identification of a novel isoform of estrogen receptor, a potential inhibitor of estrogen action, in...

JOBNAME: BBRC 219#3 PAGE: 1 SESS: 5 OUTPUT: Mon Apr 1 16:38:35 1996/xypage/worksmart/tsp000/68953f/19

Identification of a Novel Isoform of Estrogen Receptor, a PotentialInhibitor of Estrogen Action, in Vascular Smooth Muscle Cells

Satoshi Inoue,1 Shin-jiro Hoshino, Hideyuki Miyoshi, Masahiro Akishita, Takayuki Hosoi,Hajime Orimo, and Yasuyoshi Ouchi

Department of Geriatrics, Faculty of Medicine, The University of Tokyo, Hongo, Bunkyo-ku, Tokyo 113, Japan

Received January 16, 1996

Clinical and experimental studies showed that estrogen has antiatherogenic effects. We previously demon-strated that the estrogen receptor (ER) mRNA and protein are expressed in vascular smooth muscle cells(VSMC) derived from rat aorta. Here, the expression of isoforms of the ER was examined in VSMC. Reversetranscriptase-polymerase chain reaction using specific primers for rat ER cDNA was performed from RNA ofrat VSMC. This revealed the existence of ER cDNA that is shorter than the wild-type ER cDNA. Sequencingof the amplified products identified three isoforms of the ER and the wild-type ER. These ER mRNA isoformslacked the region corresponding to exon 4, exon 4 and 5, and exon 3 and 4. Therefore, they were designated asERD4 isoform, ERD4/5 isoform and ERD3/4 isoform, respectively. Chloramphenicol acetyltransferase assaywas performed with these ER isoforms constructed into the expression vector and the reporter plasmid con-taining the estrogen responsive element. The assay showed that these ER isoforms lost estrogen-dependenttransactivation activities and that ERD4/5 isoform has a inhibitory effect on normal estrogen action when it wascotransfected with the wild-type ER. These ER isoforms might be involved in the regulation of VSMC byestrogen. © 1996 Academic Press, Inc.

The protective effects of estrogen from atherosclerosis has been noticed by clinical (1, 2) andexperimental (3–5) studies. Premenopausal women have low rate of coronary artery disease com-pared with men of similar age (6). After natural or surgical menopause, the risk of atherosclerosisincreased markedly (6, 7). Estrogen replacement therapy is effective for protection of postmeno-pausal women from coronary heart disease (1, 2). However, the mechanism of estrogen action onatherosclerosis is poorly understood. Estrogen-related effects on serum lipoprotein levels mayexplain a part of these mechanisms (3, 4, 8). On the other hand, some experimental studies revealedthat the protective effect of estrogen on atherosclerosis could be observed when serum lipoproteinlevel was unchanged (5). These findings suggest that estrogen may act on the cardiovascular tissuesdirectly. To support this idea, specific binding of estrogen was shown in the heart (9), vasculartissues (10) and vascular cells (11, 12) suggesting the presence of estrogen receptor (ER).In the pathogenesis of atherosclerosis, smooth muscle cell proliferationin vivo plays a central

role (13). Atherogenic stimulations induce abnormal proliferation of the vascular smooth musclecells (VSMC), which lead to the formation of atherosclerotic lesion. Several groups have shownthat the proliferation of VSMC is inhibited by estrogen treatment (14, 15). This inhibitory effect onthe proliferation of VSMC may contributes to the anti-atherogenic effects of estrogen. We (16) andKaraset al. (17) previously reported that the functional ER was present in VSMC at both theprotein and the mRNA levels. These ER receptors may mediate the estrogen action in VSMC.Estrogen has diverse effects on various organs, tissues and cells. These effects were mainly

mediated by the ER that exists as a single gene in the mammalian genome (18). To date, severalestrogen receptor variants have been isolated in cancerous cells (19–21). These isoforms are mostlysplicing variants of the ER that lacks some exons. An isoform of the ER was reported in the normal

1 To whom correspondence should be addressed. Fax: 011-81-3-5689-2483.Abbreviations: FBS, fetal bovine serum; VSMC, vascular smooth muscle cells; ER, estrogen receptor; RT-PCR; Reverse

transcriptase-polymerase chain reaction; CAT, Chloramphenicol acetyltransferase.

