NaCT (SLC13A5) is a Na+-coupled transporter for Krebs cycle intermediates and is expressed predominantly in the liver. Human NaCT is relatively specific for citrate compared with other Krebs cycle intermediates. The transport activity of human NaCT is stimulated by Li+, whereas that of rat NaCT is inhibited by Li+. We studied the influence of Li+ on NaCTs cloned from eight different species. Li+ stimulated the activity of only NaCTs from primates (human, chimpanzee, and monkey); by contrast, NaCTs from nonprimate species (mouse, rat, dog, and zebrafish) were inhibited by Li+. Caenorhabditis elegans NaCT was not affected by Li+. With human NaCT, the Li+-induced increase in transport activity was associated with the conversion of the transporter from a low-affinity/high-capacity type to a high-affinity/low-capacity type. H+ was able to substitute for Li+ in eliciting the stimulatory effect. The amino acid Phe500 in human NaCT was critical for Li+/H+-induced stimulation. Mutation of this amino acid to tryptophan (F500W) markedly increased the basal transport activity of human NaCT in the absence of Li+, but the ability of Li+ to stimulate the transporter was almost completely lost with this mutant. Substitution of Phe500 with tryptophan in human NaCT converted the transporter from a low-affinity/high-capacity type to a high-affinity/low-capacity type, an effect similar to that of Li+ on the wild-type NaCT. These studies show that Li+-induced activation of NaCT is specific for the transporter in primates and that the region surrounding Phe500 in primate NaCTs is important for the Li+ effect.
We have recently cloned and functionally characterized a Na+-coupled transporter for citrate and other intermediates of Krebs cycle from rat, human, mouse, and Caenorhabditis elegans, and named the transporter NaCT (Na+-coupled citrate transporter) (Inoue et al., 2002a,b, 2004; Fei et al., 2004). NaCT is identified as SLC13A5 according to the Human Genome Organization nomenclature (Markovich and Murer, 2004; Pajor, 2006). In humans, SLC13A5 mRNA is expressed abundantly in the liver and to a much lesser extent in the brain and testes (Inoue et al., 2002b). In the rat, however, the transporter mRNA is expressed in the liver and testes with comparable abundance and at relatively much lower levels in the brain (Inoue et al., 2002a). There are also important differences in substrate selectivity between human and rodent NaCTs. The rat and mouse NaCTs recognize citrate with high affinity (Kt: rat, ∼20 μM; mouse, ∼40 μM), but these transporters also interact with succinate with appreciable affinity (Kt: rat, ∼170 μM; mouse, ∼40 μM) (Inoue et al., 2002a, 2004). In contrast, human NaCT exhibits relatively lower affinity for citrate (Kt, ∼600 μM) but is relatively more selective for citrate (Inoue et al., 2002b). Based on substrate specificity, NaCT is the mammalian functional ortholog of Indy (I am not dead yet) in Drosophila (Rogina et al., 2000) even though the mammalian NaCT is Na+ dependent, whereas the Drosophila Indy is Na+ independent (Inoue et al., 2002c; Knauf et al., 2002, 2006).
Among the various Krebs cycle intermediates, citrate is present at highest concentration in circulation in humans (∼160 μM). In the liver, where NaCT is expressed abundantly and is present specifically in the sinusoidal membrane (Gopal et al., 2007), the transporter mediates Na+-coupled concentrative uptake of citrate from blood into hepatocytes for subsequent use in metabolism (Inoue et al., 2003). Citrate entering into the cytoplasm of the hepatocytes via NaCT is avidly used for the synthesis of fatty acids (Inoue et al., 2003; Ganapathy and Fei, 2005). As such, deletion of the transporter in mice protects against diet-induced insulin resistance and metabolic syndrome (Birkenfeld et al., 2011). In the testes, NaCT is expressed in sperm cells (P. M. Martin and V. Ganapathy, unpublished data). Because seminal fluid contains high concentrations of citrate (∼130 mM) (Kline et al., 2006), we speculate that NaCT in spermatozoa mediates the concentrative uptake of citrate from seminal fluid into sperm cells to serve as a metabolic fuel. In the brain, NaCT is expressed specifically in neuronal cells (Wada et al., 2006; Yodoya et al., 2006), where the most likely function of the transporter is to transport citrate from extracellular medium for subsequent use in metabolism for ATP production.
The differences between human NaCT and rodent NaCTs in substrate selectivity, substrate affinity, and tissue expression pattern are relatively less pronounced compared with the difference in the influence of lithium on these transporters. The Na+-coupled transport of citrate via rodent NaCTs is inhibited by Li+ with an IC50 of ∼2 mM whereas the Na+-coupled transport of citrate via human NaCT is stimulated by Li+ with an ED50 of ∼2 mM (Inoue et al., 2003).
