Research article

Targeting and intracellular trafficking of clinically relevant hTHTR1 mutations in human cell lines
Veedamali S. Subramanian, Jonathan S. Marchant, Hamid M. Said
Clinical Science Jul 01, 2007, 113 (2) 93-102; DOI: 10.1042/CS20060331

Abstract

The micronutrient thiamine is required for normal growth and development of human tissues, and is accumulated into cells through the activity of plasma membrane thiamine transporters, e.g. hTHTR1 (human thiamine transporter 1). Recent genetic evidence has linked mutations in hTHTR1 with the manifestation of TRMA (thiamine-responsive megaloblastic anaemia), a condition also associated with diabetes mellitus, sensorineural deafness and retinal disorders. To examine how mutations in hTHTR1 impair thiamine accumulation, we have investigated the targeting and functional properties of several different hTHTR1 mutants in human cell lines derived from epithelia relevant to thiamine absorption or tissues implicated in TRMA pathology. These constructs encompassed two newly identified point mutations (P51L and T158R) and two truncations of hTHTR1 identical with those found in TRMA kindreds (W358X and Δ383fs). Our results reveal a spectrum of mutant phenotypes, underlining that TRMA can result from decreased thiamine transport activity underpinned by changes in hTHTR1 expression levels, cellular targeting and/or protein transport activity.

  • cellular targeting
  • human thiamine transporter (hTHTR)
  • micronutrient
  • thiamine
  • thiamine-responsive megaloblastic anaemia (TRMA)
  • transporter
  • vitamin

INTRODUCTION

Thiamine (vitamin B1) is a water-soluble micronutrient that cannot be synthesized de novo by humans and is therefore accumulated in target tissues by transport into the cell from the extracellular environment using carrier-mediated mechanisms. The crucial role played by thiamine in cellular metabolism is underlined by the plethora of disorders that may result from thiamine deficiency, including cardiovascular and neurological disorders [1–6]. Deficiencies in cellular accumulation also occurs in TRMA (thiamine-responsive megaloblastic anaemia), which is a rare autosomal-recessive disorder characterized by the pathological triad of megaloblastic anaemia, diabetes mellitus and sensorineural deafness [3,7,8]. Recent genetic analyses have identified the mutational locus of TRMA as the protein product of the SLC19A2 gene, encoding hTHTR1 (human thiamine transporter 1), a saturable high-affinity thiamine transporter that mediates thiamine accumulation in many tissues [2,7,9–11]. To date, 16 distinct clinical mutations have been identified in SLC19A2 comprising point mutations as well as premature truncations resulting from missense, nonsense and frame-shift mutations [2,7,10–14]. However, little is known about the cell biological effect of many of these mutations on hTHTR1 stability, localization or activity. Importantly, there is a need to resolve the cellular expression profile of TRMA-associated mutations in human tissues, using cell types relevant to thiamine homoeostasis (e.g. absorptive epithelia) or implicated in TRMA pathology (e.g. pancreas and retina).

To define the cellular expression phenotype of several clinically identified hTHTR1 mutants (P51L, T158R, W358X and Δ383fs [2,7,10–14]), we used live cell confocal imaging to resolve the distribution of hTHTR1 mutants tagged with GFP (green fluorescent protein) in five different human cell lines derived from pancreas (MIA PaCa-2), retina (ARPE-19), absorptive epithelia (HuTu-80 and Caco-2) or liver (HepG2). These specific mutants were chosen because the cellular basis of the dysfunction that produces impaired thiamine transport targeting has not been resolved previously for any of these mutants in any cell line. The results demonstrated a spectrum of cellular targeting phenotypes: the recently identified P51L mutation [13], although expressed at the cell surface, exhibited a dysregulated trafficking phenotype and impaired functionality. T158R [14] was poorly expressed after transient transfection and the majority was retained intracellularly within the endoplasmic reticulum in most, but not all, cell lines examined, a result which may have implications for the tissue-specific pathologies displayed in TRMA. Therefore different TRMA mutants can impair hTHTR1 functionality in a variety of ways such that the cellular basis of TRMA in discrete patients may result from defects in protein folding, targeting or transport activity.

MATERIALS AND METHODS

Materials

FM4-64 [N-(3-triethylammoniumpropyl)-4-(6-(4-diethylamino)phenyl)-hexatrienyl)pyridinium dibromide] was from Molecular Probes. GFP-N3, DsRed2 (Discosoma sp. red fluorescent protein 2) and DsRed2-ER (DsRed2 targeted to the endoplasmic reticulum) were from Becton Dickinson Biosciences. Plasmids encoding V1aR–GFP (rat vasopressin receptor linked to GFP) and hSVCT1–YFP [human sodium-dependent vitamin C transporter-1 linked to YFP (yellow fluorescent protein)] have been described previously [15,16]. Cell lines were obtained from A.T.C.C. DNA oligonucleotide primers were from Sigma Genosys. [3H]Thiamine (specific activity, 555 GBq/mmol) was obtained from ARC.