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS219,766–772 (1996)ARTICLE NO. 0308

7660006-291X/96 $18.00Copyright © 1996 by Academic Press, Inc.All rights of reproduction in any form reserved.


tissue, in the brain, for the first time (22). This isoform of the ER lacks exon 4 that include bothDNA-binding and estrogen-binding domains. Some of the isoforms of the ER have variant func-tions compared with the wild-type ER receptor (20, 21). It is possible that isoforms of the ER maymodulate the estrogen action in the cardiovascular system. These circumstances prompted us tostudy the existence of isoforms of the ER in VSMC and possible involvement of the isoforms inthe mechanism of estrogen action in VSMC. We performed reverse transcriptase-polymerase chainreaction (RT-PCR) using ER specific primers. Analysis of the PCR products has indicated thepresence of three ER mRNA isoforms other than the wild-type ER in VSMC. The function of theseER isoforms was studied by chloramphenicol acetyltransferase (CAT) assay and one of them wasshown to have an inhibitory effect on the wild-type ER in VSMC.

MATERIALS AND METHODS

Cell culture.Eight-week-old female rats (Nippon Bio-Supply Center) were sacrificed under ether anesthesia. VSMC werecultured from medial layer of the thoracic aorta by the method of Chamleyet al. (23) Subcultured VSMC (4–6th passage)were used in the experiments. VSMC and A10 cells (a rat aortic smooth muscle cell-line) (24) were maintained inDulbecco’s modified Eagle medium supplemented with 10% fetal bovine serum (FBS) (Cell Culture Laboratories) at 37°Cin a humidified atmosphere of 5% CO2. A10 cells were suitable for transfection experiments using calcium-phosphateprecipitation method (25).Reverse transcriptase-polymerase chain reaction (RT-PCR).PCR primers for the rat ER cDNA (18, 22) were synthesized

as E-1 (nt 611-630) 59-CTACTACCTGGAGAACGAGC-39; E-2 (nt 762-779) 59-AAGGAGACTCGCTACTGT-39; E-6 (nt1482-1501) 59-TCAAAGATCTCCACCATGCC-39; E-7 (nt 1650-1669) 59-ATCTTGTCCAGGACTCGGTG-39. RT-PCRwas performed as described previously (16). cDNA was synthesized from 10mg of total RNA of VSMC essentiallyaccording to Gubler and Hoffman (26) using the E-7 primer. Samples with or without treatment of avian myeloblastosisvirus reverse transcriptase (Seikagaku Kougyo, Tokyo, Japan) were prepared as controls. One-tenth of the resulting cDNAwas used as template DNA for the first PCR with E-1 and E-7 primers. One-tenth of the first PCR products was used astemplate DNA for the second PCR with E-2 and E-6 primers. The reaction was carried out in a final volume 20mLcontaining template DNA, 10 pmol of each primer, 200mmol/L dGTP, dATP, dTTP and dCTP, 10 mmol/L Tris-HCl (pH8.3), 50 mmol/L KCl, 1.5 mmol/L MgCl2, 0.02% gelatin and 1U Taq DNA polymerase. The temperature program for theamplification was 30 cycles of 1 min at 94°C, 1 min at 57°C, and 2 min at 72°C.DNA sequencing and plasmid construction.PCR products were cloned into PCRII plasmid (Invitrogen) according to

manufacturer’s instructions. The resulting recombinant DNA was purified and sequenced by dideoxy method (27) usingsequenase (US Biochemical). Utilizing PCR fragments derived from these DNA and full length cDNA of the rat ER (18),the wild-type ER and three ER isoforms were constructed in an expression vector PSSRa (28) as PSSRaER, PSSRaERd4,PSSRaERD4/5, and PSSRaERD3/4. The resulting plasmids were confirmed by sequencing.Chloramphenicol acetyltransferase (CAT) assay.CAT assay was performed as described (29). Briefly, 1 × 106 of A10