The stimulation of human NaCT by Li+ is associated with a marked increase in substrate affinity (Inoue et al., 2003), which interested us. Treatment with lithium in humans for affective disorders is associated with significant weight gain and increased circulating levels of triglycerides (Bergmann et al., 2007; Bardini et al., 2009). Because NaCT mediates concentrative uptake of citrate from the circulation into hepatocytes for subsequent use in cholesterol and fatty acid synthesis, we speculated that stimulation of NaCT-mediated delivery of citrate into hepatocytes for subsequent use for fat synthesis contributes to the weight gain associated with lithium therapy (Inoue et al., 2003; Ganapathy and Fei, 2005). These findings suggest that the stimulation of human NaCT by Li+ may have important clinical implications. We further explored the species-specific influence of Li+ on the function of NaCT by studying the effects of Li+ on NaCTs cloned from two nonhuman primates (monkey and chimpanzee) and from two additional nonprimate species (dog and zebrafish).
Materials and Methods
[14C]Citrate (specific radioactivity, 55 mCi/mmol) was purchased from Moravek Biochemicals (Brea, CA). Unlabeled lactate, succinate, malate, α-ketoglutarate, and citrate were purchased from Sigma-Aldrich (St. Louis, MO). Total RNA samples from dog liver, monkey (Rhesus) liver, and chimpanzee liver were used in reverse-transcription polymerase chain reaction (RT-PCR) to prepare full-length NaCT cDNA from the respective species. The cDNA for NaCTs from rat, mouse, human, and C. elegans were described in our earlier publications (Inoue et al., 2002a,b, 2004; Fei et al., 2004).
Female frogs (Xenopus laevis), used as the source of oocytes for heterologous expression of cloned NaCTs, were purchased from Xenopus 1 (Dexter, MI). The protocol for the use of frogs in these studies was approved by the Institutional Animal Care and Use Committee.
Cloning of Monkey, Chimpanzee, Dog, and Zebrafish NaCT cDNAs.
Monkey and chimpanzee NaCT cDNAs were cloned by RT-PCR using liver RNA samples from the respective species as the template. The following primers were used for both species: 5′-ATATATACCCTCCCGCGCGATG-3′ (forward) and 5′-GTCTAGACTAAGTCTCAATATGTG-3′ (reverse). The forward primer contained the start codon ATG at its 3′-end and an EcoRI site at its 5′-end. The reverse primer contained the coding sequence for the last five amino acids (Thr-His-Ile-Glu-Thr) of the carboxy terminus of human NaCT at its 3′-end and an XbaI site at its 5′-end. The restriction sites were added for subcloning the PCR products into pcDNA vector. The PCR products were first subcloned into pGEM-T Easy vector (Promega, Madison, WI), and the cDNA insert was released from the construct by EcoRI/XbaI digestion for subcloning into pcDNA vector after its linearization with EcoRI/XbaI. Both strands of the cDNAs were sequenced, and the amino acid sequences of monkey and chimpanzee NaCTs were elucidated. Dog (Canis lupus familiaris) NaCT cDNA sequence is available in the GenBank (accession no. XM_843480) . The RT-PCR primers for cloning dog NaCT cDNA were designed based on this nucleotide sequence: 5′-GATATATAGCCATGGCCTCGGCGCTCAGCTA-3′ (forward) and 5′-GCGTCTAGACTAGGTCTCGATATTCGTCACG-3′ (reverse). The PCR product was cloned into pcDNA as described above for primate NaCTs. Both strands of the cDNAs were sequenced. The deduced amino sequence of the cloned dog NaCT was identical to the sequence reported in the GenBank. The zebrafish NaCT cDNA was isolated from a zebrafish cDNA library constructed in pSPORT1 vector (Kozlowski et al., 2008). A ∼1.4-kbp fragment of the rat NaCT cDNA was used as the probe to screen the library. Both strands of the full-length clone were sequenced.
The codon-500 in human NaCT cDNA encodes phenylalanine. This codon was changed by site-directed mutagenesis to code for tyrosine (F500Y), tryptophan (F500W), alanine (F500A), valine (F500V), isoleucine (F500I), leucine (F500L), methionine (F500M), aspartate (F500D), and lysine (F500K). The QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, CA) was used to generate the mutants according to the manufacturer’s protocol, and details of the procedure have been described elsewhere (Wu et al., 1999). The entire coding region of the mutant cDNAs was sequenced to confirm the presence of the introduced mutations.
Functional Expression of NaCT cDNAs in a Mammalian Cell Line.
