Keywords

Introduction

Glycosylation is one of the most important posttranslational modifications, as glycan structures on glycoproteins and glycolipids play key roles in central biological processes such as cell fate determination/proliferation, morphogenesis/organ development, viral and bacterial infections, immune response, and cancer invasion. Glycosylation is performed in the lumens of the endoplasmic reticulum (ER) and the Golgi apparatus by a sequence of glycosyltransferase reactions (Fig. 1a). Thus, glycosyltransferase enzymes determine the glycosylation status. Glycosyltransferases transfer sugar molecules to acceptor substrates from sugar donors called nucleotide sugars , which are monosaccharides that are activated by the addition of a nucleoside mono- or diphosphate (UDP, GDP, or CMP). Most nucleotide sugars are synthesized in the cytosol; however, CMP-sialic acid is the one exception that is synthesized in the nucleus instead. After being synthesized, nucleotide sugars are transported from the cytosol into the lumen compartment by specific nucleotide-sugar transporters (NSTs). NSTs are multiple membrane-spanning proteins that transport nucleotide sugars in a coupled byproduct of a glycosyltransferase reaction and subsequent luminal nucleoside diphosphatase (NDPase) reaction inside the lumen (Fig. 1a). Therefore, NSTs also determine the glycosylation status as the supplier of donor sugars used in the glycosyltransferase reactions.

Fig. 1
figure 1

Nucleotide-sugar transporters (a) or PAPS transporters (b) transport nucleotide-sugars and PAPS from the cytosol into the ER/Golgi lumens, the location where glycosylation or sulfation takes place. Nucleotide-sugars and 3′-phosphoadenosine 5′-phosphosulfate (PAPS) are synthesized in the cytosol or the nucleus. Nucleotide-sugar transporters and PAPS transporters transport these molecules from the cytosol into the ER/Golgi lumens. Glycosyltransferases and sulfotransferases are then responsible for the transfer of sugars and sulfate molecules to acceptor substrates from these donor nucleotide sugars and PAPS, respectively. Released nucleoside diphosphates (NDPs) are hydrolyzed by nucleoside diphosphatase (NDPase) to form nucleoside monophosphates (NMPs) and inorganic phosphates (Pis). NMPs and released adenosine 3′, 5′-diphosphate (PAP) are exported in antiport with incoming nucleotide-sugars and PAPS. Pis exit via a phosphate transporter.

Full size image

Sulfation is also an important posttranslational modification of proteins. Sulfated proteins and sulfated glycans play important roles in various biological processes. For example, heparan sulfate is indispensable to signaling pathways such as the wingless, hedgehog, bone morphogenetic protein, and fibroblast growth factor (FGF) pathways, during the developmental process. Sulfations are carried out for proteins and glycans in the lumen of the Golgi apparatus by various sulfotransferases (Fig. 1b). Thus, sulfotransferases determine the sulfation status, in the same vein as glycosyltransferases act as the main driver of glycosylation. An activated form of sulfate, 3′-phosphoadenosine 5′-phosphosulfate (PAPS) , is a common sulfate donor. Sulfotransferases transfer sulfate from PAPS to a defined position on a target sugar or tyrosine (Tyr) residue. PAPS is synthesized in the cytosol and then transported by PAPS transporters (PAPSTs) into the lumen of the Golgi apparatus, where sulfation is carried out. PAPSTs are multiple membrane-spanning proteins that transport PAPS by a coupled antiport of adenosine 3′, 5′-diphosphate (PAP) (Fig. 1b). Therefore, PAPSTs also determine the sulfation status, as they supply the sulfate groups necessary as a substrate for sulfotransferase enzymes.

Both NSTs and PAPSTs are members of the solute carrier 35 (SLC35) transporter family. Nearly 200 different glycosyltransferases and 40 different sulfotransferases have been identified in humans, whereas only eight NSTs and two PAPSTs have been identified in humans (Fig. 2 and Table 1). This suggests that each of these transporters likely regulates a broader range of glycosylation or sulfation reactions, as compared to transferases, which appear to require more specificity.