cells were plated one day prior to transfection. One hour prior to transfection, the medium was replaced with phenol redfree medium containing 10% dextran charcoal-treated FBS. By calcium-phosphate precipitation method (25), cells weretransfected with 0.2mg of expression plasmids (PSSRaER, PSSRaERD4, PSSRaERD4/5 or PSSRaERD3/4), 2 mg ofvitERE-CAT reporter plasmid (30) containing the estrogen responsive element and 2mg of pCH110b-galactosidaseexpression plasmid (Pharmacia) used as an internal control to normalize for variations in transfection efficiency. In severalexperiments, indicated amount of expression plasmids for ER isoforms was added to wild-type PSSRaER expressionplasmid (0.2mg). The total amount of DNA transfected was made up to 20mg with carrier DNA pGEM3Zf(-) (Promega).After 12 hour incubation the cells were cultured further with medium change in the absence or presence of 1 × 10−8 mol/L17b-estradiol for 24 hours. The cell extracts were assayed for CAT activities as described (31). The experiment was carriedout three times and a representative pattern is shown. To determine relative intensity of the signals, a macintosh computerwith a scanner was utilized.

RESULTS

Identification of the ER Isoforms in VSMC

To investigate the expression of ER isoforms, we used two set of primers to increase thespecificity of the amplified products. E-1, E-2, E-6 and E-7 primers are located in exon 1, exon 2,exon 6 and exon 7 of the ER gene, respectively. Total RNA (10mg) from rat VSMC was convertedinto corresponding cDNA using reverse transcriptase (RT) and an ER gene-specific primer (E-7).The first PCR was performed utilizing the RT products as template DNA with E-1 and E-7 primers.

Vol. 219, No. 3, 1996 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

767


Then, the second PCR was carried out using the first PCR products as template DNA with E-2 andE-6 primers. The second PCR products were analyzed by 1% agarose gel electrophoresis (Fig. 1).As a result, other than a band corresponding to the PCR products derived from the wild-type ratcER (Fig. 1; Lane 2, 5: shown by arrowhead), shorter bands were detected in the RT-PCR productsderived from VSMC RNA (Fig. 1; lane 2: shown by arrow). There was no band detected in sampleswithout RT (Fig. 1; lane 3) and without template (Fig. 1 lane 4).Amplified PCR products were isolated from agarose gel and cloned into plasmid vector. Then,

their sequences were determined by dideoxy method. The sequence of the long band correspondedto the wild-type ER cDNA contained all exon 3, exon 4 and exon 5. The exon-exon junctionalsequence were shown in Fig. 2a. One of the sequence of the shorter bands indicated that the59-termini of exon 5 was connected to the 39-termini of exon 3 (Fig. 2b), another of the sequenceof the shorter bands indicated that the 59-termini of exon 6 was connected to the 39-termini of exon4 (Fig. 2c), and the other of the sequence of the short bands showed that the 59-termini of exon 5was connected to the 39-termini of exon 2 (Fig. 2d). Thus, they were designated as ERD4 isoform,ERD4/5 isoform and ERD3/4 isoform, respectively. A schema of the relationship of ER isoformsand functional domains of the ER is shown in Fig. 3. The exon-exon junctional sequences andcorresponding amino acids are shown for ERD4, ERD4/5 and ERD3/4 isoforms. ERD4 and ERD3/4isoforms encode the shorter protein that lack amino acids corresponding to those exons. In case ofERD4/5 isoform, a frame-shift occurred at the exon-exon junction and a premature stop codonappears. Thus, ERD4/5 isoform encodes a truncated protein.

