The members of the solute carrier 45 (SLC45) family have been implicated in the regulation of glucose homoeostasis in the brain (SLC45A1), with skin and hair pigmentation (SLC45A2), and with prostate cancer and myelination (SLC45A3). However, apart from SLC45A1, a proton-associated glucose transporter, the function of these proteins is still largely unknown, although sequence similarities to plant sucrose transporters mark them as a putative sucrose transporter family. Heterologous expression of the three members SLC45A2, SLC45A3 and SLC45A4 in Saccharomyces cerevisiae confirmed that they are indeed sucrose transporters. [14C]Sucrose-uptake measurements revealed intermediate transport affinities with Km values of approximately 5 mM. Transport activities were best under slightly acidic conditions and were inhibited by the protonophore carbonyl cyanide m-chlorophenylhydrazone, demonstrating an H+-coupled transport mechanism. Na+, on the other hand, had no effect on sucrose transport. Competitive inhibition assays indicated a possible transport also of glucose and fructose. Real-time PCR of mouse tissues confirmed mRNA expression of SLC45A2 in eyes and skin and of SLC45A3 primarily in the prostate, but also in other tissues, whereas SLC45A4 showed a predominantly ubiquitous expression. Altogether the results provide new insights into the physiological significance of SLC45 family members and challenge existing concepts of mammalian sugar transport, as they (i) transport a disaccharide, and (ii) perform secondary active transport in a proton-dependent manner.
The universal doctrine of sugar transport across mammalian cell membranes so far only allows the transport of monosaccharides either via the sodium-dependent glucose transporters (SGLTs) of the solute carrier 5 (SLC5) family or via the facilitative glucose transporters (GLUTs) of the SLC2 family [1,2]. However, the transport of disaccharides is something unheard of. With the completion of the Human Genome Project in 2003, the SLC45 family emerged as a family of putative sugar transporters. This family consists of four members, SLC45A1–SLC45A4, sharing an overall amino acid sequence identity of 20–30%. All four members belong to the major facilitator superfamily (MFS) and, surprisingly, show similarities to plant sucrose transporters. These include a predicted structure of 12 transmembrane helices, intracellular N- and C- termini and an elongated cytosolic loop between the sixth and the seventh transmembrane helix. More importantly, the region between the second and the third transmembrane helix contains a proposed sucrose transport signature .
Even though the transport of the disaccharide sucrose seemed unlikely, as many studies have ruled out the possibility of disaccharide transport across animal membranes, we recently found a sucrose transporter in Drosophila, which has been named Slc45-1 because it displays a significant similarity to the mammalian SLC45, especially to its second member SLC45A2 . Localization studies of Drosophila Slc45-1 suggested a role in substrate uptake across the apical membrane of hindgut epithelial cells and in melanin synthesis. Extensive genetic studies had identified its mammalian orthologue SLC45A2 (MATP) as the cause for oculocutaneous albinism type 4 (OCA4), a disorder of melanin biosynthesis and the second most prevalent type of OCA in Japan [5,6]. The disease is characterized by a cutaneous and ocular hypopigmentation including, among others, poor visual sharpness and eye movement disorders such as nystagmus and strabismus . The gene has also been shown to lead to melanoma susceptibility in a light-skinned population , but is also generally linked to normal skin, eye and hair colour variations in Caucasians [8,9]. The distribution of one allele encoding the L374F mutation was found to reflect hypopigmentation in Caucasian populations with frequencies highest in Germans, lowest in Indians and Bangladeshis and absent from Africans and Japanese [9,10]. This link to skin colour has also been studied in other vertebrates. An orange/red variant of the medaka fish is associated with the b allele, encoding SLC45A2 . In chicken and Japanese quail, SLC45A2 causes plumage colour variation . It is noteworthy that null mutations cause an almost complete absence of both the black eumelanin and the reddish and yellow pheomelanin, whereas some mutations are dominant and cause a specific inhibition of pheomelanin . In mouse, coat colour is defined by different underwhite (SLC45A2) alleles [6,13]. Melanosomes of underwhite mutant mice were found to be small and irregularly shaped [13,14]. Because of this, Newton et al.  suggested in 2001 a role for SLC45A2 in sugar transport, possibly regulating the volume of melanosomes by providing an osmotic potential.
