- Email: [email protected]
Inhibition of glucose transporter gene expression by antisense nucleic acids in HL-60 leukemia cells
Life Sciences, Vol. 65, No. 1, pp. 63-70, 19w Copyright 0 1999 Wsevier Scicncc Inc. Prinred in the USA. All rights resewed ocm-3205/!39/s-see front matter
PII s0024-3205(99)00219-2
INHIBITION OF GLUCOSE TRANSPORTER GENE EXPRESSION BY ANTISENSE NUCLEIC ACIDS IN HL-60 LEUKEMIA CELLS Judy Yuet-Wa Chan, Siu-Kai Kong, Yuen-Min Choy, Cheuk-Yu Lee, Kwok-Pui Fung Department of Biochemistry, The Chinese University of Hong Kong, Shatin, N.T., Hong Kong, China. (Received in final form March 12, 1999)
Summary
Glucose is the basic source of energy for mammalian cells. The energy-independent transport of glucose down its concentration gradient is mediated by the facilitative glucose transporter family (GLUT). It has long been recognised that glucose transporter genes are overexpressed in many human cancer cells, to help provide extra energy for the rapid growth of cancer cells. In the present study, antisense oligonucleotides and plasmid-derived antisense RNA against GLUT-l gene were synthesized and transfected into human leukemia HL-60 cells to investigate the effect of these antisense nucleic acids on tumour growth. Our results show that antisense nucleic acids inhibited the proliferation of HL-60 cells by 50-60% and the mRNA expression of GLUT- 1 gene was suppressed as detected by Northern hybridization. Key Worde glucosetransporter,antisensenucleic acid, HL-60 cells
The energy independent transport of glucose across the plasma membrane is carried out by members of the facilitative glucose transporter family (GLUT). This family belongs to a much larger superfamily of 12 transmembrane segment transporters (1). Recently, cDNA clones encoding six structurally-related proteins with the properties of facilitative glucose transporters have been isolated and characterized (2). The human genes encoding these proteins are named GLUT l-5 and GLUT-7. It.has long been recognized that malignant cells show increased glucose uptake and utilization when compared to their nonmalignant counterparts (3). This uptake was shown to be mediated by the facilitative glucose transporters. Of the six isoforms, GLUT- 1 appears to be the most ubiquitously distributed (4). This isoform is overexpressed in many human cancer cells. It was postulated that the overuptake of glucose may provide extra energy for cancer cell growth. Down-regulating or specifically turning off the expression of individual gene by antisense nucleotides provides an opportunity for studying the specific biological role of the gene (5-7). Corresponding author: Prof. Kwok-Pui Fung, Department of Biochemistry, Room 3 16, Basic Medical Sciences Building, The Chinese University of Hong Kong, Shatin, N.T., Hong Kong, China. Phone: (852)-26096873, FAX: (852)-26035123, E-mail: [email protected]
64
Vol. 65, No. 1, 1999
Antisense Nucleic Acids Inhibit GLUT Gene Expression
When the nucleic acid sequence of the target gene is known, molecules can be produced to bind to its complementary messenger RNA sequence and prevent translation of the mRNA. The use of antisense technology in treating cancer has been widely reported in tissue culture system (8,9) and in vivo studies (lo,1 1) Antisense DNA oligonucleotides and expressed antisense RNA which have complementary sequence against GLUT-l were synthesized. Due to the hydrophilic nature of the synthesized molecules, cationic liposome was used to mediate the transfer of antisense nucleic acids into the cells. Experimental results show that these antisense nucleic acids could inhibit the proliferation of human leukemia HL-60 cells. In addition, the mRNA level of GLUT-l was shown to be decreased when the antisense nucleic acids were taken up by the cells. Materials and Methods Cell Culture The promyelocytic leukemia cell line HL-60 was obtained from American Tissue Culture Collection. These cells were continuously maintained in suspension and showed lymphoblastlike morphology. The cells were grown in RPM1 1640 medium supplemented with 10% fetal calf serum (FCS) at 37°C in a 5% CO, incubator. Synthesis ofphosphorothioated oligonucleotiaks Phosphorothioated oligonucleotides (S-oligos) were synthesized in an automated oligonucleotide synthesizer (The Gene Assembler Special, Pharmacia Biotech). Stepwise sulphurisation was performed with the Beaucage Thiolating Reagent (Gibco BRL). S-oligos were purified by Sephadex G-25 column and sterilized by filtration through 0.22 pm filters. The sequences of the three S-oligos used in this study are shown as belows. The S-oligos were deriv’ed from the translation initiation region at the position 169 of the cDNA sequence of GLUT-l. The sense and scramble sequences were synthesized for use as controls, G169 sense 5’ GCA GCG