Departamento de Biología Celular, Facultad de Ciencias Biológicas, Center for Advanced Microscopy, CMA BIO BIO, Universidad de Concepción, Concepción, Chile
Address correspondence and reprint requests to Francisco Nualart, PhD, Laboratorio de Neurobiología Celular, Departamento de Biología Celular, Facultad de Ciencias Biológicas, Universidad de Concepción. Casilla 160-C, Chile. E-mail: email@example.com
In vitro and in vivo studies suggest that the basolateral membrane of choroid plexus cells, which is in contact with blood vessels, is involved in the uptake of the reduced form of vitamin C, ascorbic acid (AA), through the sodium-vitamin C cotransporter, (SVCT2). Moreover, very low levels of vitamin C were observed in the brains of SVCT2-null mice. The oxidized form of vitamin C, dehydroascorbic acid (DHA), is incorporated through the facilitative glucose transporters (GLUTs). In this study, the contribution of SVCT2 and GLUT1 to vitamin C uptake in human choroid plexus papilloma (HCPP) cells in culture was examined. Both the functional activity and the kinetic parameters of GLUT1 and SVCT2 in cells isolated from HCPP were observed. Finally, DHA uptake by GLUT1 in choroid plexus cells was assessed in the presence of phorbol-12-myristate-13-acetate (PMA)-activated human neutrophils. A marked increase in vitamin C uptake by choroid plexus cells was observed that was associated with superoxide generation and vitamin C oxidation (bystander effect). Thus, vitamin C can be incorporated by epithelial choroid plexus papilloma cells using the basolateral polarization of SVCT2 and GLUT1. This mechanism may be amplified with neutrophil infiltration (inflammation) of choroid plexus tumors.
In choroid plexus papilloma cells, the vitamin C transporters SVCT2 and GLUT1 are polarized to the basolateral epithelial membrane, where SVCT2 is essential for AA flux from the blood vessels into the brain. However, neutrophils, attracted by inflammation or the tumor microenvironment, can oxidize extracellular AA to DHA, thereby enabling its uptake through GLUT1. For the first time, we show the in vivo and in vitro basolateral co-distribution of functional SVCT2 and GLUT1 in epithelial cells. We postulate that patients with choroid plexus papillomas may continue to transport vitamin C from the blood to CSF. However, increased transport of oxidized vitamin C could generate pro-oxidative conditions that may help control tumor growth.
Within the ventricular cavities of the brain, the choroid plexus synthesizes and secretes cerebrospinal fluid (CSF) (Cserr 1971). It is formed from a single cell layer of epithelial (ependymal) cells surrounded by a highly vascularized connective tissue. Ependymal cells are joined by tight junctions in the apical membrane, which constitutes the blood-CSF barrier. Choroid plexus tumors comprise 0.4–0.6% of all brain tumors (Gopal et al. 2008). Most cases occur in children less than 2 years of age (Matson and Crofton 1960; Laurence 1979; Johnson 1989; D'Addario et al. 1998), but the number of adults with this type of tumors might be underestimated (McGirr et al. 1988; Wrede et al. 2005). Within this group of tumors, there are benign and malignant variants, typically classified as human choroid plexus papilloma (HCPP) and choroid plexus carcinoma (CPC), respectively. HCPP are benign intraventricular papillary neoplasms derived from choroidal plexus epithelium and formed by a delicate fibrovascular connective tissue with many projections covered by a single layer of columnar epithelial cells. HCPP cells are similar to non-neoplastic cells in that they express tight junction-associated proteins, preserving the CSF-brain barrier (Rickert and Paulus 2001). Immunohistochemical studies have shown that transthyretin (TTR) (pre-albumin) appears to be a useful marker for HCPP (Ang et al. 1990), and other antigenic determinants include enolase neuronal specific (NSE), cytokeratins, and vimentin, which are expressed by virtually all HCPP (Lach et al. 1993; Rickert and Paulus 2001).
