Expression of conventional and novel glucose transporters, GLUT1, -9, -10, and -12, in vascular smooth muscle cells

Abstract

Intimal hyperplasia is characterized by exaggerated proliferation of vascular smooth muscle cells (VSMCs). Enhanced VSMC growth is dependent on increased glucose uptake and metabolism. Facilitative glucose transporters (GLUTs) are comprised of conventional GLUT isoforms (GLUT1–5) and novel GLUT isoforms (GLUT6–14). Previous studies demonstrate that GLUT1 overexpression or GLUT10 downregulation contribute to phenotypic changes in VSMCs. To date, the expression profile of all 14 GLUT isoforms has not been fully examined in VSMCs. Using the proliferative and differentiated phenotypes of human aortic VSMCs, the present study has determined the relative abundance of GLUT1–14 mRNAs by quantitative real-time PCR analysis. Twelve GLUT mRNAs excluding GLUT7 and GLUT14 were detectable in VSMCs. In the proliferative phenotype, the relative abundance of key GLUT mRNAs was GLUT1 (∼43%) > GLUT10 (∼26%) > GLUT9 (∼13%) > GLUT12 (∼4%), whereas in the differentiated phenotype the relative abundance was GLUT10 (∼28%) > GLUT1 (∼25%) > GLUT12 (∼20%) > GLUT9 (∼14%), together constituting 86–87% of total GLUT transcripts. To confirm the expression of key GLUT proteins, immunoblot and immunocytochemical analyses were performed using GLUT isoform-specific primary antibodies. The protein bands characteristic of GLUT1, -9, -10, and -12 were detected in VSMCs in parallel with respective positive controls. In particular, GLUT1 protein expression showed different molecular forms representative of altered glycosylation. While GLUT1 protein displayed a predominant distribution in the plasma membrane, GLUT9, -10, and -12 proteins were mostly distributed in the intracellular compartments. The present study provides the first direct evidence for GLUT9 and GLUT12 expression in VSMCs in conjunction with the previously identified GLUT1 and GLUT10.

atherosclerosis, restenosis after angioplasty, and vein graft stenosis are characterized by adverse vascular events including intimal hyperplasia, which plays a critical role in luminal narrowing (45). Intimal hyperplasia results from phenotypic transition of vascular smooth muscle cells (VSMCs) from the contractile/differentiated state to the synthetic/proliferative state (37). Phenotypic switching of VSMCs to the synthetic state is facilitated by systemic and locally released trophic factors at the site of vascular lesion (37, 45). Importantly, growth factor-induced VSMC proliferation is dependent on multiple intracellular signaling events and metabolic pathways, including increased glucose consumption via glycolysis (7, 19, 20, 44, 52, 56). The initiating event toward increased glucose metabolism in proliferative VSMCs is the cellular uptake of glucose via facilitative glucose transporters. The facilitative glucose transporters are comprised of the long recognized conventional glucose transporters (GLUT1–5 isoforms) and the recently identified novel glucose transporters (GLUT6–14 isoforms; Ref. 24). Although GLUT1–5 and GLUT10 isoforms have been previously studied in VSMCs, the expression profile of all 14 GLUT isoforms (GLUT1–14) has not been fully examined.

Previous studies by several investigators have examined the mRNA expression profile of conventional glucose transporters (GLUT1–5) in arterial VSMCs from different species. Kaiser et al. (25) have shown that GLUT1 is the abundantly expressed GLUT transcript with undetectable levels of GLUT2–5 mRNAs in bovine aortic VSMCs. While GLUT4 mRNA is undetectable in bovine and rat aortic VSMCs (25, 31), the studies by Marcus et al. (34) demonstrate GLUT4 mRNA expression in smooth muscle cells from rat renal afferent microvessels. Recent studies by Liu et al. (30) reveal GLUT5 mRNA expression in rat aortic A10 VSMCs and rat aortic tissues. Thus the expression of GLUT1, -4, and -5 mRNAs has been reported in VSMCs, but the relative abundance of these GLUT transcripts remains unclear. Importantly, GLUT1 protein mediates glucose transport activity to promote glucose metabolism and VSMC proliferation, as evidenced in in vitro and in vivo studies by several investigators (1, 7, 19, 20, 25, 52). With regard to its functional relevance, Hall et al. (19) have shown that balloon injury to the rabbit carotid artery results in increased GLUT1 protein expression and glucose metabolism, thereby contributing to enhanced neointimal growth in the lesion site. In addition, smooth muscle-specific overexpression of GLUT1 in mice under the control of sm22α promoter accelerates proinflammatory lesion including VSMC proliferation upon injury to the femoral artery (1). Thus GLUT1 plays a major role in the development of intimal hyperplasia and proinflammatory lesion as revealed by studies involving animal models of arterial injury.

Of the novel glucose transporters (GLUT6–14), previous studies by Coucke et al. (8) demonstrate that loss of function mutation in GLUT10 gene results in arterial tortuosity syndrome, an autosomal recessive disorder characterized by tortuosity, aneurysm, and stenosis of large- and medium-sized arteries. In particular, GLUT10 mutation in arterial tortuosity syndrome patients results in profound downregulation of GLUT mRNA/protein in aortic VSMCs and disruption of elastic fibers in the arterial medial layer, with an accompanying upregulation of transforming growth factor-β (TGF-β) signaling presumably due to a decrease in intracellular glucose concentration (8). Recently, Lee et al. (28) have shown that GLUT10 expressed in VSMCs mediates the transport of oxidized form of vitamin C (l-dehydroascorbic acid) into the mitochondria, thereby protecting vascular cells from oxidative stress. Thus GLUT10 plays a pivotal role in the vasculature by maintaining vessel wall integrity via mechanisms beyond glucose transport. To date, the relative expression levels of other novel GLUT isoforms have not been examined in VSMCs.

The conventional and novel GLUT isoforms share common structural features that include the presence of 12 transmembrane domains, NH₂ and COOH termini facing the cytoplasmic side, and N-glycosylation sites (8, 28). Based on sequence similarity, these GLUT isoforms are categorized into three different classes such as class I, -II, and -III. Class I GLUT isoforms consist of GLUT1, -2, -3, -4, and -14. Class II isoforms consist of GLUT5, -7, -9, and -11, whereas class III isoforms consist of GLUT6, -8, -10, and -12, and -13 [H⁺/myo-inositol transporter (HMIT)]. Since the discovery of class II and class III GLUT isoforms, a number of studies have examined the expression profile and functional relevance of these novel GLUT isoforms (e.g., GLUT9 and GLUT12) in metabolically active tissues such as kidney, pancreatic islets, skeletal muscle, and/or adipose tissue (6, 14, 41, 50, 51). Notably, the 14 GLUT proteins (GLUT1–14) encoded by 14 different GLUT genes possess distinct affinities for different substrates (24). In addition to glucose, GLUT protein(s) mediate the transport of fructose, glucosamine, uric acid, l-dehydroascorbic acid, and/or myo-inositol (24), which have the potential to modulate the phenotypic characteristics of VSMCs. As an initial step, it is therefore important to determine the relative expression levels of all 14 GLUT isoforms in VSMCs.

The present study has examined the expression profile of facilitative glucose transporters, GLUT1 through GLUT14, in human aortic VSMCs under proliferative and differentiated states. To determine GLUT1–14 mRNA expression profile, quantitative real-time PCR (qPCR) analysis was performed using purified total RNA isolated from VSMCs. In parallel, the expression profile of sodium-dependent glucose transporters (17, 58), SGLT1 and SGLT2, was also determined. To confirm whether the abundantly expressed GLUT transcripts are translated to the respective GLUT proteins (GLUT1, -9, -10, and -12), immunoblotting and immunocytochemical methods were employed. Immunoblot analysis was performed using whole cell homogenates as well as subcellular fractions that included high-density microsomes (HDM), low-density microsomes (LDM), and plasma membrane (PM). The present findings demonstrate that, in addition to the previously identified GLUT1 and -10 isoforms, novel GLUT isoforms such as GLUT9 and -12 are also expressed in VSMCs at the level of mRNA and protein with distinct subcellular localization pattern.

MATERIALS AND METHODS

Materials.

The primary antibodies for GLUT1 (ab40084), GLUT10 (ab33245), and GLUT12 (ab75441) were purchased from Abcam (Cambridge, MA). The primary antibodies for GLUT9 (AP10091PU-N) and GLUT9 (BMP027) were purchased from Acris Antibodies (San Diego, CA) and MBL International (Woburn, MA), respectively. The primary antibodies for smooth muscle α-actin (ab5694) and β-actin (ab8227) were purchased from Abcam (Cambridge, MA). The primary antibodies for proliferative cell nuclear antigen, cyclin D1, and phospho-retinoblastoma protein were purchased from Cell Signaling Technology (Danvers, MA). The primary antibodies for α₁-Na⁺-K⁺-ATPase (ab7671) and sarco(endo)plasmic reticulum calcium ATPase 2 (SERCA2; ab3625) were purchased from Abcam (Cambridge, MA). The primary antibody for Golgin-97 (A-21270) was purchased from Life Technologies (Grand Island, NY). The horseradish peroxidase (HRP)-conjugated goat anti-mouse secondary antibody (sc-2005) and HRP-conjugated goat anti-rabbit secondary antibody (172–1019) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA) and Bio-Rad (Hercules, CA), respectively. Alexa Fluor 488 chicken anti-mouse IgG and Alexa Fluor 594 chicken anti-rabbit IgG were purchased from Life Technologies. Vectashield H-1200 was purchased from Vector Laboratories (Burlingame, CA). Human brain tissue membrane lysate (XBL-10166), human kidney tissue lysate (1305), and human adipose tissue lysate (XBL-11048) were purchased from ProSci (Poway, CA). Human artery tissue lysate (ab28992) and human artery tissue membrane lysate (ab28989) were purchased from Abcam. 2-deoxy-d-[2,6-³H]glucose (sp. activity: 30–60 Ci/mmol) was purchased from American Radiolabeled Chemicals (St. Louis, MO). All other chemicals were from Fisher Scientific (Fair Lawn, NJ) or Sigma Chemical (St. Louis, MO).