CTG CCA TGG 3’ G169 antisense 5’ CCA TGG CAG CGC TGC 3’ G169 scramble 5’ GGT AGC GCA GCT GCC 3’ Inhibition of cell proliferation I-IL-60 cells were resuspended in 0.1 ml of plain medium without FCS into four wells of a 96well microtiter plate at concentration of 10’ cells/ml. Sense, antisense and scrambled oligonucleotides were incubated with cationic liposome (Lipofectin Reagent, Gibco BRL) at a ratio of 1:5 (w/w) for 45 minutes and then were added to the cell suspension directly at final concentration of 10 PM. After incubation at 37’C / 5% CO, for 24 hours, 0.1 ml medium with 20% FCS were added to make the final serum concentration to be 10%. The cells were further incubated for 8, 12, 16, 20, 24, 30, 36 and 48 hours. MTT assays were performed at each time point. For the MTT assay, 50 microlitres of MTT solution (5 mg/ml) was added to each set of experiment and incubated at 37°C for 1 hour. Then, 50 ~1 of acidified (0.04 N HCl) 10% SDS was added and further incubated for 30 minutes. One hundred microlitres of isopropanol was added and incubated for 15 minutes to dissolve any crystals formed. After brief shaking, the absorbance at wavelength 540 nm was measured. The percentage cell proliferation at each time point was calculated as follows: % cell proliferation
=
Absorbance (treated sample) Absorbance (negative control)
x 100%
Vol. 65, No. 1, 1‘39
6.5
Antisense Nucleic Acids Inhibit GLUT Gene Expression
Plasmid construction and transfection
An antisense GLUT-l RNA expression plasmid (pRc-GLUTlAS) was constructed by ligating the cDNA into a plasmid vector pRc/CMV carrying the CMV promoter upstream from the multi-cloning site, the SV40 ori/early promotor, and the neo gene. Briefly, two strands of GLUT-l full-length cDNA (1476-bp) was obtained by PCR. The DNA strands was ligated between the Hind III and Not 1 sites in the multi-cloning site of pRc/CMV, one in sense and another in antisense orientation. The respective plasmid was transfected into HL-60 cells using Lipofectamine Reagent (Gibco BRL, USA) for 24 hours. Then, medium with 20% FCS was added to make the final serum concentration to be lo%, and the cells were incubated for another 8, 16, 24, 32,40, 48, 72, 96, 120 hours. rP]-labeled
oligonucleotides
uptake assay
The assay procedure was modified from Wickstrom (9). [y-“P]dCTP was purchased from Amersham. Oligonucleotides were 5’ labeled with 1 pCi of [y-32P]dCTP (3,000 Ci/mmol, Amersham) by 1 pl of T4 kinase (1 U/pi, GibcoBRL). The labeled oligonucleotides were purified by NucTrap purification column (Stratagene, U.S.A.). The column was first equilibrated by 70 pl of TE buffer (pH 8.0). [y-32P]dCTPthat had not been incorporated into the oligonucleotides was removed by applying the reaction mixture into the equilibrated column. The radiolabeled oligonucleotide was then eluted by 70 pl of TE buffer. For each time point, 5x10’ cpm of [5’-32P]oligonucleotide was added to 5~10~HL-60 cells in 4 ml of RPMI 1640 without serum. Each sample was incubated at 37°C for 0, 4, 8, 20, 26, 30 hours respectively and then sedimented at 1200 x g for 5 minutes. The supematant was removed and saved. The cells were washed twice with PBS and the wash was saved. The cell pellet was lysed with 0.1 ml of PBS with 1% SDS and then extracted with 0.1 ml of phenol. After centrifugation, the aqueous phase was removed and the phenol was extracted with another 0.1 ml of distilled water. The culture medium supematant, the cell wash and the aqueous extracts were analysed by scintillation counting. In order to determine whether the oligonucleotides could be taken up by I-IL-60 cells, oligonucleotide uptake assays were performed. Radioactivity retained by the washed cell pellets was compared with that left in the culture medium and the wash. The percentage of oligonucleotide uptake was calculated as follows: % oligonucleotide uptake =
counts per minute (cell pellet) counts per minute (cell pellet, wash and medium)
x 100%
Northern hybridization
Total RNA was isolated from HL-60 cells by TRIzol Reagent (Gibco BRL). Approximately 30-40 pg of total RNA from each sample were subjected to 1% agarose-6% formaldehyde electrophoresis and then transferred to a PVDF membrane by capillary action. cDNA probes of GLUT-l were labeled with [32P]dCTP using the Ready-To-Go DNA labeling kit (Pharmacia Biotech). Hybridization of the membrane to 32P-labeled cDNA was performed overnight in a solution of 50% formamide, 6X SSPE, 0.5% SDS, 5X Denhart’s solution and 4 mg denatured Salmon sperm DNA. Blots were washed twice each for 10 minutes in 1X SSC, 0.1% SDS and twice each for 5 minutes at 42°C in 0.2X SSC, 0.1% SDS. It was then exposed to X-ray film for 24 hours, and checked for X-ray film saturation. The amount of RNA for each treatment was normalized by reprobing the same membrane with S-actin cDNA.
