- Email: [email protected]
Estrogen receptor-mediated direct stimulation of colon cancer cell growth in vitro
ELSEVIER
Molecular and Cellular Endocrinology 105 (1994) 197-201
Estrogen receptor-mediated direct stimulation of colon cancer cell growth in vitro Xiaomeng Xu, Mary L. Thomas* Department of Pharmacology and Toxicology, University of Texas Medical Branch, Galveston, TX 77555, USA
Received 8 June 1994; accepted 19 August 1994
Abstract
In vivo and epidemiologicaldata suggest a mitogenicrole for estrogens (E) in colon cancer. The presence of estrogen receptor (ER) and ER mRNA in colonic epi~elium and coIon cancer cells, make it necessary to explore the possible direct effects of E on co1011cancer growth. In this study, a 1.5mer olig~eox~uc~eotide (oligo) antisense to the region of the ~sIation start codon of estrogen receptor mRNA inhibited ER expression in a mouse colon cancer cell line (MC-XJ), as decline by receptor binding assay. Antisense oligo ako decreased ER mRNA levels in MC-26cells. The cowl-stirnulato~ effect of E was abolished by antisense oligo treatment, demonstrating that the ER is directly involved in the regulation of colon cancer cell growth. Keywoti:
Colon cancer; Estrogen receptor; Antisense oligonucleotide; Cell growth
1. Introduction Possible roles for estrogens @) in colon cancers are suggested by several in vivo studies comparing the growth of normal colonic epi~elium and colon cancer cells in ovariectomized acd intact animals (Hoff et al., 1981; Narayan et al., 1992). Moreover, sex differences in the incidence and mortality of colon cancer have been shown in humans (Doll, 1980; McMichael et al., 1983). In addition, there are substantial similarities in the descriptive epidemiology of cancers of the large bowel and breast (Howell, 1976; Willett, 1989). For instance, within individual countries, there is a strong correlation between breast cancer and colon cancer mortality (Wynder et al., 1967); and in individual women, there is a higher than expected concurrence of breast cancer and colon cancer (Schoenberg et al., 1969). Although E has long been known to have stimulatory effects on cell proliferation in the target tissues of the breast and reproductive tract, intestinal epithelium was traditionally regarded as non-target tissue of E, and the in vivo mitogenic effects of E were considered as being mediated via indirect mechanisms. However, the presence of estrogen receptor (ER) and its mRNA in colorectal carcinomas and normal colonic epi*Corresponding
author. Tel. 409-772-9641. Fax 409-772-9642.
0303-7207/9~$07.~ SSDI
0 1994 Elsevier Science Ireland Ltd. AI1rights reserved
0303-7207(94)03381-3
thelium (McClendon et al., 1977; Alford et al., 1979; Markaverich et al., 1981; Bracali et al., 1988; Cameron et al., 1992; Singh et al., 1993; Thomas et al., 1993) indicate that it is necessary to explore possible direct effects of E on colon cancer cells. The use of the antiestrogens has been a valuable tool for determining the role of E in various ER mediated responses. However, when we carried out preliminary experiments using the antiestrogens tamoxifen and ICI 182,780 as tools to determine the specific role of estrogen in the regulation of colon cancer cell growth, we found that these agents directly inhibit the growth of cultured colon cancer cells even in the absence of E. Therefore, we have employed antisense oligodeoxynucleotides (oligo) to ER mRNA to block ER synthesis in MC-26 cells, a mouse colon cancer cell line, in order to demonstrate the ER mediated direct stimulation of MC-26 cell growth in vitro. 2. Materials and methods 2.1. Oligos A 15-mer antisense oligo (5’ GGGTCATGGTCATGG 3’) and a scrambled oligonucleotide (5’ GTGGTGGATCGTGAC 3’) were synthesized by Genosys (Woodlands, TX). The antisense oligo spanned the common translation start codon for mouse, rat and human ER mRNA. The
198
X. Xu, M.L. Thomas I Molecular and Cellular Endocrinology 105 (1994) 197-201
scrambled oligo contained the same 15 bases as the antisense oligo, but had little or no homology to any gene sequences submitted to GenBank. These oligos were similar to those which were used previously in vivo (McCarthy et al., 1993). However, the oligos were modified as phosphorothioates. They were further purified by NAP-5 column (Pharmacia, Uppsala, Sweden) before being used for cell culture.
0.66 M formaldehyde and transferred to a nylon membrane. Blots were hybridized with a random primer-labelled ER cDNA PCR fragment from rat uterus (Thomas et al., 1993), which has 95.4% sequence homology to the mouse ER cDNA sequence. The same blot was then reprobed with rat cyclophylin to normalize for the RNA loading (Danielson et al., 1988). Results of the Northern analyses were quantitated by densitometry.
2.2. Cell culture MC-26 is a 1,Zdimethylhydrazine induced mouse colon adenocarcinoma cell line that produces tumors in mice in a dose-dependent manner when injected subcutaneously (Waldrop et al., 1989). MC-26 cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM) containing 5% fetal bovine serum (FBS). For assessment of the effects of E on MC-26 cell growth, cells were plated (2000-3000 cells/well) into 24-multiwell plates in phenol red-free DMEM supplemented with 5% charcoal-treated FBS and allowed to attach for 48 h. The medium was then (day ‘0’) changed and 17/?-estradiol (E,) in absolute ethanol was added directly to the medium (final concentration of ethanol was less than 0.1%). The medium was changed and hormones added every 2 days. Control cells received equal concentrations of ethanol vehicle without hormones. On day 6 (or as described for each figure), cells were harvested by treatment with 0.025% trypsin/O.O2% EDTA and counted using a Coulter counter (Model ZF, Coulter Electronics, Hialeah, FL). Cell viability always exceeded 90% determined using trypan blue exclusion and a hemocytometer. In experiments with oligos, oligos were added as medium was changed every 2 days.