Generation of hTHTR1–GFP mutants

The QuikChange® site-directed mutagenesis kit (Stratagene) was used to introduce insertions or deletions of nucleotides into the ORF (open reading frame) of SLC19A2, cloned from a human colon cell line (Caco-2). Paired sense and antisense primer oligonucleotides encompassing the specified mutation sites (Table 1), as well as a plasmid containing hTHTR1 fused to GFP-N3 [17], were used as a template for PCR mutagenesis. Nucleotide changes in all constructs were confirmed by sequencing (Laragen).

View this table:
Table 1 Gene-specific primers used for generating mutations/truncations in SLC19A2

The required nucleotide changes in forward and reverse primers (in bold) were used for generating point mutations. Restriction sites for XhoI (italics) and BamHI (underlined) were incorporated into the SLC19A2 primer to aid subcloning into the GFP-N3 vector.

Cell culture, transient and stable transfection

Human duodenal-derived intestinal epithelial cells (HuTu-80) and human embryonic kidney cells (HEK-293 cells) were maintained in MEM (minimal essential medium). Human colon adenocarcinoma (Caco-2), humanderived liver cells (HepG2), human retinal pigmented epithelium (ARPE-19) and human pancreatic carcinoma (MIA PaCa-2) cells were maintained in DMEM (Dulbecco's modified Eagle's medium). Media was supplemented with 10% (v/v) fetal bovine serum, glutamine (0.29 g/l), sodium bicarbonate (2.2 g/l for MEM and 3.7 g/l for DMEM), penicillin (100000 units/l) and streptomycin (10 mg/l). For transient transfection, cells were grown on sterile glass-bottomed Petri dishes (MatTek) and transfected at 90% confluency with 1 μg of plasmid DNA using Lipofectamine2000™ (Invitrogen). After 24–48 h, cells were analysed by confocal microscopy. For generation of stable cell lines, transiently transfected cells were selected using G418 (0.8 mg/ml; Invitrogen) for approx. 6–8 weeks.

Uptake analysis

To determine the functionality of the recently identified mutants (P51L and T158R), we performed [3H]thiamine uptake assays on confluent monolayers as described previously [17,18].

Semi-quantitative RT-PCR (reverse transcriptase-PCR)

Total RNA (5 μg) was isolated from HuTu-80 cells and primed with oligo(dT) primers to synthesize firststrand cDNA (SuperScript™ First Strand Synthesis RT-PCR kit; Invitrogen). Gene-specific primers for hTHTR1 (forward, 5′-ATGGATGTGCCCGGCC-3′ and reverse, 5′-TGAAGTGGTTACTTGAGAACT-3′) and for β-actin (forward, 5′-TTGTAACCAACTGGGACGATATGG-3′ and reverse, 5′-GATCTTGATCTTCATGGTGCTAGG-3′) were used to examine the expression levels of hTHTR1 and β-actin mRNA using established PCR conditions [17,18].

Confocal imaging of hTHTR1 mutants

Cell monolayers grown on coverslips were imaged for construct expression using a confocal scanner (MRC1024; Bio-Rad Laboratories) attached to an upright microscope (AX70; Olympus). Fluorophores were excited using the 488-nm line from an argon ion-laser, and emitted fluorescence was monitored with a 530±20 nm band-pass (GFP) or a 620 nm long-pass (DsRed2 and FM4-64) filter. Measurements of construct-targeting polarity were made using filter-grown Caco-2 cells. Cells were seeded on to collagen-coated filters (Corning Costar) and grown post-confluence with regular changes of medium. Individual wells were transfected using Lipofectamine2000™ approx. 5 days post-confluency in serum-deficient medium [16–18]. Cells were subsequently imaged 36–48 h after transfection using a ×60 water-immersion objective lens, and fluorescence distribution was quantified using the IDL analysis package (Research Systems). Measurements of transporter distribution were analysed using one-way ANOVA and subsequent Dunnett's tests to compare experimental data with control constructs of known targeting polarity [15]. Results were considered significant at P<0.05. To quantify the number of discrete punctal structures in cells expressing hTHTR1 variants we employed a vesicle-counting algorithm. Briefly, fluorescence expression throughout single cells was captured in an axial stack (x,y,z) comprising approx. 150 planes/cell. This stack was low-pass filtered (3×3) and displayed as a 3D (three-dimensional) projection for quantification purposes using a count objects function (Metamorph). For comparison between different cells, results were corrected for cellular volume.

Flow cytometry

Flow cytometry was performed using a FACSCalibur benchtop cytometer (Becton Dickinson Biosciences). To obtain sufficient numbers of positively transfected cells for these experiments, HEK-293 cells were grown in T25 tissue culture flasks and transfected using Lipofectamine2000™ in serum-free medium (Optimem). Monolayers were trypsinized 48 h post-transfection, and cells were pelleted, resuspended and filtered (35 μm filter) in 2 ml aliquots of Ca2+-free medium at a density of 1×106 cells/ml, as described previously [19]. In all flow cytometry experiments, samples of untransfected and GFP-alone transfected HEK-293 cells were run in parallel with hTHTR1-transfected samples in order to calibrate optical parameters for identifying the intact transfected cell population.