The third member of the SLC45 family, SLC45A3, was first identified as a prostate-specific marker with a potential role in prostate cancer . Several studies demonstrated its involvement in prostate cancer as an androgen-regulated 5′ fusion partner of members of the erythroblast transformation-specific (ETS) family [16,17]. Recently, a role for SLC45A3 in myelin maintenance was discovered  and moreover, a genome-wide expression profiling revealed changes in SLC45A3 expression in human Huntington's disease brain , suggesting important roles beyond prostatic tissue. The assigned function of SLC45A3 in regulating glucose and lipid metabolism in oligodendrocytes  further supports a role of this protein in mammalian sugar transport. No roles yet are known for the fourth SLC45 member, SLC45A4. So far, a functional analysis has only been performed for the first member, SLC45A1, which was shown to transport glucose and galactose in a proton-associated manner, presumably regulating glucose homoeostasis in neuronal cells during hypercapnia .
To obtain first insights into the physiological roles of the SLC45 family, we investigated the properties of the mouse SLC45 members SLC45A2, SLC45A3 and SLC45A4, especially the transport of the disaccharide sucrose. By expressing the full-length proteins in Saccharomyces cerevisiae and measuring sucrose uptake, we provide the first evidence that SLC45A2, SLC45A3 and SLC45A4 are H+–sucrose symporters. Besides sucrose, substrate inhibition studies also suggest the transport of glucose and fructose. Since it is the first mammalian sugar transporter family (i) to transport a disaccharide, and (ii) to function in a proton-dependent manner, challenging actually two existing concepts of sugar transport, the SLC45 family represents a new, alternative sugar transporter family different from SGLTs or GLUTs.
MATERIAL AND METHODS
HOD55-4B [MATα ura3(-52 or -FS) leu2-3,112 his3-11,15; kanMX-KlPGK1p::SUC2int mal0] is derived from HD56-5A . The mal0 allele was introduced by crossing HD56-5A with AMW-13C (MATa ura3-FS trp1-FS his3-11,15 leu2-3,112 can1 cir0 mal0 GAL with FS designating frameshift mutations; a gift from Malcolm Whiteway, Montreal, Canada). After tetrad analysis, strain HOD35-1A [MATa ura3(-52 or -FS) leu2-3,112 his3-11,15 gal mal0 SUC2] was used in a cross to introduce the modified SUC2 allele. In order to obtain HOD55-4B, which only produces intracellular invertase, the KlPGK1 promoter was inserted by in vivo recombination upstream of SUC2 at its native locus, replacing a part of the original promoter and the secretion signal sequence. For this purpose, the kanMX cassette was first placed upstream of the KlPGK1 promoter. Thus, plasmid pJJH1084 was constructed by inserting the PCR fragment obtained from Kluyveromyces lactis CBS2359 genomic DNA as a template with the primer pair (5′-gcgcgtcgACATGTATAGTACGTTGCACATAG-3′ and 5′-gcgcggatccCATTTTTATTAATTCTTGATCGATTTTTTTG-3′) as a SalI/BamHI fragment into pUG6  digested with BglII/SalI. The resulting plasmid was then used as a template to amplify marker and promoter with the oligonucleotide pair 09.269/09.270 (5′-TATTCTTCTACCAAAGGCGTGCCTTTGT-TGAACTCGATCCATTATAGGCCACTAGTGGATCTG-3′ and 5′-TTGGGTGTGAAGTGGACCAAAGGTCTATCGCTAGTTT-CGTTTGTCATATTTTTATTAATTCTTGATCG-3′; sequences underlined are homologous with pJJH1084, the other sequences are homologous with the genomic SUC2 locus). Strain DHD5 (the isogenic diploid derived from HD56-5A) was then used as a recipient for the PCR product, selecting transformants for G418 resistance (YEPD + G418; procedures and media used to manipulate yeast have been described previously by Schmitz et al. ). Correct substitution of one allele at the original SUC2 upstream sequences by the kanMX-KlPGK1 promoter cassette was confirmed by PCR using flanking oligonucleotides. Haploid segregants which carried this construct after tetrad analysis were shown to lack growth on sucrose medium, in contrast with the recipient strain and SUC2 wild-type segregants. HOD55-4B was obtained from a cross of one of these segregants with HOD35-1A and then used for further analyses.