2.5. Statistics Statistical significance was determined using one-way analysis of variance with post hoc Bonferroni t-test. A P value co.05 was considered statistically significant.
2.3. ER binding assay: ER in MC-26 cells was measured using a whole cell assay as described previously (Thomas et al., 1993). Briefly, MC-26 cells were plated at 5 X lO“/well in phenol red-free DMEM supplemented with 5% charcoal-treated FBS. Two days after seeding, the medium was changed and the cells were treated with oligos for 48 h (or as stated). 48 h later, MC-26 cells were washed three times in PBS and ER binding was determined after a 2 h incubation with 1 nM [3H]E2 (Amersham, spec. act. 44.5 Ci/mmol) at room temperature. The non-specific binding was defined by using a 300-fold molar excess of diethylstilbestrol (DES). Preliminary results have shown that E2 binding in MC-26 cells was saturated at 1 nM, with Kd of approximately 0.3 nM. 2.4. Northern analysis of ER mRNA Subconfluent MC-26 cells grown in phenol red-free DMEM supplemented with 5% charcoal-treated FBS were treated wit.h 3 PM of either ER antisense oligo or scrambled oligo in the presence or absence of 1 nM Ez for 48 h. Total cellular RNA was then extracted as described by Chomczynski and Sacchi (Chomczynski et al., 1987). Total cellular RNA was separated on a 1% agarose gel containing
3. Results MC-26 cells grown under the conditions in our laboratory contain approximately 3400 specific high-affinity estrogen binding sites per cell. When MC-26 cells were treated with 1 ,uM ER antisense oligo, there was a significant decrease in cellular ER levels, as measured by ER binding assay (Fig. 1). This effect was seen as early as 6 h after antisense oligo treatment and appeared to plateau after 24 h, since treatment with antisense oligo for up to 72 h did not significantly further decrease ER binding (data not shown). Treatment with the same concentration of scrambled oligo for 48 h did not affect ER levels in MC-26 cells. Fig. 2 shows a dose dependent decrease of ER binding in MC-26 cells treated with antisense oligo for 48 h. The scrambled oligo had no effect at dose as high as 3 PM. To determine if antisense oligo also affected ER mRNA levels in MC-26 cells, the presence of ER mRNA in MC-26
* ‘t 0 antisense oligo &lscrambled oligo
*
60
0
6
12
24
-1
48
Time of treatment (hrs)
Fig. 1. Time course of ER binding decrease by rntisense oligo in MC-26 cells. MC-26 cells were treated with 1 PM of antisense oligo 6,12 24, or 48 h before ER binding was assayed. One group of cells was treated with I pM of scmmbled oligo 48 h before the binding assay. The control group received 10~1 of 10 mM sodium phosphate (pH 6.8), which was the elution buffer for NAP-5 column purification of the oligos. 1 nM [3H]E2 (44.5 Wmmol) was added to each well in the presence or absence of a 300-fold excess of unlabeled DES for 2 h at room tempemturc. Each bar represents the mean k SD for four wells expressed us 46 of control. *P < 0.05 compared to control.
X. Xu, M.L. Thomas I Molecular and Cellular Endocrinology 105 (1994) 197-201
” 3
140-
5
120-
199
150
0 control I!!!! antisense scrambled oli o P1go 0
Ocontrol lZXSEz(l nM)
8 3 : % E =%
125
100
> 4 c
75
z u 0
0.5
0.1
1
50
3
8
8
Days of Treatment
Concentration of Oligos (pM)
Fig. 2. The effect of different doses of antisense oligo on ER binding in MC-26 cells. MC-26 cells were treated with 0.1-3 PM antisense oligo for 48 h before the ER binding assay. One group of cells was treated with 3yM of scrambled oligo, and the control group received an equal volume of vehicle buffer. The binding assay was performed as described in Fig. 1. Each bar represents the mean f SD for four wells expressed as 8 of control. *P < 0.05 compared to control.
Fig. 4. E stimulation of MC-26 cell growth. MC-26 cells were treated with 1 nM % for 2-8 days. Control cells were treated with the same amount (0.1%) of ethanol vehicle alone. The medium was changed with E2 or vehicle replacement every 2 days. Each bar represents mean f SD for 4 wells expressed as % of control. The number of cells in the control groups wem 0.184, 0.615, 3.024, and 8.744 x lo5 cells/well for days 2, 4.6 and 8, respectively. *P < 0.05 compared to control.
