Volume 97, Issue 8 p. 2035-2042
Original Article
Free Access

Expression and localization of GLUT1 and GLUT12 in prostate carcinoma

Jenalle D. Chandler B.Sc. (Hons.)

Jenalle D. Chandler B.Sc. (Hons.)

Department of Medicine, St. Vincent's Hospital, The University of Melbourne, Melbourne, Australia

Search for more papers by this author
Elizabeth D. Williams Ph.D.

Elizabeth D. Williams Ph.D.

Department of Surgery, St. Vincent's Hospital, The University of Melbourne, Melbourne, Australia

Bernard O'Brien Institute of Microsurgery, Melbourne, Australia

Search for more papers by this author
John L. Slavin M.B.B.S.

John L. Slavin M.B.B.S.

Department of Pathology, St. Vincent's Hospital, Melbourne, Australia

Search for more papers by this author
James D. Best M.D.

James D. Best M.D.

Department of Medicine, St. Vincent's Hospital, The University of Melbourne, Melbourne, Australia

Search for more papers by this author
Suzanne Rogers Ph.D.

Corresponding Author

Suzanne Rogers Ph.D.

Department of Medicine, St. Vincent's Hospital, The University of Melbourne, Melbourne, Australia

Fax: 011-61-3-9288-2581

Department of Medicine, St. Vincent's Hospital, The University of Melbourne, P.O. Box 2900, Fitzroy, Victoria 3065, Australia===Search for more papers by this author
First published: 01 April 2003
Citations: 127

Abstract

BACKGROUND

Increased glucose consumption is a characteristic of malignant cells and in prostate carcinoma is associated with the proliferation of both androgen-dependent and independent cells. Transport of polar glucose across the nonpolar membrane relies on glucose transporter proteins, known as GLUTs. Increased expression of GLUT1 is a characteristic of many malignant cells. The authors characterized and cloned the cDNA for a novel glucose transporter, GLUT12, which was identified initially in malignant breast epithelial cells. To the authors' knowledge, there have been no reports on the expression of glucose transporters in the human prostate or human prostate carcinoma cells. The authors evaluated GLUT1 and GLUT12 expression in human prostate carcinoma cells.

METHODS

Reverse transcription-polymerase chain reaction was performed on total RNA extracted from cultured prostate carcinoma cells LNCaP, C4, C4-2, and C4-2B using primers to amplify GLUT1, GLUT12, or the housekeeping gene, 36B4. Total protein extracted from prostate carcinoma cell lines was assessed for GLUT12 protein by Western blot analysis. Cultured cell monolayers were incubated with antibodies to GLUT1 or GLUT12 and a peripheral Golgi protein, Golgi 58K, for detection by immunofluorescent confocal microscopy. Sections of benign prostatic hyperplasia and human prostate carcinoma were stained for immunohistochemical detection of GLUT1 and GLUT12.

RESULTS

GLUT1 and GLUT12 mRNA and protein were detected in all cell lines evaluated. Immunofluorescence staining demonstrated both GLUT1 and GLUT12 on the plasma membrane and in the cytoplasm in all cultured prostate carcinoma cell lines, with GLUT1 but not GLUT12 appearing to colocalize with the Golgi. Immunohistochemical staining of benign prostatic hyperplasia indicated expression of GLUT1 but not GLUT12. Malignant tissue stained for GLUT12 but was negative for GLUT1.

CONCLUSIONS

GLUT1 and GLUT12 are expressed in human prostate carcinoma cells. One possible rationale for the GLUT1 Golgi association is that it may supply glucose to the Golgi for byproduct incorporation into the prostatic secretory fluid. Further work will investigate the importance of glucose transport and GLUT1 and GLUT12 in prostate carcinoma cell growth. Cancer 2003;97:2035–42. © 2003 American Cancer Society.

DOI 10.1002/cncr.11293

Increased glucose consumption is a basic characteristic of malignant cells and is linked to higher energy production from aerobic glycolysis.1 There is evidence that glucose uptake is increased in prostate carcinoma cells and high rates of glucose consumption are required for the rapid proliferation of androgen-independent prostate carcinoma cells.2 Clinically, amplified glucose usage by malignant cells forms the theoretic basis for cancer detection through uptake of the glucose analoge, fluorine-18–labeled 2-deoxy-2-fluoro-D-glucose (18F-FDG) with positron emission tomography (PET).3 This technique can be used to detect prostate carcinoma and metastases in pelvic lymph nodes and to monitor the response to therapy. Previous studies have shown higher FDG avidity in prostate carcinoma with a higher Gleason score and a negative correlation between FDG uptake and hormone responsiveness.4