Fig. 2
figure 2

Phylogenetic tree of the members of the nucleotide-sugar transporter family, solute carrier 35 (SLC35) . The phylogenetic tree was created based on amino acid sequences by using the ClustalX program. Branch lengths indicate evolutionary distances between members. The scale at the top represents the evolutionary distance. The nucleotide sugars or PAPS that each transporter transfers from the cytosol to the ER or Golgi lumen are shown in Fig. 1. Members found in Homo sapiens are shown in red. At Arabidopsis thaliana, Af Aspergillus fumigatus, Bt Bos taurus, Ce Caenorhabditis elegans, Ca Candida albicans, Cgl Candida glabrata, Cl Canis lupus familiaris, Cg Cricetulus griseus, Cn Cryptococcus neoformans, Dr Danio rerio, Dm Drosophila melanogaster, Eh Entamoeba histolytica, Hs Homo sapiens, Kl Kluyveromyces lactis, Ld Leishmania donovani, Lm Leishmania major, Mm Mus musculus, Os Oryza sativa Japonica, Pp Pichia pastoris, Rn Rattus norvegicus, Sc Saccharomyces cerevisiae, Sp Schizosaccharomyces pombe, Ss Sus scrofa, Tg Toxoplasma gondii, Tb Trypanosoma brucei.

Full size image
Table 1 The members of nucleotide sugar transporter family, SLC35
Full size table

Principles

Members of NST Family SLC35

NSTs and PAPSTs belong to the NST family. The SLC35 family was initially classified based on sequence data from human NSTs. However, recently, a large number of NSTs from many other species, including plants and worms, were cloned and their activities identified. Available information about all NSTs and PAPSTs whose NST or PAPST activity has been identified is presented in Table 1. Using the amino acid sequences of these transporter molecules, their phylogenetic tree was also constructed (Fig. 2). Given that the group of GDP-Man transporters was found near the subgroups, SLC35C and SLC35D, we proposed to add this group as a new subgroup to the SLC35 family, which has not been identified in humans. In addition, it is worth mentioning that the specific NST activity of each nucleotide sugar could not be easily estimated by the phylogenetic tree using similarity between amino acid sequences. For the determination and estimation of the physiological function of each member, the identification of its transporter activity is essential.

Procedure

Identification of NST Activity

In general, two types of assays for identifying NST activity are used: a heterologous expression system (Kamiyama et al. 2003; Suda et al. 2004) and a proteoliposome system (Caffaro and Hirschberg 2006). The former method is discussed in this section. A heterologous expression system makes use of a yeast expression system for the expression of NSTs derived from various other species. This is done in order to obtain a large enough microsome fraction expressing the NST of interest to allow for the determination of its Km value. This is possible because the yeast microsome normally shows only low endogenous NST activity, except for GDP-Man transporter activity, which is highly present in yeast. The experimental procedure consists of (I) a subcellular fractionation of yeast expressing the NST of interest and (II) performing the NST activity assay.

  1. (I)

    Preparation of subcellular fractionation of yeast (Saccharomyces cerevisiae) expressing the NST of interest (Kamiyama et al. 2003, 2006; Suda et al. 2004; Goda et al. 2006; Sasaki et al. 2009):

    1. 1.

      An NST coding region is inserted into the yeast expression vector YEp352GAP-II with three copies of HA epitope tags (YPYDVPDYA) at the position corresponding to the C terminus of the NST of interest. (An HA tag can be inserted at the position corresponding to the N terminus if the NST has an ER retention signal, a dilysine motif, at its C terminus.)

    2. 2.

      The yeast strain W303-1a (MATa, ade2-1, ura3-1, his3-11, his15, trp1-1, leu2-3, leu112, and can1-100) is transformed by the lithium acetate procedure using the yeast expression vector YEp352GAP-II.

    3. 3.

      In order to select the transformants, the transformed yeast cells are grown at 30 °C in a synthetic defined medium that lacks uracil.

    4. 4.

      The cells are harvested and then washed with ice-cold 10 mM NaN3.

    5. 5.