Cell culture and treatments.

Human aortic VSMCs were purchased from Lonza (Allendale, NJ). Medium 231 and smooth muscle growth supplement (SMGS) were purchased from Life Technologies. SMGS constituents and their final concentrations after addition to medium 231 were as follows: 4.9% FBS (vol/vol), 2 ng/ml human basic fibroblast growth factor, 0.5 ng/ml human epidermal growth factor, 5 ng/ml heparin, 5 μg/ml insulin, and 0.2 μg/ml BSA. VSMCs (passages 3–6) were maintained in medium 231 containing SMGS and antibiotic/antimycotic solution in a humidified atmosphere of 95% air-5% CO₂ at 37°C. After the attainment of confluence (∼6–7 days), VSMCs were trypsinized, centrifuged, and seeded onto petri dishes or multiwell plates, as described previously (59). Subconfluent VSMCs were maintained in culture in the presence or absence of serum (SMGS) for up to 96 h as described in the legends to the respective figures. For glucose uptake studies, VSMCs were subjected to treatments with 10–100 μM cytochalasin B or 20–200 μM phloretin.

Cell counts.

Subconfluent VSMCs (initial seeding density: 7.5 × 10⁴ cells/60-mm petri dish) were maintained in culture in the presence or absence of serum for 96 h. The cells were replenished with the respective fresh media every 48 h. After the 96-h incubation period, VSMCs were trypsinized and the changes in cell number were determined using Countess Counter (Life Technologies).

Total RNA extraction/quantification and assessment of RNA quality/integrity.

Subconfluent VSMCs were maintained in culture in the presence or absence of serum for up to 96 h. Subsequently, VSMCs were washed twice with ice-cold PBS and the total RNA was extracted using RNeasy mini kit (Qiagen, Valencia, CA). The isolated RNAs were treated with RNase-free DNase I (Qiagen) to remove contamination due to genomic DNA. One microliter each of the RNA samples was applied directly onto the system optics of NanoDrop 2000 spectrophotometer (Thermo Scientific, Wilmington, DE) for RNA quantification at 260 nm followed by assessment of its purity. The purity of the RNA samples was verified by measurement of absorbance ratios at 260/280 and 260/230 nm. The absorbance ratios of the RNA samples were >2.0 at 260/280 nm and at 260/230 nm, thereby precluding contaminants due to proteins and thiocyanates/organic compounds, respectively. The integrity of the RNA was verified by conventional agarose gel electrophoresis and also by Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA). In brief, total RNA samples were subjected to denaturing electrophoresis on a 2% agarose gel and ethidium bromide staining for visual inspection of the 28S and 18S ribosomal RNA bands. These bands were then quantified by densitometry. To assess RNA integrity using Agilent 2100 Bioanalyzer, total RNA samples were subjected to microchip-based electrophoretic separation followed by laser-induced fluorescence detection. The samples with 28S/18S rRNA ratio >1.8 and an RNA integrity number (RIN) > 8.0 were considered for further analysis, as described previously (15).

cDNA synthesis and qPCR.

Three-hundred nanograms of total RNA isolated from VSMCs were reverse transcribed to cDNA by Superscript First Strand RT-PCR system (Life Technologies) using oligo(dT) primers. qPCR analysis was performed using Applied Biosystems 7900HT Fast Real-Time PCR system and QuantiTect SYBR Green PCR kit (Qiagen) with the primer sets listed in Table 1. The primers used in qPCR were designed using algorithm Primer 3. A SYBR Green master mix of the following reaction components was prepared to the indicated end-concentration: 0.3 μl forward primer (0.3 μM), 0.3 μl reverse primer (0.3 μM), 5.0 μl 2× QuantiTect SYBR Green PCR master mix (Qiagen), 3.4 μl RNase-DNase free water, and 1 μl cDNA as PCR template to obtain a final reaction volume of 10 μl. qPCR thermal conditions were as follows: 95°C for 15 min, 40 cycles of 94°C for 15 s, 53°C for 30 s, and 72°C for 30 s, followed by a default melting curve program. The specificity of qPCR products was documented with Applied Biosystems disassociation curve analysis, which resulted in single product-specific melting temperatures. In addition, electrophoresis on a 2% agarose gel resulted in a single product with the desired length for a specific primer pair. As a positive control, human reference cDNA (Clontech, Mountain View, CA) was used to verify the amplified PCR products for the chosen genes that included 14 GLUT isoforms and 2 SGLT isoforms, β₂-microglobulin (B2M) and GAPDH.

Table 1.

Primer sequence used for qPCR analysis of glucose transporter isoforms in VSMCs

Gene	Forward Primer (5′-3′)	Reverse Primer (5′-3′)	Product Size, bp
GLUT1	CTGCTCATCAACCGCAAC	CTTCTTCTCCCGCATCATCT	129
GLUT2	TACATTGCGGACTTCTGTGG	AGACTTTCCTTTGGTTTCTGG	108
GLUT3	CAGCGAGACCCAGAGATG	TTGGAAAGAGCCGATTGTAG	233
GLUT4	TGGGCTTCTTCATCTTCACC	GTGCTGGGTTTCACCTCCT	121
GLUT5	CCTTTGGGTCATCCTTCCA	ACAGACCACAGCAACGTCAA	145
GLUT6	ATGCCCTGGTCTACACATCC	GAACACGGACCCAAACCA	104
GLUT7	TCGGTGCCTACAGTTTCATC	AATGCGGTTTATCTCCACAA	113
GLUT8	GCCAAGTTCAAGGACAGCA	GACCACACCTGACAAGACCA	129
GLUT9	CCTGTTTGGAGTGATTGTGGT	CTTGCCTCGTTGTGCTTCTC	105
GLUT10	TGGTCTTTGTCAGTGCCTTC	GCTCTTCCTCGTATCTCCAC	91
GLUT11	CGTGATGGGACAGGTGGT	GCTTTCAGGGAGCAGAGG	127
GLUT12	GGGCATCGTAAATCTCAACA	CCATAAATCACAGACTCATCCA	96
HMIT	GCAGCTCTCAGGCATTAACA	ACAAGCCAGACTCCCACAAG	147
GLUT14*	CCATGTGTCTGGAATTGGTG	AGCAACTGTGATGGCAAAGA	103
SGLT1	GTCATCTCCCTCCTCACCAA	TCCAGGTCAATACGCTCCTC	95
SGLT2	TACTTCGCCATTGTGCTGTT	CTCCCGTTCCTCCTTGCT	132
B2M	GGGTTTCATCCATCCGACA	ACACGGCAGGCATACTCATC	161
GAPDH	ACCCAGAAGACTGTGGATGG	AGTAGAGGCAGGGATGATGTT	84

qPCR, quantitative real-time PCR; VSMCs, vascular smooth muscle cells; GLUT, glucose transporter; SGLT, sodium-dependent glucose transporter; HMIT, H⁺/myo-inositol transporter; B2M, β₂-microglobulin. GLUT14

^*primer pair recognizes the long isoform of the transcript.

To determine the differences in GLUT isoform mRNA expression between the proliferative and differentiation phenotypes of VSMCs, the threshold cycle (C_T) values were corrected for amplification efficiency as described previously (15) and then normalized with internal control gene (B2M). To determine the amplification efficiency of each primer pair listed in Table 1, qPCR analysis was performed using human reference cDNA as template at five serial dilutions. Each dilution was done in triplicate. The C_T values thus obtained were then plotted against the logarithmic concentration of cDNA template. The efficiency of each primer pair was calculated using the following equation: E = 10 (^−1/slope). Notably, the efficiency of all primer pairs used in the present study was in the range of 1.8 to 2.1. All measurements were done in triplicate on four independent runs.

To determine the relative abundance of GLUT isoform mRNAs within the proliferative and differentiated phenotypes of VSMCs, each of the 12 detectable GLUT mRNAs was calculated as the percentage of total GLUT transcripts.

Preparation of VSMC homogenates and solubilization of proteins using urea solubilization buffer.