Glucose uptake across the plasma membrane is considered the rate-limiting step for glucose consumption in tumor cells.5 Comprehension of the mechanisms for increased glucose uptake will be a key to understanding the biologic basis of FDG PET for detecting and monitoring prostate carcinoma. Transport of polar glucose across the nonpolar plasma membrane relies on glucose transporter proteins, known as GLUTs. Increased expression of GLUT1, a ubiquitously expressed glucose transporter, is a characteristic of many malignant cells and a correlation has been reported between immunohistochemical analysis of GLUT1 expression and FDG uptake in breast carcinoma.6 A large study of human breast carcinomas demonstrated GLUT1 expression in 42% of tumors, with increased expression in cancer cells with a higher grade and proliferative activity.7 Transformation of cultured cells with ras, src, and neu oncogenes markedly increases GLUT1 mRNA and protein levels.8 More recently, c-myc was shown to increase GLUT1 mRNA by increasing the transcription rate.9

Rogers et al.10 reported the identification of a new glucose transporter protein, GLUT12, in human breast carcinoma cells. GLUT12 belongs to the newly classified Class III sugar transporter subfamily, which also includes GLUT6 and other newly identified transporters (GLUT8, GLUT10, and HMIT).11 GLUT12 expression has been detected in the insulin-responsive tissues, skeletal muscle, and adipose tissue, as well as in the heart and small intestine.10 Subsequently, we have found GLUT12 mRNA and protein in human breast tumors (unpublished data). GLUT12 is potentially a regulator of glucose utilization in malignant cells.

This study investigates the expression and potential roles of GLUT1 and GLUT12 in prostate carcinoma cells. Although GLUT1 mRNA has been documented in a rat model of prostate carcinoma,12 there have been no reports of GLUT expression in human prostate carcinoma, in either cultured cells or tissue specimens. This study reports a preliminary evaluation of the expression and localization of GLUT1 and GLUT12 in both prostate carcinoma cell lines and primary prostate tumors.

MATERIALS AND METHODS

Cell Lines

The progressive human prostate carcinoma cell lines LNCaP, C4, C4-2, and C4-2B were purchased from Urocor (Oklahoma City, OK). Cells were maintained in T media (four parts Dulbecco Modified Eagle Medium [Life Technologies, Grand Island, NY] and one part HAM's F12 [Life Technologies]). This was supplemented with 10% fetal bovine serum (FBS; CSL Ltd., Parkville, Victoria, Australia), 5 μg/mL insulin (Sigma, St. Louis, MO), 13.65 pg/mL triiodo-thyronine (Sigma), 5 μg/mL apo-transferrin (Sigma), 0.244 μg/mL d-biotin (Sigma), 25 μg/mL adenine (Sigma), 0.1 μg/mL streptomycin (Life Technologies), and 0.1 U/mL penicillin (Life Technologies). Cells were grown at 37 °C in a humidified incubator.

Reverse Transcription-Polymerase Chain Reaction (RT-PCR)

Total RNA was extracted from cultured prostate carcinoma cells with TRIzol (Life Technologies) and 5 μg reverse transcribed with AMV reverse transcriptase and Oligo(dT)15 primers (Promega, Madison, WI). Polymerase chain reaction used the Expand High-Fidelity PCR system (Roche Diagnostics, Mannheim, Germany) and primers (Sigma Genosys, Castle Hill, Australia) to GLUT1 (5′-TTC ACT GTC GTG TCG CTG TTT G-3′ and 5′-TCA CAC TTG GGA ATC AGC CCC-3′) and GLUT12 (5′-TCC ATG GCT GGA AGT ACA T-3′ and 5′-TAA GTG TTC TGG CAC TAT C-3′). Amplification of housekeeping gene 36B4 was used as a normalizing control (5′-TGG GCT CCA AGC AGA TGC-3′ and 5′-GGC TTC GCT GGC TCC CAC-3′).13 Reactions were incubated in an automatic heat block (DNA Thermal Cycler 480, Perkin Elmer, Norwalk, CT) as follows: GLUT1, 28 cycles, denaturation 30 seconds at 94 °C, annealing 30 seconds at 61 °C, extension 30 seconds at 72 °C; GLUT12, 31 cycles, denaturation 30 seconds at 94 °C, annealing 30 seconds at 53 °C, extension 60 seconds at 72 °C; and 36B4, 20 cycles, denaturation 30 seconds at 94 °C, annealing 30 seconds at 65 °C, extension 30 seconds at 72 °C. The PCR products were run on 1% agarose gels in TBE (89mM Tris-borate, 2mM EDTA) buffer and transferred to a nylon membrane. GLUT1 and GLUT12 cDNAs were labeled with [α-32P]dCTP using the Random Primed DNA Labeling kit (Roche Diagnostics). A 36B4 oligonucleotide (5′-GTGTTCACCAAGGAGGACC-3′) was labeled with [γ32P]dATP using T4 polynucleotide kinase (Promega) as per the manufacturer's instructions.