66
Antisense
Nucleic Acids Inhibit GLUT
Vol. 65, No. 1, 1999
Gene Expression
Results Since oligonucieotides are easily degraded by nuclease present in the serum, the cells were incubated with serum free medium containing expressed antisense RNA and antisense oligonucleotides against GLUT-l for the first 24-hour uptake period. The original 10% serum concentration in the incubation medium was then resumed by adding medium with 20% serum. W.hen HL-60 cells were treated with the anti-GLUT-l oligonucleotide at a concentration of 10 ~J-Mfor 36 hours, about 25% inhibition of HL-60 cell proliferation was observed (Fig. 1). In contrast, cultures treated with either the sense or the scramble oligonucleotides grew at a similar rate to the untreated culture. For expressed antisense RNA, 55% inhibition of cell growth was observed at 72 hours after transfection (Fig. 2). Plasmid vector without insert and the sense clone were used as controls. Neither showed inhibitory effects on cell proliferation. Concerning the uptake of the nucleotides, about 3-4% of the labeled oligonucleotides were found associated with the cell pellet after 20 hours, a proportion that remained about the same up to 30 hours (Fig. 3). When liposome was used, the percentage uptake increased. So, a 24hour period was set as the transfection time in our studies. m-60 cells were treated with either antisense RNA or antisense oligonucleotides. RNA was then extracted from these cells and Northern hybridization was performed to detect GLUT-l mRNA. The result showed that GLUT-l mRNA levels were downregulated by 20% compared with control by the antisense oligonucleotides (Fig. 4). When the cells were treated with antisense RNA, 40% decrease in GLUT-l mRNA levels compared with control were observed (Fig. 5)
65
J:-~-0
l_&---
10
---+-~
20
incubation
_’
+~~
t
30
40
50
Time
(hours)
1 60
Fig. 1 The effect of different oligonucleotides on HL-60 ceil proliferation The data are expressed as mean t S D (n=5)
Antisense Nucleic Acids Inhibit GLUT
Vol. 65, No. 1, 1999
67
Gene Expression
120 110 %
100 SO
c L e
80
L %
70
5:
60 50 40 30 0
20
40
60
00
lncubatlon
Tlme
100
120
140
(hours)
Fig. 2 The effect on HL-60 cell proliferation by incubating with the expressed nucleic acid for different time intervals. The data are expressed as mean f SD. (n=5).
6
4 10
15
Incubation
20
Time
25
30
(hours)
Fig. 3 The percentage uptake of oligonucleotide in HL-60 cells. The data are expressed as mean k S.D. (n=5).
35
Antisense Nucleic Acids Inhibit GLUT Gene Expression
68
Vol. 65, No. 1, 1999
Fig. 4 Northern analysis of GLUT-l mRNA expression in HL-60 cells. Lane 1: negative control; Lane 2: sense; Lane 3: antisense oligonucleotide; Lane 4: scramble control
2
3
-28s
2.8kbw
-18s
Fig. 5 Northern analysis of GLUT- 1 mRNA expression in HL-60 cells.Lane 1: negative control; Lane 2: sense; Lane 3 : expressed antisense RNA; Lane 4: vector control.
