refer to the SMCT described herein as SMCT1. The present study was undertaken to investigate the handling of ketone bodies (β-D-HB and acetoacetate) by SMCT1, and to determine the expression pattern of the transporter in the brain.

Materials and methods

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References


[14C]β-d-Hydroxybutyrate (specific activity, 55 mCi/mmol) was from American Radiolabeled Chemicals (St. Louis, MO, USA). Mouse monoclonal anti-NeuN antibody, mouse monoclonal anti-glial fibrillary acidic protein (GFAP) and anti-MCT1 antibody were from Chemicon International (Temecula, CA, USA). Anti-MCT2 antibody was obtained from Abcam (Cambridge, MA, USA). Goat anti-rabbit IgG coupled to Alexa Fluor 568, goat anti-mouse IgG coupled to Alexa Fluor 488 and Vectashield Hardset Mounting Media containing Hoechst 33342 (a nuclear stain) were from Molecular Probes (Carlsbad, CA, USA).

Subcloning of the human SMCT1 coding region in pSPORT1 vector

The previously published human SMCT1 cDNA clone consisted of the coding region inserted into the oocyte expression vector pGH19 (Miyauchi et al. 2004). This construct was not suitable for heterologous expression of the transporter in mammalian cells using the vaccinia virus expression technique because the procedure requires a vector with the T7 promoter, so we subcloned the coding region of human SMCT1 cDNA into pSPORT1 vector for this purpose. As the cloned human SMCT1 cDNA did not have the 5′- and 3′-untranslated regions, we flanked the coding region with the 5′- and 3′-untranslated regions derived from mouse SMCT1 cDNA (Gopal et al. 2004) to stabilize the mRNA. This was done by replacing the coding region of mouse SMCT1 cDNA in the pSPORT1-mouse SMCT1 cDNA plasmid with the coding region of human SMCT1 cDNA. This construct was used for expression of human SMCT1 in mammalian cells.

Functional expression of human SMCT1 in a mammalian cell line

The vaccinia virus expression system was used to express the cloned human SMCT1 heterologously in the human retinal pigment epithelial cell line, HRPE, as described previously (Gopal et al. 2004, 2005). The uptake of radiolabeled β-D-HB was measured in these cells 15 h after transfection. The transport buffer, in most cases, was 25 mm HEPES/Tris (pH 7.5) containing 140 mm NaCl, 5.4 mm KCl, 1.8 mm CaCl2, 0.8 mm MgSO4 and 5 mm glucose. Transport measurements were made in parallel at 37°C in vector-transfected cells and in SMCT1 cDNA-transfected cells to account for endogenous transport activity. SMCT1-specific transport was determined by subtracting the transport values measured in vector-transfected cells from the transport values measured in cDNA-transfected cells. The Na+ dependence of the uptake process was investigated by comparing the uptake of β-D-HB measured in the presence and absence of Na+. The Na+-free uptake buffer was prepared by replacing NaCl iso-osmotically with N-methyl-d-glucamine (NMDG) chloride. The expression level of the cloned SMCT1 in this system was low and therefore, cDNA-induced transport activity was monitored with a 15 min incubation period, even though these experimental conditions did not yield linear uptake rates. As this experimental system did not allow us to monitor linear rates of SMCT1-mediated uptake, only limited studies were performed with this approach as detailed kinetic analysis was not possible.

Functional analysis of human SMCT1 in Xenopus oocytes

The human SMCT1 cDNA, subcloned in the oocyte expression vector pGH19, was expressed heterologously in Xenopus oocytes by cRNA injection (Miyauchi et al. 2004). Capped cRNA from SMCT1 cDNA was synthesized using the mMESSAGE mMACHINE kit (Ambion Inc., Austin, TX, USA). Mature oocytes (stage IV or V) from Xenopus laevis were injected with 50 ng cRNA. Water-injected oocytes served as controls. The oocytes were used for electrophysiological studies 3–6 days after cRNA injection. Electrophysiological studies were performed by the two-microelectrode voltage-clamp method (Gopal et al. 2004, 2005; Miyauchi et al. 2004). Oocytes were perfused with an NaCl-containing buffer (100 mm NaCl, 2 mm KCl, 1 mm MgCl2, 1 mm CaCl2 and 10 mm HEPES/Tris, pH 7.5) followed by the same buffer containing different substrates. The membrane potential was clamped at − 50 mV. The differences between the steady-state currents measured in the presence and absence of substrates were considered as the substrate-induced currents. In the analysis of the saturation kinetics of substrate-induced currents, the kinetic parameter Km (i.e. the substrate concentration necessary for the induction of half-maximal current) was calculated by fitting the values of the substrate-induced currents to the Michaelis–Menten equation. The Na+-activation kinetics of substrate-induced currents was analyzed by measuring the substrate-specific currents in the presence of increasing concentrations of Na+, and the data for the Na+-dependent currents were analyzed according to the Hill equation to determine the Hill coefficient (h, the number of Na+ ions involved in the activation process) and Km for Na+ (i.e. the concentration of Na+ necessary for half-maximal activation). As the expression levels varied significantly from oocyte to oocyte, kinetic analyses were carried out by normalizing the expression levels to eliminate the variations among different oocytes. This was done by taking the maximally-induced SMCT1-specific currents in each kinetic experiment in individual oocytes as 1. The kinetic parameters were determined using the computer program Sigma Plot, version 6.0 (SPSS, Inc., Chicago, IL, USA).