RESULTS

Overview of the introduced mutations within hTHTR1

Figure 1 shows a schematic diagram of the hTHTR1 protein to highlight the sites of missense, nonsense and frame-shift mutations identified in TRMA patients within the predicted 12 transmembrane-spanning regions of hTHTR1. To investigate the effect of four of these TRMA mutations on hTHTR1 cellular targeting, protein expression and function, we analysed transfected cells first at the population level by flow cytometry, then at the single cell level in human cell lines using confocal microscopy and finally using functional uptake studies with [3H]thiamine. Results from all these analyses are described in the following sections and are summarized in Table 2.

Figure 1
Figure 1 Schematic representation of mutational loci within hTHTR1

The sequence of hTHTR1 (the transmembrane regions are shaded and the extracellular regions are underlined) is shown overlaid with the locations of all of the TRMA-associated mutations (boxed) identified to date. These encompass point mutations (P51L, D93H, S143F, T158R and G172D), truncation mutants (E66X, R162X, W250X and W358X) and frame-shift mutants (ins 81fs97, Δ95fs127, Δ142fs239, Δ241fs259, Δ294fs313, Δ383fs385 and 408+1 splice). The four mutants investigated in the present study are highlighted (black boxes).

View this table:
Table 2 Summary of functional, expression and targeting data with hTHTR1 constructs

Expression data were from flow cytometry assays (Figure 2) expressed relative to results with hTHTR1–GFP. Functional data ([3H]thiamine uptake) were performed in stable HuTu-80 cells and are expressed relative to results with hTHTR1–GFP. [3H]Thiamine accumulation could not be resolved over control levels using transient transfection methods. Targeting data are summarized in this Table owing to the infeasibility of presenting images for all of the constructs in every cell line and the distributions are as follows: (bold rectangle) co-localization with FM4-64, a plasma membrane marker; (shaded) co-localization with DsRed2-ER, an endoplasmic reticulum marker; (shaded bold rectangle) mixed phenotype; and (x) no detectable expression. Polarity data were collected in filter-grown Caco-2 cells using methods reported previously [16,17]. n.d., not-detectable. n/a, not applicable.

Flow cytometry analysis of hTHTR1 mutants

In order to quantify the relative levels of expression of individual constructs, we performed flow cytometry. For these experiments we used HEK-293 cells, owing to the high transfection efficiency attainable in this cell line (approx. 70%). Two measurements were obtained. First, a comparison of the number of fluorescent events relative to the total number of cells counted yielded a crude index of transfection efficiency. This value was expected to be similar between constructs under the control of the same promoter, provided that the individual fusion proteins were equally tolerated after cellular expression. However, the fluorescence of the four mutant constructs were distributed into two distinct groups (Figure 2A): one group exhibited comparable transfection efficiency with that observed with full-length type hTHTR1–GFP (approx. 60–80%, n=3 independent transfections), whereas the second group, comprising the nonsense mutant W358X and frame-shift mutant Δ383fs, displayed significantly lower transfection efficiency (Figure 2A). Secondly, comparison of the mean fluorescence intensities between cells positively transfected with mutant constructs and wild-type hTHTR1–GFP-expressing cells provided a measure of the expression efficiency of individual mutants (Figure 2B). Samples fell into three groupings: (i) P51L was expressed at a similar level to wild-type (100±1%); (ii) T158R expression was considerably less (approx. 50% of wild-type), despite similar transfection efficiencies (Figure 2A), and (iii) both truncation constructs which exhibited poor transfection efficiency and poor expression (W358X and Δ383fs) had low fluorescence intensities compared with wild-type. These results predict a variable degree of impairment of hTHTR1 expression resulting from specific mutations. To define the cellular nature of these defects more precisely, we proceeded to examine the cellular distribution of each mutant using live cell confocal imaging.

Figure 2
Figure 2 Flow cytometry analyses of hTHTR1–GFP mutants

(A) Comparison of the number of fluorescent events relative to the total number of HEK-293 cells counted, a crude index of transfection efficiency. (B) Mean total fluorescence of HEK-293 cells transfected with various mutant constructs, expressed as a percentage of that obtained with the full-length hTHTR1–GFP construct (mean of at least three independent experiments). Values are means±S.E.M.