Vol. 65, No. 1, 1999
Antisense Nucleic Acids Inhibit GLUT Gene Expression
Discussion The diversity of glucose transporters expressed in mammalian tissues reflects the importance of glucose as a source of metabolic energy, and different glucose transporters have evolved to facilitate dietary uptake and tissue-specific utilization. The presence ,of GLUT-l mRNA in almost all human tissues suggests that this isoform might mediate basal glucose uptake. Significant increases in mRNA for GLUT-l in cancers of the esophagus, colon, pancreas, and many other malignancies suggest its importance in tumor biology. In our study, the expression of GLUT-l was suppressed both by antisense oligonucleotides and by expressed antisense RNA. For antisense oligonucleotides designed around the translation start codon, a transient inhibition of HL-60 cell proliferation of about 2530% could be detected by MTT assay. Maximum inhibition appeared at about 36 hours after the oligonucleotides were taken up by the cells (i.e. 60 hours after addition of oligonucleotide to the cells). Results from Northern analysis showed a 20% decrease in GLUT-l mRNA compared with control by treatment of antisense oligonucleotides. The specificity of this inhibition was demonstrated by the fact that neither the sense oligomer nor the scramble oligonucleotides had any effect on cell proliferation. The sequence of scramble oligonucleotide does not match with other cDNA sequence when compared with the sequence data in Genebank. This scramble oligonucleotide therefore did not hybridize to any DNA strand within cells and produce non-specific antisense effects. Where inhibition of cell proliferation by antisense oligonucleotides is concerned, only about 30% of inhibition was achieved and the effect was transient. In our study, we also used a mammalian expression vector, pRc/CMV, to express antisense RNA. After cloning the sense and antisense DNA strands into the vector, large scale plasmid preparation was performed. Untreated vectors were used as controls for the nonspecific effects of transformation or plasmid integration. Results showed more than 50% cell proliferationn inhibition as detected by MTT assay. By Northern analysis, the expression of GLUT-l mRNA was decreased. When comparing the efficiencies of oligonucleotides. and expressed antisense RNA in inhibiting GLUT 1 gene expression, expressed antisense RNA produced much greater effects in inhibiting HL-60 proliferation and suppressing the expression of GLUT-l mRNA levels in the cells. This might be due to the greater number of antisense strands produced by the vector than by the oligonucleotides because of the probable digestion of oligonucleotides by DNase present in the medium. The antisense of effect produced by antisense oligonucleotides was transient and the effect was observed within 36 hours after oligonucleotide uptake. For expressed nucleotides, the effect could be observed only after 72 hours. More time was required for the vector to express the antisense nucleotides. The next step will be in vivo experiments as a prelude to possible applications in humans. As GLUT-l appears in almost all cell types, targeting the antisense nucleic acids to specific cancer cells without affecting normal cells will be a challenging task. One method would be to link the liposome to an antibody specific to the antigen on cancer cells (12). Oligonucleotide uptake efficiency can be further improved by using liposomes of different formulations. The nature and proportion of ionic compound and lipid in liposome can be varied to give maximum oligonucleotide uptake depending on cancer cell type and oligonucleotide modification.
69
70
Antisense Nucleic Acids Inhibit GLUT Gene Expression
Vol. 65, No. 1, 1999
Acknowledgements This study was supported by direct grant and earmarked grants from Research Council,, Hong Kong (CUHK 4060/97M and CUHK 4148/98M)
Grants
References 1. G.I. BELL, CF. BURANT, J. TAKEDA AND G.W. GOULD, J. Biol. Chem. 268 1916119164 (1993). 2. M. MUECKLER, Eur. J. Biochem. 219 713-725 (1994). 3. T. YAMAMOTO, Y. SEINO, H. FUKUMOTO, G. KOH, H. YANO, N. INAGAKI, Y. YAMADA, K. INOUE, T. MANABE AND H. IMURA, Biochem. Biophys. Res. Commun. 170 223-230 (1990). 4. M. YOUNES, L.V. LECHAGO, J.R. SOMOANO, M. MOSHARAF AND J. LECHAGO, Cancer Res. 56 1164- 1167 (1996). 5. F.K. ASKARI AND W.M. MCDONNELL, New Engl. J. Med. 334 3 16-3 18 (1996). 6. R.W. WAGNER, Nature 372 333-335 (1994). 7. M.K. GHOSH AND J.S. COHEN, Prog. Nucleic Acid Res. Mol. Biol. 42 79-126 (1992). 8. M. GOSLING, S. AKHTAR, J.W. SMITH AND D.R. POYNER, Biochem. Biophys. Res. Commun. 227 15-19 (1996) 9. E. WICKSTROM, Antisense RNA and DNA, J.A.H. Murray (Ed), 3 17-334, Wiley-L& New York (1992). 10. R. OBERBAUER, G.F. SCHREINER J. BIBER, H. MURER AND T.W. MEYER, Proc. Natl. Acad. Sci. 93 4903-4906 (1996). 11. A. VALERA, G. SOLANES, J. FERNANDEZ-ALVAREZ, A. PUJOL, J. FERRER G. ASINS, R. GOMIS AND F. BOSCH, J. Biol. Chem. 269 28543-28546 (1994). 12. K. MARUYAMA, T. TAKIZAWA, T. YUDA, S.J. KENNEL, L. HUANG AND M. IWATSURU, Biochim. Biophys. Acta 1234 74-80 (1995).