Data analysis

Experiments with HRPE cells were repeated three times with three independent transfections, and transport measurements were made in duplicate in each experiment. Electrophysiological measurements of substrate-induced currents were repeated at least three times with separate oocytes. Data are presented as means ± SEM of these replicates.

In situ hybridization

To localize the mRNA transcript encoding SMCT1 in brain, in situ hybridization was performed on sections of mouse brain as described previously (Seth et al. 2001; Inoue et al. 2002). Tissue sections were hybridized with the digoxigenin-labeled antisense probe (1 µg/mL) and incubated overnight at 58°C. For immunological detection of the probe, the anti-digoxigenin antibody, conjugated to alkaline phosphatase, was diluted 1 : 5000, and slides were incubated with this antibody for 2 h at 25°C. The color reaction was developed in nitro blue tetrazolium/5-bromo-4-chloro-3-indolyl phosphate. In all cases, some cryosections were hybridized with the sense (negative control) riboprobe to determine non-specific binding. Preparation of the antisense and sense riboprobes for mouse SMCT1 has been described previously (Gopal et al. 2004).

Preparation of polyclonal antibody against mouse SMCT1

A rabbit polyclonal antibody against mouse SMCT1 was generated using a commercial firm (Biosynthesis, Inc., Lewisville, TX, USA). The anti-peptide antibody was raised against the peptide sequence ELNFTDHSGKINGTRL, which corresponds to residues 596–611 of mouse SMCT1. To determine the specificity of the anti-SMCT1 antibody, the vaccinia virus expression system was used to express the cloned mouse SMCT1 heterologously in HRPE cells. Immunofluorescence experiments were performed in parallel in vector-transfected cells and in SMCT1 cDNA-transfected cells. HRPE cells do not express SMCT1 (Gopal et al. 2004) and therefore, vector-transfected cells served as a negative control. Normal rabbit pre-immune serum was used as an additional negative control. Furthermore, the antibody was neutralized with an excess of the antigenic peptide and used as a negative control in immunofluorescence studies with mouse brain sections.

Immunofluorescence localization

Immunofluorescence methods were used to localize SMCT1 protein in mouse brain. Sagittal cryosections of mouse brain were fixed in ice-cold acetone for 5 min, washed with 0.01 m phosphate-buffered saline (PBS) (pH 7.4) and blocked with 1× Power Block (Biogenex, San Ramon, CA, USA) for 10 min at room temperature (22°C). For double-labeling experiments, sections were then incubated overnight, at 4°C, with the primary polyclonal antibody against SMCT1 at a concentration of 1 : 250 and monoclonal anti-NeuN, a neuronal cell body marker, at a dilution of 1 : 100, or with polyclonal anti-SMCT1 antibody (1 : 250) and monoclonal anti-GFAP, a marker for glial cells, at a dilution of 1 : 100. Negative control sections were treated identically except that PBS was substituted in place of primary antibodies for overnight incubation. Sections were rinsed and incubated for 1 h with goat anti-rabbit IgG coupled to Alexa Fluor 568 and goat anti-mouse IgG coupled to Alexa Fluor 488, both at a dilution of 1 : 1000. Similar techniques were used to assess the co-localization of SMCT1 with either MCT1 or MCT2. Coverslips were mounted with Vectashield Hardset mounting medium with Hoechst 33342 (a nuclear stain), and sections were examined by epifluorescence using the Zeiss Axioplan 2 microscope (Carl Zeiss Inc., Oberkochen, Germany) and photographed with a Spot Camera (AxioCam HRM; Carl Zeiss Inc.) using Spot Software Version 2.2.