Confocal microscopy of hTHTR1 mutants

To resolve the cellular localization of individual constructs, we transfected individual mutants into five different human cell lines: a retinal epithelial cell line (ARPE-19 cells), a hepatically derived cell line (HepG2), a pancreatic-derived cell line (MIA PaCa-2), duodenally derived intestinal epithelial cells (HuTu-80) and a colonic-derived epithelial cell line (Caco-2). These cell lines represent two tissue types implicated in TRMA pathology (MIA PaCa-2 and ARPE-19) as well as three epithelially derived cell lines involved in thiamine absorption/transport which are known to express endogenous hTHTR1 protein [17,18,20]. In all five cell lines, wild-type hTHTR1–GFP protein was targeted predominantly to the plasma membrane, but was also evident within a variety of intracellular vesicular structures (Figures 3A and 3B). Cell-surface localization of hTHTR1–GFP was confirmed by the high degree of co-localization in live cells co-stained with a red plasma membrane marker FM4-64 (>70%), but a much lower degree of co-localization with either the cytosolic marker DsRed2 (<20%) or an endoplasmic-reticulum-targeted fluorescent protein construct DsRed2-ER (<10%; Figure 3B). This plasma-membrane targeting of the full-length hTHTR1–GFP construct was also consistent with functional data (see below), in which [3H]thiamine uptake was significantly (P<0.05) higher in HuTu-80 cells overexpressing hTHTR1–GFP (212±10 fmol·mg−1 of protein·7 min−1) compared with control cells expressing GFP alone (approx. 115±10 fmol·mg−1 of protein·7 min−1). The relatively low extent (approx. 2-fold) of potentiation of the kinetics of [3H]thiamine uptake in cells overexpressing hTHTR1 deserves comment. First, native HuTu-80 cells have significant basal thiamine uptake, owing to the presence of the other thiamine transporter subtype (hTHTR2), which is strongly expressed in these cells [21]. RNA interference experiments have also underlined the contribution of both thiamine transporters in basal thiamine uptake [18]. Secondly, a 2–3-fold potentiation of the uptake rate following overexpression is consistently observed with other SLC19A family members, including hRFC (human reduced folate carrier; SLC19A1 [15]). This may indicate a common regulation of the transport kinetics of these proteins, despite their divergent substrate specificity. Finally, it is possible that the level of potentiation correlates better with the amount of overexpressed protein that is trafficked to the cell surface, i.e. overexpression results in only a 2-fold increase in hTHTR1 in the plasma membrane. Finally, hTHTR1–GFP was predominantly expressed at the basolateral membrane domain (66.6±5.8% basolateral, n=28 cells) in filter-grown Caco-2 cells which form fully polarized monolayers in culture post-confluency, as shown in the present study through the differential targeting of known apical (hSVCT1) and basolateral constructs (V1aR; Figure 3C). This distribution is consistent with its speculated role in vectorial transport of thiamine across polarized epithelia [15,18]. In summary, these results confirm the functionality and appropriate cellular targeting of the wild-type hTHTR1–GFP fusion protein. Using these same methods, comparisons were made with each of the hTHTR1 mutants, and these results have been organized into the following two sections below: (i) clinical missense mutations (P51L and T158R), and (ii) clinical truncations (W358X and Δ383fs).

Figure 3
Figure 3 Cellular distribution of hTHTR1–GFP

(A) Confocal images (x,y) of hTHTR1 localization in human cell lines of different tissue origin (ARPE-19, HepG2, MIA PaCa-2 and HuTu-80). (B) Dual-emission imaging of hTHTR1–GFP-expressing HuTu-80 cells that were co-stained with FM4-64 (top panel) or co-transfected with DsRed2 (middle panel) or DsRed2-ER (bottom panel). Each set of images shows the green fluorescence channel (GFP alone; upper left), red fluorescence channel (lower left) and an overlay between these two labels (depicted in yellow; right). (C) Polarized expression of hTHTR1–GFP in filter-grown Caco-2 cells shown in a lateral section (x,y; top left) and axial section (z; bottom left), relative to the distributions (right) of known basolaterally targeted (V1aR; distribution, 28.2±3.7% basolaterally; n=17 cells) and apically targeted (hSVCT1; distribution, 90.0±6.4% apically; n=8 cells) proteins. Methods for quantification of live cell construct distribution have been reported previously [17,18].

Clinical missense mutations

The two clinical missense mutants (P51L and T158R) were transiently transfected into the same human cell lines and lateral (x,y) confocal images were captured 24–48 h later. The newly identified point mutation P51L [13] was resolved at the cell surface, as well as within many intracellular vesicular structures in all of the cell lines tested (Figure 4A and Table 2). However, in comparison with cells transfected with hTHTR1–GFP, P51L-expressing cells appeared by eye to contain a greater number of intracellular vesicles (Figure 4A), a difference that was best demonstrated in confocal images of a focal plane just superficial to the coverslip. This phenotype was reproducible in all of the human cell lines tested (Figure 4B), and there was little co-localization of these intracellular structures with the endoplasmic reticulum (Figure 4B). With both wild-type and P51L constructs, the intracellular vesicular population appeared to be heterogeneous, comprising small dense vesicles as well as larger ‘hollow’ vesicular structures (Figure 4C). To quantify this difference better, we employed a vesicle-counting algorithm applied to 3D projections of axial (x,y,z) stacks collected from single cells expressing either hTHTR1–GFP or hTHTR1[P51L]–GFP respectively (Figure 4D). This analysis confirmed that P51L-expressing cells contained a significantly (P<0.05) higher number of vesicular structures than wild-type cells (approx. 1.3–1.7-fold, depending on dimensions of search criteria, n=40 cells; Figure 4D). Axial confocal sections of polarized Caco-2 monolayers revealed the cell-surface distribution of P51L (68.2±7.9% basolateral, n≥10 cells and n≥3 transfections) was similar to that observed with wild-type hTHTR1–GFP (66.6±5.8% basolateral). Finally, functional uptake of P51L was studied in stably transfected HuTu-80 cells. Uptake of [3H]thiamine was significantly (P<0.05) inhibited in HuTu-80 cells expressing the P51L mutant when compared with wild-type cells (139±10 and 212±10 fmol·mg−1 of protein·7 min−1 respectively; Figure 4E). The lower levels of [3H]thiamine uptake observed with this mutant were not due to a failure in transcription, as semi-quantitative RT-PCR measurements demonstrated similar levels of transporter mRNA in these cells (Figure 4E, inset).