The human retinal pigment epithelial cell line (HRPE) was used for heterologous expression of the cloned NaCT cDNAs. The vaccinia virus expression technique was employed for this purpose, and the details of the procedure were published previously (Inoue et al., 2002a,b, 2004). The activity of NaCT was monitored by measuring [14C]citrate uptake. Uptake measurements were performed at 37°C in a NaCl-containing buffer, pH 7.5 (composition: 25 mM HEPES/Tris buffer, 140 mM NaCl, 5.4 mM KCl, 1.8 mM CaCl2, 0.8 mM MgSO4, and 5 mM glucose).
The uptake of citrate in control cells transfected with vector alone was very low, and the uptake increased in NaCT cDNA-transfected cells by a minimum of 25-fold. With each of the NaCTs studied here, the cDNA-specific citrate uptake was linear at least up to 45 minutes. In experiments in which the cation dependence of NaCT activity was studied, NaCl in the uptake buffer was replaced iso-osmotically with LiCl, KCl, or N-methyl-d-glucamine (NMDG) chloride.
To investigate the dependence of NaCT activity on Cl−, NaCl in the uptake buffer was replaced iso-osmotically with sodium gluconate; in addition, KCl was replaced with potassium gluconate, and CaCl2 with calcium gluconate. The kinetic parameters Michaelis constant (Kt) and maximal velocity (Vmax) were determined from saturation kinetics of citrate uptake.
To investigate the influence of Li+ on NaCT activity, citrate uptake was measured in the presence of NaCl with varying concentrations of LiCl. Equimolar concentrations of NMDG chloride were added for controls to adjust for osmolality.
Functional Expression of Rat and Human NaCTs in X. laevis Oocytes.
Capped cRNA was prepared from the cloned rat or human NaCT cDNAs using the mMESSAGE mMACHINE kit (Ambion, Austin, TX). Functional expression in X. laevis oocytes was performed as described elsewhere (Inoue et al., 2004) by microinjection of cRNA (50 ng/oocyte) into mature oocytes. Uninjected oocytes were used as negative controls.
Electrophysiologic studies were performed 4–6 days after the cRNA injection using the two-microelectrode voltage-clamp technique. Oocytes were perifused with a NaCl-containing buffer (100 mM NaCl, 2 mM KCl, 1 mM MgCl2, 1 mM CaCl2, 3 mM HEPES, 3 mM 4-morpholineethanesulfonic acid, and 3 mM Tris, pH 7.5), followed by the same buffer containing citrate. The membrane potential was clamped at −50 mV. The currents measured in the absence of citrate were subtracted from the currents measured in the presence of citrate to calculate citrate-specific currents. In uninjected oocytes, the citrate-specific currents were undetectable. The citrate-induced currents in NaCT-expressing oocytes were taken as a measure of NaCT activity. The stimulation of NaCT activity by Li+ was studied by analyzing the citrate-induced currents in the NaCl-containing buffer with varying concentrations of LiCl; equimolar concentrations of NMDG chloride were added to the buffer to adjust for osmolality. The kinetic parameters Kt and Vmax were determined from the saturation kinetics of citrate-induced currents.
Experiments with the mammalian cell expression system were repeated at least 3 times with independent transfections, and uptake measurements were made in duplicate in each experiment. Experiments with the oocyte expression system were also repeated at least 3 times with different sets of oocytes. The kinetic constants (Kt and Vmax) were calculated independently each time the experiment was repeated. The kinetic parameters were calculated using the commercially available computer program SigmaPlot, version 6.0 (SPSS, Chicago, IL). We used the unpaired two-sample t test to determine whether the difference in kinetic parameters between any given pair of experimental conditions was statistically significant; P < 0.05 was considered statistically significant. Data are given as mean ± S.E.
Cloning and Functional Characterization of Chimpanzee, Rhesus Monkey, Dog, and Zebrafish NaCTs.
We cloned chimpanzee NaCT cDNA from liver RNA, and the protein predicted by the sequence of the coding region of the cDNA consisted of 568 amino acids (GenBank accession no. HM_998308), which was equal to the number of amino acids in human NaCT. Sequence comparison between human and chimpanzee NaCTs showed 99% identity. There were only three amino acid differences between the transporters from the two species. The Rhesus monkey (Macaca mulatta) NaCT cDNA was cloned from liver, and the sequence predicted a protein consisting of 568 amino acids (GenBank accession no. HM_998307) , similar to human and chimpanzee NaCTs. Sequence comparison between human and Rhesus monkey NaCTs showed 97% identity. There were only 16 amino acid differences between the transporters from the two species. The dog NaCT cDNA was cloned from liver, and the predicted amino acid sequence was 100% identical to the sequence already reported in the GenBank. Sequence comparison between human and dog NaCTs showed 84% identity and 91% similarity. We cloned zebrafish NaCT cDNA from a cDNA library prepared using poly(A) RNA isolated from whole zebrafish. The cDNA is 2551-bp long with an open reading frame coding for 582 amino acids ( GenBank accession no. HM_998309). Sequence comparison between human and zebrafish NaCTs showed 61% identity and 78% similarity.