cells treated with oligos was examined by Northern blot analyses (Fig. 3). As has been observed before in MC-26 cells without oligo treatment (unpublished data), there was a small increase (approximately 20%) in ER mRNA level in cells treated with scrambled oligo in the presence of & compared to that in cells treated with scrambled oligo in the absence of EZ. This is probably due to E upregulation of its own receptors. However, in cells treated with antisense oligo, the upregulation of ER mRNA level by E2 was not
seen. Treatment of MC-26 cells with 3pM of antisense oligo for 48 h resulted in decrease in ER mRNA levels compared to cells treated with the same concentration of scrambled oligo. Treatment of MC-26 cells with E2 stimulates cell growth. The maximum growth stimulation was achieved at 1 nM of E2 treatment (data not shown). The effect was most prominent after 6 days of continuous exposure to E2 (Fig. 4). ER antisense oligo was used in order to determine whether ER was involved the growth-stimulatory effects of E in MC-26 cells. The growth stimulatory effect of 1 nM E2 was completely blocked by 1 FM of antisense oligo (Fig. 5). Although statistically not significant, there was a
10
-0
Cyclophilin
I
ONo E2 6S3E2(1 nM)
-
Oligo: It,:
ERlcyclophilin
4
ratio:
SC
ASN
-+-
+
1.0 1.2 0.8 0.7 Control
Fig. 3. Northern analysis of the effect of antisense oligo on ER mRNA levels in MC-26 cells. MC-26 cells were treated with 3pM of either antisense oligo (ANS) or scrambled oligo (SC) in the presence (+) or absence (-) of I nM E2 for 48 h. Total cellular RNA was extracted and Northern blot analysis was performed. Each lane contained 25 pg of total RNA. The blot was hybridized with a 32P-labelled ER cDNA and reprobed with cyclophylin cDNA as described in Section 2. The ER mRNA was normalized to cyclophylin mRNA by densitometry, and the ER to cyclophylin ratio is indicated under each lane by designating the ratio from the cells treated with SC in the absence of E2 as 1.O.
ANS Oligo
SC Oligo
Fig. 5. Inhibition of the E growth-stimulatory effect by antisense oligo. MC-26 cells were treated with 1PM of either antisense (ANS) oligo or scrambled (SC) oligo in the presence or absence of 1 nM E2 for 6 days. Control cells were treated with the same amount of ethanol (0.1%) and elution buffer (10~1) or 1 nM E2 and elution buffer. The medium was changed with q and oligos, or ethanol and oligo. or vehicle replacement every 2 days. Each bar represents cell number mean i SD for 4 wells. *P < 0.05 compared to control without E2. **f < 0.05 compared to control with I nM E2.
200
X. Xu. M.L. Thomas I Molecular and Cellular Endocrinology 105 (1994) 197-201
small inhibition of cell growth in cells treated with antisense oligo, either with or without E2 addition. This could be due to the presence of residue E in the culture medium. However, the scrambled oligo had no effect on either basal or Ez stimulated cell growth. Consistent with the effects of the ER antisense oligo, the anti-estrogen, ICI 182,780 (10 nM) was also able to completely block the growthstimulatory effect of 1 nM Ez (data not shown). However, ICI 182,780 alone exhibited statistically significant concentration-dependent growth-inhibitory effects on MC-26 cells. 4. Discussion We have demonstrated that an antisense oligo targeting ER mRNA inhibited expression of ER in MC-26 cells as measured by receptor binding assay. The inhibition of ER synthesis and inhibition of Ez-stimulated MC-26 cell growth suggest that the ER-mediated growth-stimulatory effect of E is directly involved in the regulation of cell growth in this colon cancer cell line. This may be one of the mechanisms responsible for the mitogenic effect of E observed in colon cancer in vivo, and play a role in the epidemiological similarities between breast cancer and colon cancer. The mechanism of action of phosphorothioate oligonucleotides has been investigated in a number of studies. In addition to common mechanisms of action of modified and unmodified antisense oligonucleotides, unlike other modified oligonucleotides (e.g. a-oligonucleotides and methylphosphonates), phosphorothioates can act by promotion of mRNA degradation by an RNase H-dependent mechanism (Furdon et al., 1989; Crooke, 1992). On the other hand, unlike unmodified oligonucleotides, phosphorothioate oligos are nuclease-resistant (Campbell et al., 1990). Moreover, they appear to be adequately internalized in cells (Gao et al., 1989; Crook, 1991). For example, studies employing a 28-mer phosphorothioate deoxycytidine that was uniformly labeled with 35S demonstrated that when HeLa cells were incubated with 1 PM of the drug, significant intracellular concentrations were achieved. Cellular uptake reached a plateau in 6 h (Gao et al., 1989). In our experiments, the effect of antisense oligo to decrease ER binding appeared in less than 6 h after treatment. This could be due to the fact that ER is a rapidly turning over protein. Although not characterized in colon cancer cells, ER half-life has been well described to be of approximately 4 h in breast cancer and uterine cells in vitro and in vivo (Eckert et al., 1984; Nardulli et al., 1986). As determined by Northern analysis, antisense oligo treatment blocked the upregulation of ER mRNA level by E2 and it also slightly decreased MC-26 cell ER mRNA levels compared to cells treated with the same concentration of scrambled oligo. This effect was more prominent in the presence of E. These data suggest that E upregulation of ER mRNA levels may be ER-mediated and that E upregulation of its own receptor could be one of the mechanisms