Figure 4
Figure 4 Distribution of the hTHTR1[P51L]–GFP mutation in human cell lines

(A) Confocal (x,y) images of HuTu-80 cells expressing P51L (upper cell) and wild-type hTHTR1–GFP (lower cell) captured in a lateral (x,y) section within a confocal plane just superficial to the coverslip. (B) Expression of hTHTR1[P51L]–GFP in ARPE-19, HepG2, MIA PaCa-2 and HuTu-80 cells. In HuTu-80 cells, DsRed2-ER (red) was co-expressed with hTHTR1 (green). (C) Enlarged section of an HuTu-80 cell revealing a multitude of intracellular vesicular structures containing P51L. (D) Left-hand panel, 3D projections of an hTHTR1[P51L]–GFP-expressing cell constructed from a axial stack of x,y images taken at 0.08 μm increments (inset). Images were captured before (upper cell) and after (lower cell) the application of the particle identification algorithm to show voxels that were scored positively (orange). Right-hand panel, comparison of the number of objects identified in hTHTR1[P51L]–GFP compared with hTHTR1–GFP cells using a variety of different search criteria (puncta search size represented by different colours). On the left (filled squares), the ratio of particles in hTHTR1[P51L]–GFP cells compared with hTHTR1–GFP cells is shown for individual measurements (different colours). The average from each experiment (n>10 field measurements) using a specific search criterion is shown as an arrow (right) with the same colour as the individual data. Therefore, over a range of search criteria, hTHTR1[P51L]–GFP-containing cells contained more intracellular particles (approx. 1.3–1.7-fold; black line). (E) Uptake of [3H]thiamine in control cells and GFP-, hTHTR1–GFP-, hTHTR1[P51L]–GFP (P51L)- and hTHTR1[T158R]–GFP (T158R)-expressing stable HuTu-80 cells. Results are means±S.E.M from five different experiments. Inset, semi-quantitative RT-PCR products from HuTu-80 cells expressing the indicated constructs, as well as β-actin controls (lower band).

Finally, in contrast with the results with P51L, T158R was retained within intracellular membranes in the majority of cell lines examined (ARPE-19, MIA PaCa-2 and HepG2; Figure 5A). In a minority (approx. 10%) of HepG2 cells, T158R was evident at the cell surface, although the predominate phenotype was that of intracellular retention (Figure 5A). However, in HuTu-80 cells transiently or stably transfected with the same construct, most cells expressed T158R at the cell surface (Figure 5A), suggesting a higher tolerability of this mutant for trafficking in this cell line. However, even though this mutant was expressed in the plasma membrane in HuTu-80 cells, functional uptake assays revealed a significantly (P<0.05) impaired [3H]thiamine accumulation in stable HuTu-80 cells expressing T158R compared with wild-type-expressing cells (119±10 and 212±10 fmol·mg−1 of protein·7 min−1 respectively; Figure 4E).

Figure 5
Figure 5 Distribution of the hTHTR1[T158R]–GFP mutation in human cell lines

(A) Targeting of T158R in human cell lines. T158R was expressed intracellularly in ARPE-19 and MIA PaCa-2 cells. A similar phenotype was seen in HepG2 cells, although a small proportion of cells exhibited cell-surface targeting (arrow). In HuTu-80 cells, the majority of cells expressed T158R at the cell surface. (B) Theoretical models leading to tissue-specific phenotype of TRMA. Graphical representation of possible imbalances in supply (i, ii and iii; left) or demand (iv; right) that could lead to the specific tissue-specific pathologies observed in TRMA patients. (i) Alternative route for thiamine uptake. Affected tissues (trapezoid; top) are exclusively dependent on functional hTHTR1 (black square). Impairment of hTHTR1 function (dashed arrow) results in insufficient thiamine uptake (pale background) to meet tissue need for thiamine. In unaffected tissues (circle; below), an alternative transporter (open square, e.g. hTHTR2, passive uptake by a low-affinity carrier) ensures an adequate intracellular thiamine concentration (grey background). (ii) Variation in hTHTR1 expression between tissues. Higher expression of a poorly functional transporter is sufficient to meet metabolic demands in unaffected tissues, whereas lower expression levels of hTHTR1 in affected tissues make these tissue uniquely sensitive to impairment of hTHTR1 function. (iii) Tissues display differential capacity to synthesize/target mutant hTHTR1 to the cell surface. (iv) TRMA-affected tissues have uniquely high demands for thiamine (dark-grey background) compared with other cell types (grey background) and are therefore very sensitive to decreased intracellular thiamine concentrations resulting from mutational impairment of hTHTR1. Obviously, none of these possibilities are mutually exclusive, but serve as a simplification for discussion.