We have previously characterized the transport function of rat, mouse, and human NaCTs (Inoue et al., 2002a,b, 2004). Here, we examined the functional features of the newly cloned chimpanzee, monkey, dog, and zebrafish NaCTs using a mammalian cell expression system. NaCTs from all these four species mediated citrate uptake (Supplemental Fig. 1). In each case, the cDNA-induced citrate uptake was obligatorily dependent on Na+ (Supplemental Fig. 2). Substrate selectivity was then examined by monitoring the abilities of unlabeled citrate, succinate, α-ketoglutarate, malate, and lactate to compete with [14C]citrate (10 μM) for the cDNA-induced uptake process (Supplemental Fig. 3). At a concentration of 2.5 mM, citrate, succinate, and malate caused significant inhibition of [14C]citrate uptake induced by chimpanzee and monkey NaCT cDNAs. Citrate was the most potent inhibitor. α-Ketoglutarate and lactate had little or no effect. Similar results were obtained with dog and zebrafish NaCT cDNAs except that the inhibition with citrate, succinate, and malate was much greater than that seen with chimpanzee and monkey NaCTs. In addition, α-ketoglutarate, which had little effect on citrate uptake mediated by chimpanzee and monkey NaCTs, was a potent inhibitor of citrate uptake mediated by dog and zebrafish NaCTs. We also found that unlabeled citrate inhibited [14C]citrate uptake mediated by dog and zebrafish NaCTs much more than [14C]citrate uptake mediated by chimpanzee and monkey NaCTs.
Differential Influence of Li+ on Primate versus Nonprimate NaCTs.
We studied the effects of Li+ on the transport function of NaCTs from eight different species (primates: human, chimpanzee, and monkey; nonprimates: mouse, rat, dog, zebrafish, and C. elegans). We have shown previously that Li+ stimulates human NaCT, inhibits mouse, rat, and zebrafish NaCTs, and has no effect on C. elegans NaCT (Inoue et al., 2003). These studies were done with a single concentration of Li+ (2 mM).
In the present study, we examined the dose-response relationship for the influence of Li+ on NaCTs from different species (Fig. 1). The activities of the primate NaCTs were stimulated by Li+ (Fig. 1A). In contrast, the activities of nonprimate NaCTs were inhibited by Li+ (Fig. 1B). C. elegans NaCT was not affected by Li+ over a concentration range of 1–20 mM. At a concentration of 2.5 mM, Li+ increased the activities of chimpanzee and human NaCTs approximately 3-fold. Monkey NaCT was stimulated 2-fold under similar conditions. With nonprimate NaCTs, 2.5 mM Li+ caused 30%–40% inhibition.
We have previously shown that Li+ stimulates human NaCT by converting the transporter from a low-affinity/high-capacity type to a high-affinity/low-capacity type (Inoue et al., 2003). A similar phenomenon occurs with the other two primate NaCTs (Fig. 2). For chimpanzee NaCT-mediated citrate uptake, the Kt and Vmax values were 820 ± 117 μM and 1162 ± 115 pmol/106 cells per minute in the absence of Li+; the corresponding values in the presence of 10 mM Li+ were 36 ± 8 μM and 291 ± 28 pmol/106 cells per min (Fig. 2A). For monkey NaCT-mediated citrate uptake, the Kt and Vmax values were 224 ± 14 μM and 640 ± 18 pmol/106 cells per minute in the absence of Li+; the corresponding values in the presence of 10 mM Li+ were 53 ± 10 μM and 199 ± 21 pmol/106 cells per minute (Fig. 2B).
We also studied the differential influence of Li+ on human NaCT versus rat NaCT using the electrophysiologic technique in the X. laevis oocyte expression system. We found that perifusion of human NaCT- or rat NaCT-expressing oocytes with citrate in the presence of Na+ induced marked inward currents and that such currents were not detectable when Na+ in the perifusion buffer was replaced with Li+. This shows that Li+ does not substitute for Na+ to support the activity of NaCT. However, in the presence of Na+, Li+ increased the magnitude of Na+/citrate-induced currents for human NaCT but decreased the magnitude of Na+/citrate-induced currents for rat NaCT (Fig. 3).
Analysis of the dose-response relationship revealed that the stimulation for human NaCT was approximately 2.5-fold at 2.5 mM Li+. The magnitude of stimulation increased further with increasing concentrations of Li+; the stimulation observed at 20 mM Li+ was approximately 6-fold. For rat NaCT, the inhibition was ∼40% at 2.5 mM Li+, and it increased further with increasing concentrations of Li+ (∼85% inhibition at 20 mM Li+).