of E stimulated cell growth in MC-26 cells. However, we did not directly address this by measuring ER levels in cells treated with oligos in the presence of E, due to the fact that the presense of E2 in the medium interferes the radioligand binding assay. Like most antisense oligo studies, we used the antisense oligo targeting the AUG translation initiation codon of ER mRNA. Targeting the AUG codon would predictably mask the ribosome recognition site and prevent the formation of the translation complex. However, since antisense oligo also decreased ER mRNA levels in MC-26 cells as determined by Northern analysis, the RNase H dependent mechanism seems to also play a role in antisense oligonucleotide inhibition of ER expression. It is possible that both RNase H-dependent and -independent mechanisms are involved in the inhibition of expression in MC-26 cells. Inhibition of ER expression plateaued 24-48 h after antisense oligo treatment. Longer than 48 h treatment with antisense oligo did not produce a further decrease in ER levels as measured by ER binding. However, antisense oligo treatment for 6 days completely blocked the growth stimulatory effect of E, in MC-26 cells. This suggests that E stimulation of MC-26 cell growth is ER-dependent and that it is necessary to have a sufficient number of ERs available to mediate this response. Based on the fact that the scrambled oligo had no effect on either receptor binding or MC-26 cell growth, the effect of antisense oligo is most probably not due to a toxic or non-specific effect of the oligos. In summary, we have demonstrated that antisense oligonucleotides are capable of specifically inhibiting the expression of ER in a cultured colon cancer cell line. Inhibition of expression with antisense oligo was monitored by receptor binding assay. Furthermore, inhibition of ER expression abolished the growth stimulatory effect of E in MC-26 cells, demonstrating that E is directly involved in the regulation of colon cancer cell growth. We believe that the ER mRNA antisense oligo approach is more specific than antiestrogens in demonstrating ER mediated actions, particularly for cell growth experiments, because it has been demonstrated that antiestrogens may have both ERdependent and -independent anti-growth activities. Therefore, antisense oligos may be useful as research tools and therapeutic agents for ER positive cancers such as breast cancer, and according to this study, colon cancer as well. Acknowledgement The MC-26 mouse colon cancer cell line was obtained from the laboratory of Dr. Courtney Townsend at the University of Texas Medical Branch in Galveston. References Alford, T.C., Do, H.M. and Gelihoed, G.W. ( 1979) Cancer 43,980-981. Brclcoli, G., Caracino, A.M.,
Rossodivita, F., Bianchi, C., Loli, M.G. and
Bracali, M. (1988) lnt. 1. Biol. Markers 3,41--28.
X. Xu, ML. Thomas I Molecuhr and Cellular Endocrinology 105 /1994) 197-201
Cameron, B.L., Butler, J.A.. Rutgers, J., Vargas, H.I., Purtell, M. and Sheppard, B. (1992) Am. Surg. 58,758-760. Campbell, J.M., Bacon, T.A. and Wickstrom, l?. (1990) J. Biochcm. Biophys, Methods 20,259-263. Chomczynski, P. and Sac&i, N. (1987) Anal. Biochem. 162,156159. Crook, R.M. (1991) Anticancer Drug Des. 6,609-646. Crooke. ST. (1992) Annu. Rev. Pharmacol. Toxicol. 32,329-76. Danielson, P.E., Forss-Petter. S., Brow, M.A. et al. (1988) DNA 4, 261267. Doll, R. General epidemiologic considerations in aetiology of colorectal cancer, in Colorectal Cancer: Prevention, Epidemiology and Scmening (Winawer, S., Schottenfeld, D. and Sherlock, P. eds.), Raven Press, New York, 1980, pp. l-12. Eckert, R.L., Mullick, A., Rorke, E.A. and Ka~nellen~gen, B.S. (1984) ~d~~nolo~ 114.629-637. Furdon, P., Dominski, 2. and Kole, R. (1989) Nucleic Acids Res. 17, 9193-9204. Gao, W., Stein, C.A., Cohen, J.S., Dutschman, G.E. and Cheng, C.-Y. (1989) J Biol Chem. 264, 11521-l 1526. Hoff, M.B., Chang, W.W.L. and Mak, K.M. (1981) Virchows Arch. (Cell Pathol.) 35263-273. Howell, M.A. (1976) J. Chron. Dis. 29.243-261.
201
Ma&cave&h, B.M., Williams, M., U~hu~h, S. and Clark, J.H. (1981) Endocrinology lOQ,63-69. McCarthy, M.M., Schlenker, B.H. and Pfaff, D.W. (1993) Endocrinology 133.433-439. McClendon, J.E., Appleby, D. and Claudon, D.B. (1977) Arch. Surg. 112,140-141. McMichael, A.J. aud Potter, J.D. (1983) Am. 3. Epidemiol. 118, 620627. Narayan, S., Rajukumar, G., Prouix, H. and Singh, P. (1992) Gastroenterology 103, 1823-1832. Nardutli, A.M. and Katzenellenbogen, B.S. (1986) Endocrinology 119, 20382046. Schoenberg, B.S., Greenberg, R.A. and Eisenberg, H. (1969) 1. Nan. Cancer Inst. 43, 15-32. Singh, S., Sheppard, MC. and Langman. M.J.S. (1993) Gut 34,611-615. Thomas, M.L., Xu, X., Norfleet, A.M. and Watson, C.S. (1993) Bndocrinology 132,426-430. Waldrop, R.D., Saydjari, R., Rubin, N.H., Rayford, P.L., Townsend, Jr., C.M. and Thompson, J.C. (1989) Life Sci. 45,737-744. Willett, WC. (1989) Nature 338.389-394. Wynder, E.F. and Shigematsu. T. (1967) Cancer 20, 1520-1561.