PII s0024-3205(99)00219-2
INHIBITION OF GLUCOSE TRANSPORTER GENE EXPRESSION BY ANTISENSE NUCLEIC ACIDS IN HL-60 LEUKEMIA CELLS Judy Yuet-Wa Chan, Siu-Kai Kong, Yuen-Min Choy, Cheuk-Yu Lee, Kwok-Pui Fung Department of Biochemistry, The Chinese University of Hong Kong, Shatin, N.T., Hong Kong, China. (Received in final form March 12, 1999)
Summary
Glucose is the basic source of energy for mammalian cells. The energy-independent transport of glucose down its concentration gradient is mediated by the facilitative glucose transporter family (GLUT). It has long been recognised that glucose transporter genes are overexpressed in many human cancer cells, to help provide extra energy for the rapid growth of cancer cells. In the present study, antisense oligonucleotides and plasmid-derived antisense RNA against GLUT-l gene were synthesized and transfected into human leukemia HL-60 cells to investigate the effect of these antisense nucleic acids on tumour growth. Our results show that antisense nucleic acids inhibited the proliferation of HL-60 cells by 50-60% and the mRNA expression of GLUT- 1 gene was suppressed as detected by Northern hybridization. Key Worde glucosetransporter,antisensenucleic acid, HL-60 cells
The energy independent transport of glucose across the plasma membrane is carried out by members of the facilitative glucose transporter family (GLUT). This family belongs to a much larger superfamily of 12 transmembrane segment transporters (1). Recently, cDNA clones encoding six structurally-related proteins with the properties of facilitative glucose transporters have been isolated and characterized (2). The human genes encoding these proteins are named GLUT l-5 and GLUT-7. It.has long been recognized that malignant cells show increased glucose uptake and utilization when compared to their nonmalignant counterparts (3). This uptake was shown to be mediated by the facilitative glucose transporters. Of the six isoforms, GLUT- 1 appears to be the most ubiquitously distributed (4). This isoform is overexpressed in many human cancer cells. It was postulated that the overuptake of glucose may provide extra energy for cancer cell growth. Down-regulating or specifically turning off the expression of individual gene by antisense nucleotides provides an opportunity for studying the specific biological role of the gene (5-7). Corresponding author: Prof. Kwok-Pui Fung, Department of Biochemistry, Room 3 16, Basic Medical Sciences Building, The Chinese University of Hong Kong, Shatin, N.T., Hong Kong, China. Phone: (852)-26096873, FAX: (852)-26035123, E-mail: [email protected]
64
Vol. 65, No. 1, 1999
Antisense Nucleic Acids Inhibit GLUT Gene Expression
When the nucleic acid sequence of the target gene is known, molecules can be produced to bind to its complementary messenger RNA sequence and prevent translation of the mRNA. The use of antisense technology in treating cancer has been widely reported in tissue culture system (8,9) and in vivo studies (lo,1 1) Antisense DNA oligonucleotides and expressed antisense RNA which have complementary sequence against GLUT-l were synthesized. Due to the hydrophilic nature of the synthesized molecules, cationic liposome was used to mediate the transfer of antisense nucleic acids into the cells. Experimental results show that these antisense nucleic acids could inhibit the proliferation of human leukemia HL-60 cells. In addition, the mRNA level of GLUT-l was shown to be decreased when the antisense nucleic acids were taken up by the cells. Materials and Methods Cell Culture The promyelocytic leukemia cell line HL-60 was obtained from American Tissue Culture Collection. These cells were continuously maintained in suspension and showed lymphoblastlike morphology. The cells were grown in RPM1 1640 medium supplemented with 10% fetal calf serum (FCS) at 37°C in a 5% CO, incubator. Synthesis ofphosphorothioated oligonucleotiaks Phosphorothioated oligonucleotides (S-oligos) were synthesized in an automated oligonucleotide synthesizer (The Gene Assembler Special, Pharmacia Biotech). Stepwise sulphurisation was performed with the Beaucage Thiolating Reagent (Gibco BRL). S-oligos were purified by Sephadex G-25 column and sterilized by filtration through 0.22 pm filters. The sequences of the three S-oligos used in this study are shown as belows. The S-oligos were deriv’ed from the translation initiation region at the position 169 of the cDNA sequence of GLUT-l. The sense and scramble sequences were synthesized for use as controls, G169 sense 5’ GCA GCG CTG CCA TGG 3’ G169 antisense 5’ CCA TGG CAG CGC TGC 3’ G169 scramble 5’ GGT AGC GCA GCT GCC 3’ Inhibition of cell proliferation I-IL-60 cells