Expression of SLC45 members in
Full-length clones were obtained by PCR using mouse cDNA as a template and the following primers SLC45A2: 5′-ATG-AGTGGAAGCAATGGGCCG-3′ (forward) and 5′-CTAAT-CTACATATCTTACGAAGAG-3′ (reverse), SLC45A3: 5′-ATGATCCAGAGGCTGTGGGC-3′ (forward) and 5′-CTACA-CTGAGTATTTGGCCAAGTCG-3′ (reverse), and SLC45A4: 5′-ATGAAAATGGCTCCGCAGAATG-3′ (forward) and 5′-TCACACCATGGACTCCGTCTC-3′ (reverse). PCR products were ligated into the pDR196 expression vector  via the SpeI and/or XhoI restriction site and transformed into Escherichia coli DH5α cells for replication.
A total of 1 μg of plasmid DNA was transformed into the yeast strain HOD55-4B via electroporation. Transformants were grown in minimal medium containing 0.67% Difco yeast nitrogen base (YNB) without amino acids, supplemented with auxotrophic requirements and glucose or sucrose at 2%.
For growth tests, yeast cultures were grown to a D600 between 0.6 and 1, washed three times with sterile water and used to inoculate fresh minimal medium containing 2% sucrose to a D600 of 0.01. Cells were grown at constant shaking at 30°C and cell growth was assayed by measuring D600 values at several time points.
Yeast colonies were grown in minimal YNB medium supplemented with auxotrophic requirements and 2% sucrose (BioXtra ≥99.5% purity, Sigma–Aldrich) or 2% glucose to late exponential growth phase. Cells were collected by centrifugation, washed first in water, followed by 25 mM sodium phosphate (pH 5.7, except where otherwise indicated) and resuspended in 1 ml of the same buffer to an D600 of 20. Cells were kept on ice until shortly before the assay. They were prewarmed to 30°C for 1 min just before the uptake was started by adding [14C]sucrose (final concentration 10 mM and 0.0025 μCi/μl, or 0.005 μCi/μl for measuring sucrose concentration dependence; specific activity of [14C]sucrose stock 600 mCi/mmol, purchased from Hartmann Analytic). After indicated time periods, cells were collected by vacuum filtration on a 0.45 μm cellulose acetate filter (Whatman) and washed with ice-cold water. Accumulated [14C]sucrose was measured by scintillation counting (Beckman LS 6500).
For substrate competition assays, putative substrates were added to a final concentration of 40 mM at the same time as sucrose. The uptake of [14C]sucrose was analysed after 10 min of incubation in 25 mM sodium phosphate (pH 5.7) and compared with uptake rates without additional sugars. All substrates tested were obtained from Sigma–Aldrich with a purity ≥99%, except for glucose (≥99.5%) and mannitol (≥98%), which were from Serva.
Three female B57C6 mice were killed by cervical dislocation, and tissues were collected and frozen in liquid nitrogen. Male tissues were obtained from normal house mice. RNA was isolated using TRIzol Reagent (Ambion) according to the manufacturer's protocol. Any residual DNA traces were removed using the RNAse-Free DNase Set (Qiagen). Concentration and purity were verified by measuring UV absorbance using the Implen NanoPhotometer®. cDNA was reverse-transcribed from total mRNA using the ThermoScript RT-PCR System (Invitrogen) with Oligo dT Primers. Real-time PCR was done in triplicates using Maxima SYBR Green (Thermo Scientific) according to the manufacturer's instructions using 0.3 μM of the following specific primers: SLC45A2, 5′-GCCGACTGACACCCATACC-3′ (forward) and 5′-CTGTGCATGACAAGTCTCCC-3′ (reverse); SLC45A3, 5′-CACCCTACGCCGACTCTTTG-3′ (forward) and 5′-CTCCCACGAAGTCCGTGTAGA-3′ (reverse); SLC45A4, 5′-AGTCAGAGCACGAGCTGTC-3′ (forward) and 5′-AGATG-GAGGGTTCGATGTCAT-3′ (reverse); and eukaryotic elongation factor 2 (Eef2), 5′-GCTTCCCTGTTCACCTCTGA-3′ (forward) and 5′-CGGATGTTGGCTTTCTTGTC-3′ (reverse). Primer pairs were chosen from the database PrimerBank  and qPrimerDepot .