VSMCs were washed twice with ice-cold PBS. With the use of a cell scraper, the cells were collected in ice-cold PBS containing protease inhibitor cocktail (Thermo Scientific). The cell suspension was then centrifuged at 600 g for 10 min (4°C) to pellet the cells. The pelleted cells were homogenized in 250 μl of HES buffer (20 mM HEPES, 1 mM EDTA, and 250 mM sucrose, pH 7.4) containing protease inhibitor cocktail, as described previously with minor modifications (11, 14, 50). In particular, the cells were homogenized using an electric-powered instrument (model no. JR4000; Arrow Engineering, Hillside, NJ) and a prechilled glass-glass homogenizer with eight strokes three times by intermittently placing the glass tubes on ice. The homogenates were mixed with urea solubilization buffer (8 M urea, 50 mM Tris, 5% SDS, and 350 mM dithiothreitol) at a fixed ratio of 2.5:1, respectively. The above mixture was allowed to stand at room temperature for 1 h with intermittent vortexing every 10 min to ensure complete solubilization of proteins. Proteins were quantified using 2-D Quant kit (GE Healthcare).

Preparation of subcellular fractions from VSMCs and solubilization of proteins using urea solubilization buffer.

VSMC homogenates were prepared as described above and centrifuged at 600 g for 10 min at 4°C to remove nuclei, cell debris, and unbroken cells (36). Postnuclear supernatants thus obtained were subjected to differential centrifugation to obtain subcellular fractions that included PM, HDM, and LDM, as described previously with minor modifications (14, 26, 36, 50). In brief, the supernatants were centrifuged at 14,000 g for 15 min at 4°C to obtain the pellet (P1) and the supernatant (S1). The pellet (P1) was resuspended in HES buffer containing protease inhibitor cocktail and overlaid on an equal volume of 38% sucrose cushion, followed by centrifugation at 105,000 g (max) for 1 h at 4°C using TLS-55 swinging bucket rotor (Beckman Coulter, Brea, CA). The PM fraction, obtained at the interface of sample buffer and sucrose cushion, was suspended in HES buffer and then centrifuged at 50,000 g (max) for 30 min at 4°C. The pellet containing the PM fraction was resuspended in HES buffer using pestle motor (Kimble Chase, Vineland, NJ). The supernatant (S1) remaining after the 14,000 g spin was centrifuged at 50,000 g (max) for 30 min at 4°C to obtain the pellet (P2) and the supernatant (S2). The pellet (P2) containing the HDM fraction was resuspended in HES buffer using pestle motor. The supernatant (S2) was centrifuged at 225,000 g (max) for 75 min at 4°C to obtain the pellet (P3). The pellet (P3) containing the LDM fraction was resuspended in HES buffer using pestle motor. The PM, HDM, and LDM fractions resuspended in HES buffer were then solubilized using urea solubilization buffer, as described above. Proteins were quantified using 2-D Quant kit (GE Healthcare). Enrichment of membrane markers in the subcellular fractions was determined by immunoblot blot analysis using primary antibodies specific for Na⁺/K⁺-ATPase protein (PM fraction), SERCA2, an endoplasmic reticulum membrane protein (HDM fraction), and Golgin-97, a Golgi membrane protein (LDM fraction; Refs. 14, 46, 50).

Immunoblot analysis.

The solubilized proteins from VSMC homogenates (20 μg protein each) and subcellular fractions (4 μg protein each) were subjected to electrophoresis using precast 4–12% NuPage mini-gels (Life Technologies). The resolved proteins were then transferred to nitrocellulose membranes (Hybond C; GE Healthcare). Subsequently, the membranes were blocked in 5% nonfat milk and probed with the respective primary antibodies including anti-GLUT1, ab40084, 1:500; anti-GLUT9, AP10091PU-N, 1:500; anti-GLUT10, ab33245, 1:1,000; and anti-GLUT12, ab75441, 1:500. After extensive washes, the immunoreactivity was detected using HRP-conjugated goat anti-mouse secondary antibody (sc-2005, 1:2,000) or goat anti-rabbit secondary antibody (172–1019, 1:2,000; Bio-Rad) followed by enhanced chemiluminescence (Amersham Biosciences), as described previously (59). The protein bands were quantified using Bio-Rad GS-800 calibrated densitometer.

Immunocytochemistry and confocal microscopy.

Subconfluent VSMCs grown on glass coverslips were incubated in serum-free medium for 48 h, and the immunolocalization of different GLUT isoforms was analyzed. The cells were washed in PBS once and fixed with 4% paraformaldehyde in PBS for 10 min at room temperature (RT) followed by three washes in PBS. The cells were then incubated in buffer containing 1% BSA/0.3M glycine in 0.1% PBS-Tween for 1 h to permeabilize the cells and to block nonspecific protein-protein interactions. Next, the cells were incubated with the respective primary antibodies (anti-GLUT1, ab40084, 1:100; anti-GLUT9, BMP027, 1:50; anti-GLUT10, ab33245, 1:100; and anti-GLUT12, ab75441, 1:50) for 2 h at RT in buffer containing 1% BSA in 0.1% PBS-Tween, followed by three washes for 5 min each in the same buffer. The cells were counterstained with Alexa Fluor 488 chicken anti-mouse IgG or Alexa Fluor 594 chicken anti-rabbit IgG (1:500 dilution) for 1 h at RT followed by three washes. The coverslips were mounted on glass slides with Vectashield containing DAPI (Vector Labs) to stain the nucleus. Fluorescence images were taken in a Zeiss confocal microscope (LSM-510 Laser module) using Zen 2009 software.

Glucose uptake studies.

Subconfluent VSMCs were maintained in culture in the presence or absence of serum for 48 h. VSMCs were then washed twice with Krebs-Ringer-HEPES (KRH) buffer that consists of 105 mM NaCl, 5 mM KCl, 1.4 mM CaCl₂, 1 mM KH₂PO₄, 1.4 mM MgSO₄, 10 mM NaHCO₃, and 15 mM HEPES, pH 7.4. Next, the cells were incubated in 2 ml of KRH buffer containing unlabeled 2-deoxy-d-glucose (0.2 mM) and 2-deoxy-d-[2,6-³H]glucose (1 μCi/ml) for 30 min. The cells were then washed three times with ice-cold PBS and lysed with 1% SDS. Glucose uptake was determined using liquid scintillation counter (25, 29). Proteins were quantified using Biorad DC assay reagent. 2-deoxy-glucose uptake in VSMCs was calculated as picomoles per milligram protein per minute. To correct for noncarrier-mediated uptake, 2-deoxy-glucose uptake studies were performed in the absence or presence of 10–100 μM cytochalasin B that is known to inhibit glucose uptake by binding to the inner surface of GLUT1 (53). In parallel, 2-deoxy-glucose uptake studies were also performed in the absence or presence of 20–200 μM phloretin that is known to inhibit glucose uptake by binding to the external glucose binding site of GLUT1 (53).

Statistical analyses.

Results are expressed as means ± SE values. Statistical analyses of the data were performed by one-way repeated-measures ANOVA followed by Bonferroni t-test. Values of P < 0.05 were considered statistically significant.

RESULTS

Phenotypic switching of VSMCs from the proliferative to differentiated state.

VSMCs possess the unique property of phenotypic plasticity that allows reversible transition from the contractile phenotype to the synthetic phenotype upon arterial injury in vivo (37). Notably, under in vitro conditions, VSMCs in culture acquire the synthetic phenotype due to the presence of serum trophic factors in the culture medium (21, 37). To obtain synthetic/proliferative and contractile/differentiated phenotypes for the present study, subconfluent VSMCs were therefore maintained in culture in the presence and absence of serum trophic factors, as described in previous studies (21). As shown in Fig. 1A, serum deprivation for 96 h resulted in a significant decrease in VSMC proliferation by 76.6 ± 3.9%. The decrease in VSMC proliferation was associated with diminished expression levels of proliferative markers such as proliferative cell nuclear antigen, cyclin D1, and phospho-retinoblastoma protein (Fig. 1B). Importantly, the diminished expression of proliferative markers was accompanied by enhanced expression of the contractile protein, smooth muscle α-actin (Fig. 1B). The proliferative and differentiated phenotypes of VSMCs thus obtained were then used to determine the mRNA and protein expression profile of glucose transporters.

Fig. 1.

Effects of serum deprivation on vascular smooth muscle cell (VSMC) proliferation, proliferative markers, and contractile protein expression. Subconfluent VSMCs were maintained in culture in the presence and absence of serum trophic factors for 48 or 96 h as shown in A and B. A: cell number was determined using Countess automated cell counter. Data shown in the bar graph are the means ± SE values. *P < 0.05, compared with proliferative phenotype; n = 4. B: VSMC lysates were subjected to immunoblot analysis using primary antibodies specific for proliferative cell nuclear antigen (PCNA), cyclin D1, phospho-retinoblastoma protein (p-Rb), and smooth muscle (SM) α-actin. β-actin was used as an internal control. Data shown are representative of 3–4 separate experiments.

Integrity of RNAs isolated from the proliferative and differentiated phenotypes of VSMCs.