Glucose utilization by tumor cells is greater than benign cells, suggesting increased glucose transport across the plasma membrane mediated through increased expression of facilitative glucose transporters (GLUTs) (Godoy et al. 2006). Fourteen members of the GLUT family have been described as: GLUT1 to GLUT12, GLUT14, and a proton-coupled myoinositol transporter (HMIT or GLUT13) (Joost and Thorens 2001). Among all of the isoforms, only GLUT1–5 have been functionally characterized in detail (Mueckler et al. 1985; Kayano et al. 1988, 1990; Fukumoto et al. 1989; Nualart et al. 1999; Watanabe et al. 1999) as GLUT6–12 were only recently identified using homology searches of EST databases (Doege et al. 2001; Nualart et al. 2009). GLUT14 appears to be a duplication of the GLUT3 gene (Wu and Freeze 2002). GLUT1, GLUT3, and GLUT4 transport dehydroascorbic acid (DHA) (Vera et al. 1993; Rumsey et al. 1997, 2000), the oxidized form of vitamin C and one of the two major water soluble antioxidants in human cells (Vera et al. 1993; Maulen et al. 2003; Nualart et al. 2003).
Although nervous tissue has been shown to attain vitamin C concentrations that rank among the highest of human tissue (Kratzing and Kelly 1982; Kratzing et al. 1982; Milby et al. 1982a,b), there is a considerable lack of information related to the expression and function of glucose and vitamin C transporters in human choroid plexus, and to our knowledge, no information regarding their expression and function in choroid plexus tumors. Vitamin C can exist in two biological forms, the oxidized form, DHA, and the reduced form, ascorbic acid (AA). As previously mentioned, DHA uptake is mediated by GLUT1, GLUT3, and GLUT4 (Vera et al. 1993; Rumsey et al. 2000) whereas high affinity sodium-dependent l-AA transport is mediated by the sodium-ascorbic acid cotransporters (SVCTs), SVCT1 and SVCT2 (Faaland et al. 1998; Daruwala et al. 1999; Rajan et al. 1999; Tsukaguchi et al. 1999). SVCT1 is a 604 amino acid protein mainly detected in the liver, ovary, prostate, small intestine, and kidney (Daruwala et al. 1999; Wang et al. 1999, 2000). SVCT2 is a 592 amino acid protein that shares 65% homology with SVCT1 and has been detected mainly in the brain and eyes (Rajan et al. 1999; Tsukaguchi et al. 1999). However, in situ hybridization analysis indicates that normal cells of the choroid plexus express SVCT2 mRNA (Tsukaguchi et al. 1999). Consistent with this observation, kinetic analyses revealed that isolated choroid plexus cells transport AA (Spector and Lorenzo 1973). However, the expression of SVCT2 protein within choroid plexus has yet to be determined. In addition, choroid plexus cells express the facilitative glucose transporter, GLUT1 (Dwyer and Pardridge 1993; Kumagai et al. 1994; Kurosaki et al. 1995; Cornford et al. 1998). Over-expression of GLUT(s), specifically GLUT1, by tumor cells has been associated with high levels of glucose and DHA uptake (Spielholz et al. 1997), which is generated through extracellular oxidation of AA by superoxides produced by stromal cells, a mechanism known as the ‘bystander effect’ (Nualart et al. 2003).
In this study, the expression and function of glucose and vitamin C transporters in HCPP cells was determined. Our results indicated that HCPP cells express GLUT1 and transport glucose with high efficiency. Moreover, HCPP cells transported AA through a mechanism that involves SVCT2, DHA, and GLUT1 participation through a bystander effect in the presence of a reactive environment.
Materials and methods
Primary cultures of HCPP cells
HCPP tissues were obtained from the Department of Pathology at the University of Concepción (UCO) in accordance with the accepted standards of the ethics committee on the use of human specimens at UCO 2012101H. Briefly, tissue specimens were digested for 10 min with trypsin 0.35% (w/v) diluted in 0.1 M phosphate buffer (pH 7.4) and homogenized using a fire-polished Pasteur pipette. The digested cell extract was plated at a density of 300 000 cells/mL and cultured in Dulbecco's minimal essential medium F12 media (Gibco, Rockville, MD, USA) supplemented with 10% bovine fetal serum (Gibco), 2 mM l-glutamine (Nalgene, Rochester, NY, USA), 0.2 mM insulin (Sigma, St. Louis, MO, USA) and 1% penicillin/streptomycin (Gibco). Cell cultures were maintained in 5% CO2 in a humidified environment at 37°C. The isolated cells were cultured for no more than six passages.