Reactions were performed in a thermal cycler (iCycler, Bio-Rad Laboratories) according to the following protocol: 95°C for 10 min, and 35 cycles at 95°C for 30 s, 57°C for 20 s, and 72°C for 30 s. The specificity of the primers was verified by agarose gel electrophoresis and melting curve analysis. Initially, PCR products were confirmed by sequencing. Mean normalized expression levels of target genes were calculated from cycle thresholds (CT) values as described by Simon  using Eef2 as a reference gene .
SLC45A2, SLC45A3 and SLC45A4 are functional sucrose transporters
To ascertain whether sucrose is a substrate for mammalian SLC45A2, SLC45A3 and SLC45A4, the proteins were expressed in a yeast strain incapable of sucrose uptake, HOD55-4B. This strain is deficient for yeast sucrose transporters and lacks the leader sequence of SUC2, an extracellular invertase mediating the hydrolysis of sucrose into glucose and fructose. Without the leader sequence, SUC2 remains intracellular (SUC2int), enabling the yeast to metabolize sucrose once it is transported into the cells. In principle, the yeast strain resembles SuSy7, a commonly used system for the characterization of plant sucrose transporters [29,30], with the advantage of SUC2int being chromosomally integrated with constitutive expression. A growth test of HOD55-4B yeast cells transformed with the expression vector pDR196 revealed absolutely no growth on sucrose as the sole carbon source, demonstrating that no invertase leaks to the outside (Figure 1A). In contrast, HOD55-4B yeast cells transformed with SLC45A2, SLC45A3 and SLC45A4 were able to grow on sucrose-containing minimal medium, clearly demonstrating sucrose uptake.
The Slc45 family members A2-A4 transport sucrose
To confirm these results and analyse transport properties of these proteins in more detail, [14C]sucrose uptake was measured. Initially, we grew yeast cultures for uptake assays under glucose conditions and observed transport rates that were 8–12-fold higher compared with yeast transformed with the empty vector (Supplementary Figure S1). Later on, taking advantage of using the strain HOD55-4B, we switched to growing yeast transformed with SLC45A2, SLC45A3 and SLC45A4 under sucrose conditions. This procedure should prevent the transporters from being recycled back from the plasma membrane under glucose repression and put additional selection pressure on cells expressing the transporters. Doing so quadrupled transport, suggesting that carbon catabolite repression indeed had affected transport activity. This transport of sucrose increased linearly over time up to 10 min for SLC45A2 and SLC45A3, and at least 2 min for SLC45A4 (Figure 1B). Because of that, further [14C]sucrose-uptake experiments were performed for 2 min (SLC45A4) or for 5 min (SLC45A2 and SLC45A3) of incubation.
Concentration-dependent uptake assays revealed a half-maximal [14C]sucrose uptake at ∼7.4±0.8 mM for SLC45A2, 4.4±0.9 mM for SLC45A3 and 6.3±0.9 mM for SLC45A4 (Figure 1C). These values resemble the affinities for sucrose of moderate/low-affinity sucrose transporters from plants, such as StSUT4 (potato)  or PmSUC3 (Plantago major) , with apparent Km values of ∼5 mM. Compared with the only other known animal sucrose transporter from Drosophila Slc45-1, which has an apparent Km of ∼16 mM, SLC45A2, SLC45A3 and SLC45A4 are roughly three times more sensitive to sucrose .