We then analyzed the influence of Li+ on the substrate affinity of human NaCT (Fig. 4). In the absence of Li+, the citrate saturation kinetics for human NaCT conformed to the Michaelis-Menten equation describing a single saturable system; the Eadie-Hofstee plot was linear and the Kt for citrate was 6.6 ± 0.2 mM (Fig. 4A). In the presence of Li+ (10 mM), the citrate-induced currents increased in magnitude, and the currents were saturable with increasing concentrations of citrate.
Interestingly, the saturation kinetics did not conform to the Michaelis-Menten equation describing a single saturable system. The Eadie-Hofstee plot was curvilinear, indicating the presence of more than one saturable system (Fig. 4B). Indeed, the data fit to the Michaelis-Menten equation describing a combination of two saturable systems. The Kt values for citrate for the two systems were 8.1 ± 1.7 mM (low-affinity system) and 0.69 ± 0.04 mM (high-affinity system).
We then evaluated the influence of Li+ on the kinetic parameters of rat NaCT using the oocyte expression system by following its transport activity electrophysiologically. The citrate-induced currents were decreased in the presence of Li+, but the currents were saturable with increasing concentrations of citrate both in the absence and presence of Li+ (Fig. 5). Interestingly, the influence of Li+ on rat NaCT was not associated with any significant change in the affinity for citrate. The Kt value for citrate was 44 ± 7 μM in the absence of Li+; the value did not change in the presence of 10 mM Li+ (43 ± 4 μM).
Influence of H+on the Transport Activity of Human NaCT.
Because lithium (atomic number 3) is closest to hydrogen (atomic number 1) in the periodic table and also because Li+ substitutes for H+ in certain transporters (e.g., Na+/H+ exchanger), we asked whether H+ is capable of stimulating Na+-coupled citrate uptake mediated by human NaCT. To address this question, we first analyzed the influence of pH on the transport function of human NaCT using the mammalian cell expression system (Fig. 6A).
We found the extracellular pH to have a profound effect on NaCT activity. Increasing the concentration of H+ in the uptake buffer enhanced the activity of human NaCT markedly. Changing the pH from 8.5 to 6.5 increased the activity 7-fold. The optimum pH for human NaCT was 6.5. Further acidification of the uptake buffer decreased the activity.
Because extracellular pH is also expected to change the ionic state of citrate, it is difficult to interpret the observed H+-induced increase in transport activity. Therefore, we compared the magnitude of Li+ (10 mM)-induced stimulation of human NaCT at three different H+ concentrations: 3.16 nM (pH 8.5), 31.6 nM (pH 7.5), and 316 nM (pH 6.5) (Fig. 6B). The stimulation was 4.7-fold at pH 8.5, 3.6-fold at pH 7.5, and 1.6-fold at pH 6.5.
Role of Phe500 in Human NaCT in Li+-Induced Stimulation of Transport Activity.
Phe500 in human NaCT is critical for Li+-induced stimulation of transport activity (Inoue et al., 2003). Substitution of leucine for phenylalanine at this position (F500L) stimulates the transport activity and also almost completely abolishes Li+ stimulation (Inoue et al., 2003). We evaluated the influence of nine different amino acid substitutions for Phe500 in human NaCT on the transport activity and also on Li+-induced stimulation (Table 1).
Substitution with charged amino acids aspartate (F500D) or lysine (F500K) decreased the transport activity markedly, but the remaining residual activity was still stimulated by Li+. Substitution with tyrosine (F500Y) also decreased the transport activity, and the ability of Li+ to stimulate the remaining residual activity was still intact. Upon substitution with other neutral amino acids, the influence on basal activity varied between almost no effect (F500A) and a robust stimulation (8.3-fold stimulation for F500W).
Interestingly, the greater the basal activity, the lesser was the Li+-induced stimulation. The wild-type human NaCT was stimulated 6-fold by 10 mM Li+, but the activity of the F500W mutant was not at all affected. There was a reciprocal correlation between the basal activity and the magnitude of Li+ stimulation for all amino acid substitutions (r2 = −0.92).
Comparison of the Effects of Li+ between Wild-Type and F500W Mutant Human NaCTs.
The primary mechanism by which Li+ stimulates human NaCT is by increasing the substrate affinity. Because the F500W mutant human NaCT has greatly increased basal activity and is not stimulated by Li+, we asked whether the substitution of phenylalanine at position 500 with tryptophan alters the transporter’s affinity for citrate by converting the transporter from a low-affinity type to a high-affinity type. To address this question, we analyzed the kinetic parameters of citrate uptake by wild-type and F500W mutant human NaCTs using the mammalian cell expression system in the absence and presence of Li+ (10 mM).