HCPP specimens were fixed by immersion in Bouin's solution and embedded in paraffin (Nualart et al. 2012b). Tissue sections (7 μm) were obtained and mounted on poly-l-lysine-coated glass slides. For immunocytochemistry, the HCPP cells were fixed with 4% paraformaldehyde in Phosphate buffered saline (PBS) or Bouin′s solution for 30 min at 4°C, washed with PBS, and incubated in PBS containing 1% bovine serum albumin and 0.2% Triton X-100 for 5 min at 22°C. HCPP tissue sections or cells were next incubated at 22°C overnight with primary antibodies, including those specific for TTR (Dako, Carpinteria, CA, USA), a marker of ependymal-derived tumor epithelial cells, GLUT1–5 (Alpha Diagnostic, San Antonio, TX, USA); (Godoy et al. 2006), SVCT1 (D-19; Santa Cruz Biotechnology, Palo Alto, CA, USA), SVCT2 (S19; Santa Cruz Biotechnology) and anti-ZO-1 (Santa Cruz Biotechnology); (Garcia et al. 2005; Nualart et al. 2012b). The reaction was detected using a secondary anti-rabbit, mouse, or goat IgG antibody and a PAP complex. The enzymatic activity was developed using diaminobenzidine and H2O2 according to standard procedures (Godoy et al. 2006; Castro et al. 2008). For confocal laser scanning microscopy (CLSM), tissue sections were incubated for 2 h at 22°C with Cy2 or Cy3-conjugated affinity purified donkey anti-rabbit IgG (1 : 200; Jackson ImmunoResearch, West Grove, PA, USA) or with anti-goat IgG (1 : 150; Jackson ImmunoResearch). Tissue sections were treated with propidium iodine and RNase to stain the nuclei. Negative controls for GLUT(s) and SVCT(s) were provided using primary antibodies pre-absorbed with the respective peptides or using pre-immune serum (Nualart et al. 2012a). Fluorescence was detected with a confocal laser scanning microscope (Zeiss LSM 780; Carl Zeiss, Munich, Germany).
Poly (A+) RNA was isolated from HCPP cells using the Oligotex direct kit (Qiagen, Valencia, CA, USA). The RNA (0.5–1 μg) was incubated in 20 μL of a reaction mix containing 10 mM Tris pH 8.3, 50 mM KCl, 5 mM MgCl2, RNase inhibitor 20 U, 1 mM dNTPs, 2.5 μM random hexanucleotides, and 50 units MuLV reverse transcriptase (Perkin Elmer, Boston, MA, USA) for 10 min at 23°C followed by 30 min at 42°C and 5 min at 94°C. Parallel reactions were performed in the absence of reverse transcriptase to control for the presence of contaminant DNA. For amplification, a cDNA aliquot was added to a 12 μL mix containing 20 mM Tris, pH 8.4, 50 mM KCl, 1.6 mM MgCl2, 0.4 mM dNTPs, 0.04 units Taq DNA polymerase (Gibco BRL), and 0.4 μM primers. The mix was incubated at 94°C for 5 min, 94°C for 30 s, 55°C for 90 s, and 72°C for 135 s for 35 cycles. PCR products were separated using 1.2% agarose gel electrophoresis and visualized using ethidium bromide. Primer sequences for amplification included the following: human SVCT1, forward primer 5′-GGTGCCAAATATGAGAGGG-3′ and reverse primer 5′-GCCCCTGAACACCTCTCATA-3′ (expected product size: 450 bp); human SVCT2, forward primer 5′- TTCTGTGTGGGAATCACTAC-3′ and reverse primer 5′-ACCAGAGAGGCCAATTAGGG-3′ (expected product size: 353 bp); bovine SVCT2, forward primer 5′-ACGTTTGGATGCAGGTTACCC-3′ and reverse primer 5′-TGAAGCAGAGCAGCCAGGATAC-3′ (expected product size: 517 bp); bovine TTR, forward primer 5′-CTCGCTGGACTGGTGTTTGT-3′ and reverse primer 5′-GGGGAGATGCCAAGTGACT-3′ (expected product size: 287 bp); human GLUT1, forward primer 5′-TGAACCTGCTCCCCTTC-3′ and reverse primer 5′-GCAGCTTCTTTAGCACA-3′ (expected product size: 399 bp); human GLUT2, forward