were resuspended in 0.1 ml of plain medium without FCS into four wells of a 96well microtiter plate at concentration of 10’ cells/ml. Sense, antisense and scrambled oligonucleotides were incubated with cationic liposome (Lipofectin Reagent, Gibco BRL) at a ratio of 1:5 (w/w) for 45 minutes and then were added to the cell suspension directly at final concentration of 10 PM. After incubation at 37’C / 5% CO, for 24 hours, 0.1 ml medium with 20% FCS were added to make the final serum concentration to be 10%. The cells were further incubated for 8, 12, 16, 20, 24, 30, 36 and 48 hours. MTT assays were performed at each time point. For the MTT assay, 50 microlitres of MTT solution (5 mg/ml) was added to each set of experiment and incubated at 37°C for 1 hour. Then, 50 ~1 of acidified (0.04 N HCl) 10% SDS was added and further incubated for 30 minutes. One hundred microlitres of isopropanol was added and incubated for 15 minutes to dissolve any crystals formed. After brief shaking, the absorbance at wavelength 540 nm was measured. The percentage cell proliferation at each time point was calculated as follows: % cell proliferation
=
Absorbance (treated sample) Absorbance (negative control)
x 100%
Vol. 65, No. 1, 1‘39
6.5
Antisense Nucleic Acids Inhibit GLUT Gene Expression
Plasmid construction and transfection
An antisense GLUT-l RNA expression plasmid (pRc-GLUTlAS) was constructed by ligating the cDNA into a plasmid vector pRc/CMV carrying the CMV promoter upstream from the multi-cloning site, the SV40 ori/early promotor, and the neo gene. Briefly, two strands of GLUT-l full-length cDNA (1476-bp) was obtained by PCR. The DNA strands was ligated between the Hind III and Not 1 sites in the multi-cloning site of pRc/CMV, one in sense and another in antisense orientation. The respective plasmid was transfected into HL-60 cells using Lipofectamine Reagent (Gibco BRL, USA) for 24 hours. Then, medium with 20% FCS was added to make the final serum concentration to be lo%, and the cells were incubated for another 8, 16, 24, 32,40, 48, 72, 96, 120 hours. rP]-labeled
oligonucleotides
uptake assay
The assay procedure was modified from Wickstrom (9). [y-“P]dCTP was purchased from Amersham. Oligonucleotides were 5’ labeled with 1 pCi of [y-32P]dCTP (3,000 Ci/mmol, Amersham) by 1 pl of T4 kinase (1 U/pi, GibcoBRL). The labeled oligonucleotides were purified by NucTrap purification column (Stratagene, U.S.A.). The column was first equilibrated by 70 pl of TE buffer (pH 8.0). [y-32P]dCTPthat had not been incorporated into the oligonucleotides was removed by applying the reaction mixture into the equilibrated column. The radiolabeled oligonucleotide was then eluted by 70 pl of TE buffer. For each time point, 5x10’ cpm of [5’-32P]oligonucleotide was added to 5~10~HL-60 cells in 4 ml of RPMI 1640 without serum. Each sample was incubated at 37°C for 0, 4, 8, 20, 26, 30 hours respectively and then sedimented at 1200 x g for 5 minutes. The supematant was removed and saved. The cells were washed twice with PBS and the wash was saved. The cell pellet was lysed with 0.1 ml of PBS with 1% SDS and then extracted with 0.1 ml of phenol. After centrifugation, the aqueous phase was removed and the phenol was extracted with another 0.1 ml of distilled water. The culture medium supematant, the cell wash and the aqueous extracts were analysed by scintillation counting. In order to determine whether the oligonucleotides could be taken up by I-IL-60 cells, oligonucleotide uptake assays were performed. Radioactivity retained by the washed cell pellets was compared with that left in the culture medium and the wash. The percentage of oligonucleotide uptake was calculated as follows: % oligonucleotide uptake =
counts per minute (cell pellet) counts per minute (cell pellet, wash and medium)
x 100%
Northern hybridization
Total RNA was isolated from HL-60 cells by TRIzol Reagent (Gibco BRL). Approximately 30-40 pg of total RNA from each sample were subjected to 1% agarose-6% formaldehyde electrophoresis and then transferred to a PVDF membrane by capillary action. cDNA probes of GLUT-l were labeled with [32P]dCTP using the Ready-To-Go DNA labeling kit (Pharmacia Biotech). Hybridization of the membrane to 32P-labeled cDNA was performed overnight in a solution of 50% formamide, 6X SSPE, 0.5% SDS, 5X Denhart’s solution and 4 mg denatured Salmon sperm DNA. Blots were washed twice each for 10 minutes in 1X SSC, 0.1% SDS and twice each for 5 minutes at 42°C in 0.2X SSC, 0.1% SDS. It was then exposed to X-ray film for 24 hours, and checked for X-ray film saturation. The amount of RNA for each treatment was normalized by reprobing the same membrane with S-actin cDNA.