Molecular and Cellular Endocrinology 105 (1994) 197-201
Estrogen receptor-mediated direct stimulation of colon cancer cell growth in vitro Xiaomeng Xu, Mary L. Thomas* Department of Pharmacology and Toxicology, University of Texas Medical Branch, Galveston, TX 77555, USA
Received 8 June 1994; accepted 19 August 1994
Abstract
In vivo and epidemiologicaldata suggest a mitogenicrole for estrogens (E) in colon cancer. The presence of estrogen receptor (ER) and ER mRNA in colonic epi~elium and coIon cancer cells, make it necessary to explore the possible direct effects of E on co1011cancer growth. In this study, a 1.5mer olig~eox~uc~eotide (oligo) antisense to the region of the ~sIation start codon of estrogen receptor mRNA inhibited ER expression in a mouse colon cancer cell line (MC-XJ), as decline by receptor binding assay. Antisense oligo ako decreased ER mRNA levels in MC-26cells. The cowl-stirnulato~ effect of E was abolished by antisense oligo treatment, demonstrating that the ER is directly involved in the regulation of colon cancer cell growth. Keywoti:
Colon cancer; Estrogen receptor; Antisense oligonucleotide; Cell growth
1. Introduction Possible roles for estrogens @) in colon cancers are suggested by several in vivo studies comparing the growth of normal colonic epi~elium and colon cancer cells in ovariectomized acd intact animals (Hoff et al., 1981; Narayan et al., 1992). Moreover, sex differences in the incidence and mortality of colon cancer have been shown in humans (Doll, 1980; McMichael et al., 1983). In addition, there are substantial similarities in the descriptive epidemiology of cancers of the large bowel and breast (Howell, 1976; Willett, 1989). For instance, within individual countries, there is a strong correlation between breast cancer and colon cancer mortality (Wynder et al., 1967); and in individual women, there is a higher than expected concurrence of breast cancer and colon cancer (Schoenberg et al., 1969). Although E has long been known to have stimulatory effects on cell proliferation in the target tissues of the breast and reproductive tract, intestinal epithelium was traditionally regarded as non-target tissue of E, and the in vivo mitogenic effects of E were considered as being mediated via indirect mechanisms. However, the presence of estrogen receptor (ER) and its mRNA in colorectal carcinomas and normal colonic epi*Corresponding
author. Tel. 409-772-9641. Fax 409-772-9642.
0303-7207/9~$07.~ SSDI
0 1994 Elsevier Science Ireland Ltd. AI1rights reserved
0303-7207(94)03381-3
thelium (McClendon et al., 1977; Alford et al., 1979; Markaverich et al., 1981; Bracali et al., 1988; Cameron et al., 1992; Singh et al., 1993; Thomas et al., 1993) indicate that it is necessary to explore possible direct effects of E on colon cancer cells. The use of the antiestrogens has been a valuable tool for determining the role of E in various ER mediated responses. However, when we carried out preliminary experiments using the antiestrogens tamoxifen and ICI 182,780 as tools to determine the specific role of estrogen in the regulation of colon cancer cell growth, we found that these agents directly inhibit the growth of cultured colon cancer cells even in the absence of E. Therefore, we have employed antisense oligodeoxynucleotides (oligo) to ER mRNA to block ER synthesis in MC-26 cells, a mouse colon cancer cell line, in order to demonstrate the ER mediated direct stimulation of MC-26 cell growth in vitro. 2. Materials and methods 2.1. Oligos A 15-mer antisense oligo (5’ GGGTCATGGTCATGG 3’) and a scrambled oligonucleotide (5’ GTGGTGGATCGTGAC 3’) were synthesized by Genosys (Woodlands, TX). The antisense oligo spanned the common translation start codon for mouse, rat and human ER mRNA. The
198
X. Xu, M.L. Thomas I Molecular and Cellular Endocrinology 105 (1994) 197-201
scrambled oligo contained the same 15 bases as the antisense oligo, but had little or no homology to any gene sequences submitted to GenBank. These oligos were similar to those which were used previously in vivo (McCarthy et al., 1993). However, the oligos were modified as phosphorothioates. They were further purified by NAP-5 column (Pharmacia, Uppsala, Sweden) before being used for cell culture.
0.66 M formaldehyde and transferred to a nylon membrane. Blots were hybridized with a random primer-labelled ER cDNA PCR fragment from rat uterus (Thomas et al., 1993), which has 95.4% sequence homology to the mouse ER cDNA sequence. The same blot was then reprobed with rat cyclophylin to normalize for the RNA loading (Danielson et al., 1988). Results of the Northern analyses were quantitated by densitometry.
2.2. Cell culture MC-26 is a 1,Zdimethylhydrazine induced mouse colon adenocarcinoma cell line that produces tumors in mice in a dose-dependent manner when injected subcutaneously (Waldrop et al., 1989). MC-26 cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM) containing 5% fetal bovine serum (FBS). For assessment of the effects of E on MC-26 cell growth, cells were plated (2000-3000 cells/well) into 24-multiwell plates in phenol red-free DMEM supplemented with 5% charcoal-treated FBS and allowed to attach for 48 h. The medium was then (day ‘0’) changed and 17/?-estradiol (E,) in absolute ethanol was added directly to the medium (final concentration of ethanol was less than 0.1%). The medium was changed and hormones added every 2 days. Control cells received equal concentrations of ethanol vehicle without hormones. On day 6 (or as described for each figure), cells were harvested by treatment with 0.025% trypsin/O.O2% EDTA and counted using a Coulter counter (Model ZF, Coulter Electronics, Hialeah, FL). Cell viability always exceeded 90% determined using trypan blue exclusion and a hemocytometer. In experiments with oligos, oligos were added as medium was changed every 2 days.