primer 5′-CCACAGGTAATAATATC-3′ and reverse primer 5′-CTCGCACACCAGACAGG-3′ (expected product size: 583 bp); human GLUT3, forward primer 5′-AAGCATAACTATAATGG-3′ and reverse primer 5′-GGTCTCCTTAGCAGGCT-3′ (expected product size: 416 bp); human GLUT4, forward primer 5′-CAGAAGGTGATTGAACA-3′ and reverse primer 5′-CAGGTAGCACTGTGAGG-3′ (expected product size: 493 bp); and human GLUT5, forward primer 5′-GAATTCATGGAAGACTT-3′ and reverse primer 5′-GCCATCTACGTTTGCAA-3′ (expected product size: 393 bp); β-actin, forward primer 5′-GCTCGTCGTCGACAACGGCTC-3′ and reverse primer 5′-CAAACATGATCTGGGTCATCTTCTC-3′ (expected product size: 353 bp). PCR conditions for each set of primers were optimized using RNA extracted from human GLUT1–5 sequences cloned in pcDNA3 plasmids.
Primary cultures of HCPP cells grown in monolayers were used for kinetic assays. Cell cultures were carefully selected under the microscope to ensure that only plates showing uniformly growing cells were used. All uptake experiments were performed in six-well plates containing approximately 400 000 cells/well. Briefly, HCPP cells were incubated in incubation buffer (15 mM Hepes, 135 mM NaCl, 5 mM KCl, 1.8 mM CaCl2, 0.8 mM MgCl2) for 10 min at 22°C. Glucose uptake assays were performed in a final volume of 0.5 mL of incubation buffer containing 0.1 μCi 2-deoxi-d-[1,2-(N)3H] glucose (specific activity 26.3 Ci/mmol; DuPont NEN, Boston, MA, USA) and 0.2–10 mM 2-deoxy-d-glucose (2-DOG) (Garcia et al. 2003). AA uptake was performed using 0.1–0.4 μCi l-14C-l-AA (specific activity 8.2 mCi/mmol; DuPont NEN) in incubation buffer containing 0.1 mM dithiothreitol (DTT). Transport of AA in the absence of sodium ions was accomplished by replacing the NaCl in the incubation buffer with choline chloride. Different preparations of AA were oxidized to DHA by addition of 0.02 units of ascorbate oxidase (Sigma-Aldrich, St. Louis, MO, USA) per ml of 1 μM AA to a final concentration of 100 μM. The affinity constant (Km) was calculated using the Lineweaver-Burk analysis. Data represent means ± SD of three experiments with each determination performed in triplicate. As a control, inhibition of the glucose transporter activity in HCPP cells was performed by pre-incubating these cells with cytochalasin B (Sigma-Aldrich) for 10 min prior to analysis of glucose transport.
These assays were performed according to previous studies (Nualart et al. 2003). Briefly, polymorphonuclear granulocytes were obtained using Polymorphoprep solution (Axis Shield PoC AS, Oslo, Norway) from the whole blood of the same individual in each experiment. The isolated human neutrophils were maintained in Iscove's Modified Dulbecco's Media supplemented with 10% Fetal bovine serum. For the co-culture experiments, HCPP cells were grown on six-well plates (2 × 105 cells/well), and after adherence, 1 × 105 neutrophils were per well. The co-cultures (HCPP + neutrophils) were activated using 1 μM phorbol-12-myristate-13-acetate (PMA), and radiolabeled AA was added to the wells for the uptake assays. After uptake, the adherent cells were washed, isolated by scraping, and processed for scintillation counting. As a control, inhibition of the glucose transporter activity in HCPP cells was performed by pre-incubating the cells with cytochalasin B (cyt B) for 15 min before the glucose transport assay. Superoxide dismutase (400 U) or catalase (130 U) (Sigma-Aldrich) enzymes were utilized to confirm neutrophil production of superoxide anions or H2O2, respectively.