Clinical truncation mutations

The cellular distribution of clinical nonsense (W358X) and frame-shift (Δ383fs) mutants was examined after transient transfection. For both constructs, very little fluorescence could be resolved in any of the five cell lines tested, despite expression of control constructs (GFP and hTHTR1–GFP) transfected in parallel. Even at the highest levels of laser illumination, GFP fluorescence signals were weak and intracellular (Table 2).

DISCUSSION

The aim of the present study was to investigate the effect of newly identified clinically relevant mutations within hTHTR1 on the expression of this thiamine transporter in live cells. We generated four discrete hTHTR1 mutants and analysed their expression profile in five different human cell lines to gain insight into the cellular impact of known clinically relevant residues within this important micronutrient transporter.

At least sixteen distinct mutations have been identified from patients with TRMA, encompassing missense mutations (e.g. P51L and T158R), as well as nonsense or frame-shift mutations [7,12,14]. In the present study, we have described for the first time the cellular effects of the recently discovered P51L and T158R point mutations [13,14], as well as the W358X truncation and Δ383 frame-shift mutation [10,22]. Furthermore, by employing live cell imaging methods, we can now resolve that different cellular targeting phenotypes are associated with the loss of thiamine transport activity. Although P51L is targeted with similar expression efficiency (Figure 2B) and polarity to wild-type hTHTR1, thiamine transport is impaired (Figure 4E). In contrast, the T158R point mutation was retained within the endoplasmic reticulum in several cell types in a manner consistent with clinical point mutations occurring within many other nutrient transporters [23,24].

It is not currently understood why certain tissues are selectively impaired in TRMA, beyond the general assumption that a mutation in hTHTR1 results in thiamine supply being insufficient to meet the metabolic need of these specific tissues (Figure 5B). Theoretically, one can envisage several mechanisms, all of which could contribute to this equation, including the absence of alternative pathways for thiamine uptake [e.g. hTHTR2; Figure 5B(i)], low expression levels of partially active hTHTR1 in affected tissues [low thiamine uptake capacity; Figure 5B(ii)], impaired cell-surface targeting [Figure 5B(iii)], as well as uniquely high thiamine demand [Figure 5B(iv)]. The relative importance of these possibilities has yet to be dissected experimentally, although emphasis has been placed on the absence of alternative thiamine transport pathways in the literature [13]. In the present study, we would highlight that the T158R mutant had a higher efficiency at reaching the cell surface in some cell types (HuTu-80, HepG2 and Caco-2) compared with others (ARPE-19 and MIA PaCa2). Although academic for this transporter (as it is non-functional in [3H]thiamine accumulation assays; Figure 4E), it is worth highlighting that cell-type-specific processing of transporters may represent a contributory factor in the manifestation of the tissue-specific pathologies associated with TRMA. For example, if certain cell types are better able to traffic transporters with residual, but significant, functionality to the cell surface (e.g. as observed in the present study in gut epithelia), but other cell types cannot (as observed in the present study in pancreatic and retinal cell lines), then these tissues might exhibit differential abilities to buffer a macroscopic thiamine deficiency [see Figure 5B(iii)]. Obviously, this must be a generalized phenotype (i.e. hold for several different mutants), but it could be a contributory mechanism that allows some tissues to be more resistant to the effects of impaired hTHTR1 functionality.

Both truncation mutants (W358X and Δ383fs) were poorly tolerated by cells (Table 2), and very little expression was present/observed intracellularly (cytoplasmic or membrane-bound). These observations confirm and extend our previous analyses of targeting determinants within hTHTR1, where we demonstrated a requirement for both N-terminal sequence and integrity of the transmembrane polypeptide backbone for cell-surface expression of hTHTR1 [17]. Notably, premature truncation of hTHTR1 between the sixth and seventh transmembrane domains {hTHTR1[1–240]–GFP} in a region where three clinical relevant truncations are localized (after residues 239, 250 and 259) resulted in intracellular retention within the endoplasmic reticulum [18]. In the present study, we have analysed the phenotype of clinical mutants with less extensive truncations (at residues 358 and 383), and both mutants were poorly expressed (Figure 2). These results confirm the importance of the sequence within the polypeptide backbone for cellsurface expression of hTHTR1 and, given the tolerance of cytoplasmic tail deletions (residues 479–497; [17]), highlight the role of the sequence within the region spanning residues 383–479 as an important determinant for targeting hTHTR1 to the cell surface.