Establishment of mixed cultures of cortical neurons and astrocytes from rat brain

Mixed cultures of cortical neurons and astrocytes from rat brain were established by a procedure described previously for mesencephalic neurons (Prasad and Amara 2001). Sprague–Dawley rat pups (2–4 days old) were anesthetized by intraperitoneal injection of ketamine HCl (3 mg/pup). The cortex was dissected out and incubated in a dissociation medium, containing 20 U/mL activated papain, at 34–36°C under continuous oxygenation for 2 h. Tissue was then dissociated with a fire-polished Pasteur pipette in minimum essential medium. Dissociated cells were plated on glass coverslips previously coated with 100 µg/mL poly d-lysine and 5 µg/mL laminin at a density of approximately 150 000 cells/cm2. Neuronal medium (50% minimum essential medium, 39% Ham's-F12 medium, 10% heat-inactivated horse serum, 1% heat-inactivated newborn calf serum, 0.45%d-glucose, 5 pg/mL insulin and 0.1 mg/mL apotransferrin), conditioned overnight over glia, was used to maintain the cultures. These cultures were used for immunofluorescence studies as described above for mouse brain sections.


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References

Transport of β-d-hydroxybutyrate via human SMCT1 in a mammalian cell line

To investigate the ability of human SMCT1 to transport β-D-HB, we first used a mammalian cell expression system. We expressed the cDNA in HRPE cells, using the vaccinia virus expression technique, and compared the uptake of β-D-HB in vector-transfected cells and in cDNA-transfected cells (Fig. 1a). The uptake of β-D-HB was significantly higher in cDNA-transfected cells than in control cells (approximately twofold) when monitored in the presence of Na+. This increase was not observed when the uptake was measured in the absence of Na+. The cDNA-induced β-D-HB uptake was inhibited markedly by various monocarboxylates known to be substrates for human SMCT1 (Fig. 1b). At a concentration of 5 mm, pyruvate, nicotinate, l-lactate, propionate and butyrate inhibited the cDNA-induced uptake of β-D-HB by > 80%.


Figure 1.  Transport of β-D-HB by human SMCT1 in a mammalian cell expression system. HRPE cells were transfected with either pSPORT1 vector alone or pSPORT1-human SMCT1 cDNA construct. (a) Uptake of β-D-HB (30 µm) was measured in vector-transfected cells (black bars) and hSMCT1 cDNA-transfected cells (shaded gray bars) in the presence of either NaCl or NMDG chloride. The time of incubation was 15 min. (b) Inhibition of Na+-dependent uptake of [14C]β-D-HB (30 µm; 15 min incubation) by various unlabeled short-chain fatty acids (5 mm). Data are given as percentage of control uptake (100%) measured in the absence of inhibitors.

Download figure to PowerPoint

Transport of β-d-hydroxybutyrate via human SMCT1 in Xenopus oocytes

Even though the mammalian cell expression system clearly indicated that the cloned human SMCT1 is able to transport β-D-HB in an Na+-dependent manner, the cDNA-induced uptake activity was small (approximately twofold). This low activity made it difficult to carry out detailed kinetic analysis of the transport function due to inability to monitor the initial rates of cDNA-induced transport rates. Therefore, we employed the Xenopus laevis oocyte expression system. As a positive control, we used l-lactate as a substrate in the initial studies. This monocarboxylate is a transportable substrate for human SMCT1 as it has been shown previously that exposure of oocytes expressing the cloned human SMCT1 to this compound induces Na+-dependent inward currents under voltage-clamp conditions (Miyauchi et al. 2004). Similar results were obtained in our study. Exposure of human SMCT1-expressing oocytes to 0.5 mm l-lactate induced marked inward currents (Fig. 2). Such currents were not detectable in water-injected oocytes (data not shown). The l-lactate-induced currents in human SMCT1-expressing oocytes were obligatorily dependent on the presence of Na+. There was no involvement of external Cl. We then examined whether β-D-HB and acetoacetate were recognized as substrates by human SMCT1 by monitoring the inward currents in human SMCT1-expressing oocytes upon exposure to these ketone bodies in the presence of Na+ (Fig. 2). At a concentration of 0.5 mm, both β-D-HB and acetoacetate induced inward currents in SMCT1-expressing oocytes. These currents were not detectable in water-injected oocytes (data not shown). As seen with l-lactate, the currents induced by the ketone bodies were obligatorily dependent on the presence of Na+, with no involvement of Cl. While the general characteristics of the induced currents were similar for both β-D-HB and acetoacetate in terms of Na+ dependence and Cl independence, the magnitude of the induced currents was significantly greater with β-D-HB than with acetoacetate when examined at 0.5 mm (133 ± 44 nA for β-D-HB vs. 87 ± 10 nA for acetoacetate in three different oocytes, p < 0.01). Figure 3 shows the voltage dependence of the inward currents induced by β-D-HB and l-lactate in SMCT1-expressing oocytes. Hyperpolarization of the oocyte membrane enhanced the magnitude of induced currents, as expected from the electrogenic nature of SMCT1. We also examined the influence of extracellular pH on the transport activity. A change of pH from 7.5 to 6 reduced l-lactate-induced currents only to a small extent (about 15%), whereas the effect was more pronounced in the case of β-D-HB (about 30%).