Antibodies

The rabbit polyclonal GLUT1 antibody (Chemicon International, Temecula, CA) directed at the 12 terminal amino acids of the C-terminus of GLUT1 was used at a dilution of 1:150. A GLUT12 polyclonal antibody, R1396, raised in rabbit against the last 16 amino acids of the C-terminus of GLUT12, was affinity purified and used at concentrations of 10 and 50 μg/mL.10, 14 Nonimmune rabbit serum was utilized as a negative control (1:150). A peripheral Golgi marker, monoclonal anti-Golgi 58K (Sigma), was diluted 1:50. Secondary antibodies employed were horseradish peroxidase (HRP)-conjugated swine antirabbit immunoglobulin (IgG) (1:1000; Dako Corporation, Carpinteria, CA), biotinylated swine antirabbit IgG (1:200; Dako), [fluorescein isothiocyanate] (FITC)-conjugated antirabbit IgG (1:100; Dako), FITC-conjugated antimouse IgG (1:100; Dako), and Alexa Fluor 568 antirabbit IgG (1:1000; Molecular Probes, Eugene, OR). All antibodies were diluted in phosphate-buffered saline (PBS) with 5% FBS, apart from the HRP-conjugated antibody which was diluted in Tris-buffered saline (TBS).

Western Blot Analysis

Detection of GLUT12 protein by Western blot analysis has been described previously.10, 15 In brief, the protein was extracted from cultured prostate carcinoma cells with TRIzol (Life Technologies) 40 μg protein precipitated in a × 5 volume of cold acetone. Pellets were resuspended in loading buffer containing 0.72 M β-mecaptoethanol (Sigma), analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) on a 10% gel, and electrophorectically transferred to PDVF filters (Roche Diagnostics). Filters were incubated with 5% blocking solution (5% skim milk powder in TBS) for 1 hour. Filters were then immunoblotted with the affinity-purified R1396 antibody preincubated for 1 hour on ice with 100 μg/mL of competitive peptide (the peptide used to generate the R1396 antibody) or with the affinity-purified R1396 antibody alone. Filters were washed with 0.1% Tween-20 (Sigma) in TBS and blocked in 2% blocking solution before incubation with the secondary antibody. Filters were washed and labeled proteins were detected using SuperSignal West Femto Substrate (Pierce Chemical, Rochester, NY) diluted 1:5 in dH2O and autoradiography film (Amersham Pharmacia Biotech, Little Chalfont, U.K.).

Immunofluorescence Studies

Cultured cells were fixed with acetone, incubated in 100 mM glycine for 15 minutes, permeabilized with 0.1% Triton X-100 (Ajax Chemicals, Sydney, New South Wales, Australia) for 15 minutes, and blocked in 10% normal swine serum (NSS) (Institute of Medical and Veterinary Science, Adelaide, Australia) and 5% FBS in PBS for 30 minutes. Overnight incubation at 4 °C with antibodies to GLUT1 or GLUT12 and Golgi 58K was followed by several washes with PBS containing 0.1% Tween-20 (PBST) and by incubation for 1 hour in the dark with the appropriate secondary antibodies. Nonimmune rabbit serum was used as a negative control. Nuclei were stained with either propidium iodide (PI; 20 μL of 250 ng/mL PI [Sigma] in 0.1% Triton X-100, 2 μL 500 μg/mL RNase [Roche Diagnostics], and 660 μL PBST), or TOTO-3 (Molecular Probes) diluted 1:1000 in PBST with 2.5 μg/mL RNase. After washing with PBST, sections were mounted in fluorescent mounting medium (Dako) with glass coverslips and examined with a Bio-Rad MCR 1024 inverting laser scanning confocal microscope using Lasersharp 2000 software (Bio-Rad, Hercules, CA).