In summary, the results of the present study have demonstrated a spectrum of cellular targeting phenotypes, underlining that TRMA can result from defects in transport activity underpinned by changes in hTHTR1 expression level, protein folding, cellular targeting and/or protein transport activity.

Acknowledgments

This work was supported by the Department of Veterans Affairs, and NIH (National Institutes of Health) grants DK-56061, DK-58057 (H.M.S), DK-71538 (V.S.S) and NS-046783 (J.S.M.).

Abbreviations: DMEM, Dulbecco's modified Eagle's medium; DsRed2, Discosoma sp. red fluorescent protein 2; DsRed2-ER, DsRed2 targeted to the endoplasmic reticulum; FM4-64, N-(3-triethylammoniumpropyl)-4-(6-(4-diethylamino)phenyl)-hexatrienyl)pyridinium dibromide; GFP, green fluorescent protein; HEK-293 cell, human embryonic kidney cell; hSVCT1, human sodium-dependent vitamin C transporter 1; hTHTR, human thiamine transporter; MEM, minimal essential medium; RT-PCR, reverse transcriptase-PCR; TRMA, thiamine-responsive megaloblastic anaemia; 3D, three-dimensional; V1aR, rat vasopressin receptor

References

  1. ↵
    1. Abboud M. R.,
    2. Alexander D.,
    3. Najjar S. S.
    (1985) Diabetes mellitus, thiamine-dependent megaloblastic anemia, and sensorineural deafness associated with deficient α-ketoglutarate dehydrogenase activity. J. Pediatr. 107:537–541.
    OpenUrlCrossRefMedlineWeb of Science
  2. ↵
    1. Diaz G. A.,
    2. Banikazemi M.,
    3. Oishi K.,
    4. Desnick R. J.,
    5. Gelb B. D.
    (1999) Mutations in a new gene encoding a thiamine transporter cause thiamine-responsive megaloblastic anaemia syndrome. Nat. Genet. 22:309–312.
    OpenUrlCrossRefMedlineWeb of Science
  3. ↵
    1. Mandel H.,
    2. Berant M.,
    3. Hazani A.,
    4. Naveh Y.
    (1984) Thiamine-dependent beriberi in the thiamine-responsive anemia syndrome. N. Engl. J. Med. 311:836–838.
    OpenUrlCrossRefMedlineWeb of Science
    1. Villa V.,
    2. Rivellese A.,
    3. Di Salle F.,
    4. Iovine C.,
    5. Poggi V.,
    6. Capaldo B.
    (2000) Acute ischemic stroke in a young woman with the thiamine-responsive megaloblastic anemia syndrome. J. Clin. Endocrinol. Metab. 85:947–949.
    OpenUrlCrossRefMedlineWeb of Science
    1. Berdanier C. D.
    (1998) Advanced Nutritional Micronutrients (CRC, New York).
  6. ↵
    1. Victor M.,
    2. Adams R. D.,
    3. Collins G. H.
    (1989) The Wernicke–Korsakoff Syndrome and Related Neurological Disorders Due to Alcoholism and Malnutrition (Davis Publications, Philadelphia, PA).
  7. ↵
    1. Raz T.,
    2. Labay V.,
    3. Baron D.,
    (2000) The spectrum of mutations, including four novel ones, in the thiamine-responsive megalobastic anemia gene SLC19A2. Hum. Mutat. 16:37–42.
    OpenUrlCrossRefMedlineWeb of Science
  8. ↵
    1. Rogers L. E.,
    2. Porter F. S.,
    3. Sidbury J. B.
    (1969) Thiamine-responsive megaloblastic anemia. J. Pediatr. 74:494–504.
    OpenUrlCrossRefMedlineWeb of Science
  9. ↵
    1. Dutta B.,
    2. Huang W.,
    3. Molero M.,
    (1999) Cloning of the human thiamine transporter, a member of the folate transporter family. J. Biol. Chem. 274:31925–31929.
    OpenUrlAbstract/FREE Full Text
  10. ↵
    1. Fleming J. C.,
    2. Tartaglini E.,
    3. Steinkamp M.,
    4. Schorderit D. F.,
    5. Cohen N.,
    6. Neufeld E. J.
    (1999) The gene mutated in thiamine-responsive anemia with diabetes and deafness (TRMA) encodes a functional thiamine transporter. Nat. Genet. 22:305–308.
    OpenUrlCrossRefMedlineWeb of Science
  11. ↵
    1. Labay V.,
    2. Raz T.,
    3. Baron D.,
    (1999) Mutations in SLC19A2 cause thiamine-responsive megaloblastic anaemia associated with diabetes mellitus and deafness. Nat. Genet. 22:300–304.
    OpenUrlCrossRefMedlineWeb of Science
  12. ↵
    1. Neufield E. J.,
    2. Fleming J. C.,
    3. Tartaglini E.,
    4. Steinkamp M. P.
    (2001) Thiamine-responsive megaloblastic anemia syndrome: a disorder of high-affinity thiamine transport. Blood Cell Mol. Dis. 27:135–138.
    OpenUrlCrossRefMedlineWeb of Science
  13. ↵