Figure 2.  Transport of ketone bodies via human SMCT1 in the Xenopus laevis oocyte expression system. Oocytes expressing human SMCT1 were perfused with 0.5 mmβ-D-HB and acetoacetate at pH 7.5 in the presence of Na+ and Cl (NaCl), in the absence of Na+ but in the presence of Cl (– Na+), or in the presence of Na+ but in the absence of Cl (– Cl). The substrate-induced inward currents were monitored using the two-microelectrode voltage-clamp technique. The oocyte membrane potential was clamped at − 50 mV. Lactate was used as the known substrate for human SMCT1 for comparison. Water-injected oocytes were used in similar experiments, and there was no detectable current with any of the substrates either in the presence or absence of Na+ (data not shown).

Download figure to PowerPoint


Figure 3.  Voltage dependence of currents induced by l-lactate and β-D-HB in SMCT1-expressing oocytes. Oocytes expressing human SMCT1 were exposed to 0.5 mm l-lactate or β-D-HB, and the substrate-induced currents were monitored at different testing membrane potentials using the two-microelectrode voltage-clamp technique.

Download figure to PowerPoint

The induced currents were saturable for β-D-HB and acetoacetate and conformed to Michaelis–Menten kinetics (data not shown). The values for Km were 1.4 ± 0.1 mm for β-D-HB and 0.21 ± 0.04 mm for acetoacetate (Table 1). The difference between these two ketone bodies in terms of Km was statistically significant (p < 0.001). As the expression levels varied from oocyte to oocyte, kinetic analysis was carried out by normalizing the expression levels. When the actual currents induced by the saturating concentrations of the two ketone bodies were considered, there was a statistically significant difference between the values (445 ± 93 nA for β-D-HB vs. 117 ± 20 nA for acetoacetate in three different oocytes; p < 0.001).

Table 1.   Michaelis constants (Km) for various monocarboxylates and ketone bodies for transport via human SMCT1
β-d-Hydroxybutyrate1442 ± 124
β -l-Hydroxybutyrate2327 ± 169
d-Lactate1088 ± 68
l-Lactate184 ± 8
Pyruvate387 ± 43
Acetoacetate213 ± 39
α-Ketoisocaproate209 ± 27

We then compared the affinities of the d- and l-isomers of β-HB and lactate by saturation analysis of the inward currents induced by increasing concentrations of these compounds in SMCT1-expressing oocytes. The data are presented in Table 1. The d- as well as the l-isomers of β-HB were recognized as transportable substrates by SMCT1 with comparable affinities (Km value for the β-L-HB was 2.3 ± 0.2 mm compared with the corresponding value of 1.4 ± 0.1 mm for β-D-HB). In contrast, there was a marked difference in affinities between l-lactate and d-lactate, even though both isomers were transported via SMCT1. The Km value for l-lactate (0.18 ± 0.01 mm) was about fivefold lower than that for d-lactate (1.09 ± 0.07 mm), indicating a preferential recognition of the l-isomer by the transporter. The Km values for pyruvate and acetoacetate for transport via SMCT1 were comparable with that of l-lactate. In addition to β-D-HB and acetoacetate, there is a third keto acid which is physiologically important to brain metabolism. A significant amount of glutamate/glutamine is synthesized in astrocytes via transamination between leucine and α-ketoglutarate, a process in which α-ketoisocaproate is released as a product (Yudkoff 1997). This keto acid has been shown to accumulate in neuroblastoma cells by a Na+-dependent mechanism (Bachowska-Mac et al. 1997). Therefore, we examined whether this keto acid is a transportable substrate for SMCT1. Exposure of SMCT1-expressing oocytes to α-ketoisocaproate induced marked inward currents in a Na+-dependent manner, indicating transport via the transporter. The induced currents were saturable with a Km value of 0.21 ± 0.03 mm (Table 1).

As β-D-HB and acetoacetate are present as monovalent anions at pH 7.5, the induction of inward currents by these substrates indicates co-transport of two or more than two Na+ per transport cycle. To confirm this, we analyzed the Na+-activation kinetics for β-D-HB-induced currents (Fig. 4). The dependence of β-D-HB-induced currents on the concentration of Na+ was sigmoidal, suggesting involvement of multiple Na+ ions in the activation process. Analysis of the data by the Hill equation gave a value of 1.9 ± 0.1 for the Hill coefficient (h), which indicates a Na+ : β-D-HB stoichiometry of 2 : 1 for the activation process. The concentration of Na+ necessary for half-maximal activation of β-D-HB (5 mm)-induced currents (Km) was 20 ± 1 mm.