Transport of sucrose by SLC45A2, SLC45A3 and SLC45A4 is coupled to an H+ gradient
To assess whether transport is coupled to protons as is the case for plant sucrose transporters as well as for two other members of the SLC45 family, the mammalian SLC45A1  and the insect Slc45-1 , we checked transport in the presence of the specific protonophore carbonyl cyanide m-chlorophenylhydrazone (CCCP) which has been shown to cause depolarization of the yeast plasma membrane [33,34]. Figure 2(A) shows a significant reduction in sucrose uptake by SLC45A2, SLC45A3 and SLC45A4 when CCCP was added, indicating proton-coupled sucrose transport. Although the inhibition by CCCP reduced sucrose transport by SLC45A2 and SLC45A4 to similar values of slightly more than 30%, its effect on SLC45A3 was bigger, leading to a significantly stronger reduction down to 18%. We do not know the reason for this differential effect of 50 μM CCCP which, however, has also been observed for several plant H+/sucrose transporters expressed in yeast cells [29,32,35,36]. So, for instance, we neither know the copy numbers of the three transporters in the yeast plasma membrane nor their turnover, and we do not know what occurs, after loss of the proton motive force induced by CCCP, when only the sucrose gradient remains to drive this sugar across the membrane. To check the pH-dependence of transport, we used two buffers with optimal buffering capacities at different pH values and found a preference for a slightly acidic pH, whereas transport rates decreased at more acidic and at neutral to slightly alkaline pH values (Figure 2B). This profile is consistent with our suggestion that the SLC45 transporters make use of the proton motive force. Noticeably, transport rates were identical in the presence of Na+ and K+ (Figure 2C). This result excludes a Na+-dependent mechanism which is the norm for most animal sugar symporters, and strengthens our conclusion that sucrose transporters of the SLC45 family are H+ symporters just like the evolutionarily closely related plant sucrose transporters .
Sucrose transport is coupled to a proton gradient
To gain initial knowledge about the substrate specificities, [14C]sucrose was competed with a 4-fold excess of unlabelled sugars representing potential substrates for the transporters (Table 1). The addition of non-labelled sucrose led to a 75% decrease of detected radioactivity, due to a lower specific radioactivity. Out of 14 further sugars tested, only glucose, fructose, mannose and 2-deoxyglucose (2-DOG) caused a reduced transport of [14C]sucrose. Galactose, often transported by GLUTs and by SGLT1, did not significantly compete, unlike maltose, a disaccharide structurally similar to sucrose and transported by the yeast endogenous maltose transporter MALx1 . Raffinose, an alternative substrate for the sucrose-hydrolysing enzyme invertase, did also not cause a change in transport activity. Interestingly, there was no significant decrease of sucrose transport observed with methyl α-D-glucopyranoside (α-MDG), a competitor of glucose transport in SGLTs, nor with 3-O-methylglucoside (3-O-MG), a competitor of glucose transport in both SGLTs and GLUTs. 2-DOG, on the other hand, an inhibitor of glycolysis and a specific competitor for GLUTs, inhibited [14C]sucrose uptake by SLC45A2 as well as by SLC45A3 and SLC45A4. Reduced sucrose transport by the presence of glucose, fructose or mannose could be an indication that they themselves are transported. On the other hand, these three monosaccharides are the preferred carbon sources for yeast, and thus reduced sucrose transport may be an effect of carbon catabolite repression induced by them. We therefore also competed these sugars with [14C]sucrose in yeast cells grown in glucose to have catabolite-repressing conditions before the uptake assays already. Under these conditions, glucose is known to activate transport rates in StSut1, a specific plant H+–sucrose symporter , by way of energizing the plasma membrane . Our results also showed an activation of Sut1 (Table 2). In contrast, transport rates of SLC45A2, SLC45A3 and SLC45A4 were still inhibited by glucose and fructose and to a small extent by mannose. Compared with the experiments without catabolite-repressing conditions, inhibition rates were not as strong, which may be due to a simultaneous activation of these transporters. Because of the strong effects of glucose, fructose and mannose on various signalling pathways and protein expression in yeast, we cannot exclude the possibility that these sugars also affect sucrose transport via a transport-independent mechanism. Nevertheless, our results indicate, especially when considering the different behaviour compared with that of SUTs, that glucose and fructose (and perhaps mannose) may be transported substrates of SLC45A2, SLC45A3 and SLC45A4.