The Kt and Vmax values for the wild-type human NaCT were 586 ± 47 μM and 738 ± 38 pmol/106 cells per minute in the absence of Li+ (Fig. 7A). The corresponding values in the presence of Li+ were 62 ± 8 μM and 210 ± 15 pmol/106 cells per minute (Fig. 7B), indicating that Li+-induced stimulation of wild-type NaCT transport activity is associated with the conversion of the transporter from a low-affinity/high-capacity type to a high-affinity/low-capacity type. The Kt and Vmax values for the F500W mutant human NaCT were 23 ± 3 μM and 170 ± 11 pmol/106 cells per minute in the absence of Li+ (Fig. 7C). The corresponding values in the presence of Li+ were 20 ± 5 μM and 56 ± 5 pmol/106 cells per minute (Fig. 7D).
These data show that substitution of Phe500 with tryptophan converts the transporter from a low-affinity/high-capacity type to a high-affinity/low-capacity type even in the absence of Li+. The presence of Li+ elicits only a minor influence on the kinetic parameters of the mutant human NaCT, indicating that the F500W mutant human NaCT behaves in the absence of Li+ kinetically similar to wild-type human NaCT in the presence of Li+.
Comparison of the Effects of H+ between Wild-Type and F500W Mutant Human NaCTs.
Because Li+ and H+ elicit similar influence on the transport function of human NaCT, we analyzed the impact of various amino acid substitutions at position 500 in human NaCT on the stimulatory effect of H+. We compared the transport activities of wild-type and various mutant human NaCTs at pH 8.5 and pH 6.5 (Fig. 8).
As seen with Li+-induced stimulation, the amino acid substitutions that increased the basal transport activity significantly decreased the magnitude of H+-induced stimulation. The activity of wild-type human NaCT increased 3- to 4-fold when the extracellular pH was changed from 8.0 to 6.5. In contrast, the F500W mutant human NaCT, which had almost 10 times higher basal activity compared with wild-type human NaCT, responded only minimally when the extracellular pH was changed from 8.0 to 6.5. The other amino acid substitutions (F500L, F500M, F500V, and F500I), which increased the basal transport activity at an intermediate level, also showed significantly attenuated response to H+-induced stimulation.
We then analyzed the kinetic parameters of citrate uptake by wild-type and F500W mutant human NaCTs under identical conditions using the mammalian cell expression system at pH 8.5 and pH 6.5. The Kt and Vmax values for the wild-type human NaCT were 2132 ± 169 μM and 872 ± 56 pmol/106 cells per minute at pH 8.5 (Fig. 9A). The corresponding values at pH 6.5 were 256 ± 18 μM and 573 ± 28 pmol/106 cells per minute (Fig. 9B), indicating that H+-induced stimulation of wild-type NaCT transport activity is associated with the conversion of the transporter from a low-affinity/high-capacity type to a high-affinity/low-capacity type.
The Kt and Vmax values for the F500W mutant human NaCT were 57 ± 4 μM and 147 ± 6 pmol/106 cells per minute at pH 8.5 (Fig. 9C). The corresponding values at pH 6.5 were 18 ± 5 μM and 147 ± 13 pmol/106 cells per minute (Fig. 9D). It is thus evident that H+ elicits only a minor influence on the kinetic parameters of the F500W mutant human NaCT because the substitution of phenylalanine at this position with tryptophan itself increases the basal activity of the transporter by increasing its substrate affinity.
Citrate is a critical metabolite at the center of many important metabolic pathways. Apart from its mitochondrial role in the Krebs cycle (tricarboxylic acid cycle) and the cytoplasmic role as the source of acetyl CoA for the synthesis of fatty acids and cholesterol, it is also an important regulator of glycolysis by its ability to inhibit phosphofructo-1-kinase. The commonly accepted view regarding citrate in the cytoplasm is that after synthesis inside the mitochondria, citrate exits the mitochondrial matrix via the mitochondrial citrate carrier (SLC25A1) if appropriate metabolic conditions exist in the cell. It is generally considered that the mitochondrial citrate is the only source of citrate in the cytoplasm. Even though the levels of citrate in blood are significant (∼160 μM), it is generally not taken into account as a possible source of cytoplasmic citrate in tissues that use citrate for the synthesis of fatty acids and cholesterol.
The discovery of NaCT as a plasma membrane transporter that is capable of uphill transfer of citrate from the circulation into the cytoplasm of the cells (Inoue et al., 2002a,b, 2004) indicates that the blood-borne citrate is also likely an important source of cytoplasmic citrate in mammalian cells. Liver is the major organ where the activities of the cytoplasmic metabolic pathways involving citrate are robust, and this is the tissue where NaCT is expressed most abundantly (Inoue et al., 2002a,b, 2004). The sinusoidal location of the transporter in the hepatocytes (Gopal et al., 2007) is appropriate for this postulated role.