Figure 4.  Na+ activation kinetics for human SMCT1 in the Xenopus oocyte expression system with β-D-HB as the substrate. Oocytes expressing human SMCT1 were perfused with 5 mmβ-D-HB in the presence of increasing concentrations of Na+ (10–100 mm), and substrate-specific currents were monitored at − 50 mV using the two-microelectrode voltage-clamp technique. Na+-dependent currents were used for kinetic analysis. Osmolality of the perfusion buffer was maintained by iso-osmotic substitution of NaCl with NMDG chloride. Because of variable expression levels among different oocytes, the current induced in the presence of 100 mm Na+ in each oocyte was taken as 1 to correct for the variations. The inset shows the Hill plot.

Download figure to PowerPoint

Expression of SMCT1 in mammalian brain

Functional studies of human and mouse SMCT1 have shown that the transporter is capable of mediating Na+-coupled active uptake of l-lactate (Coady et al. 2004; Gopal et al. 2004, 2005; Miyauchi et al. 2004) and ketone bodies (present study). As l-lactate and ketone bodies serve as crucial energy substrates for neurons under various physiological and pathological conditions, we examined the expression of this transporter in the mouse brain by in situ hybridization and immunofluorescence. Positive signals with an antisense riboprobe were detected throughout the brain, including the cortex, hippocampus, cerebellum and pituitary gland (Fig. 5). The hybridization signals with the antisense probe were specific, because a sense probe did not yield detectable signals (Fig. 5c). At higher magnification of the cortex and hippocampal regions (Fig. 5b), it was evident that the expression of SMCT1 mRNA was restricted primarily to neuronal cell bodies.


Figure 5.  Distribution of SMCT1 mRNA in mouse brain as assessed by in situ hybridization. (a) Hybridization of a sagittal section of mouse brain with an antisense riboprobe specific for mouse SMCT1. (b) Higher magnification of the regions representing the cortex and hippocampus hybridized with the antisense riboprobe. (c) Negative control hybridized with a sense riboprobe.

Download figure to PowerPoint

We then investigated the expression pattern of SMCT1 protein in the brain by immunofluorescence analysis. Exposure of the brain sections to an antibody specific for SMCT1 (red) demonstrated an expression pattern of SMCT1 protein similar to that of SMCT1 mRNA (Figs 6a and b). To determine whether the transporter is expressed in neurons or glial cells, we focused on the cortical and hippocampal regions and performed co-localization studies with neuronal markers (NeuN, green) (Fig. 6a) and glial markers (GFAP, green) (Fig. 6b). Both in the cortex and in the hippocampus, SMCT1 protein co-localized with NeuN; no such co-localization was evident between SMCT1 and GFAP. These data show clearly that SMCT1 protein is expressed specifically in neurons, with no detectable expression in glial cells. We confirmed the specificity of the antibody by two different approaches. HRPE cells do not express SMCT1 constitutively (Gopal et al. 2004) and accordingly, there was no immunopositive signal with the antibody when the cells were transfected with vector alone. However, the antibody gave a positive signal in cells transfected with mouse SMCT1 cDNA (Fig. 6c). We also assessed the specificity of the antibody by neutralization with the antigenic peptide. The positive signal detected with the antibody in brain sections disappeared when the antibody was neutralized with an excess of the antigenic peptide (Fig. 6d).


Figure 6.  Immunofluorescence localization of SMCT1 in mouse brain (a, b) and specificity of the anti-SMCT1 antibody (c, d). (a) Saggital cryosections of mouse brain were incubated with an antibody against SMCT1 (red) (i, iv) and an antibody against NeuN (green), a neuronal cell body marker (ii, v). The merged images of positive signals for SMCT1 and NeuN are shown in (iii) and (vi). (b) Saggital cryosections of mouse brain were incubated with an antibody against SMCT1 (red) (vii, x) and an antibody against GFAP (green), a glial cell marker (viii, xi). The merged images of positive signals for SMCT1 and GFAP are shown in (ix) and (xii). (c) HRPE cells were transfected with either vector alone (vector/HRPE) or human SMCT1 cDNA (SMCT1/HRPE), and the anti-SMCT1 antibody was then used to detect the heterologously expressed SMCT1 protein (red). (d) Saggital sections of mouse brain were incubated with either anti-SMCT1 antibody or with the antibody that had been neutralized with an excess of the antigenic peptide. The secondary antibody was coupled to Alexa Fluor 488 (green). Hoechst 33342 (blue) was used as a nuclear counterstain.

Download figure to PowerPoint

To substantiate the neuron-specific localization of SMCT1 further, we performed co-localization studies with MCT1, which is expressed in non-neuronal cells, and MCT2, which is expressed predominantly in neuronal cells (Figs 7a and b). We found no overlap between the expression of SMCT1 and that of MCT1; in contrast, there was a significant overlap between the expression of SMCT1 and MCT2.