|Sugar added .||SLC45A2 .||SLC45A3 .||SLC45A4 .|
|Sugar added .||SLC45A2 .||SLC45A3 .||SLC45A4 .|
|Sugar added .||SLC45A2 .||SLC45A3 .||SLC45A4 .||Sut1 .|
|Sugar added .||SLC45A2 .||SLC45A3 .||SLC45A4 .||Sut1 .|
To test directly an uptake of monosaccharides, SLC45A2, SLC45A3 and SLC45A4 were transformed into the yeast strain EBY-VW4000, which is deficient for all endogenous hexose transporters  but can easily survive on maltose. However, complementation analysis did not show any growth on agar plates containing glucose, fructose or mannose (Supplementary Figure S2). There are two possible explanations for this result: either the Slc45 proteins do not transport these monosaccharides or their expression in EBY-VW4000 strain was not successful. To check this point, we created the strain EBY-SUC2int by cloning the intracellular invertase gene into the EBY-VW4000 strain, which allows yeast to grow on sucrose if this sugar is transported. Supplementary Figure S3 shows the growth of EBY-SUC2int cells transformed with SLC45A2, SLC45A3, SLC45A4 or the plant sucrose transporter SUT1. Again, we could not detect any growth on glucose. The failure of the cells, also of those transformed with SUT1, to grow on sucrose, ultimately indicated that the transporters were not functionally expressed in both EBY yeast strains. A similar problem had also been described for the mammalian sugar transporters GLUT1 and GLUT4 which only became functionally expressed at the plasma membrane of EBY-VW4000 cells after additional spontaneous mutations were introduced . Unfortunately, mutagenesis was not successful in our case (results not shown).
Tissue expression of SLC45A2, SLC45A3 and SLC45A4 mRNA
To know where sugar transport by the SLC45 family could play a physiological role, we isolated mRNA from different mouse tissues and tested expression by real-time PCR (Figure 3). Remarkably, expression of SLC45A3 and SLC45A4 was found in all tissues. Especially SLC45A4 showed an abundant ubiquitous expression. The expression of SLC45A3 was, as expected, highest in the prostate. However, far from being prostate-specific, we found significant expression in other tissues as well, which is in good agreement with the tissue expression profile from the TiGER database . In contrast with SLC45A4 and SLC45A3, expression of SLC45A2 was much more confined to eyes and skin, showing only weak or no expression in other tissues, which is also in accordance with its role in pigmentation [8,42].
Expression of SLC45A2, SLC45A3 and SLC45A4 in various mouse tissues
In the present paper, we describe an alternative sugar transporter family, the solute carrier family 45, with characteristics different from all other known mammalian sugar transporters. Besides their property of being H+ symporters, a further hallmark is the transport of the disaccharide sucrose. Whether monosaccharides such as glucose, fructose or mannose are transported too remains to be seen. If these monosaccharides are substrates, the affinity of the transporters appears to be in a similar range as for sucrose because inhibition of [14C]sucrose uptake by a 4-fold excess of the monosaccharides did not differ essentially from inhibition by an excess of unlabelled sucrose (see Table 1).
Even though all four members of the SLC45 family contain the signature of sucrose transporters, SLC45A1 seems to be the only member of this family with different substrate specificity, because glucose and galactose, but not fructose or sucrose, uptake was detected in transfected COS-7 cells . However, glucose uptake significantly increased under acidic conditions and transport was inhibited by CCCP, suggesting an H+-coupled transport. A potential uptake of sucrose and of fructose had been tested only at a pH of 7.5, with no significant increase in transport. In case the affinity for sucrose and fructose may be lower than for glucose and galactose, there might still be a possibility of sucrose transport by SLC45A1 under more acidic conditions.