All values are expressed as the means ± SEM. Statistical comparisons between groups were performed using one-way anova followed by Bonferroni's multiple comparison test. All analyses were performed using GraphPad Prism 4.0 Software (GraphPad Software Inc., San Diego, CA, USA).
Isolated cells from normal choroid plexus maintain GLUT1 and SVCT2 expression
To standardize the detection of vitamin C and GLUTs in choroid plexus cells in situ and in vitro, their expression in bovine brain tissue was initially analyzed. The bovine plexus presents a high degree of cellularity (Fig. 1a and b), which maintains polarization when isolated, as can be observed with SEM (Fig. 1a and b). Immunohistochemical analysis showed basolateral polarization of GLUT1 (Fig. 1f–h). SVCT2 polarization was less evident, as observed by its apparent intracellular localization (Fig. 1c–e). RT-PCR analysis of isolated cells confirmed expression of TTR (Fig. 1i), a specific marker for ependymal-derived tumor epithelial cells of the choroid plexus, as well as SVCT2 (Fig. 1j) and GLUT1 (data not shown). In isolated cells cultured at a high density, intracellular localization of TTR and SVCT2 (Fig. 1k–l), and basolateral polarization of GLUT1 (Fig. 1m) were observed. These results confirmed that bovine choroid plexus cells maintain the expression of GLUT1 and SVCT2 expression in vitro.
Expression of GLUT1 and SVCT2 in HCPP tissues
Magnetic resonance image (MRI) revealed a lobular and irregular mass located at the left lateral ventricle that distorted the shape of its horn in a 22-year-old patient, which was subsequently diagnosed as HCPP (Fig. 2a). Resected tissue sections were stained using Gallego's stain to analyze tumor tissue architecture and cell morphology (Fig. 2b and c). As expected, the HCPP showed an extensive papillary growth (Fig. 2b), with multiple papillae covered by single layer of tumor epithelial cells protruding to the center of the lateral ventricle (Fig. 2c). Immunofluorescence analysis of TTR, (Fig. 1d–f) revealed its cytoplasmic localization within tumor epithelial cells (Fig. 2f). In parallel, analysis of GLUT1–5, SVCT1 and SVCT2 expression in HCPP tissues by immunohistochemistry revealed positive immunostaining for only GLUT1 and SVCT2 (data not shown). To gain insight into the cellular distribution of GLUT1 and SVCT2 in HCPP tissues, expression of these proteins was analyzed using immunohistochemistry (Fig. 2g–i). Whereas GLUT1 was preferentially localized to the plasma membrane and the cytoplasm of HCPP cells (Fig. 2g), SVCT2 was primarily observed in the cytoplasm of HCPP cells (Fig. 2h). Co-localization of both proteins was evident (Fig. 2i, yellow immunostaining). These results indicated that HCPP cells have the potential to incorporate and metabolize glucose as well as AA and DHA.
Primary cultures of HCPP cells maintain in vitro expression of GLUT1 and SVCT2
To date, the ability of HCPP cells to incorporate glucose or vitamin C has not been assessed. To gain insights into the molecular mechanisms that mediate glucose or vitamin C transport in HCPP cells, we developed protocols to isolate and culture HCPP cells using proteolytic digestion (Fig. 3). Immunocytochemistry and differential interference contrast imaging revealed that more than 95% of the isolated cells expressed cytoplasmic TTR when cultured in vitro (Fig. 3a–c), confirming the identity of the HCPP cell cultures in vitro. Consistent with the in vivo data, HCPP cells expressed GLUT1 (Fig. 3d and g), but not GLUT2, 3, 4, 5 (data not shown). In addition, HCPP cells expressed SVCT2 (Fig. 3e and h, lanes 1–4), but not SVCT1 (Fig. 3h, lanes 5–7). Furthermore, ZO-1 expression was observed in HCPP cells (Fig. 3f). RT-PCR analysis revealed the absence of an amplification product in samples in which the cDNA synthesis step was performed as well as in the absence of reverse transcriptase, indicating the absence of contamination with genomic DNA (data not shown). These results confirmed expression of GLUT1 and SVCT2 in HCPP cells and validated their potential to incorporate glucose and vitamin C in vitro.