66
Antisense
Nucleic Acids Inhibit GLUT
Vol. 65, No. 1, 1999
Gene Expression
Results Since oligonucieotides are easily degraded by nuclease present in the serum, the cells were incubated with serum free medium containing expressed antisense RNA and antisense oligonucleotides against GLUT-l for the first 24-hour uptake period. The original 10% serum concentration in the incubation medium was then resumed by adding medium with 20% serum. W.hen HL-60 cells were treated with the anti-GLUT-l oligonucleotide at a concentration of 10 ~J-Mfor 36 hours, about 25% inhibition of HL-60 cell proliferation was observed (Fig. 1). In contrast, cultures treated with either the sense or the scramble oligonucleotides grew at a similar rate to the untreated culture. For expressed antisense RNA, 55% inhibition of cell growth was observed at 72 hours after transfection (Fig. 2). Plasmid vector without insert and the sense clone were used as controls. Neither showed inhibitory effects on cell proliferation. Concerning the uptake of the nucleotides, about 3-4% of the labeled oligonucleotides were found associated with the cell pellet after 20 hours, a proportion that remained about the same up to 30 hours (Fig. 3). When liposome was used, the percentage uptake increased. So, a 24hour period was set as the transfection time in our studies. m-60 cells were treated with either antisense RNA or antisense oligonucleotides. RNA was then extracted from these cells and Northern hybridization was performed to detect GLUT-l mRNA. The result showed that GLUT-l mRNA levels were downregulated by 20% compared with control by the antisense oligonucleotides (Fig. 4). When the cells were treated with antisense RNA, 40% decrease in GLUT-l mRNA levels compared with control were observed (Fig. 5)
65
J:-~-0
l_&---
10
---+-~
20
incubation
_’
+~~
t
30
40
50
Time
(hours)
1 60
Fig. 1 The effect of different oligonucleotides on HL-60 ceil proliferation The data are expressed as mean t S D (n=5)
Antisense Nucleic Acids Inhibit GLUT
Vol. 65, No. 1, 1999
67
Gene Expression
120 110 %
100 SO
c L e
80
L %
70
5:
60 50 40 30 0
20
40
60
00
lncubatlon
Tlme
100
120
140
(hours)
Fig. 2 The effect on HL-60 cell proliferation by incubating with the expressed nucleic acid for different time intervals. The data are expressed as mean f SD. (n=5).
6
4 10
15
Incubation
20
Time
25
30
(hours)
Fig. 3 The percentage uptake of oligonucleotide in HL-60 cells. The data are expressed as mean k S.D. (n=5).
35
Antisense Nucleic Acids Inhibit GLUT Gene Expression
68
Vol. 65, No. 1, 1999
Fig. 4 Northern analysis of GLUT-l mRNA expression in HL-60 cells. Lane 1: negative control; Lane 2: sense; Lane 3: antisense oligonucleotide; Lane 4: scramble control
2
3
-28s
2.8kbw
-18s
Fig. 5 Northern analysis of GLUT- 1 mRNA expression in HL-60 cells.Lane 1: negative control; Lane 2: sense; Lane 3 : expressed antisense RNA; Lane 4: vector control.