Tissue Samples

Three samples of benign prostatic hyperplasia (BPH) and three primary prostate carcinoma biopsy samples (all paraffin embedded) were obtained from the Department of Pathology, St. Vincent's Hospital, Melbourne. This study was approved by the St. Vincent's Hospital Melbourne Human Ethics Committee (070/02).

Immunohistochemistry Studies

R1396 antibody was preincubated overnight on ice with 100 μg/mL of competitive peptide or an unrelated noncompetitive peptide. Paraffin-embedded BPH and malignant prostate biopsy specimens were sequentially sectioned onto aminopropyltriethoxysilane-coated microscope slides and deparaffinized. Endogenous peroxidase was inhibited by incubation in 5% hydrogen peroxide in methanol for 30 minutes. Sections were blocked with 10% NSS in PBS containing 5% FBS for 30 minutes and incubated overnight with antibody to GLUT1, GLUT12, or nonimmune serum at 4 °C. Washes with PBST were followed by incubation with a biotinylated secondary antibody. Complexes were visualized using the avidin-biotin complex Vectastain (Vecta Laboratories, Inc., Burlingame, CA), chromagen 3,3′-diaminobenzidine (Sigma), and hematoxylin counterstaining.

RESULTS

The LNCaP human prostate carcinoma cell line is well characterized and has several derivatives (including C4, C4-2, and C4-2B) that correlate with an increase in tumor progression and loss of androgen responsiveness.16 PCR primers were designed to specifically amplify GLUT1, GLUT12, and 36B4. The number of cycles used was chosen from preliminary experiments that determined the linear phase of amplification (data not shown). Amplification by RT-PCR of RNA from cultured cell monolayers detected GLUT1 and GLUT12 mRNA in all cell lines evaluated. Figure 1A is a representative Southern blot analysis of RT-PCR products. Similar results were obtained for one to two PCR reactions performed using triplicate RT reactions.

Details are in the caption following the image

(A) Representative Southern blot analysis of the expression of GLUT1, GLUT12, and 36B4 mRNA in the prostate carcinoma cell lines LNCaP, C4, C4-2, and C4-2B. Reverse transcription-polymerase chain reaction was performed on total RNA extracted from cultured cell monolayers, with products transferred to a nylon membrane for cDNA labeling. (B) Western blot analysis of the expression of GLUT12 protein in total protein extracted from the cultured prostate carcinoma cells. Filters were probed with affinity-purified antibody to GLUT12 in the presence or absence of competitive peptide.

To demonstrate the specificity of the R1396 GLUT12 polyclonal antibody and to confirm expression of GLUT12 in prostate carcinoma cell lines, Western blot analyses were performed. When total protein extracts were immunoblotted with affinity-purified R1396 antiserum, a protein species of approximately 60 kilodalton was detected in all cell lines (Fig. 1B). Preincubation with the peptide used to raise the antibody completely inhibited the signal, confirming antibody specificity.

Initial immunofluorescence experiments involved single staining for either the GLUT1 or GLUT12 protein in prostate carcinoma cell monolayers. Probing with the primary antibody was followed by incubation with FITC-labeled anti-rabbit secondary antibody, and nuclei were stained with PI. GLUT1 was detected on the plasma membrane and in the cytoplasm, whereas GLUT12 was found to be perinuclear, cytoplasmic, and on the plasma membrane of cells from all lines evaluated (Fig. 2). To assess further the cytoplasmic localization of the two glucose transporters, a cytoplasmic peripheral Golgi marker, anti-Golgi 58K monoclonal antibody, was used in conjunction with GLUT1 or GLUT12 antibodies. Nuclei were stained with TOTO-3. GLUT12 did not appear to associate directly with the Golgi (Fig. 3). Rather, cytoplasmic staining of GLUT12 was adjacent to the Golgi. However, staining did suggest colocalization of GLUT1 and Golgi 58K. Results were uniform for all cell lines studied.

Details are in the caption following the image

Immunofluorescent staining of the prostate carcinoma cell line C4-2. (A) GLUT1. (B) GLUT12. (C) Nonimmune control. Localization of the glucose transporters GLUT1 and GLUT12 stained green with fluorescein isothiocyanate-labeled secondary antibody. Nuclei were stained red with propidium iodide. All cell lines showed similar results (Original magnification ×600).