Subconfluent VSMCs were maintained in culture in the presence and absence of serum for 4 days to obtain the proliferative and differentiated phenotypes, respectively. Total RNAs isolated from VSMCs were then subjected to purification steps, as described under materials and methods. Previous studies by Pfaffl and co-workers (15) have shown that verification of RNA integrity is critically important for the qPCR analysis of target mRNAs. As an initial step toward quantifying the relative expression levels of GLUT mRNAs in VSMCs, we therefore verified the quality and integrity of total RNAs. Electrophoretic separation of purified total RNAs on a 2% agarose gel showed that the 28S/18S rRNA ratios were 1.87 ± 0.03 and 1.83 ± 0.03 for the proliferative and differentiated phenotypes, respectively (n = 4). In addition, electrophoretic separation of total RNA samples using the Agilent Bioanalyzer revealed that the 28S/18S rRNA peak ratio values (Fig. 2, A and B) were comparable to those observed using agarose gel electrophoresis. Furthermore, analysis of RNA samples using Bioanalyzer showed that the RIN values were 9.1 ± 0.2 and 8.2 ± 0.1 for the proliferative and differentiated phenotypes, respectively (n = 4). Together, these data demonstrate that the integrity of RNA is preserved (RIN > 8.0) during the extraction procedures. The isolated total RNA therefore served as an optimal template for the quantification of GLUT mRNAs in VSMCs.

Fig. 2.

Electropherogram of total RNAs isolated from the proliferative and differentiated phenotypes of VSMCs. Subconfluent VSMCs were maintained in culture in the presence or absence of serum trophic factors for 96 h. Total RNAs isolated from VSMCs were then purified and examined using Agilent 2100 Bioanalyzer, as described in materials and methods. Electrophoretic runs shown in A and B are representative of 4 separate experiments. FU on y-axis refers to fluorescence units, and [s] on x-axis refers to time in seconds.

Verification of amplified PCR products of glucose transporter primers using human reference cDNA.

Table 1 shows the list of primer pairs used for the detection of all 14 GLUT mRNAs, 2 SGLT mRNAs, B2M mRNA, and GAPDH mRNA. As described in materials and methods, qPCR analysis was performed using each of these primer pairs and human reference cDNA. The amplified PCR products were then run on a 2% agarose gel. Figure 3 shows the appearance of a single band for each gene at the expected product size. These data validate the use of chosen primer pairs for the detection and quantification of GLUT and SGLT mRNAs in VSMCs.

Fig. 3.

Verification of amplified PCR products of the primer pairs specific for GLUT1–14, SGLT1–2, B2M, and GAPDH. Quantitative PCR analysis was performed using human reference cDNA as the template and each of the primer pairs listed in Table 1. The amplified PCR products were then verified by agarose gel electrophoresis. Lane 1: Qiagen 50-bp DNA ladder; lane 2: GLUT1 (129 bp); lane 3: GLUT2 (108 bp); lane 4: GLUT3 (233 bp); lane 5: GLUT4 (121 bp); lane 6: GLUT5 (145 bp); lane 7: GLUT6 (104 bp); lane 8: GLUT7 (113 bp); lane 9: GLUT8 (129 bp); lane 10: GLUT9 (105 bp); lane 11: GLUT10 (91 bp); lane 12: GLUT11 (127 bp); lane 13: GLUT12 (96 bp); lane 14: HMIT (147 bp); lane 15: GLUT14 (103 bp); lane 16: SGLT1 (95 bp); lane 17: SGLT2 (132 bp); lane 18: B2M (161 bp); lane 19: GAPDH (84 bp); and lane 20: Qiagen 50-bp DNA ladder. GLUT, glucose transporter; SGLT, sodium-dependent glucose transporter; HMIT, H⁺/myo-inositol transporter; B2M, β₂-microglobulin.

Normalization of glucose transporter mRNAs in the proliferative and differentiated phenotypes of VSMCs.

Previous studies using 3T3 fibroblasts have shown that serum trophic factors enhance GAPDH mRNA but not B2M mRNA expression levels (48). In addition, Hastings et al. (22) have used B2M as an internal control for the qPCR analysis of several genes including differentiation genes in VSMCs. In the present study, we therefore verified the suitability of GAPDH vs. B2M gene as an internal control to normalize GLUT and SGLT mRNA expression in VSMCs. Equal amounts of total RNAs isolated from the proliferative and differentiated phenotypes of VSMCs were subjected to qPCR analysis using GAPDH- and B2M-specific primer pairs. In the proliferative phenotype, GAPDH mRNA expression level was 2.5 ± 0.2-fold higher compared with differentiated phenotype. In contrast, B2M mRNA expression levels did not differ significantly between the proliferative and differentiated phenotypes. Together, these data precluded the use of GAPDH as an internal control in VSMCs. Hence, we used B2M as an internal control to normalize GLUT and SGLT mRNA expression levels in the present study. Accordingly, GLUT and SGLT mRNAs were quantified and expressed as transcripts per 10⁵ B2M transcripts in VSMCs.

Abundance of total GLUT mRNAs in the proliferative and differentiated phenotypes of VSMCs.

Recent studies have reported the abundance of total GLUT mRNAs that belong to all three classes in striated muscles such as skeletal muscle (51) and cardiac muscle (2) but not in smooth muscle cells. In the present study, subconfluent VSMCs were maintained in culture in the presence and absence of serum trophic factors for 96 h to obtain proliferative and differentiated phenotypes, respectively. Total RNAs isolated from VSMCs were then subjected to qPCR analysis using specific primers, as described in the legend to Fig. 4. The abundance of total GLUT1–14 mRNAs was examined in VSMCs in parallel with SGLT1–2 mRNAs. In the proliferative phenotype of VSMCs, GLUT and SGLT mRNAs together constituted a total of 6,399 ± 145 transcripts/10⁵ B2M transcripts (n = 4). However, in the differentiated phenotype, GLUT and SGLT mRNAs together constituted a total of 3,289 ± 69 transcripts/10⁵ B2M transcripts (n = 4). Thus there was a significant decrease in the expression of total glucose transporter transcripts by 49% (from 6,399 ± 145 to 3,289 ± 69 transcripts/10⁵ B2M transcripts) upon phenotypic switching of VSMCs from the proliferative to differentiated state. Of the 14 GLUT isoforms and two SGLT isoforms, the transcripts for GLUT7, GLUT14, and SGLT2 were undetectable in both the proliferative and differentiated phenotypes. With regard to SGLT1, there were 0.57 transcript/10⁵ B2M transcripts and 1.76 transcripts/10⁵ B2M transcripts in the proliferative and differentiated phenotypes, respectively. Together, these data reveal that 12 GLUT isoforms (excluding GLUT7 and GLUT14) and SGLT1 contribute to the abundance of total glucose transporter transcripts in the proliferative and differentiated phenotypes of VSMCs.

Fig. 4.

GLUT isoform mRNA expression levels in the proliferative and differentiated phenotypes of VSMCs. Subconfluent VSMCs were maintained in culture in the presence and absence of serum trophic factors for 96 h to obtain proliferative and differentiated phenotypes, respectively. Total RNAs isolated from VSMCs were then subjected to qPCR analysis using primers specific for each of the 14 GLUT isoforms, as described in materials and methods. Pie charts shown in A and B illustrate the relative abundance of each GLUT isoform mRNA within the proliferative and differentiated phenotypes, respectively. Bar graphs shown in C–E compare the relative expression levels of GLUT isoform mRNAs between the proliferative (□) and differentiated (■) phenotypes. C: comparison of GLUT1, -2, -3, -4, and -14 mRNAs that belong to class I. D: comparison of GLUT5, -7, -9, and -11 mRNAs that belong to class II. E: comparison of GLUT6, -8, -10, and -12, and HMIT mRNAs that belong to class III. The GLUT isoform mRNAs were expressed as means ± SE values of transcripts/10⁵ B2M transcripts. Data shown are representative of 4 separate experiments. *P < 0.05, compared with proliferative phenotype.

Relative abundance of individual GLUT isoform mRNA within the proliferative and differentiated phenotypes of VSMCs.

The relative abundance of GLUT1–14 mRNAs has not been fully examined in VSMCs. The present study has therefore examined the relative expression levels of GLUT1–14 mRNAs in human aortic VSMCs by qPCR analysis using GLUT isoform-specific primer pairs. As shown in Fig. 4A, the pie chart shows the relative abundance of each GLUT isoform mRNA within the proliferative and differentiated phenotypes of VSMCs. In the proliferative phenotype, the relative abundance of GLUT isoform mRNA was in the following order: GLUT1 > GLUT10 > GLUT9 > GLUT5 > HMIT (GLUT13) > GLUT12 > GLUT6 > GLUT11 > GLUT3 > GLUT8 > GLUT4 > GLUT2. In particular, the relative abundance of GLUT1, GLUT10, GLUT9, GLUT5, HMIT, and GLUT12 was 43.2 ± 0.96, 26.1 ± 0.74, 13.2 ± 0.44, 4.2 ± 0.36, 3.9 ± 0.19, and 3.8 ± 0.32%, respectively. These data show that GLUT1 is the predominant isoform that contributes to the high abundance (∼43%) of GLUT transcripts in the proliferative phenotype. In addition, the mRNA levels of glucose transporters such as GLUT1, GLUT10, GLUT9, and GLUT12 together constitute ∼86% of total GLUT transcripts in the proliferative phenotype of VSMCs.