SLC45A3 and SLC45A4 may be involved in sugar-reabsorption processes
Transport of the disaccharide sucrose is the most remarkable feature of the SLC45 family, challenging classical models of sugar transport across cell membranes in animals. So far, sugar transport in animals has been dominated by the SLC2 and SLC5 families. The SLC5 family contains the Na+ symporters, SGLTs, whereas the SLC2 family contains the facilitative sugar transporters, GLUTs. In the classical model of intestinal sugar transport, SGLT1 is responsible for glucose and galactose and GLUT5 for fructose absorption at the apical side of enterocytes, whereas GLUT2 transports glucose, galactose and fructose out of the cells into the blood . This model is slowly being extended into a more complex one. In 2000, Kellett and Helliwell  proposed that a second ‘diffusive’ component, which for a long time had been dismissed as simple, perhaps paracellular, diffusion was due to the recruitment of GLUT2 to the apical membrane. In 2004, another facilitative transporter, GLUT7, showing high affinity for glucose and fructose, was discovered in the apical membrane . Studies on GLUT2 deficient mice or humans suggested that still other mechanisms based on microsomal membrane traffic exist to transport sugars [46,47]. Mannose transport across enterocytes also seems to be mediated by at least two different transport systems . In the apical membrane of rat small intestine and kidney cortex, a Na+-coupled and electrogenic transport of mannose was found, which differed from that mediated by SGLT1 . An Na+-independent transport component was described in the basolateral membrane of Caco-2 cells . The process of intestinal sugar transport continues to be a controversial topic, leaving room for other, perhaps entirely new, sugar transporters. However, one thing that has not been questioned is the uptake of disaccharides, as sucrose had never been shown to be directly transported, but has to be hydrolysed by sucrases into glucose and fructose first . Nevertheless, as sucrose is the major product of photosynthesis, it occurs in every fruit and vegetable, and it seems likely that some of the disaccharide molecules escape being hydrolysed by sucrases in the apical membranes of the small intestine enterocytes. The reason that no disaccharide transport had been detected yet may be due to the second remarkable nature of the SLC45 family: the Na+-independent, but H+-coupled, transport of sugars. Experiments conducted at a pH of ∼7.5 would not detect significant transport rates.
Although nutrient uptake in mammalian cells is mostly energized by a Na+ gradient, it is energized by a transmembrane electrochemical proton gradient in bacteria, yeast and plants. Likewise, the H+-coupled peptide transporters were not discovered until the mid-1990s, greatly altering the conception of absorption thought to happen only via Na+-coupled transport of amino acids . Later it was discovered that in mammals di- and tri-peptides are transported this way by the SLC15 family in the small intestine and in the kidney proximal tubule, but also some amino acids, vitamins, organic acids and some trace elements are transported by various H+-dependent transport systems . Furthermore, an H+-driven sugar transport through the Na+–glucose co-transporter SGLT1 was revealed in 1994 . This confirmed earlier observations of H+-coupled glucose transport in rabbit brush border membrane vesicles and toad small intestine . In contrast with SGLT1, however, the SLC45 family does not show transport of Na+. Overall, H+-driven peptide or sugar transporters seem to be physiologically relevant in the small intestine and the kidney. Beside the acidification of the duodenum and the proximal jejunum by the acidic chyme from the stomach, a slightly acidic up to neutral microenvironment exists on the epithelial surface (about pH 6.3 in the proximal jejunum and pH 7.0 in the distal ileum of rat small intestine) . Together with an inside negative membrane potential, even such a low H+ gradient may quite effectively drive transport.
Our data from real-time PCR indicate the presence of SLC45A3 and SLC45A4 in the intestine primarily in the duodenum, where a low-affinity, high-capacity, H+-coupled sugar transporter would make the most sense because sucrose has probably not been fully hydrolysed yet in this part. If and how much the members of the SLC45 family contribute to intestinal sugar absorption needs further investigation. It is also designative that expression of human SLC45A4 in the TiGER database is highest in colon, kidney, small intestine and brain . A further indication could be the fact that the Drosophila Slc45-1 orthologue of the mammalian SLC45 family is located in the apical membrane of hindgut epithelial cells where the pH of the intestinal lumen is slightly acidic, which also supports the hypothesis that Slc45-1 is involved in nutrient uptake . Although a function of sucrose in the blood is not known, endogenous sucrose was detected in human serum of fasting specimen at approximately 75 μM in a study by Vinet et al. . The authors could not explain where this should come from, but excluded it as an artefact.
We also found expression of SLC45A3 and SLC45A4 in the colon, which is slightly acidic with a pH of ∼6.7 . As one function of the colon is to extract water from waste, the role of an H+/sugar transporter in this region of the gut may be to contribute to the osmotic gradient across the epithelium. The establishment of an osmotic gradient is also important in the kidney, another tissue where expression of SLC45A3 and SLC45A4 was detected. The kidney also plays a very important role in glucose homoeostasis. Since about 90% of the glucose filtrated from the blood is reabsorbed via the high-capacity low-affinity sodium–glucose co-transporter SGLT2 in the early proximal tubule, many SGLT2 inhibitors have been developed in an attempt to treat Type 2 diabetes [58–60]. Paradoxically, these inhibitors only inhibit 30–50% of glucose reabsorption. Although many explanations have been offered, none seemed quite sufficient. It would be interesting to know whether the members of the SLC45 family play a role in this respect.