2.5. Statistics Statistical significance was determined using one-way analysis of variance with post hoc Bonferroni t-test. A P value co.05 was considered statistically significant.
2.3. ER binding assay: ER in MC-26 cells was measured using a whole cell assay as described previously (Thomas et al., 1993). Briefly, MC-26 cells were plated at 5 X lO“/well in phenol red-free DMEM supplemented with 5% charcoal-treated FBS. Two days after seeding, the medium was changed and the cells were treated with oligos for 48 h (or as stated). 48 h later, MC-26 cells were washed three times in PBS and ER binding was determined after a 2 h incubation with 1 nM [3H]E2 (Amersham, spec. act. 44.5 Ci/mmol) at room temperature. The non-specific binding was defined by using a 300-fold molar excess of diethylstilbestrol (DES). Preliminary results have shown that E2 binding in MC-26 cells was saturated at 1 nM, with Kd of approximately 0.3 nM. 2.4. Northern analysis of ER mRNA Subconfluent MC-26 cells grown in phenol red-free DMEM supplemented with 5% charcoal-treated FBS were treated wit.h 3 PM of either ER antisense oligo or scrambled oligo in the presence or absence of 1 nM Ez for 48 h. Total cellular RNA was then extracted as described by Chomczynski and Sacchi (Chomczynski et al., 1987). Total cellular RNA was separated on a 1% agarose gel containing
3. Results MC-26 cells grown under the conditions in our laboratory contain approximately 3400 specific high-affinity estrogen binding sites per cell. When MC-26 cells were treated with 1 ,uM ER antisense oligo, there was a significant decrease in cellular ER levels, as measured by ER binding assay (Fig. 1). This effect was seen as early as 6 h after antisense oligo treatment and appeared to plateau after 24 h, since treatment with antisense oligo for up to 72 h did not significantly further decrease ER binding (data not shown). Treatment with the same concentration of scrambled oligo for 48 h did not affect ER levels in MC-26 cells. Fig. 2 shows a dose dependent decrease of ER binding in MC-26 cells treated with antisense oligo for 48 h. The scrambled oligo had no effect at dose as high as 3 PM. To determine if antisense oligo also affected ER mRNA levels in MC-26 cells, the presence of ER mRNA in MC-26
* ‘t 0 antisense oligo &lscrambled oligo
*
60
0
6
12
24
-1
48
Time of treatment (hrs)
Fig. 1. Time course of ER binding decrease by rntisense oligo in MC-26 cells. MC-26 cells were treated with 1 PM of antisense oligo 6,12 24, or 48 h before ER binding was assayed. One group of cells was treated with I pM of scmmbled oligo 48 h before the binding assay. The control group received 10~1 of 10 mM sodium phosphate (pH 6.8), which was the elution buffer for NAP-5 column purification of the oligos. 1 nM [3H]E2 (44.5 Wmmol) was added to each well in the presence or absence of a 300-fold excess of unlabeled DES for 2 h at room tempemturc. Each bar represents the mean k SD for four wells expressed us 46 of control. *P < 0.05 compared to control.
X. Xu, M.L. Thomas I Molecular and Cellular Endocrinology 105 (1994) 197-201
” 3
140-
5
120-
199
150
0 control I!!!! antisense scrambled oli o P1go 0
Ocontrol lZXSEz(l nM)
8 3 : % E =%
125
100
> 4 c
75
z u 0
0.5
0.1
1
50
3
8
8
Days of Treatment
Concentration of Oligos (pM)
Fig. 2. The effect of different doses of antisense oligo on ER binding in MC-26 cells. MC-26 cells were treated with 0.1-3 PM antisense oligo for 48 h before the ER binding assay. One group of cells was treated with 3yM of scrambled oligo, and the control group received an equal volume of vehicle buffer. The binding assay was performed as described in Fig. 1. Each bar represents the mean f SD for four wells expressed as 8 of control. *P < 0.05 compared to control.
Fig. 4. E stimulation of MC-26 cell growth. MC-26 cells were treated with 1 nM % for 2-8 days. Control cells were treated with the same amount (0.1%) of ethanol vehicle alone. The medium was changed with E2 or vehicle replacement every 2 days. Each bar represents mean f SD for 4 wells expressed as % of control. The number of cells in the control groups wem 0.184, 0.615, 3.024, and 8.744 x lo5 cells/well for days 2, 4.6 and 8, respectively. *P < 0.05 compared to control.