Inhibitory Effects of the ERD4/5 Isoform on the Wild-Type ER Action

To examine the possible function of these ER isoforms, we performed CAT assay using ERisoform expression plasmids. It was shown that the wild-type ER expression plasmid have estrogendependent transactivating activity and that ERD4, ERD4/5 and ERD3/4 isoforms have lost suchactivities (Fig. 4). To examine the effect of these isoforms on function of the wild-type ER, eachER isoform expression plasmid was cotransfected with the wild-type ER expression plasmid (Fig.5A). In case of ERD4/5 the activation of ERE-tk-CAT by wild-type ER was significantly decreasedwhen 50 times amount of isoform ER was added. To confirm this observation, the wild-type ERexpression plasmid was cotransfected with increasing amount of ERD4/5 isoform expressionplasmid (Fig. 5B). The activation of ERE-tk-CAT by the wild-type ER was decreased dosedependently with the ERD4/5 isoform. Relative values of acetylated chloramphenicol signals wereshown in Fig 5B.

FIG. 1. Estrogen receptor (ER) isoforms detected in VSMC. Reverse transcriptase-polymerase chain reaction (RT-PCR)products were analyzed by 1% agarose gel electrophoresis. RT-PCR samples of the VSMC (VSMC) with (lane 2; RT+) andwithout (lane 3; RT-) reverse transcriptase, and without template (lane 4; NT) and PCR products from the rat ER cDNA(lane 5; rat cER) are shown. Lane 1 (M) shows thel/Hind III size marker.


768


DISCUSSION

In the previous study (16), we demonstrated the presence of an estrogen receptor in vascularsmooth muscle cells at the mRNA level by Northern blot analysis and RT-PCR. The presence ofthe ER protein was confirmed by immunocytochemistry utilizing anti-ER antibody. We examined

FIG. 3. A schematic representation of the estrogen receptor (ER) isoforms and functional domains of the ER. Theposition of the primers used in this experiment are shown by arrowheads. The exon–exon junctional sequences andcorresponding amino acids of the ER isoforms are shown. Note that a frame-shift is occurred and premature stop codonappears encoding the truncated protein only in the ERD4/5 isoform.

FIG. 2. The sequence analysis of the reverse transcriptase-polymerase chain reaction products. Autoradiogram ofexon–exon junctional sequence determined by dideoxy method for the wild-type estrogen receptor (ER) (a), ERD4 (b),ERD4/5 (c) and ERD3/4 (d) are shown.


769


the regulation of several estrogen responsive genes and found that the transcripts of c-fos proto-oncogene was regulated by estrogen in these cells. Karaset al. (17) also described that humanvascular smooth muscle cells contain the ER both at the mRNA and the protein levels. They haveshown by luciferase assay that the ER has estrogen dependent transactivation activity in those cells.These reports suggested that estrogen can act on the vascular smooth muscle cells directly and that

FIG. 5. Inhibitory effect of the ERD4/5 isoform on the wild-type estrogen receptor (ER) action. (A) The wild-type ERexpression plasmid PSSRaR was cotransfected with indicated amount (the same or 50 times) of PSSRaERD4 (D4),PSSRaERD4/5 (D4/5) or PSSRaERD3/4 (D3/4). After culturing in the absence (−) or presence (+) of 10 nmol/L 17b-estradiol (E2), chloramphenicol acetyltransferase (CAT) assay was performed. (B) The wild-type ER expression plasmidPSSRaER was cotransfected with indicated amount (1×, 3× 9× 27× and 81×) of PSSRaERD4/5 (D4/5). After culturing inthe absence (−) or presence (+) of 10 nmol/liter 17b-estradiol (E2), CAT assay was performed. Relative values (transfectionwith the wild-type ER under the presence of E2 as 100) of acetylated cloramphenicol signals were shown at the top.

FIG. 4. Chloramphenicol acetyltransferase (CAT) assay analysis of three isoforms of the estrogen receptor (ER) mRNA.The wild-type and isoform ER expression plasmids, PSSRaER (ER), PSSRaERD4 (D3/4), PSSRaERD4/5 (D4/5),PSSRaERD3/4 (D3/4) or PSSRa (vector) were transfected into A10 vascular smooth muscle cells with the vitERE-tk-CATreporter plasmid. After culturing in the absence (−) or presence (+) of 10 nmol/L 17b estradiol (E2), CAT assay wasperformed.