      The cells are converted into spheroplasts by incubation at 37 °C for 30 min in spheroplast buffer (1.4 M sorbitol, 50 mM potassium phosphate (pH 7.5), 10 mM NaN3, 40 mM 2-mercaptoethanol, and 1 mg/g of cells of zymolyase 100T (Seikagaku Kogyo Co. Ltd.; Tokyo, Japan)).

    6. 6.

      The spheroplasts are centrifuged into a pellet at 4 °C and washed twice with 1.0 M ice-cold sorbitol to remove traces of zymolyase.

    7. 7.

      The spheroplasts are suspended in ice-cold lysis buffer (0.8 M sorbitol in 10 mM triethanolamine (pH 7.2), 5 μg/ml pepstatin A, and 1 mM phenylmethylsulfonyl fluoride) and homogenized using a Dounce homogenizer.

    8. 8.

      The lysate is centrifuged at 1,000 g for 10 min to remove unlysed cells and cell wall debris.

    9. 9.

      The supernatant is centrifuged at 10,000 g for 15 min at 4 °C to obtain the P10 membrane fraction pellet.

    10. 10.

      The supernatant is centrifuged further at 100,000 g to obtain the P100 membrane fraction pellet.

    11. 11.

      Each membrane fraction pellet is used for the NST activity assay.

    The expression level of HA-tagged NST in each membrane fraction is determined by western blot analysis using an anti-HA mouse monoclonal antibody (Santa Cruz Biotechnology, Inc.; Santa Cruz, CA, USA).

  2. (II)

    The NST activity assay (Kamiyama et al. 2003, 2006; Suda et al. 2004; Goda et al. 2006; Sasaki et al. 2009):

    1. 1.

      Each of the pellets described above (100 μg protein) is incubated at appropriate temperature for 5 min in 100 μl of reaction buffer (20 mM Tris–HCl (pH 7.5), 0.25 M sucrose, 5.0 mM MgCl2, 1.0 mM MnCl2, and 10 mM 2-mercaptoethanol) that contains 1 μM radiolabeled nucleotide sugars or PAPS substrate.

    2. 2.

      The reaction is stopped by adding 1 ml of stop buffer (20 mM Tris–HCl (pH 7.5), 0.25 M sucrose, 5.0 mM MgCl2).

    3. 3.

      The radioactivity incorporated in the microsomes is trapped by a 0.45-μm nitrocellulose filter. Then, the filter is washed with 10 ml of stop buffer and trapped radioactivity is measured by using a liquid scintillation counter. The amount of incorporated radioactivity is calculated as the difference from a background value obtained by the same assay at corresponding temperature at 0 min for each sample.

Functional Analysis of NSTs by RNA Interference

RNA interference (RNAi) is widely used for the functional analysis of a broad range of biologically important molecules. The introduction and expression of double-stranded RNAs into a cell results in the knockdown of expression of a targeted gene through interference of this exogenously introduced RNA molecule with the RNA molecule of the target gene (Kamiyama et al. 2003, 2011; Goda et al. 2006; Sasaki et al. 2009). NST genes are no exception and can be targeted with this same scheme. In embryonic stem cells, short hairpin RNAs (shRNAs) expressed using shRNA expression vectors and 21–25-nt synthetic small interfering RNAs (siRNAs) can be used with effective knockdown efficiency (Sasaki et al. 2009). However, in mature mammalian cell lines, synthetic siRNAs are more useful than shRNAs expressed using expression vectors (Kamiyama et al. 2006, 2011). In order to eliminate off-target effects, at least two different siRNA sequences should be used. As long as the sequences used for knockdown are in the 3′ noncoding regions, rescue experiments can be carried out by the over expression of knockdown genes. The siRNA-induced knockdown of NSTs had a delayed effect on glycosylation compared to knockdown of genes coding for core proteins and glycosyltransferases. Effect of synthetic siRNAs cannot be kept long. Thus, it is recommended that the transfection of synthetic siRNAs should be repeated three times. For detailed descriptions of experimental conditions, refer to the references (Kamiyama et al. 2003, 2006, 2011; Goda et al. 2006; Sasaki et al. 2009).