Figure 7.  Immunofluorescence localization of SMCT1, MCT1 and MCT2 in mouse brain. (a) Sections of mouse brain were incubated with an antibody against SMCT1 (green) and an antibody against MCT1 (red). (b) Sections of mouse brain were incubated with an antibody against SMCT1 (red) and an antibody against MCT2 (green). (c, d) Immunofluorescence localization of SMCT1 in mixed cultures of cortical neurons and astrocytes from rat brain. The mixed cultures of rat brain neurons and astrocytes were incubated with an antibody against SMCT1 (red) (i, v), NeuN (green) (ii) and GFAP (green) (vi). Hoechst 33342 (blue) was used as a nuclear counterstain (vii). The merged image of positive signals for SMCT1 and NeuN is shown in (iii). The merged image of positive signals for SMCT1, GFAP and Hoechst 33342 is shown in (viii). A negative control with pre-immune serum is shown in (iv).

Download figure to PowerPoint

To corroborate the findings using the mouse brain sections with respect to the neuron-specific expression of the transporter, we analyzed the expression of the transporter in mixed cultures of rat cortical neurons and astrocytes. Again, co-localization studies were performed with these cultures using antibodies specific for SMCT1 (red), NeuN (green) or GFAP (green) (Figs 7c and d). SMCT1 was detected only in cells that were also positive for NeuN. SMCT1 in these cells was localized to axonal processes whereas NeuN (a marker for neuronal cell body) was detected in the cell bodies. Incubation of similar cultures with antibodies specific for SMCT1 and GFAP revealed no co-localization between positive signals for SMCT1 and the glial cell marker. Localization of SMCT1 in these cultures was completely independent of localization for GFAP.


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References

These studies establish, for the first time, the functional identity of human SMCT1 as a Na+-coupled transporter not only for l-lactate but also for the ketone bodies β-D-HB and acetoacetate. Similarly to the transport of l-lactate via SMCT1, the transport of the ketone bodies via the transporter is also Na+-dependent and electrogenic. The Na+ : substrate stoichiometry of 2 : 1 for the ketone bodies suggests that the transporter can mediate the concentrative uptake of these metabolic intermediates into cells very effectively coupled to the sodium motive force. These findings are of biological significance because SMCT1 is the only transporter, thus far known, that can transport l-lactate and ketone bodies into cells in a Na+-coupled manner. This has significant physiological and pathological implications for the survival and function of tissues which depend on these metabolites for generation of energy. Arguably, the most important of such tissues is the brain. In the brain, neurons are the major users of metabolic energy and therefore, they would benefit the most from such an energy-coupled active uptake system for these metabolic intermediates. Our studies show for the first time that SMCT1 is expressed in the brain in a neuron-specific manner. Therefore, it is most likely that SMCT1 plays a significant role in the energetics of neurons in the central nervous system. MCT2, which is expressed in neurons, is also capable of transporting l-lactate and ketone bodies by an H+-coupled and electroneutral mechanism. MCT2 possesses higher affinities for l-lactate and ketone bodies compared with MCT1 and MCT4. The Michaelis constants (Km) for l-lactate and β-D-HB for transport via MCT2 are approximately 0.7 mm and 1.2 mm, respectively (Lin et al. 1998; Broer et al. 1999). SMCT1 has a Km value of about 0.2 mm for l-lactate and about 1.5 mm for β-D-HB (Miyauchi et al. 2004; present study). Therefore, we speculate that SMCT1 and MCT2 may complement each other in the maintenance of the energy status of the neurons by providing l-lactate and β-D-HB via two different mechanisms. Further work is needed to assess precisely the relative contributions of MCT2 and SMCT1 to the provision of these energy substrates to neurons under various physiological and pathological conditions.

The finding that α-ketoisocaproate is a transportable substrate for SMCT1 may be relevant to the handling of this keto acid in the brain. Astrocytes are capable of glutamate synthesis for subsequent conversion into glutamine that is then released into the extracellular medium. The released glutamine is taken up by neurons where it serves as an immediate precursor for the generation of glutamate. Synthesis of glutamate within astrocytes occurs via transamination between leucine and α-ketoglutarate, a process which results in the generation of α-ketoisocaproate. This keto acid is then released from the astrocytes for subsequent uptake into neurons for conversion back into leucine. A Na+-coupled transport system has been described for this keto acid in neuroblastoma cells (Bachowska-Mac et al. 1997), but the molecular identity of the transport system remains unknown. Interestingly, other α-keto acids that arise from transamination of valine and isoleucine are able to compete with α-ketoisocaproate for uptake into these cells, whereas β-hydroxybutyrate and lactate are without any effect. Our studies show that SMCT1 expressed in neurons can recognize α-ketoisocaproate as well as β-hydroxybutyrate and lactate. The substrate selectivity of the transport system described in neuroblastoma cells thus seems to be different from that of SMCT1. However, it has to be noted that, under the experimental conditions used for uptake measurements, the rate of conversion of the transported α-ketoisocaproate into leucine seems to be very fast in these cells compared with the rate of its entry (Bachowska-Mac et al. 1997), which might make interpretation of the data difficult. Further work is needed to determine whether SMCT1 is indeed responsible for the uptake of α-ketoisocaproate reported in these cells.