In the differentiated phenotype of VSMCs (Fig. 4B), the relative abundance of GLUT isoform mRNA was in the following order: GLUT10 > GLUT1 > GLUT12 > GLUT9 > HMIT (GLUT13) > GLUT5 > GLUT11 > GLUT3 > GLUT8 > GLUT6 > GLUT4 > GLUT2. Specifically, the relative abundance of GLUT10, GLUT1, GLUT12, GLUT9, HMIT, and GLUT5 was 27.8 ± 0.56, 24.6 ± 0.55, 19.8 ± 0.62, 14.1 ± 0.39, 5.1 ± 0.10, and 2.9 ± 0.25%, respectively. Unlike the GLUT mRNA expression profile in the proliferative phenotype, GLUT10 is the predominant isoform that contributes to the high abundance (∼28%) of GLUT transcripts in the differentiated phenotype. Furthermore, the mRNA levels of glucose transporters such as GLUT10, GLUT1, GLUT12, and GLUT9 together constitute ∼87% of total GLUT transcripts in the differentiated phenotype of VSMCs.

Thus the key GLUT transcripts (GLUT1, GLUT9, GLUT10, and GLUT12) together comprise ∼86–87% of total GLUT transcripts within the proliferative or differentiated phenotypes of VSMCs. It should be noted that GLUT5 isoform preferentially transports fructose in humans (42), whereas HMIT (GLUT13) is a proton-coupled myo-inositol transporter (24). Both GLUT5 and HMIT together constitute ∼8% of total GLUT transcripts in both the phenotypes of VSMCs.

Comparison of class I, -II, and -III GLUT mRNA expression between the proliferative and differentiated phenotypes of VSMCs.

As illustrated in Fig. 4, C, D, and E, the present study compares the mRNA expression levels of GLUT1–14 isoforms (class I, class II, and class III GLUTs) between the proliferative and differentiated phenotypes of VSMCs. Of the five GLUT isoforms (GLUT1–4, -14) that belong to class I category (Fig. 4C), GLUT1 mRNA showed the predominance in expression in both the proliferative and differentiated phenotypes. Notably, phenotypic switching to the differentiated state resulted in significant diminution of GLUT1 mRNA expression by 70.7 ± 1% (P < 0.05), compared with proliferative phenotype. The observed modest expression levels of GLUT3 mRNA did not differ significantly between the two VSMC phenotypes. The marginal expression levels of GLUT4 mRNA seen in the proliferative phenotype diminished significantly upon phenotypic switching to the differentiated state. While GLUT2 mRNA expression was at extremely low levels, GLUT14 mRNA was undetectable in both the VSMC phenotypes.

Of the four GLUT isoforms (GLUT5, -7, -9, and -11) that belong to class II category (Fig. 4D), GLUT9 mRNA was the predominant transcript in both the proliferative and differentiated phenotypes of VSMCs. In addition, there was a significant decrease in GLUT9 mRNA expression by 45 ± 3.7% in the differentiated phenotype (P < 0.05), compared with proliferative phenotype. GLUT5 mRNA was the next abundant isoform in the proliferative phenotype, and there was a significant decrease in the expression of this transcript by 63.8 ± 3.7% in the differentiated phenotype (P < 0.05). GLUT5 mRNA expression was followed by modest expression levels of GLUT11 mRNA, which did not differ significantly between the two VSMC phenotypes. Furthermore, GLUT7 mRNA was undetectable in both the VSMC phenotypes.

Of the five GLUT isoforms (GLUT6, -8, -10, -12, and -13) that belong to class III category (Fig. 4E), GLUT10 mRNA showed the predominance in expression followed by GLUT12 in the proliferative and differentiated phenotypes. Importantly, phenotypic switching to the differentiated state led to a significant decrease in GLUT10 mRNA and an increase in GLUT12 mRNA expression compared with proliferative phenotype. While GLUT10 mRNA expression decreased by 44.9 ± 2.6% (P < 0.05), GLUT12 mRNA showed an increase in expression by ∼2.8-fold (P < 0.05) in the differentiated phenotype. In addition, there were significant decreases in the expression of GLUT13 (HMIT) and GLUT6 mRNAs by 32.1 ± 6.1% (P < 0.05) and 85.1 ± 1.6% (P < 0.05), respectively, upon phenotypic switching to the differentiated state. The observed modest expression levels of GLUT8 mRNA did not differ significantly between the two VSMC phenotypes.

Together, these data reveal that phenotypic transition from the proliferative to differentiated state results in the differential regulation of key GLUT mRNAs whereby GLUT1 (class I), GLUT9 (class II), and GLUT10 (class III) mRNAs are downregulated with an accompanying upregulation of GLUT12 mRNA (class III).

Time-dependency studies demonstrate downregulation of GLUT1, GLUT9, and GLUT10 mRNAs with an accompanying upregulation of GLUT12 mRNA upon phenotypic switching of VSMCs to the differentiated state.

To determine time-dependent changes in the expression of key GLUT mRNAs (GLUT1, -9, -10, and -12), subconfluent VSMCs were maintained in the presence or absence of serum trophic factors for 8–96 h. Figure 5A shows that serum deprivation led to significant downregulation of GLUT1 mRNA expression at the 8-h time point and the extent of GLUT1 downregulation remained essentially the same for up to 96 h. The diminutions in GLUT1 mRNA expression levels occurred to the extent of 74.9 ± 0.1, 75 ± 4.6, 72 ± 1.2, and 71 ± 1.0% at the 8, 24, 48, and 96-h time points (P < 0.05), respectively, compared with proliferative phenotype. In parallel, serum deprivation significantly diminished GLUT9 mRNA expression levels by 29.3 ± 4.4, 27.2 ± 10.1, 47.1 ± 1.1, and 45 ± 3.7% at 8, 24, 48, and 96-h time points (P < 0.05), respectively (Fig. 5B). Furthermore, serum deprivation significantly diminished GLUT10 mRNA expression levels by 24.9 ± 7.9, 40.9 ± 18.2, 38.4 ± 3.5, and 44.9 ± 2.6% at the 8, 24, 48, and 96-h time points (P < 0.05), respectively (Fig. 5C). In contrast, serum deprivation resulted in significant upregulation of GLUT12 mRNA expression in VSMCs for up to 96 h with maximal effects occurring at the 8-h time point. The increases in GLUT12 mRNA expression levels occurred to the extent of 193.7 ± 26.9, 198.2 ± 14.5, 247.0 ± 6.3, and 275.9 ± 26.7% at the 8, 24, 48, and 96-h time points (P < 0.05), respectively (Fig. 5D). Thus time-dependency studies provide further evidence for the differential regulation of key GLUT mRNAs (GLUT1, -9, -10, and -12) during phenotypic transition of VSMCs from the proliferative to differentiated state.

Fig. 5.

Time-dependent changes in GLUT1, GLUT9, GLUT10, and GLUT12 isoform mRNA expression in VSMCs upon phenotypic switching to the differentiated state. Subconfluent VSMCs were maintained in culture in the presence or absence of serum trophic factors for 8–96 h. The total RNAs isolated from VSMCs were then subjected to qPCR analysis using primers specific for GLUT isoforms. A–D: mRNA values for GLUT1, GLUT9, GLUT10, and GLUT12 isoforms at different time intervals in both the proliferative (○) and differentiated (●) phenotypes of VSMCs. GLUT isoform mRNAs were expressed as means ± SE values of transcripts/10⁵ B2M transcripts. Data shown in the linear graphs are representative of 3 separate experiments. *P < 0.05, compared with proliferative phenotype at the respective time points.

Expression of key GLUT proteins, GLUT1, -9, -10, and -12, in VSMCs as revealed by immunoblot analysis using whole cell homogenates.

It is well known that the expression levels of mRNA do not always closely correlate with the respective protein levels. To determine whether the downregulation and/or upregulation of key GLUT mRNAs observed in the present study is reflected by corresponding changes in GLUT protein expression upon phenotypic transition to the differentiated state, immunoblot analysis was performed with whole cell homogenates using GLUT isoform-specific primary antibodies.

As shown in Fig. 6A, GLUT1 protein appeared as two distinct forms at ∼51- and ∼47-kDa regions especially in the proliferative phenotype of VSMCs. Upon transition to the differentiated phenotype by serum deprivation for 48 and 96 h, there was a profound reduction in the expression of GLUT1 protein bands at ∼51- and ∼47-kDa regions. In addition, the differentiated phenotype of VSMCs showed marked increases in the expression of GLUT1 protein at ∼43- to 45-kDa regions (n = 4). As a positive control for GLUT1, human brain tissue membrane lysate was used.

Fig. 6.

Comparison of GLUT1, -9, -10, and -12 protein expression between the proliferative and differentiated phenotypes of VSMCs. Subconfluent VSMCs were maintained in culture in the presence or absence of serum trophic factors for 48 and 96 h. Whole cell lysates were then subjected to immunoblot analysis using primary antibodies specific for GLUT1 (A), GLUT9 (B), GLUT10 (C), and GLUT12 (D) proteins. As positive controls (Con), the lysates from human brain tissue membrane (GLUT1), human kidney tissue (GLUT9), and human adipose tissue (GLUT10 and GLUT12) were used. In parallel, human artery tissue membrane lysate (ATM) and human artery tissue lysate (AT) were used. β-actin was used as an internal control. Data shown are representative of 3 to 4 separate experiments.