770


the ER demonstrated in vascular smooth muscle cells can mediate estrogen action in the cardio-vascular system.In the present study, we detected three isoforms of ER in the rat VSMC by RT-PCR. One is the

isoform that lacks exon 4. Another isoform lacks both exon 4 and exon 5, and the other lacks exon3 and exon 4. Although it is difficult to compare accurately the amount of mRNA by RT-PCR, thesignal of shorter bands are comparable with that of the long band corresponding to the wild-typeER cDNA. From six clones derived from shorter bands, we obtained 3 clones of ERD4, 2 clonesof ERD4/5 and 1 clone of ERD3/4. Thus, we estimate that the existence of the ERD4/5 isoform,that was identified for the first time, would not be negligible.In cancerous cells, several isoforms have been demonstrated, such as, ERD3, ERD4, ERD5,

ERD7 isoforms. On the other hand, only few reports showed estrogen receptor isoforms in normalorgans and tissues. Skipperet al. (22) has been reported the existence of ERD4 in the rat brain. Werecently found ERD4 and ERD3/4 isoforms in the rat bone tissues (32). However, the function ofthese isoforms were still unknown. It has been reported that ERD3 isoform in the human breastcancer cells inhibits estrogen-dependent transcription activation when it is cotransfected with thewild-type ER and reporter plasmid (20). Both exon 3 and exon 4 contain a part of the DNA bindingdomain, and both exon 4 and exon 5 contain a part of the estrogen-binding domain (22, 33).Therefore, these three isoforms may lose binding activities to estrogen responsive elements andestrogen. CAT assay clearly demonstrated that these isoforms lost estrogen dependent transacti-vation activity via the estrogen responsive element. Furthermore, CAT assay cotransfected with thewild-type ER revealed one of these isoform, ERD4/5, has an inhibitory effect on normal estrogenaction with a dominant negative fashion. It is notable that ERD4/5 encodes truncated proteinbecause of a frame-shift occurred at the exon-exon junction. These results suggest that the truncatedprotein compete with the wild-type ER having the normal function.The data presented in this report are interesting because estrogen has bidirectional effects in

various organs, tissues and cells. For example, estrogen promote the growth of breast cancerMCF-7 cells, whereas estrogen inhibit the growth of VSMC (Akishitaet al., submitted). It ispossible that the differential effects of estrogen are regulated by the receptor isoforms, such asERD4/5 isoform that is a potential inhibitor of the wild-type ER function. Alternatively, theinhibitory effect of estrogen action by the ER isoform might be involved in pathophysiologicalconditions. Further studies should be required to clarify the roles of these ER isoforms in normaland atherosclerotic arteries.

ACKNOWLEDGMENTS

We thank Ms. M. Watanabe, Ms. H. Yamaguchi and Ms. M. Goto for technical assistance. This work was supported bygrants from Foundation and the Ministry of Education, Science and Culture, Japan.

REFERENCES

1. Bush, T. L., Barrett-Conner, E., Cowan, L. D., Criqui, M. H., Wallace, R. B., Suchindran, C. M., Tyroler, H. A., andRifkind, B. M. (1987)Circulation 75, 1102–1109.

2. Stampfer, M. J., Colditz, G. A., Willett, W. C., Manson, J. E., Rosner, B., Speizer, F. E., and Hennekens, C. H. (1991)N. Engl. J. Med.325,756–762.

3. Kushwaha, R. S., and Hazzard, W. R. (1981)Metabolism30, 359–366.4. Adams, M. R., Clarkson, T. B., Kaplan, J. R., and Koritnik, D. R. (1989)Transplant. Proc.21, 3662–3664.5. Cheng, L. P., Kuwahara, M., Jacobsson, J., and Foegh, M. L. (1991)Transplantation52, 967–972.6. Kannel, W. B., Hjortland, M. C., McNamara, P. M., and Gordon, T. (1976)Ann. Intern. Med.85, 447–452.7. Wittman, J. C. M., Grobbee, D. E., Kok, F. J., Hoffman, A., and Valkenburg, H. A. (1989)Br. Med. J.198,642–644.8. Walsh, B. W., Schiff, I., Rosner, B., Greenberg, L., Ravnikar, V., and Sacks, F. M. (1991)N. Engl. J. Med.325,

1196–1204.9. Stumph, W. E., Sar, M., and Aumüller, G. (1977)Science196,319–321.10. Horwitz, K. B., and Horwitz, L. D. (1982)J. Clin. Invest.69, 750–758.11. Colburn, P., and Buonassisi, V. (1978)Science201,817–819.