The potential role of NaCT in channeling circulating citrate into important biochemical pathways in the liver may have biologic consequences. NaCT was discovered in mammals as the result of the discovery of a citrate transporter in Drosophila, which has a critical role as an important determinant of life span in this organism (Rogina et al., 2000). The Drosophila transporter, known as Indy, is expressed most abundantly in fat body, the organ equivalent to the liver in mammals (Rogina et al., 2000). Even though the role of Indy as a life span–determining gene has been questioned (Toivonen et al., 2007), many studies have shown convincingly that the transporter does indeed play a role as a modulator of lifespan not only in Drosophila (Marden et al., 2003; Neretti et al., 2009; Wang et al., 2009), but also in C. elegans (Fei et al., 2004).
Even though the relevance of NaCT to life span in mammals remains to be determined, it is likely that this transporter plays an essential role in important metabolic pathways. This is supported by metabolic studies in Slc13a5-null mice, which show that the absence the transporter protects against diet-induced adiposity, insulin resistance, and the metabolic syndrome (Birkenfeld et al., 2011).
In the present study, we cloned NaCT from four additional species (chimpanzee, monkey, dog, and zebrafish) and characterized their function. There is a single entry in the GenBank (accession no. XM_511987) predicting the amino acid sequence of chimpanzee (Pan troglodytes) NaCT. The predicted protein consists of 475 amino acids, 93 amino acids shorter than human NaCT cloned in our laboratory (GenBank accession no. XM_137672). This indicates that the chimpanzee NaCT in the GenBank database (XM_511987) represents a truncated protein. This was confirmed by the sequence of the chimpanzee NaCT cloned in the present study. There are four entries in the GenBank (accession nos. XM_2763975, XM_2763976, XM_2763977, and XM_2763978) predicting the amino acid sequence of marmoset (Callithrix jacchus) NaCT, each prediction giving a different number of amino acids (571, 554, 528, or 525) in the protein. There was no information on monkey NaCT. The present study is the first to report on the sequence of monkey NaCT. The sequence of dog NaCT already exists in GenBank, and the dog NaCT cDNA cloned in the present study is identical to the one present in the GenBank.
We analyzed the transport function of the newly cloned chimpanzee, monkey, dog, and zebrafish NaCTs and compared the functional features among the four species. Even though NaCTs from all four species were absolutely Na+ dependent, chimpanzee and monkey NaCTs were relatively more selective for citrate than dog and zebrafish NaCTs, and chimpanzee and monkey NaCTs showed lower affinity for citrate than dog and zebrafish NaCTs. Li+ stimulated the activity of only chimpanzee and monkey NaCTs as was the case with human NaCT; in contrast, Li+ inhibited the activity of dog and zebrafish NaCTs.
With human NaCT expressed in X. laevis oocytes, two kinetically distinguishable transport components were evident in the presence of Li+. The low-affinity system apparently corresponded to the system in the absence of Li+ whereas the high-affinity system apparently represented the Li+-stimulated system. The presence of Li+ did convert human NaCT from a relatively low-affinity system to a high-affinity system, but the conversion was not homogeneous. The high-affinity population had bound Li+ with consequent change in citrate affinity whereas the low-affinity population apparently did not have bound Li+ and hence showed no change in citrate affinity.
These findings are in contrast to the results from the mammalian cell expression system where the conversion of human NaCT in the presence of Li+ from a low-affinity system to a high-affinity system was complete. Another notable difference between the oocyte expression system and the mammalian cell expression system was in the Michaelis constant for citrate. In the absence of Li+, the Kt for citrate was 0.60 ± 0.07 mM in the mammalian cell expression system; the value was almost 10 times higher (6.6 ± 0.2 mM) in the oocyte expression system.
The transport function of rat NaCT was inhibited by Li+. The inhibition was associated solely with a decrease in the maximal velocity of the transport system. This contrasted the influence of Li+ on human NaCT where the influence of Li+ was on the affinity for citrate by converting the transporter from a low-affinity/high-capacity type to a high-affinity/low-capacity type. Another notable difference between human NaCT and rat NaCT when analyzed using the mammalian cell expression system and the oocyte expression system was in the effect of the expression system on substrate affinity. With human NaCT, the affinity for citrate was 10 times poorer in the oocyte expression system than in the mammalian cell expression system. With rat NaCT, however, the decrease in citrate affinity was only 2-fold.
In the present study, we also compared the effects of Li+ and H+ on the activity of human NaCT. In the presence of Na+, increasing the concentrations of H+ increased the activity of the transporter as was seen with Li+. There was a competition between Li+ and H+ to elicit the stimulatory effect. The ability of Li+ to stimulate the transport activity of human NaCT decreased with increasing [H+] because the transporter was activated by H+ itself. This difference in Li+-induced stimulation was independent of the pH-induced changes in the relative amounts of differently charged ionic species of citrate. This suggests that H+ is able to substitute for Li+ for the stimulation of human NaCT.