Figure 6, B, C, and D, shows the expression of novel GLUT proteins, GLUT9, -10, and -12 in VSMCs. Each of these GLUT proteins appeared as a single band, the expression of which did not differ significantly between the proliferative and differentiated phenotypes (n = 3 to 4). In parallel, the positive controls for GLUT9 (human kidney tissue lysate) and GLUT10/-12 (human adipose tissue lysate) showed the expression of respective GLUT isoform proteins.

Furthermore, the expression of GLUT1, -9, -10, and -12 proteins in vascular tissues was confirmed using human arterial tissue and membrane lysates (Fig. 6, A–D).

It should be noted that the relative levels of GLUT1, -9, -10, and -12 proteins within the proliferative or differentiated phenotypes of VSMCs cannot be directly deduced from the data shown in Fig. 6, A–D, since these findings have been observed using antibodies that exhibit differences in immunoreactivity.

Subcellular distribution of key GLUT proteins, GLUT1, -9, -10, and -12, in VSMCs as revealed by immunoblot analysis using HDM, LDM, and PM fraction.

In classical insulin-responsive tissues/cells (e.g., skeletal muscle and adipose tissue), the significance of GLUT4 translocation/redistribution toward glucose transport has been extensively studied using subcellular fractions that include PM and intracellular compartments (HDM, enriched in endoplasmic reticulum proteins; and LDM, enriched in Golgi proteins; Refs 9, 36). As an initial step toward understanding the contribution of key GLUT proteins toward glucose transport in VSMCs, it is therefore important to determine the distribution of GLUTs in these subcellular fractions. Immunoblot analysis using GLUT isoform-specific primary antibodies (Fig. 7, A–D) reveals the distribution of GLUT1, -9, -10, and -12 proteins in HDM, LDM, and PM fractions from the proliferative and differentiated phenotypes of VSMCs.

Fig. 7.

Comparison of GLUT1, -9, -10, and -12 protein expression in the subcellular fractions obtained from proliferative and differentiated phenotypes of VSMCs. Subconfluent VSMCs were maintained in culture in the presence or absence of serum trophic factors for 96 h. Whole cell homogenates were subjected to differential centrifugation to obtain subcellular fractions, which included high-density microsomes (HDM), low-density microsomes (LDM), and plasma membrane (PM) as described in materials and methods. Subcellular fractions were then subjected to immunoblot analysis using primary antibodies specific for GLUT1 (A), GLUT9 (B), GLUT10 (C), and GLUT12 (D). Bar graphs for GLUT isoform proteins (E, F, G, and H) in the respective subcellular fractions are representative of 3–4 separate experiments. In the proliferative phenotype (□), the fold changes in GLUT isoform protein expression in the LDM and PM fractions were calculated with reference to HDM-localized GLUT protein (normalized to 1). Percent changes in subcellular GLUT isoform protein expression between the proliferative (□) and differentiation (■) phenotypes were then determined. *P < 0.05, compared with HDM-localized GLUT isoform protein in the proliferative phenotype. #P < 0.05, compared with the respective subcellular GLUT protein in the proliferative phenotype. To determine the enrichment of HDM, LDM, and PM fractions with the appropriate membrane markers, immunoblot analysis was performed using primary antibodies specific for sarco(endo)plasmic reticulum calcium ATPase 2 (SERCA2), Golgin-97, and Na⁺-K⁺ ATPase proteins, respectively (I–K).

In the proliferative phenotype of VSMCs, GLUT1 protein showed modest and marginal distributions in HDM and LDM fractions, respectively, with a predominant localization in the PM fraction (Fig. 7A). Compared with GLUT1 in HDM (normalized to 1), the distribution of GLUT1 protein in LDM and PM fractions was 0.1-fold (P < 0.05) and 5.2-fold (P < 0.05), respectively, in the proliferative phenotype (Fig. 7E). Upon transition to the differentiated phenotype by serum deprivation for 96 h (Fig. 7, A and E), there was a significant increase in GLUT1 protein in the HDM by ∼65% (P < 0.05) with a concomitant decrease in GLUT1 protein in the PM fraction by ∼39.5% (P < 0.05), suggesting diminished trafficking of GLUT1 to the PM under these conditions. In conformity with the whole cell homogenate data, GLUT1 localized in the PM displayed more pronounced protein bands at the higher molecular weight regions especially in the proliferative phenotype, compared with differentiated phenotype.

In contrast to GLUT1, the novel GLUT isoforms (GLUT9, -10, and -12) displayed predominant distributions in the intracellular compartments (HDM and/or LDM) in both the proliferative and differentiated phenotypes of VSMCs (Fig. 7, B–D). While GLUT9 and GLUT10 showed modest distributions in the PM fraction, the presence of GLUT12 in the PM fraction was in negligible amounts. In the proliferative phenotype of VSMCs, compared with GLUT9 in HDM (normalized to 1), the distribution of GLUT9 protein in LDM and PM fractions was 1.2- and 0.15-fold (P < 0.05), respectively (Fig. 7, B and F). In addition, compared with GLUT10 in HDM (normalized to 1), the distribution of GLUT10 protein in LDM and PM fractions was 1.7-fold (P < 0.05) and 0.2-fold (P < 0.05), respectively (Fig. 7, C and G). Furthermore, compared with GLUT12 in HDM (normalized to 1), the distribution of GLUT12 protein in the LDM fraction was 3.4-fold (P < 0.05) with a negligible amount in the PM fraction (P < 0.05; Fig. 7, D and H). Upon transition to the differentiated phenotype, there were no significant changes in the distribution of GLUT9, -10, and -12 proteins in the HDM, LDM, or PM fractions.

Figure 7, I–K, shows the enrichment of HDM, LDM, and PM fractions from differentiated VSMCs with the respective subcellular markers. The HDM fraction contained ∼64% of the endoplasmic reticulum marker (SERCA2), whereas the LDM fraction contained ∼81% of the Golgi marker (Golgin-97). In addition, PM fraction contained ∼78% of Na⁺/K⁺ ATPase.

Together, subcellular fractionation studies provide evidence for GLUT1 localization predominantly in the PM especially in the proliferative phenotype. With regard to novel GLUT isoforms, GLUT9, -10, and -12 are distributed mostly in the intracellular compartments including HDM enriched in SERCA2 and/or LDM enriched in Golgin-97.

Distinct differences in the localization of key GLUT proteins, GLUT1, -9, -10, and -12, in VSMCs.

As shown in Fig. 8, A–F, immunofluorescence studies using GLUT isoform-specific primary antibodies revealed a distinct pattern of distribution for GLUT1, -9, -10, and -12 proteins in VSMCs. Compared with control conditions (Fig. 8A), inclusion of the primary antibody specific for GLUT1 showed immunoreactivity for the endogenously expressed GLUT1 protein (Fig. 8B) in VSMCs. In particular, GLUT1 was localized in the PM and intracellular compartments with a marked accumulation in the perinuclear region showing concentric and/or crescent patterns in VSMCs, as described previously for GLUT4 in L6 myoblasts (13). Furthermore, compared with control conditions (Fig. 8C), the use of isoform-specific primary antibodies revealed immunoreactivity for the endogenously expressed novel GLUT proteins such as GLUT9, -10, and -12. GLUT9 protein expression was localized mostly in the perinuclear region with a concentric/crescent pattern and to some extent in other intracellular compartments (Fig. 8D). Most of the GLUT10 protein displayed distinct juxtanuclear localization with a crescent pattern in VSMCs (Fig. 8E). In addition, GLUT12 immunoreactivity was found to be distributed in the intracellular compartments (Fig. 8F). Together, immunofluorescence studies using GLUT isoform-specific primary antibodies demonstrate the distinct subcellular localization of endogenously expressed GLUT1, -9, -10, and -12 proteins in VSMCs. To further characterize the compartment-specific localization pattern of key GLUT isoforms, colocalization studies should be performed using organelle-specific primary antibodies.

Fig. 8.

Confocal immunofluorescence detection of GLUT1, -9, -10, and -12 proteins in VSMCs. Subconfluent VSMCs were serum deprived for 48 h and then subjected to immunocytochemical analysis followed by confocal microscopy, as described in materials and methods. A: as a negative control for GLUT1 immunoreactivity, GLUT1 primary antibody was not included. B: GLUT1 immunoreactivity was detected using mouse GLUT1 primary antibody. C: as a negative control for GLUT9, GLUT10, and GLUT12 immunoreactivity, the primary antibody for these GLUT proteins was not included. D–F: GLUT9, GLUT10, and GLUT12 immunoreactivity was detected using the respective rabbit primary antibodies. Images shown were captured at a magnification of ×63 (scale bar = 10 μm) and are representative of 3 separate experiments.

Enhanced glucose uptake in VSMC proliferative phenotype compared with differentiated phenotype.

To determine the extent to which glucose uptake is altered in VSMCs upon phenotypic transition to the differentiated state, 2-deoxy-d-[³H] glucose uptake studies were performed as described in materials and methods. As shown in Fig. 9, there was an increase in glucose uptake by 45.7 ± 2.6% in the proliferative phenotype compared with differentiated phenotype. In addition, cytochalasin B at 10–100 μM concentrations diminished glucose uptake by ∼94–96 and ∼90–91% in the proliferative and differentiated phenotypes, respectively. Furthermore, phloretin treatment resulted in inhibition of glucose uptake at higher concentrations since its IC₅₀ value is ∼110-fold higher than the IC₅₀ value for cytochalasin B (53).