771


12. Nakao, J., Chang, W.-C., Murota, S.-I., and Orimo, H. (1981)Atherosclerosis38, 75–80.13. Ross, R. (1993)Nature362,801–809.14. Fischer-Dzoga, K., Wissler, R. W., and Vesselinovitch, D. (1983)Exp. Mol. Pathol.39, 355–363.15. Vargas, R., Wroblewska, B., Rego, A., Hatch, J., and Ramwell, P. W. (1993)Br. J. Pharmacol.109,612–617.16. Orimo, A., Inoue, S., Ikegami, A., Hosoi, T., Akishita, M., Ouchi, Y., Muramatsu, M., and Orimo, H. (1993)Biochem.

Biophys. Res. Commun.195,730–736.17. Karas, R. H., Patterson, B. L., and Mendelsohn, M. E. (1994)Circulation 89, 1943–1950.18. Koike, S., Sakai, M., and Muramatsu, M. (1987)Nucleic Acids Res.15, 2499–2513.19. Dotzlaw, H., Alkhalaf, M., and Murphy, L. C. (1991)Mol. Endcrinol.6, 773–785.20. Wang, Y., and Miksicek, R. J. (1991)Mol. Endcrinol.5, 1707–1715.21. Fuqua, S. A. W., Chamnes, G. C., and McGuire, W. L. (1993)J. Cell. Biochem.51, 135–139.22. Skipper, J. J., Young, L. J., Bergeron, J. M., and Tetzlaff, M. T. (1992)Proc. Natl. Acad. Sci. USA.90, 7172–7175.23. Chamley, J. H., Campbell, G. R., McConnell, J. D., and Gröschel-Stewart, U. (1977)Cell Tiss. Res.177,503–522.24. Kimes, B. W., and Brandt, B. L. (1976)Exp. Cell Res.98, 349–366.25. Graham, F. L., and van der Eb, A. J. (1973)Virology 52, 456–467.26. Gubler, V., and Hoffman, B. J. (1983)Gene25, 456–467.27. Sanger, F., Nicklen, S., and Coulson, A. R. (1977)Proc. Natl. Acad. Sci. USA.74, 5463–5467.28. Takebe, Y., Seiki, M., Fujisawa, J.-I., Hoy, P., Yokote, K., Arai, K.-I., Yoshida, M., and Arai, N. (1988)Mol. Cell. Biol.

8, 466–472.29. Inoue, S., Orimo, A., Hosoi, T., Kondo, S., Toyoshima, H., Kondo, T., Ikegami, A., Ouchi, Y., Orimo, H., and

Muramatsu, M. (1993)Proc. Natl. Acad. Sci. USA.90, 11117–11121.30. Inoue, S., Kondo, S., Hashimoto, M., Kondo, T., and Muramatsu, M. (1991)Nucleic Acids Res.19, 4091–4096.31. Gorman, C. M., Moffat, L. F., and Howard, B. H. (1982)Mol. Cell. Biol. 2, 1044–1051.32. Hoshino, S., Inoue, S., Hosoi, T., Saito, T., Ikegami, A., Kaneki, M., Ouchi, Y., and Orimo, H. (1995)Calcif. Tissue

Int. 57, 466–468.33. Ponglikitmongkol, M., Green, S., and Chambon, P. (1988)EMBO J.7, 3385–3388.


772

identification of a novel isoform of estrogen receptor, a potential inhibitor of estrogen action, in...

Documents