    1. Lagarde W. H.,
    2. Underwood L. E.,
    3. Moats-Staats B. M.,
    4. Calikoglu A. S.
    (2004) Novel mutation in the SLC19A2 gene in an African-American female with thiamine-responsive megaloblastic anemia syndrome. Am. J. Med. Gen. 125:A299–A305.
    OpenUrl
  14. ↵
    1. Ricketts C. J.,
    2. Minton J. A.,
    3. Samuel J.,
    (2006) Thiamine-responsive megaloblastic anemia syndrome: long-term follow-up and mutation analysis of seven families. Acta Paediatr. 95:99–104.
    OpenUrlCrossRefMedlineWeb of Science
  15. ↵
    1. Boulware M. J.,
    2. Subramanian V. S.,
    3. Said H. M.,
    4. Marchant J. S.
    (2003) Polarized expression of members of the solute carrier SLC19A gene family of water soluble multivitamin transporters: implications for physiological function. Biochem. J. 376:43–48.
    OpenUrlCrossRefMedlineWeb of Science
  16. ↵
    1. Subramanian V. S.,
    2. Marchant J. S.,
    3. Boulware M. J.,
    4. Said H. M.
    (2004) A C-terminal region dictates the apical plasma membrane targeting of the human sodium-dependent vitamin C transporter-1 in polarized epithelia. J. Biol. Chem. 279:27719–27728.
    OpenUrlAbstract/FREE Full Text
  17. ↵
    1. Subramanian V. S.,
    2. Marchant J. S.,
    3. Parker I.,
    4. Said H. M.
    (2003) Cell biology of the human thiamine transporter-1 (hTHTR1): intracellular trafficking and membrane targeting mechanisms. J. Biol. Chem. 278:3976–3984.
    OpenUrlAbstract/FREE Full Text
  18. ↵
    1. Said H. M.,
    2. Balamurugan K.,
    3. Subramanian V. S.,
    4. Marchant J. S.
    (2004) Expression and functional contribution of the human thiamin transporter-2 (hTHTR-2) in thiamin absorption in the human intestine. Am. J. Physiol. Gastrointest. Liver Physiol. 286:G491–G498.
    OpenUrlAbstract/FREE Full Text
  19. ↵
    1. Marchant J. S.,
    2. Subramanian V. S.,
    3. Parker I.,
    4. Said H. M.
    (2002) Intracellular trafficking and membrane targeting mechanisms of the human reduced folate carrier in mammalian epithelial cells. J. Biol. Chem. 277:33325–33333.
    OpenUrlAbstract/FREE Full Text
  20. ↵
    1. Said H. M.,
    2. Reidling J. C.,
    3. Ortiz A.
    (2002) Cellular and molecular aspects of thiamin uptake by human liver cells: studies with cultured HepG2 cells. Biochim. Biophys. Acta 1567:106–112.
    OpenUrlMedline
  21. ↵
    1. Subramanian V. S.,
    2. Marchant J. S.,
    3. Said H. M.
    (2006) Targeting and trafficking of the human thiamine transporter-2 in epithelial cells. J. Biol. Chem. 281:5233–5245.
    OpenUrlAbstract/FREE Full Text
  22. ↵
    1. Scharfe C.,
    2. Hauschild M.,
    3. Klopstock T.,
    (2000) A novel mutation in the thiamine responsive megaloblastic anemia gene SLC19A2 in a patient with deficiency of respiratory chain complex I. J. Med. Genet. 37:669–673.
    OpenUrlAbstract/FREE Full Text
  23. ↵
    1. Martin M. G.,
    2. Lostao M. P.,
    3. Turk E.,
    4. Lam J.,
    5. Wright E. M.,
    6. Kreman M.
    (1997) Compound missense mutations in the sodium/D-glucose cotransporter result in trafficking defects. Gastroenterol. 112:1206–1212.
    OpenUrlCrossRefMedlineWeb of Science
  24. ↵
    1. Stein M.-P.,
    2. Wandinger-Ness A.,
    3. Roitbak T.
    (2002) Altered trafficking and epithelial cell polarity in disease. Trends Cell Biol. 12:374–381.
    OpenUrlCrossRefMedlineWeb of Science
View Abstract
Back to top

July 2007

Volume: 113 Issue: 2

Clinical Science: 113 (2)

Actions

Email
Targeting and intracellular trafficking of clinically relevant hTHTR1 mutations in human cell lines
Veedamali S. Subramanian, Jonathan S. Marchant, Hamid M. Said
Clinical Science Jul 2007, 113 (2) 93-102; DOI: 10.1042/CS20060331
Permalink: Copy
del.icio.us logo Digg logo Reddit logo Technorati logo Twitter logo CiteULike logo Facebook logo Mendeley logo
Citation Tools
Print
Alerts
Request Permissions
Save to my folders

View Full PDF

Jump To

Keywords

cellular targeting
human thiamine transporter (hTHTR)
micronutrient
thiamine
thiamine-responsive megaloblastic anaemia (TRMA)
transporter
vitamin