Vol. 65, No. 1, 1999
Antisense Nucleic Acids Inhibit GLUT Gene Expression
Discussion The diversity of glucose transporters expressed in mammalian tissues reflects the importance of glucose as a source of metabolic energy, and different glucose transporters have evolved to facilitate dietary uptake and tissue-specific utilization. The presence ,of GLUT-l mRNA in almost all human tissues suggests that this isoform might mediate basal glucose uptake. Significant increases in mRNA for GLUT-l in cancers of the esophagus, colon, pancreas, and many other malignancies suggest its importance in tumor biology. In our study, the expression of GLUT-l was suppressed both by antisense oligonucleotides and by expressed antisense RNA. For antisense oligonucleotides designed around the translation start codon, a transient inhibition of HL-60 cell proliferation of about 2530% could be detected by MTT assay. Maximum inhibition appeared at about 36 hours after the oligonucleotides were taken up by the cells (i.e. 60 hours after addition of oligonucleotide to the cells). Results from Northern analysis showed a 20% decrease in GLUT-l mRNA compared with control by treatment of antisense oligonucleotides. The specificity of this inhibition was demonstrated by the fact that neither the sense oligomer nor the scramble oligonucleotides had any effect on cell proliferation. The sequence of scramble oligonucleotide does not match with other cDNA sequence when compared with the sequence data in Genebank. This scramble oligonucleotide therefore did not hybridize to any DNA strand within cells and produce non-specific antisense effects. Where inhibition of cell proliferation by antisense oligonucleotides is concerned, only about 30% of inhibition was achieved and the effect was transient. In our study, we also used a mammalian expression vector, pRc/CMV, to express antisense RNA. After cloning the sense and antisense DNA strands into the vector, large scale plasmid preparation was performed. Untreated vectors were used as controls for the nonspecific effects of transformation or plasmid integration. Results showed more than 50% cell proliferationn inhibition as detected by MTT assay. By Northern analysis, the expression of GLUT-l mRNA was decreased. When comparing the efficiencies of oligonucleotides. and expressed antisense RNA in inhibiting GLUT 1 gene expression, expressed antisense RNA produced much greater effects in inhibiting HL-60 proliferation and suppressing the expression of GLUT-l mRNA levels in the cells. This might be due to the greater number of antisense strands produced by the vector than by the oligonucleotides because of the probable digestion of oligonucleotides by DNase present in the medium. The antisense of effect produced by antisense oligonucleotides was transient and the effect was observed within 36 hours after oligonucleotide uptake. For expressed nucleotides, the effect could be observed only after 72 hours. More time was required for the vector to express the antisense nucleotides. The next step will be in vivo experiments as a prelude to possible applications in humans. As GLUT-l appears in almost all cell types, targeting the antisense nucleic acids to specific cancer cells without affecting normal cells will be a challenging task. One method would be to link the liposome to an antibody specific to the antigen on cancer cells (12). Oligonucleotide uptake efficiency can be further improved by using liposomes of different formulations. The nature and proportion of ionic compound and lipid in liposome can be varied to give maximum oligonucleotide uptake depending on cancer cell type and oligonucleotide modification.
69
70
Antisense Nucleic Acids Inhibit GLUT Gene Expression
Vol. 65, No. 1, 1999
Acknowledgements This study was supported by direct grant and earmarked grants from Research Council,, Hong Kong (CUHK 4060/97M and CUHK 4148/98M)
Grants
References 1. G.I. BELL, CF. BURANT, J. TAKEDA AND G.W. GOULD, J. Biol. Chem. 268 1916119164 (1993). 2. M. MUECKLER, Eur. J. Biochem. 219 713-725 (1994). 3. T. YAMAMOTO, Y. SEINO, H. FUKUMOTO, G. KOH, H. YANO, N. INAGAKI, Y. YAMADA, K. INOUE, T. MANABE AND H. IMURA, Biochem. Biophys. Res. Commun. 170 223-230 (1990). 4. M. YOUNES, L.V. LECHAGO, J.R. SOMOANO, M. MOSHARAF AND J. LECHAGO, Cancer Res. 56 1164- 1167 (1996). 5. F.K. ASKARI AND W.M. MCDONNELL, New Engl. J. Med. 334 3 16-3 18 (1996). 6. R.W. WAGNER, Nature 372 333-335 (1994). 7. M.K. GHOSH AND J.S. COHEN, Prog. Nucleic Acid Res. Mol. Biol. 42 79-126 (1992). 8. M. GOSLING, S. AKHTAR, J.W. SMITH AND D.R. POYNER, Biochem. Biophys. Res. Commun. 227 15-19 (1996) 9. E. WICKSTROM, Antisense RNA and DNA, J.A.H. Murray (Ed), 3 17-334, Wiley-L& New York (1992). 10. R. OBERBAUER, G.F. SCHREINER J. BIBER, H. MURER AND T.W. MEYER, Proc. Natl. Acad. Sci. 93 4903-4906 (1996). 11. A. VALERA, G. SOLANES, J. FERNANDEZ-ALVAREZ, A. PUJOL, J. FERRER G. ASINS, R. GOMIS AND F. BOSCH, J. Biol. Chem. 269 28543-28546 (1994). 12. K. MARUYAMA, T. TAKIZAWA, T. YUDA, S.J. KENNEL, L. HUANG AND M. IWATSURU, Biochim. Biophys. Acta 1234 74-80 (1995).