Functional characterization of glucose transport in HCPP cells
GLUT1 function was validated in primary cultures of HCPP cells by measuring the cellular transport of 2-DOG and 3-O-Methylglucose (3-OMG) (Fig. 4). Time course analyses of 2-DOG (0.2 mM, Fig. 4a) and 3-OMG (0.2 and 30 mM, Fig. 4e and f) indicated that HCPP cells take up glucose at a constant rate of approximately 1.4 nmol/106 cells/min for at least 1.5 min when using 2-DOG (Fig. 4a) and at a rate of approximately 0.05 and 45 nmol/106 cells/min when using 0.2 and 30 mM 3-OMG, respectively (Fig. 4e and f). Dose-response experiments showed that glucose transport approached saturation at approximately 6 mM when using 2-DOG (Fig. 4b) and approximately 10 mM when using 3-OMG (Fig. 4g) in HCPP cells. Analysis of the dose-response data (data from panels b and g) using the Lineweaver–Burk equation generated a straight line that is indicative of the presence of a single functional component with an apparent Km 4 mM and Vmax of 120 nmol/106 cells/min for 2-DOG (Fig. 4c) and an apparent Km of 5 mM and Vmax of 14 nmol/106 cells/min for 3-OMG (Fig. 4h). To confirm participation of GLUTs, various concentrations of cytochalasin B (0.005–30 μM), a competitive inhibitor of glucose transport (King et al. 1991), was utilized to inhibit 2-DOG transport (Fig. 4d, white circles). Transport of 2-DOG was inhibited in a dose-dependent manner by cytochalasin B, with an IC50 of approximately 0.1 μM. To control for the effect of cytochalasin B on cellular actin microfilaments, the effect of cytochalasin E, another member of the cytochalasin group that has similar effects on actin microfilaments but does not inhibit glucose transport, was also determined. As expected, incubation with cytochalasin E did not alter 2-DOG transport in HCPP cells (Fig. 4d, black circles). Taken together, these results indicated that HCPP cells incorporate glucose through GLUT1.
Functional characterization of vitamin C transport in HCPP cells
The capacity of HCPP cells to transport AA and DHA in vitro was next determined. AA transport assays confirmed the previous RT-PCR and immunostaining analyses indicating that primary cultures of HCPP cells expressed an AA transporter with functional properties similar to those of SVCT2 (Fig. 5). Time course analysis of AA uptake revealed that the HCPP cells take up AA at a constant rate of 20 pmol/106 cells/min for at least 30 min (Fig. 5a, white circles), which was almost completely inhibited when sodium ions were replaced with choline chloride in the transport buffer (Fig. 5a, black circles). Dose-response analyses showed that transport of AA approached saturation at 40–60 μM AA (Fig. 5b), and analysis of the transport data (from panel b) using the Lineweaver–Burk equation generated a straight line that is indicative of the presence of a single functional component with an apparent Km of 42 μM and a Vmax of 27 pmol/106 cells/min (Fig. 5c).
DHA transport was also analyzed in primary cultures of HCPP cells. As shown in Fig. 5d, HCPP cells incorporated DHA at a rate that was fivefold higher at 5 min than observed for AA (Fig. 5d). DHA uptake analyses revealed that HCPP cells take up DHA at a constant rate of 0.7 pmol/106 cells/min for at least 2 min (Fig. 5d, white triangles), and dose-response analyses indicated that saturation at a concentration of approximately 2 mM DHA (Fig. 5e). Analysis of the transport data (from panel e) using the Lineweaver–Burk equation generated a straight line that is indicative of the presence of a single functional component with an apparent Km of 1.9 mM and a Vmax of 18 nmol/106 cells/min (Fig. 5f). As expected, DHA transport was inhibited in a dose-dependent manner by cytochalasin B (Fig. 5g, black circles), with an IC50 of 0.08 μM, but not by cytochalasin E (Fig. 5g, white circles). Moreover, DHA transport was inhibited in a dose-dependent manner by increasing concentrations of 2-DOG with an IC50 of approximately 1 mM (Fig. 5h). Taken together these results indicate that HCPP cells transport AA using SVCT2 and DHA using GLUT1.