Results

Determination of Km Values

The Km values of human FRC1 (Hs SLC35D2/HFRC1) for nucleotide sugars were determined by using the methods described above (Suda et al. 2004). Hs SLC35D2/HFRC1 is the human ortholog of Drosophila Frc (Dm FRC) and C. elegans SQV-7, which showed multi-substrate specific NST activity (Fig. 2). Hs SLC35D2/HFRC1 showed transport activity for UDP-GlcNAc and UDP-Glc and their Km values were measured to be 8.0 and 2.1 μM, respectively. Km values of Dm FRC for UDP-GlcNAc were 3 μM or 7.8 μM, where S. cerevisiae or Leishmania microsomal vesicles was used for overexpressing HA-tagged Dm FRC, respectively (Selva et al. 2001; Muraoka et al. 2007). The Km values for UDP-GlcNAc were not very different between human and Drosophila.

The Km values of PAPST1 and PAPST2 from various species for PAPS were also determined by using the method described above. Apparent Km values of 0.8 μM (Kamiyama et al. 2003), 1.2 μM (Kamiyama et al. 2003), 1.54 μM (Sasaki et al. 2009), and 4.03 μM (Dejima et al. 2010) were determined for human PAPST1 (Hs SLC35B2/PAPST1), Drosophila SLL (Dm SLL), mouse PAPST1 (Mm SLC35B2/PAPST1), and C. elegans PST1 (Ce M03F8.2/PST-1), respectively. The Km values of PAPST1 for PAPS were not significantly different between species. The Km values of 2.2 μM (Kamiyama et al. 2006), 2.3 μM (Goda et al. 2006), and 1.49 μM (Sasaki et al. 2009) were measured for human PAPST2 (Hs SLC35B3/PAPST2), Drosophila PAPST2 (Dm PAPST2), and mouse PAPST2 (Mm SLC35B3/PAPST2), respectively. The PAPST2 of each species showed a similar Km value as its PAPST1.

Functional Analysis by RNAi

NSTs or PAPSTs determine the status of glycosylation or sulfation by supplying the donor substrates to glycosyltransferases or sulfotransferases, respectively. To study the role of PAPSTs in colorectal cancer, synthetic siRNAs were used in colorectal cancer cell lines as a knockdown for human PAPST1 and PAPST2 (Kamiyama et al. 2011). In this study, the siRNAs were transfected three times, once every 3 days, in order to obtain optimal interference and inhibition. Knockdown of either of the two PAPST genes reduced the amount of sulfation of glycans on cellular proteins, FGF signaling, and also cellular proliferation. This result indicates that PSPTs play a role in the proliferation of colorectal cancer cells by controlling the sulfation state of their target molecules.

Comments

Many NSTs have been cloned to date, and their activity has also been identified (Fig. 2 and Table 1). Several NSTs can transport the same nucleotide sugar. For example, four types of human NSTs have been identified to be able to transport UDP-GlcNAc, namely, Hs SLC35A3/UGTrel2, Hs SLC35B4/huYEA4, Hs SLC35D1/UG, and Hs SLC35D2/HFRC1. Hs SLC35A3/UGTrel2, Hs SLC35B4/huYEA4, and Hs SLC35D2/HFRC1 are localized in the Golgi apparatus, while Hs SLC35D1/UGTrel7 is localized in the ER (Table 1). The purpose of such redundancy is not yet clear. One possibility is that specific glycosyltransferases bind to each of these NSTs and use them differently. Redundancy can also be seen in “PAPSTs,” such as PAPST1 and PAPST2, both of which are localized in the Golgi apparatus of humans and other species (Table 1). In plants, there are many UDP-Gal transporters in different SLC35 subgroups, SLC35B, SLC35D, and SLC35E, which are yet other examples of redundancy for which the purpose is not clear (Fig. 2 and Table 1). This unsolved issue of redundancy might be important for our understanding of the basic mechanisms underlying the glycosylation and sulfation systems.