SLC45A2: an important player in melanin synthesis?
As mentioned above, SLC45A2 was mostly confined to eyes and skin. Extensive genetic studies on SLC45A2 provide a very good idea of its role in the production of melanin. From mutations in the mouse gene underwhite we know that hypopigmentation correlates with melanosome size, shape, melanin content and maturity . In 1998, Lehman et al.  noted irregularly shaped and unevenly pigmented melanosomes of underwhite mutant mice. These findings indicate a dysfunction in osmotic regulation in melanosomes, which may be especially necessary as melanin is a very large polymer, synthesized from tyrosine molecules. The transport of sugars could thus provide compatible osmolytes during melanin synthesis.
The pH of melanosomes is also critical for melanin synthesis, since tyrosinase, the key enzyme for melanization, is inactive at acidic pH values . On the other hand SLC45A2 co-transports protons, and thus it may counteract melanization by contributing to the acidification of melanosomes. This acidification, however, could be antagonized by a secondary active H+ export, especially since Smith et al.  described the expression of five different NHEs in melanosomal membranes which may expel protons into the cytosol. We propose that the rate at which substrates are transported by SLC45A2 may affect the amount of melanin that can be produced due to changes in osmolarity and pH, thus providing differential molecular mechanisms to mediate human skin colour variations. Remarkably, the melanosomes of Caucasian melanocytes are acidic with a mostly inactive tyrosinase, whereas those of black skin are more neutral containing a fully active tyrosinase . Given that SLC45A2 would also be more readily transporting protons and sugar into melanosomes of neutral pH, its activity would correlate with the activity of tyrosinase. Interestingly, Fuller et al.  observed a fairly stable modification of tyrosinase increasing its activity upon treating Caucasian melanocytes with ammonium chloride, which led to a more alkaline milieu in the melanosomes.
A role for the SLC45 family in cancer?
An increased sugar uptake is a hallmark of malignant cancer cells, and a deregulated GLUT and SGLT expression has been observed in many cancer types. Recent studies showed that unrelated mutations that drive cancer share the common effect of increased glucose uptake and metabolism, stressing the key role of nutrient uptake in tumourigenesis [63–65]. In this regard, the role of the SLC45 family in cancer may be a special one: as tumour cells often become highly glycolytic, they extrude large amounts of protons and lactate produced by glycolysis and the extracellular pH becomes acidic . As the SLC45 family transports sugars in an H+-dependent manner, cancer cells benefit from increased sugar uptake enormously by being able to channel more sugars into glycolysis. Hence, the SLC45 family may represent an important target against cancer by limiting the energy access of tumour cells. Indeed, expression profiles link all four members of the SLC45 family to various cancer types (https://www.oncomine.org), including an up-regulation of SLC45A2 in melanoma and of SLC45A3 in prostate cancer [67,68]. Interestingly, up-regulation of SLC45A2 in eight of 11 melanoma cell lines was independent of melanization , suggesting an additional role to its involvement in producing protective pigmentation.
We thank Sabine Heuer for excellent technical assistance.
carbonyl cyanide m-chlorophenylhydrazone
eukaryotic elongation factor 2
major facilitator superfamily
sodium-dependent glucose transporter
yeast nitrogen base
Rabea Bartölke, Helmut Wieczorek and Olga Vitavska designed and analysed the experiments which were performed by Rabea Bartölke. The paper was written by Rabea Bartölke, Helmut Wieczorek and Olga Vitavska. Jürgen J. Heinisch created the yeast strain HOD55-4B and advised on yeast metabolism. Helmut Wieczorek and Olga Vitavska conceived and supervised the project.
This work was supported by the Deutsche Forschungsgemeinschaft [grant numbers VI 673/2-1, WI 698/7-1] and by the Lichtenberg Ph.D. programme.