HCPP cells accumulate vitamin C through a bystander effect
The validity of the ‘bystander effect’ concept was examined in an in vitro model consisting of activated human neutrophils (the reactive cells) co-cultured with primary cultures of HCPP cells (the bystander cells) (Nualart et al. 2003). As depicted in Fig. 6a, adherent HCPP cell monolayers can be easily separated from human neutrophils that grow in suspension, permitting the determination of the content of vitamin C (AA) in each cell type separately (Fig. 6a). As expected, uptake of AA in untreated HCPP cells incubated alone was observed because of the presence of the AA transporter, SVCT2, in these cells (Fig. 6b, white circles). Treatment of HCPP cells with PMA (Fig. 6b, black circles) or co-culture with non-activated human neutrophils (Fig. 6b, white triangles) did not alter AA uptake. However, HCPP cells incubated in the presence of AA and co-cultured with PMA-activated human neutrophils showed a considerable increase in vitamin C uptake (Fig. 6b, black triangles). HCPP cells contained more than 4 nmol of vitamin C per million cells, an amount 15-fold higher than that accumulated by PMA-treated HCPP cells incubated alone (compare black circles with black triangles in panel b). This uptake of vitamin C by stimulated HCPP cells (HCPP cells in presence of AA and PMA-activated human neutrophils) was completely inhibited by cytochalasin B but not by cytochalasin E, confirming the participation of the glucose transporter in this process (Fig. 6c). HCPP cells did not produce superoxide anions when cultured under similar conditions (data not shown). In parallel experiments, superoxide dismutase but not catalase abolished, up to at least 50%, of the vitamin C uptake by stimulated HCPP cells, indicating that superoxide production by human neutrophils is central to vitamin C uptake by the bystander cells (Fig. 6d). Replacing the sodium chloride in the medium with choline chloride did not affect the uptake of vitamin C by stimulated HCPP cells (data not shown). Taken together, these results indicate that HCPP cells transport the DHA that is generated by oxidation of AA by locally activated neutrophil cells (Fig 6a). Moreover, the inhibitory effect of cytochalasin B, the sodium independence of the overall uptake process, and the expression pattern of GLUTs in HCPP cells, are consistent with the bystander cells taking up DHA through GLUT1.
Choroid plexus tumors are most common in the first year of life and are quite large at the time of diagnosis. Surgical resection is, until now, the treatment of choice for this type of tumor, and many children require diversion of the CSF despite removal of the tumor. Unfortunately, little is known about the biology of this type of neoplasm. In this study, the expression and function of GLUT1 and SVCT2 was observed in bovine and human choroid plexus tissues and in isolated bovine and human HCPP cells. In situ analysis of GLUT1 detected its basolateral polarization. In addition, primary cultures of HCPP cells maintained the capacity to transport glucose and DHA in vitro. Distribution of SVCT2 within the cytoplasm and cellular membrane was also maintained in vitro. In addition, the kinetic parameters of glucose and DHA transport were determined.
Studies of rabbit choroid plexus four decades ago defined active and saturable ascorbate transport, reporting an apparent affinity constant of 44 μM (Spector and Lorenzo 1973). Using polarized cultures of porcine epithelial cells from the choroid plexus, an affinity constant of 67 μM was determined for the transport of ascorbate (Hakvoort et al. 1998). In this study, we observed ascorbate transport in HCPP cells that was active and highly dependent on extracellular sodium concentration, with a Km of 42 μM, which is similar to what has been described for different types of cells that functionally express SVCT2 (Rajan et al. 1999; Castro et al. 2001; Maulen et al. 2003; May et al. 2006; Godoy et al. 2007; Qiu et al. 2007).
A second general mechanism of vitamin C capture consists of incorporating DHA through facilitative GLUT transporters (Vera et al. 1993; Rumsey et al. 2000). There is ample description of GLUT1 localization in capillaries that form the Blood-brain barrier (BBB) and in the basolateral membrane of epithelial cells of the choroid plexus. In rats and mice injected with radioactive DHA and subsequently analyzed with autoradiography, a rapid accumulation of radioactivity was observed throughout the brain (Hammarstrom 1966; Agus et al. 1997), suggesting that the BBB is involved in the entrance of oxidized vitamin C to the brain. However, radioactivity also remained in the blood for several hours, indicating that the radioactivity observed in the brain could be the vitamin C in the blood of the capillaries that form the BBB or to the DHA that enters endothelial cells of the BBB through GLUT1. To date, the transfer of DHA from the blood to the cerebral parenchyma through the BBB has not been demonstrated.