Details are in the caption following the image

Immunofluorescent staining of the prostate carcinoma cell line C4 with antibody to GLUT1 or GLUT12 and with the Golgi marker, Golgi 58K. GLUT1 and GLUT12 are stained secondarily with anti-rabbit-Alexa 598 (red), Golgi with antimouse fluorescein isothiocyanate (green), and nuclei with TOTO-3 (blue). All cell lines showed similar results. Arrows indicate GLUT12 staining adjacent to the Golgi (Original magnification ×600).

For a preliminary assessment of glucose transporter expression in human BPH and malignant tumor biopsy specimens, we immunohistochemically stained paraffin-embedded sections for GLUT1 and GLUT12. In BPH samples, GLUT1 but not GLUT12 protein expression was evident in all three biopsy specimens evaluated (Fig. 4A, B). There was positive GLUT12 staining of the muscle surrounding the hyperplasic ducts. Two of three malignant tumor biopsy specimens were negative for GLUT1 (Fig. 4D), whereas one section stained very weakly the epithelium of cancerous ducts (results not shown). As GLUT1 is present in erythrocytes,17 successful staining of sections for GLUT1 was verified by positive staining of red blood cells. Staining of malignant ducts and surrounding tissue was strong for GLUT12 using 50 μg/mL R1396 in all biopsy specimens (Fig. 4G). In one section, BPH located in tissue close to the cancerous ducts stained weakly for GLUT12. GLUT12 staining was R1396 concentration dependent, with weaker staining observed using 10 μg/mL R1396. The specificity of GLUT12 staining was confirmed by the demonstration of reduced immunohistochemical staining following preincubation of affinity-purified R1396 antibody with competitive peptide (results not shown). No staining was observed when sections were incubated in the presence of nonimmune rabbit serum (negative control; Fig. 4C,E,H).

Details are in the caption following the image

Immunohistochemical staining of the glucose transporters GLUT1 and GLUT12 in benign and malignant prostate biopsy specimens. (A) Staining of benign tissue for GLUT1. (B) Staining of benign tissue for GLUT12. (C) Nonimmune control for A and B. (D) Staining of malignant tissue for GLUT1. (E) Nonimmune control for D. (F) Hematoxylin and eosin (H & E) staining of section adjacent to that shown in D and E. (G) Staining of malignant tissue for GLUT12. (H) Nonimmune control for G. (I) H & E staining of section adjacent to that shown in G and H (Original magnification ×200).

DISCUSSION

In contrast to the many publications on glucose transporters and breast carcinoma, there have been no reports of GLUT expression in either cultured human prostate carcinoma cells or tissue specimens. We hypothesized that GLUT1 was likely to be present in human prostate carcinoma cells as it has been found in many other malignant epithelial cells. The presence of GLUT1 mRNA has been documented in a rat model of prostate carcinoma, correlating with FDG uptake.12 We previously identified a novel glucose transporter, GLUT12, in cultured breast carcinoma cells and malignant breast tissue specimens. GLUT12 is a potential regulator of glucose uptake in malignant cells. An initial evaluation by RT-PCR of a prostate carcinoma cell line indicated the presence of GLUT12 mRNA. The current study evaluated the expression and localization of GLUT1 and GLUT12 in human prostate carcinoma cells.

GLUT1 is expressed consistently in human breast carcinoma cell lines with a correlation between in vitro invasive potential of the cell lines and GLUT1 protein expression.18 The human prostate carcinoma cell line LNCaP and its derivatives (C4, C4-2, and C4-2B) correlate to an increase in tumor progression and androgen independence from LNCaP to C4-2B. Our studies using RT-PCR confirmed the presence of GLUT1 and GLUT12 mRNA in these cell lines. Although RNA extraction and RT were performed several times, no relationship could be established linking tumor progression, the degree of androgen responsiveness, and the level of GLUT mRNA expression using this method. More readily quantifiable techniques such as Northern blot analysis or Real-Time PCR could allow further evaluation of any possible correlation.

There are potential discrepancies between mRNA and protein expression for both GLUT1 and GLUT12. We found GLUT1 mRNA and protein in all the cultured prostate carcinoma cell lines, although protein was weakly expressed in only one of the primary tumor sections evaluated. This result may reflect differences in the characteristics of cultured cells from those in vivo. Northern blot analysis has shown the expression of GLUT12 mRNA in human prostate cells,10 but GLUT12 protein was not expressed in the normal prostate tissue samples evaluated in the current study. Further studies will be required to determine the relationship between mRNA and protein levels of glucose transporters in benign and malignant prostate tissue specimens.