The relevance of ketone bodies to neuronal function has become increasingly apparent in recent years. Ketone bodies are an important source of metabolic energy for the brain in the neonatal period when fat-rich and ketogenic breast milk constitutes the primary caloric source in the diet (Morris 2005). In addition to the physiological significance of a ketogenic diet in the neonatal period, such diets appear to have a therapeutic potential in the management of epilepsy (Swink et al. 1997; Greene et al. 2003; Stafstrom and Bough 2003). Furthermore, ketone bodies have been shown to offer protection to neurons against various insults, such as hypoxia (Masuda et al. 2005) and excitotoxicity (Massieu et al. 2003), as well as in neurodegenerative disorders such as Parkinson's disease (Tieu et al. 2003). Ketone bodies also support synaptic function during development (Izumi et al. 1998) and improve cognitive function in patients with Alzheimer's disease (Reger et al. 2004). Such biological and therapeutic roles of ketone bodies are undoubtedly related to the ability of these metabolites to serve as an important energy source for the neurons. SMCT1, expressed in neurons, may play a critical role in the efficient entry of these metabolites into neurons for subsequent oxidation to produce the metabolic energy.


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References
  • Bachowska-Mac M., Nehlig A., Nalecz M. J. and Nalecz K. A. (1997) Transport of alpha-ketoisocaproate in neuroblastoma NB-2a cells. Biochem. Biophys. Res. Commun. 237, 6367.
  • Broer S., Broer A., Schneider H. P., Stegen C., Halestrap A. P. and Deitmer J. W. (1999) Characterization of the high-affinity monocarboxylate transporter MCT2 in Xenopus laevis oocytes. Biochem. J. 341, 529535.
  • Coady M. J., Chang M. H., Charron F. M., Plata C., Wallendorff B., Sah J. F., Markowitz S. D., Romero M. F. and Lapointe J. Y. (2004) The human tumour suppressor gene SLC5A8 expresses a Na+-monocarboxylate cotransporter. J. Physiol. 557, 719731.
  • Cremer J. E. (1982) Substrate utilization and brain development. J. Cereb. Blood Flow Metab. 2, 394407.
  • Ganapathy V., Gopal E., Miyauchi S. and Prasad P. D. (2005) Biological functions of SLC5A8, a candidate tumour suppressor. Biochem. Soc. Trans. 33, 237240.
  • Gerhart D. Z., Enerson B. E., Zhdankina O. Y., Leino R. L. and Drewes L. R. (1997) Expression of monocarboxylate transporter MCT1 by brain endothelium and glia in adult and suckling rats. Am. J. Physiol. 273, E207E213.
  • Gjedde A. and Crone C. (1975) Induction processes in blood–brain transfer of ketone bodies during starvation. Am. J. Physiol. 229, 11651169.
  • Gopal E., Fei Y. J., Sugawara M., Miyauchi S., Zhuang L., Martin P., Smith S. B., Prasad P. D. and Ganapathy V. (2004) Expression of slc5a8 in kidney and its role in Na+-coupled transport of lactate. J. Biol. Chem. 279, 4452244532.
  • Gopal E., Fei Y. J., Miyauchi S., Zhuang L., Prasad P. D. and Ganapathy V. (2005) Sodium-coupled and electrogenic transport of B-complex vitamin nicotinic acid by slc5a8, a member of the Na/glucose co-transporter gene family. Biochem. J. 388, 309316.
  • Greene A. E., Todorova M. T. and Seyfried T. N. (2003) Perspectives on the metabolic management of epilepsy through dietary reduction of glucose and elevation of ketone bodies. J. Neurochem. 86, 529537.
  • Gupta N., Martin P. M., Prasad P. D. and Ganapathy V. (2006) SLC5A8 (SMCT1)-mediated transport of butyrate forms the basis for the tumor suppressive function of the transporter. Life Sci. 78, 24192425.
  • Guzman M. and Blazquez C. (2004) Ketone body synthesis in the brain: possible neuroprotective effects. Prostaglandins Leukot. Essent. Fatty Acids 70, 287292.
  • Halestrap A. P. and Meredith D. (2004) The SLC16 gene family — from monocarboxylate transporters (MCTs) to aromatic amino acid transporters and beyond. Pflugers Arch. 447, 619628.
  • Hawkins R. A., Williamson D. and Krebs H. A. (1971) Ketone-body utilization by adult and suckling rat brain in vivo. Biochem. J. 122, 1318.