The choroid plexus can transport DHA through GLUT1 when AA is oxidized to DHA in the connective tissue of the choroid plexus or under physiopathological conditions (genesis of oxidants). However, capture of DHA in the choroid plexus by GLUT1 was not assessed prior to this study. HCPP cells captured DHA with a Km of 1.9 mM, similar to the Km reported in oocytes of Xenopus laevis microinjected with GLUT1 mRNA and in HL-60 cells (Vera et al. 1993; Rumsey et al. 1997). Surprisingly, in a competition assay with 5 mM 2-DOG, we observed a 40% rate of 50 μM DHA capture, demonstrating that it is feasible that plexus cells capture low concentrations of DHA even in the presence of high glucose concentrations.
DHA capture is also favored when local increases in DHA concentration occur. During oxidative burst, neutrophils produce large quantities of free radicals that lead to the oxidation of extracellular ascorbate (Fig. 7). Locally generated DHA can also be used by other adjacent cell types or by bystander cells. This has been observed in tumor cells that incorporate DHA produced during the activation of oxidant-generating cells (Nualart et al. 2003). We tested this effect using co-cultures of HCPP cells and PMA-activated neutrophils and found a pronounced increase in vitamin C capture. At 40 min, the vitamin C uptake rate was 10-fold higher than that of the HCPP control cells and was inhibited by 80% with cytochalasin B, a GLUT1 inhibitor, and by 50% with superoxide dismutase. These results indicate that extracellular oxidation of vitamin C by free radicals, including superoxide, and its subsequent uptake by HCPP cells (bystander cells) using the GLUT1 transporter could represent an alternative mechanism that favors the transport of vitamin C to CSF (Fig. 7).
DHA may also transduce intracellular signals once taken up by HCPP cells. In HeLa cells, DHA inhibits nuclear factor kappa B (NFκB) transcription factor activation that is mediated by pro-inflammatory cytokines, such as TNF-α (Carcamo et al. 2002). Inhibition of NFκB by DHA occurs through its suppression of free radicals, which can activate NFκB, and through direct inhibition of IκB kinase (Carcamo et al. 2002, 2004). Considering the tumoral origin of papilloma cells, the effect of DHA on the NFκB pathway could regulate apoptosis, favoring the survival of these tumor cells. Although the bystander effect has been proposed for tumor cells that over-express GLUTs (Nualart et al. 2003), the basolateral localization of GLUT1 in epithelial cells of the choroid plexus and the presence of defense cells in the stroma of the normal plexus suggest that the bystander effect may also occur under physiological conditions (Fig. 7).
In summary, RT-PCR, western blot and immunocytochemical analyses revealed the expression and localization of SVCT2 and GLUT1 vitamin C transporters in the choroid plexus (Fig. 7). Moreover, primary HCPP cell cultures incorporated vitamin C, ascorbate, and DHA. Increased DHA capture by HCPP cells was detected in the presence of activated neutrophils and as a result of the generation of superoxide by the neutrophils and oxidation of vitamin C within the medium, a mechanism that has been termed the bystander effect. Consequently, we have demonstrated the basolateral polarization of SVCT2 in choroid plexus epithelial cells, which may permit the incorporation of ascorbate from the blood to the plexus in the first step in the transfer and accumulation of vitamin C in the central nervous system. Furthermore, these cells capture DHA (Fig. 7), which is favored in inflammatory or tumor microenvironments.
Considering these results, patients with choroid plexus papilloma tumors may continue to transport vitamin C from the blood to CSF. However, increased transport of oxidized vitamin C (i.e., DHA) could generate pro-oxidative conditions that may help to control tumor growth.
The authors thank Ximena Koch from Universidad de Concepción for her technical support. This study was supported partially by grants from FONDECYT, Chile (1100396) and DIUC from Universidad de Concepción, Chile.
Conflict of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.