The salient feature of the present study is the discovery that the stimulatory effect of Li+ is unique only to primate NaCTs. Li+ inhibits nonprimate NaCTs. The stimulatory effect of Li+ on human NaCT is potent; the stimulation is more than 2-fold at a Li+ concentration of 1 mM. These findings have clinical and therapeutic implications. Lithium is used widely in the treatment of affective disorders such as bipolar disorder. The plasma concentrations of Li+ in patients receiving lithium treatment are in the range of 0.8–2.0 mM (Sproule, 2002). Therefore, the activity of NaCT in patients treated with lithium is likely to be increased manyfold, enhancing the use of circulating citrate for the synthesis of fatty acids and cholesterol in the liver. Considering the fact that lithium is prescribed also in the form of lithium citrate, patients who take lithium in this form will not only have Li+ in the circulation to activate NaCT in the liver but also have increased levels of citrate in the circulation as the substrate for NaCT.
The finding that Li+ activates the human transporter by increasing the affinity for citrate is also physiologically relevant. Citrate is present in the circulation only at subsaturating concentrations (Kt for citrate in the mammalian cell expression system: ∼600 μM; Kt for citrate in the oocyte expression system: ∼6 mM; plasma concentration of citrate: ∼160 μM). Therefore, the Li+-induced increase in the affinity will certainly enhance the entry of circulating citrate into liver cells under physiologic conditions.
The difference in substrate affinity for human NaCT between the mammalian cell expression system and the oocyte expression system is interesting. We have shown previously a similar difference between HepG2 cells and the mammalian cell expression system (Inoue et al., 2002b; Gopal et al., 2007). The human NaCT cDNA was isolated from a HepG2 cell cDNA library (Inoue et al., 2002b). When this cDNA was expressed in HRPE cells, the affinity for citrate was ∼650 μM (Inoue et al., 2002b). But when the activity of the constitutively expressed NaCT was monitored in HepG2 cells, the affinity for citrate was ∼5 mM (Gopal et al., 2007). When the same transporter was expressed in oocytes in the present study, the affinity for citrate was ∼6 mM. The reasons for this difference are not known, but we speculate that posttranslational modifications of the transporter may have a role.
The heterologous expression in HRPE cells is transient, and the activity of the transporter is measured within 12 hours of transfection with the transporter cDNA, a time frame that may not be conducive for effective posttranslational modifications. In contrast, NaCT is expressed constitutively in HepG2 cells; even though we have no information on the half-life of the transporter protein in these cells, it is possible that the transporter resides in these cells long enough for post-translational modifications. In the oocyte expression system, the transporter activity is monitored 4 to 6 days after injection of the transporter cRNA, which may allow sufficient time for post-translational modifications. This may be the reason why the affinity for citrate is similar for the human NaCT in the oocyte expression system and in HepG2 cells. The validity of this explanation remains to be determined in future studies.
The activation of human NaCT with therapeutically relevant concentrations of Li+ has significant clinical implications. One of the unwanted side effects of lithium therapy in humans is the weight gain associated with an increase in the circulating levels of triglycerides and cholesterol (Bergmann et al., 2007; Bardini et al., 2009). Because NaCT facilitates the use of circulating citrate in liver cells for fatty acid and cholesterol synthesis, it is plausible that the Li+-induced activation of this transporter underlies this side effect. Unfortunately, this cannot be tested in animal studies using mice or rats because rodent NaCTs are inhibited, not activated, by Li+. The present findings that the Li+-induced activation is specific for primate NaCTs preclude the use of mice or rats as suitable animal models to examine this mechanism. Nonhuman primates can be used for this purpose. Alternatively, transgenic mice expressing human NaCT in the liver instead of mouse NaCT (i.e., “humanized” mice) need to be generated to determine whether activation of NaCT is the reason for the elevated levels of circulating triglycerides and cholesterol associated with lithium therapy.
The finding that substitution of phenylalanine at position 500 in human NaCT with other amino acids has profound influence on the transporter activity also has biologic significance. The effects of these substitutions on the functional activity of NaCT are profound. Substitution of Phe500 with selective neutral amino acids increases the transporter activity, whereas substitution with charged amino acids decreases the activity. It is conceivable that polymorphisms at this particular position may have profound effects in humans with regard to NaCT activity and consequently may significantly influence the lipid profile.
Participated in research design: Gopal, Babu, Prasad, Ganapathy.
Conducted experiments: Gopal, Babu, Ramachandran, Prasad.
Performed data analysis: Gopal, Babu, Ramachandran, Bhutia, Prasad, Ganapathy.
Wrote or contributed to the writing of the manuscript: Ganapathy.
- human retinal pigment epithelial cell line
- Na+-coupled citrate transporter
- reverse-transcription polymerase chain reaction
- solute carrier
- Copyright © 2015 by The American Society for Pharmacology and Experimental Therapeutics