Tumor tissue specimens are often associated with the overexpression of glucose transporter proteins, particularly GLUT1. Several large studies of GLUT1 protein expression in epithelial cancer tissue specimens, including ovarian and gastric tumors, have reported an association between an increase in immunohistochemical detection of GLUT1 protein with malignancy progression.19, 20 In breast carcinoma cells, GLUT1 was detected in less than 50% of samples.7 In the small number of prostate cancers evaluated in this study, very weak GLUT1 staining was observed in only one of the sections. A larger numbers of samples will be required to determine the true prevalence of GLUT1 expression in prostate carcinoma cells. Cancer cells can also be associated with abnormal expression of glucose transporter proteins that are not present in the corresponding benign cells. GLUT3 and GLUT5 can be expressed in cancer cells, differing from the usual tissue specific distribution.21, 22 In this study, we detected the presence of GLUT12 protein in malignant prostate tissue specimens using immunohistochemical staining of sectioned biopsies. GLUT12 was also found in BPH cells surrounding cancerous ducts, but not in BPH cells from normal prostate tissue specimens. It is possible that cells adjacent to the cancerous tissue are already undergoing transformation. This may result in an increase in GLUT12 expression in an attempt to accommodate the increased requirement for glucose by the transformed cells.

GLUT1 but not GLUT12 protein was expressed in BPH. Normal human prostate secretory epithelial cells form high levels of citrate from aspartic acid and glucose, but malignant cells are reported to have a relatively low rate of production.23 This decrease in citrate synthesis could be attributed to the dedifferentiation of the cancerous cell, and may correlate with a reduction in GLUT1 expression from normal to malignant prostate tissue. Another possible explanation is that glucose is preferentially utilized by other metabolic processes in the proliferating cell and is not available for citrate synthesis.

As both GLUT1 and GLUT12 are expressed in cultured prostate carcinoma cells, we evaluated the cellular localization of both transporters in these cells. GLUT1 is localized at the plasma membrane in most normal and malignant cells, where it is responsible for basal glucose uptake. An exception to this localization is in the lactating mammary gland, in which GLUT1 is present at the Golgi apparatus and is likely to play a major role in supplying glucose to this organelle.24 Mammary cells are the only known cells in which the Golgi requires free glucose, which subsequently is used in the production of lactose. We used Golgi marker 58K, a peripheral membrane protein exposed on the cytoplasmic side of the Golgi, to assist in determining the cytoplasmic localization of GLUT1 and GLUT12 in cultured prostate carcinoma cells. GLUT1 was found on the plasma membrane and staining also suggested association with the Golgi. The secretory product of the normal prostate makes up approximately 75% of the seminal fluid and is rich in citric acid as well as hydrolytic enzymes that contain glycoproteins.25 Colocalization of GLUT1 and Golgi 58K suggests that GLUT1 may be transporting glucose to the Golgi for byproduct incorporation into the prostatic secretory fluid, analogous to its role in the mammary gland.

GLUT12 was detected in the cytoplasm, the perinuclear region, and on the plasma membrane of cultured prostate carcinoma cells, but staining did not indicate colocalization with the Golgi. In the breast carcinoma cell line MCF-7, GLUT12 is localized to a perinuclear region and there is evidence that this localization is altered when cells are grown in the presence of insulin.10 GLUT12 localization in prostate carcinoma cells in the presence and absence of insulin was not determined in the current study. Given the number of other growth additives required in the culture media to maintain the cell lines studied, there were many factors that could have influenced basal GLUT12 localization.

This study presents the first report of glucose transporter protein expression in the human prostate and prostate carcinoma. In addition, we have shown a potentially novel colocalization of GLUT1 with a Golgi marker in cultured prostate carcinoma cells. Our results indicate that GLUT1 and GLUT12 are expressed in, and may contribute to meeting the elevated energy requirements of proliferating prostate carcinoma cells. We cannot exclude the possibility that GLUT12 may facilitate transport of other hexoses and molecules. For example, GLUT1 transports dehydroascorbic acid, the oxidized form of vitamin C,26 although this function is considered secondary to its role as a glucose transporter.

Evaluation of a large number of prostate biopsy specimens will be conducted to provide a better understanding of both GLUT1 and GLUT12 protein expression in benign and malignant prostate tissue and to determine whether expression is altered with an increase in tumor aggression. A significant change in GLUT1 or GLUT12 protein expression may prove a beneficial diagnostic marker of early malignancy.

Acknowledgements

The authors thank Maria Macheda, Susan Docherty, and Peter Tonoli for their assistance and advice.