cells treated with oligos was examined by Northern blot analyses (Fig. 3). As has been observed before in MC-26 cells without oligo treatment (unpublished data), there was a small increase (approximately 20%) in ER mRNA level in cells treated with scrambled oligo in the presence of & compared to that in cells treated with scrambled oligo in the absence of EZ. This is probably due to E upregulation of its own receptors. However, in cells treated with antisense oligo, the upregulation of ER mRNA level by E2 was not
seen. Treatment of MC-26 cells with 3pM of antisense oligo for 48 h resulted in decrease in ER mRNA levels compared to cells treated with the same concentration of scrambled oligo. Treatment of MC-26 cells with E2 stimulates cell growth. The maximum growth stimulation was achieved at 1 nM of E2 treatment (data not shown). The effect was most prominent after 6 days of continuous exposure to E2 (Fig. 4). ER antisense oligo was used in order to determine whether ER was involved the growth-stimulatory effects of E in MC-26 cells. The growth stimulatory effect of 1 nM E2 was completely blocked by 1 FM of antisense oligo (Fig. 5). Although statistically not significant, there was a
10
-0
Cyclophilin
I
ONo E2 6S3E2(1 nM)
-
Oligo: It,:
ERlcyclophilin
4
ratio:
SC
ASN
-+-
+
1.0 1.2 0.8 0.7 Control
Fig. 3. Northern analysis of the effect of antisense oligo on ER mRNA levels in MC-26 cells. MC-26 cells were treated with 3pM of either antisense oligo (ANS) or scrambled oligo (SC) in the presence (+) or absence (-) of I nM E2 for 48 h. Total cellular RNA was extracted and Northern blot analysis was performed. Each lane contained 25 pg of total RNA. The blot was hybridized with a 32P-labelled ER cDNA and reprobed with cyclophylin cDNA as described in Section 2. The ER mRNA was normalized to cyclophylin mRNA by densitometry, and the ER to cyclophylin ratio is indicated under each lane by designating the ratio from the cells treated with SC in the absence of E2 as 1.O.
ANS Oligo
SC Oligo
Fig. 5. Inhibition of the E growth-stimulatory effect by antisense oligo. MC-26 cells were treated with 1PM of either antisense (ANS) oligo or scrambled (SC) oligo in the presence or absence of 1 nM E2 for 6 days. Control cells were treated with the same amount of ethanol (0.1%) and elution buffer (10~1) or 1 nM E2 and elution buffer. The medium was changed with q and oligos, or ethanol and oligo. or vehicle replacement every 2 days. Each bar represents cell number mean i SD for 4 wells. *P < 0.05 compared to control without E2. **f < 0.05 compared to control with I nM E2.
200
X. Xu. M.L. Thomas I Molecular and Cellular Endocrinology 105 (1994) 197-201
small inhibition of cell growth in cells treated with antisense oligo, either with or without E2 addition. This could be due to the presence of residue E in the culture medium. However, the scrambled oligo had no effect on either basal or Ez stimulated cell growth. Consistent with the effects of the ER antisense oligo, the anti-estrogen, ICI 182,780 (10 nM) was also able to completely block the growthstimulatory effect of 1 nM Ez (data not shown). However, ICI 182,780 alone exhibited statistically significant concentration-dependent growth-inhibitory effects on MC-26 cells. 4. Discussion We have demonstrated that an antisense oligo targeting ER mRNA inhibited expression of ER in MC-26 cells as measured by receptor binding assay. The inhibition of ER synthesis and inhibition of Ez-stimulated MC-26 cell growth suggest that the ER-mediated growth-stimulatory effect of E is directly involved in the regulation of cell growth in this colon cancer cell line. This may be one of the mechanisms responsible for the mitogenic effect of E observed in colon cancer in vivo, and play a role in the epidemiological similarities between breast cancer and colon cancer. The mechanism of action of phosphorothioate oligonucleotides has been investigated in a number of studies. In addition to common mechanisms of action of modified and unmodified antisense oligonucleotides, unlike other modified oligonucleotides (e.g. a-oligonucleotides and methylphosphonates), phosphorothioates can act by promotion of mRNA degradation by an RNase H-dependent mechanism (Furdon et al., 1989; Crooke, 1992). On the other hand, unlike unmodified oligonucleotides, phosphorothioate oligos are nuclease-resistant (Campbell et al., 1990). Moreover, they appear to be adequately internalized in cells (Gao et al., 1989; Crook, 1991). For example, studies employing a 28-mer phosphorothioate deoxycytidine that was uniformly labeled with 35S demonstrated that when HeLa cells were incubated with 1 PM of the drug, significant intracellular concentrations were achieved. Cellular uptake reached a plateau in 6 h (Gao et al., 1989). In our experiments, the effect of antisense oligo to decrease ER binding appeared in less than 6 h after treatment. This could be due to the fact that ER is a rapidly turning over protein. Although not characterized in colon cancer cells, ER half-life has been well described to be of approximately 4 h in breast cancer and uterine cells in vitro and in vivo (Eckert et al., 1984; Nardulli et al., 1986). As determined by Northern analysis, antisense oligo treatment blocked the upregulation of ER mRNA level by E2 and it also slightly decreased MC-26 cell ER mRNA levels compared to cells treated with the same concentration of scrambled oligo. This effect was more prominent in the presence of E. These data suggest that E upregulation of ER mRNA levels may be ER-mediated and that E upregulation of its own receptor could be one of the mechanisms