Fig. 9.

Changes in glucose uptake in the proliferative and differentiated phenotypes of VSMCs. Subconfluent VSMCs were maintained in culture in the presence (□) or absence (■) of serum trophic factors for 48 h. VSMCs were then subjected to 2-deoxy-d-[³H] glucose uptake studies as described in materials and methods. Data shown in the bar graphs are the mean ± SE values; n = 3. *P < 0.05, compared with proliferative phenotype; #,†P < 0.05, compared with the respective untreated VSMCs.

DISCUSSION

Studies by several investigators have underscored the functional relevance of GLUT1 (class I GLUT isoform) and GLUT10 (class III GLUT isoform) toward phenotypic modulation of VSMCs in different species (1, 8, 19, 28). For instance, the detrimental effects of GLUT1 overexpression in VSMCs have been evidenced by enhanced glucose uptake/metabolism and exaggerated VSMC proliferation and intimal hyperplasia (1, 19). In addition, the beneficial effects of GLUT10 expression have been evidenced by loss-of-function studies. While GLUT10 mutation results in arterial tortuosity syndrome in vulnerable human subjects (8), GLUT10 gene-silencing diminishes the transport of oxidized form of ascorbic acid into the mitochondria thereby inducing oxidative stress in VSMCs (28). The present study demonstrates for the first time that, in addition to the previously identified GLUT1 and GLUT10 isoforms, novel GLUT isoforms such as GLUT9 (class II) and GLUT12 (class III) are also expressed in VSMCs at the mRNA and protein level. Given the importance of GLUT9 toward uric acid reabsorption in the kidney (6, 12) and of GLUT12 toward glucose uptake in the skeletal muscle/adipose tissue (41, 50, 51), the current observations of GLUT9 and -12 expression in VSMCs in conjunction with GLUT1 and -10 may provide new insights into the glucose-dependent and glucose-independent roles of key GLUT isoform(s) in vascular health and disease.

Previously, Stuart et al. (51) have examined the relative abundance of GLUT mRNAs, GLUT1–12, in human skeletal muscle biopsy samples. The principal findings from these studies include the predominance in the expression of GLUT4, -5, and -12 mRNAs with marginal expression of GLUT1 mRNA (GLUT4 > GLUT5 > GLUT12 >> GLUT1), and all three key GLUT isoforms together constitute ∼98% of the total detectable GLUT transcripts in skeletal muscle. The high degree of metabolic plasticity in terminally differentiated skeletal muscle has been evidenced by upregulation of GLUT4 mRNA, protein, and transport activity by exercise and downregulation by inactivity or metabolic disease (10, 39). Intriguingly, smooth muscle cells possess the unique property of phenotypic plasticity characterized by reversible transition between the differentiated/contractile phenotype and the proliferative/synthetic phenotype (37, 49). Hence, we determined the mRNA expression profile of all 14 GLUT isoforms, GLUT1–14, using the proliferative and differentiated phenotypes of VSMCs. Of importance, the abundantly expressed GLUT mRNAs in VSMCs include GLUT1, -9, -10, and -12, and all four key GLUT isoforms together constitute ∼86–87% of the total GLUT transcripts within the proliferative or differentiated phenotype of VSMCs. As illustrated in the present study, the rank order for the relative abundance of key GLUT mRNAs in the proliferative phenotype was GLUT1 (∼43%) > GLUT10 (∼26%) > GLUT9 (∼13%) > GLUT12 (∼4%), whereas the rank order in the differentiated phenotype was GLUT10 (∼28%) > GLUT1 (∼25%) > GLUT12 (∼20%) > GLUT9 (∼14%). It is noteworthy that smooth muscle cells and skeletal muscle exhibit unique differences in the GLUT isoform mRNA expression profile, especially with regard to GLUT1, -9, -10, and -4. While VSMCs show robust expression levels of GLUT1 mRNA as evidenced in the present study and previous studies (25, 31), skeletal muscle has marginal expression levels of GLUT1 transcript (51). In addition, VSMCs exhibit marked expression of GLUT9 and -10 mRNAs, whereas in skeletal muscle the expression of GLUT9 and -10 mRNAs is either undetectable (51) or at marginal levels (35, 40). Furthermore, skeletal muscle that accounts for the majority of basal and insulin-induced glucose disposal from the systemic circulation exhibits robust expression levels of GLUT4 mRNA (51), whereas the primary cultures of aortic VSMCs have marginal (present study) or undetectable levels (25, 31) of GLUT4 mRNA. Nevertheless, VSMCs show significant mRNA expression for GLUT12 isoform that has recently been shown to play an adjunct role in skeletal muscle in promoting insulin-induced glucose uptake and metabolism (41, 50) reminiscent of the thoroughly characterized GLUT4. Thus the smooth muscle-rich blood vessel that supplies glucose to the skeletal muscle may be developmentally programmed to have its own complement of GLUT transcripts to maintain vascular tone and vessel wall integrity.

The role of transcription factors in maintaining the appropriate expression levels of GLUT genes as a function of cellular phenotype has been previously studied in striated muscles such as skeletal muscle and cardiac muscle (47, 55) but not in smooth muscle (49). For instance, during skeletal myoblast differentiation to mature myotubes, MyoD myogenic factor decreases the expression of Sp1 transcription factor thereby resulting in diminished GLUT1 promoter activity and transcription (55). In addition, adult heart exhibits a diminished Sp1 level that downregulates GLUT1 promoter activity and transcription. (47). Notably, downregulation of GLUT1 mRNA is associated with upregulation of GLUT4 mRNA in terminally differentiated skeletal and cardiac myocytes (47). As illustrated in the present study, phenotypic switching of VSMCs to the differentiated state downregulates GLUT1, -9, and -10 mRNAs by ∼45–70% with an accompanying upregulation of GLUT12 mRNA by ∼2.8-fold, suggesting the existence of transcriptional control mechanisms for all four key GLUT isoforms in VSMCs. Although increases in Sp1 (32) and GLUT1 (1, 19) expression have been observed in neointimal smooth muscle cells, the potential for Sp1 or other key transcription factors to regulate the transcriptional activation of GLUT1 promoter has not yet been examined in VSMCs. Further studies are clearly warranted that should examine the critical roles of transcription factors/cofactors toward inducing or repressing GLUT1, -9, -10, and -12 promoter activity in VSMCs.

The expression of GLUT isoform mRNA does not necessarily imply its translation to the protein. For instance, human adipose tissue that exhibits significant levels of GLUT3 mRNA fails to show the expression of GLUT3 protein (57), an isoform predominantly expressed in brain neurons that have a constant demand for glucose (33). To confirm whether the abundantly expressed GLUT transcripts, GLUT1, -9, -10, and -12, are translated to proteins in VSMCs, we therefore performed immunoblot analysis using VSMC homogenates and immunocytochemical analysis using paraformaldehyde-fixed VSMCs. The observations from both experimental approaches provide evidence for the expression of all four key GLUT proteins in VSMCs. Consistent with previous studies (19, 25, 52), GLUT1 protein is detected in VSMCs but it appears as an ∼47-/∼51-kDa protein in the proliferative phenotype and as ∼43- to 45-kDa protein in the differentiated phenotype. At this juncture, it is important to note that previous studies by Simpson and co-workers (33, 54) demonstrate the existence of two distinct forms of GLUT1 proteins (55 and 45 kDa) in the brain representing highly glycosylated and less glycosylated forms of GLUT1. These findings support the notion that the endogenously expressed GLUT1 protein in VSMCs has the potential to undergo posttranslational modification such as glycosylation upon exposure to serum trophic factors, thereby contributing to the appearance of different molecular masses of GLUT1. With regard to GLUT10 isoform, the present study demonstrates the endogenous expression of GLUT10 protein in the primary cultures of VSMCs, and these data provide further evidence for the recently reported GLUT10 protein expression in rat aortic A10 VSMCs after ectopic overexpression of GLUT10/EGFP fusion constructs (28). Furthermore, the present study demonstrates the expression of GLUT9 and GLUT12 proteins for the first time in VSMCs. In conjunction with the previously reported GLUT1 and GLUT10 immunolocalization in VSMCs (1, 8, 28), parallel analysis of GLUT1, -9, -10, and -12 immunoreactivity in the present study reveals a distinct pattern of immunolocalization for all four key GLUT isoforms in VSMCs, as illustrated in Fig. 8. Together, mRNA and protein expression studies support the existence of transcriptional and translational control mechanisms for GLUT1, -9, -10, and -12 isoforms in VSMCs.

Previous studies using L6 skeletal myoblasts demonstrate a decrease in the molecular mass of GLUT1 protein upon differentiation to myotubes (36). In particular, the size of the GLUT1 protein decreases from 53 to 47 kDa during myogenesis and this has been attributed to posttranslation modification involving altered glycosylation. In the present study using VSMC homogenates, phenotypic switching from the proliferative to differentiated state results in a profound reduction of GLUT1 protein expression at the ∼47-/∼51-kDa region with a marked increase in GLUT1 protein expression at the ∼43- to 45-kDa region. The observed shift in the molecular masses from ∼47-/∼51-kDa to ∼43- to 45-kDa region may be due to diminished glycosylation of GLUT1 upon transition of VSMCs to the differentiated phenotype. Notably, Asano et al. (4) have previously reported that glycosylation-defective GLUT1 remains localized in the intracellular compartment, as evidenced in studies using GLUT1 cDNA engineered to lack N-glycosylation consensus site. It is conceivable from the subcellular fractionation studies that highly glycosylated GLUT1, as it occurs in the proliferative phenotype, facilitates its trafficking to the PM to enhance glucose uptake and proliferation in VSMCs.