  • Hawkins R. A., Mans A. M. and Davis D. W. (1986) Regional ketone body utilization by rat brain in starvation and diabetes. Am. J. Physiol. 250, E169E178.
  • Inoue K., Zhuang L., Maddox D. M., Smith S. B. and Ganapathy V. (2002) Structure, function, and expression pattern of a novel sodium-coupled citrate transporter (NaCT) cloned from mammalian brain. J. Biol. Chem. 277, 3946939476.
  • Izumi Y., Ishii K., Katsuki H., Benz A. M. and Zorumski C. F. (1998) beta-Hydroxybutyrate fuels synaptic function during development. Histological and physiological evidence in rat hippocampal slices. J. Clin. Invest. 101, 11211132.
  • Laffel L. (1999) Ketone bodies: a review of physiology, pathophysiology and application of monitoring to diabetes. Diabetes Metab. Res. Rev. 15, 412426.
  • Leino R. L., Gerhart D. Z., Duelli R., Enerson B. E. and Drewes L. R. (2001) Diet-induced ketosis increases monocarboxylate transporter (MCT1) levels in rat brain. Neurochem. Int. 38, 519527.
  • Lin R. Y., Vera J. C., Chaganti R. S. K. and Golde D. W. (1998) Human monocarboxylate transporter 2 (MCT2) is a high affinity pyruvate transporter. J. Biol. Chem. 273, 28 95928 965.
  • Massieu L., Haces M. L., Montiel T. and Hernandez-Fonseca K. (2003) Acetoacetate protects hippocampal neurons against glutamate-mediated neuronal damage during glycolysis inhibition. Neuroscience 120, 365378.
  • Masuda R., Monahan J. W. and Kashiwaya Y. (2005) d-beta-hydroxybutyrate is neuroprotective against hypoxia in serum-free hippocampal primary cultures. J. Neurosci. Res. 80, 501509.
  • Mitchell G. A., Kassovska.-Bratinova S., Boukaftane Y., Robert M. F., Wang S. P., Ashmarina L., Lambert M., Lapierre P. and Potier E. (1995) Medical aspects of ketone body metabolism. Clin. Invest. Med. 18, 193216.
  • Miyauchi S., Gopal E., Fei Y. J. and Ganapathy V. (2004) Functional identification of SLC5A8, a tumor suppressor down-regulated in colon cancer, as a Na+-coupled transporter for short-chain fatty acids. J. Biol. Chem. 279, 13 29313 296.
  • Morris A. A. (2005) Cerebral ketone body metabolism. J. Inherit. Metab. Dis. 28, 109121.
  • Pellerin L. and Magistretti J. (2004) Neuroenergetics: calling upon astrocytes to satisfy hungry neurons. Neuroscientist 10, 5362.
  • Pellerin L., Halestrap A. P. and Pierre K. (2005) Cellular and subcellular distribution of monocarboxylate transporters in cultured brain cells and in the adult brain. J. Neurosci. Res. 79, 5564.
  • Pierre K. and Pellerin L. (2005) Monocarboxylate transporters in the central nervous system: distribution, regulation and function. J. Neurochem. 94, 114.
  • Prasad B. M. and Amara S. G. (2001) The dopamine transporter in mesencephalic cultures is refractory to physiological changes in membrane voltage. J. Neurosci. 21, 75617567.
  • Reger M. A., Henderson S. T., Hale C., Cholerton B., Baker L. D., Watson G. S., Hyde K., Chapman D. and Craft S. (2004) Effects of beta-hydroxybutyrate on cognition in memory-impaired adults. Neurobiol. Aging 25, 311314.
  • Seth P., Ganapathy M. E., Conway S. J., Bridges C. D., Smith S. B., Casellas P. and Ganapathy V. (2001) Expression pattern of the type 1 sigma receptor in the brain and identity of critical anionic amino acid residues in the ligand-binding domain of the receptor. Biochim. Biophys. Acta 1540, 5967.
  • Srinivas S. R., Gopal E., Zhuang L., Itagaki S., Martin P. M., Fei Y. J., Ganapathy V. and Prasad P. D. (2005) Cloning and functional identification of slc5a12 as a sodium-coupled low-affinity transporter for monocarboxylates (SMCT2). Biochem. J. 392, 655664.
  • Stafstrom C. E. and Bough K. J. (2003) The ketogenic diet for the treatment of epilepsy: a challenge for nutritional neuroscientists. Nutr. Neurosci. 6, 6779.
  • Swink T. D., Vining E. P. and Freeman J. M. (1997) The ketogenic diet: 1997. Adv. Pediatr. 44, 297329.
  • Tieu K., Perier C., Caspersen C. et al. (2003) d-beta-hydroxybutyrate rescues mitochondrial respiration and mitigates features of Parkinson disease. J. Clin. Invest. 112, 892901.
  • Yudkoff M. (1997) Brain metabolism of branched-chain amino acids. Glia 21, 9298.