of E stimulated cell growth in MC-26 cells. However, we did not directly address this by measuring ER levels in cells treated with oligos in the presence of E, due to the fact that the presense of E2 in the medium interferes the radioligand binding assay. Like most antisense oligo studies, we used the antisense oligo targeting the AUG translation initiation codon of ER mRNA. Targeting the AUG codon would predictably mask the ribosome recognition site and prevent the formation of the translation complex. However, since antisense oligo also decreased ER mRNA levels in MC-26 cells as determined by Northern analysis, the RNase H dependent mechanism seems to also play a role in antisense oligonucleotide inhibition of ER expression. It is possible that both RNase H-dependent and -independent mechanisms are involved in the inhibition of expression in MC-26 cells. Inhibition of ER expression plateaued 24-48 h after antisense oligo treatment. Longer than 48 h treatment with antisense oligo did not produce a further decrease in ER levels as measured by ER binding. However, antisense oligo treatment for 6 days completely blocked the growth stimulatory effect of E, in MC-26 cells. This suggests that E stimulation of MC-26 cell growth is ER-dependent and that it is necessary to have a sufficient number of ERs available to mediate this response. Based on the fact that the scrambled oligo had no effect on either receptor binding or MC-26 cell growth, the effect of antisense oligo is most probably not due to a toxic or non-specific effect of the oligos. In summary, we have demonstrated that antisense oligonucleotides are capable of specifically inhibiting the expression of ER in a cultured colon cancer cell line. Inhibition of expression with antisense oligo was monitored by receptor binding assay. Furthermore, inhibition of ER expression abolished the growth stimulatory effect of E in MC-26 cells, demonstrating that E is directly involved in the regulation of colon cancer cell growth. We believe that the ER mRNA antisense oligo approach is more specific than antiestrogens in demonstrating ER mediated actions, particularly for cell growth experiments, because it has been demonstrated that antiestrogens may have both ERdependent and -independent anti-growth activities. Therefore, antisense oligos may be useful as research tools and therapeutic agents for ER positive cancers such as breast cancer, and according to this study, colon cancer as well. Acknowledgement The MC-26 mouse colon cancer cell line was obtained from the laboratory of Dr. Courtney Townsend at the University of Texas Medical Branch in Galveston. References Alford, T.C., Do, H.M. and Gelihoed, G.W. ( 1979) Cancer 43,980-981. Brclcoli, G., Caracino, A.M.,
Rossodivita, F., Bianchi, C., Loli, M.G. and
Bracali, M. (1988) lnt. 1. Biol. Markers 3,41--28.
X. Xu, ML. Thomas I Molecuhr and Cellular Endocrinology 105 /1994) 197-201
Cameron, B.L., Butler, J.A.. Rutgers, J., Vargas, H.I., Purtell, M. and Sheppard, B. (1992) Am. Surg. 58,758-760. Campbell, J.M., Bacon, T.A. and Wickstrom, l?. (1990) J. Biochcm. Biophys, Methods 20,259-263. Chomczynski, P. and Sac&i, N. (1987) Anal. Biochem. 162,156159. Crook, R.M. (1991) Anticancer Drug Des. 6,609-646. Crooke. ST. (1992) Annu. Rev. Pharmacol. Toxicol. 32,329-76. Danielson, P.E., Forss-Petter. S., Brow, M.A. et al. (1988) DNA 4, 261267. Doll, R. General epidemiologic considerations in aetiology of colorectal cancer, in Colorectal Cancer: Prevention, Epidemiology and Scmening (Winawer, S., Schottenfeld, D. and Sherlock, P. eds.), Raven Press, New York, 1980, pp. l-12. Eckert, R.L., Mullick, A., Rorke, E.A. and Ka~nellen~gen, B.S. (1984) ~d~~nolo~ 114.629-637. Furdon, P., Dominski, 2. and Kole, R. (1989) Nucleic Acids Res. 17, 9193-9204. Gao, W., Stein, C.A., Cohen, J.S., Dutschman, G.E. and Cheng, C.-Y. (1989) J Biol Chem. 264, 11521-l 1526. Hoff, M.B., Chang, W.W.L. and Mak, K.M. (1981) Virchows Arch. (Cell Pathol.) 35263-273. Howell, M.A. (1976) J. Chron. Dis. 29.243-261.
201
Ma&cave&h, B.M., Williams, M., U~hu~h, S. and Clark, J.H. (1981) Endocrinology lOQ,63-69. McCarthy, M.M., Schlenker, B.H. and Pfaff, D.W. (1993) Endocrinology 133.433-439. McClendon, J.E., Appleby, D. and Claudon, D.B. (1977) Arch. Surg. 112,140-141. McMichael, A.J. aud Potter, J.D. (1983) Am. 3. Epidemiol. 118, 620627. Narayan, S., Rajukumar, G., Prouix, H. and Singh, P. (1992) Gastroenterology 103, 1823-1832. Nardutli, A.M. and Katzenellenbogen, B.S. (1986) Endocrinology 119, 20382046. Schoenberg, B.S., Greenberg, R.A. and Eisenberg, H. (1969) 1. Nan. Cancer Inst. 43, 15-32. Singh, S., Sheppard, MC. and Langman. M.J.S. (1993) Gut 34,611-615. Thomas, M.L., Xu, X., Norfleet, A.M. and Watson, C.S. (1993) Bndocrinology 132,426-430. Waldrop, R.D., Saydjari, R., Rubin, N.H., Rayford, P.L., Townsend, Jr., C.M. and Thompson, J.C. (1989) Life Sci. 45,737-744. Willett, WC. (1989) Nature 338.389-394. Wynder, E.F. and Shigematsu. T. (1967) Cancer 20, 1520-1561.