The regulation of GLUT1 and GLUT9 proteins has been recently studied in mouse uterine endometrial stromal cells, where GLUT1 protein expression is upregulated without a significant change in GLUT9 expression upon differentiation to decidual cells in the presence of progesterone and estrogen (16). In the present study using VSMC homogenates or subcellular fractions, there were no significant changes in GLUT9 protein expression or its subcellular distribution between the proliferative and differentiated phenotypes. Nevertheless, GLUT9 expression and/or function may be modulated by diabetes or high glucose conditions. For instance, streptozotocin-diabetic mice exhibit increased expression of GLUT9 protein in gluconeogenic tissues such as liver and kidney (27). In pancreatic β-cells, GLUT9 may function as a glucose sensor and participate in glucose-induced insulin secretion, thereby playing a supportive role to the thoroughly characterized GLUT2 in this regard (14). Furthermore, GLUT9 functions as a uric acid transporter and facilitates uric acid reabsorption in the kidney (3, 12). Given the importance of uric acid in enhancing VSMC proliferation (43) and of the potential link between hyperuricemia and hypertension (23), future studies should examine the functional relevance of GLUT9 toward the development of vascular complications including hypertension under diabetic and nondiabetic states.

Although the relationship between GLUT1 and GLUT10 protein expression has remained unclear thus far, emerging evidence from recent studies with VSMCs suggest that overexpression of GLUT1 or downregulation of GLUT10 results in vascular remodeling with an accompanying upregulation of TGF-β signaling (1, 28). Importantly, the GLUT10 gene was first identified as a putative candidate for susceptibility loci associated with noninsulin-dependent diabetes mellitus within the human genome (35). Subsequent studies by Coucke et al. (8) demonstrate the functional relevance of GLUT10 mutation toward the development of vascular complications characterized by arterial tortuosity syndrome. Although GLUT10 loss in VSMCs may result in decreased intracellular glucose concentration (8), recent studies by Lee et al. (28) reveal that GLUT10 protein mediates the transport of oxidized form of vitamin C (l-dehydroascorbic acid) into the mitochondria of VSMCs, thereby protecting vascular cells from oxidative stress. With regard to its regulation, insulin stimulation does not affect the mitochondrial localization of GLUT10 protein in GLUT10-overexpressing A10 VSMCs (28). In the present study, phenotypic switching of VSMCs to the differentiated state did not produce significant changes in GLUT10 protein expression or its subcellular distribution. Together, these findings suggest that GLUT10 protein expression may not be regulated by insulin or serum trophic factors. Future studies should examine the likely alterations in GLUT10 expression and its function in VSMCs in conjunction with GLUT1 in the context of diabetic vascular complications.

Recent studies using endometrial stromal cells demonstrate that upregulation of GLUT1 protein is associated with almost a complete loss of GLUT12 protein expression upon differentiation to decidual cells in the presence of progesterone and estrogen (16). In the present study using VSMCs, phenotype switching to the differentiated state did not lead to significant changes in GLUT12 protein expression or its subcellular distribution. Importantly, GLUT12 that is expressed in insulin-responsive tissues such as skeletal muscle and adipose tissue has recently been shown to promote glucose uptake and metabolism, thus playing an adjunct role for the thoroughly characterized GLUT4 isoform in these tissues (41, 50). For instance, studies by Stuart et al. (50) demonstrate that acute insulin treatment induces translocation of GLUT12 to the PM via PI 3- kinase pathway in the skeletal muscle, as revealed by hyperinsulinemic-euglycemic clamp studies in human subjects. In addition, Purcell et al. (41) have shown that in GLUT12 transgenic mice insulin stimulation increases glucose clearance rate in skeletal muscle and adipose tissue thereby increasing whole body insulin sensitivity. Most recent studies by Aerni-Flessner et al. (2) demonstrate that in GLUT4 knockout mice there is a compensatory increase in GLUT12 expression in the ventricular myocardium, but it is not clear whether GLUT12 expression is also increased in arterial smooth muscle under these conditions. Previously, Atkins et al. (5) have shown that DOCA-salt hypertension downregulates the expression of GLUT4 protein in rat carotid artery and aorta, thereby enhancing arterial contractility. Furthermore, chronic diminution and absence of GLUT4 protein expression, as evidenced in the aortas from angiotensin II hypertensive mice and GLUT4 knockout mice, respectively, are associated with enhanced vasoreactivity (38). However, the likely compensatory role of GLUT12 in the regulation of smooth muscle contractility remains unclear in GLUT4-deficient states. Future studies should therefore utilize GLUT12 transgenic mice to determine the regulatory effects of this novel GLUT isoform on vasoreactivity under normotensive and hypertensive states.

Consistent with previous studies (25, 29), the role of GLUT1 toward glucose uptake in VSMCs is supported by the significant inhibitory effects observed with cytochalasin B and phloretin, which bind to the inner surface of GLUT1 and to the external glucose binding site of GLUT1, respectively (53). Although GLUT9 has been shown to promote the uptake of glucose, fructose, and uric acid in different cell types (3, 24, 53), glucose transport activity of this GLUT isoform is not subject to inhibition by cytochalasin B (6). Notably, glucose uptake is nearly abolished in the presence of cytochalasin B in the present study, suggesting that GLUT9 is likely to transport substrates other than glucose in VSMCs. Moreover, the potential intracellular localization of GLUT10 may preclude the participation of this GLUT isoform toward glucose uptake in VSMCs, as described previously by Lee et al. (28). Since GLUT12 has been shown to promote glucose uptake similar to GLUT4 in skeletal muscle and adipose tissue (41, 50, 51), the GLUT12 isoform may also contribute to phloretin-sensitive glucose uptake in VSMCs. Nevertheless, GLUT9, GLUT10, and GLUT12 proteins that are expressed in VSMCs display a predominant intracellular localization (HDM and/or LDM) with modest to negligible levels in the PM in both the proliferative and differentiated phenotypes. In addition, there are no significant changes in the distribution/expression of GLUT9, GLUT10, and GLUT12 proteins in the PM fractions between the proliferative and differentiated phenotypes. Together, these data suggest that increased distribution/localization of GLUT1 in the PM contributes to enhanced glucose transport in VSMC proliferative phenotype, compared with differentiated phenotype.

Of the six different SGLT isoforms (SGLT1–6), the present study using VSMCs has determined the mRNA expression levels of classical sodium-dependent glucose transporters, SGLT1 and SGLT2, which are known to be expressed predominantly in the small intestine and kidney, respectively (58). While SGLT1 mRNA is expressed at marginal levels, SGLT2 mRNA expression is undetectable in VSMCs. With regard to SGLT3–6, previous studies have shown that these SGLT isoforms do not mediate glucose transport (18, 58). For instance, SGLT3 appears to function as a glucose sensor whereas SGLT4/-5 primarily transports mannose and fructose. In addition, SGLT6 functions as a sodium/myoinositol cotransporter. Future studies should determine the expression and likely biological roles of SGLT3–6 isoforms in VSMCs.

In conclusion, VSMCs in culture may provide a useful in vitro model system to understand the functional consequences of dysregulated expression and transport activity of key GLUT isoforms such as GLUT1, -9, -10, and -12 in arterial smooth muscle under pathophysiological conditions. Given the fact that GLUT1 overexpression or GLUT10 mutation results in enhanced TGF-β signaling in VSMCs (1, 28), future studies should determine the likely regulatory effects of GLUT9 and/or GLUT12 on TGF-β signaling and proinflammatory lesion in the vessel wall. In addition to the aforementioned key GLUT isoforms, the potential roles of GLUT5 and GLUT13 (HMIT) toward modulating VSMC phenotype warrant further study but the modestly expressed transcripts for these isoforms need to be confirmed at the protein level. Notably, GLUT isoforms exhibit unique differences in the affinity toward various substrates including glucose, fructose, glucosamine, uric acid, l-dehydroascorbic acid, and myo-inositol. Future studies involving overexpression or downregulation of biologically important GLUT isoforms in VSMCs will therefore provide new insights into the transport of novel substrates beyond glucose that may contribute to beneficial or detrimental effects in the vessel wall.

GRANTS

This work was supported by National Heart, Lung, and Blood Institute Grant R01-HL-097090 and University of Georgia Research Foundation (to L. Segar) and Jane Barsumian/Mary Lyons Trust (to J. Jun).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Author contributions: R.P. and N.P. performed experiments; R.P., N.P., and L.S. analyzed data; R.P., N.P., J.Y.J., and L.S. interpreted results of experiments; R.P. and N.P. prepared figures; R.P., N.P., J.Y.J., and L.S. edited and revised manuscript; R.P., N.P., J.Y.J., and L.S. approved final version of manuscript; L.S. conception and design of research; L.S. drafted manuscript.