Wild type and alternatively spliced estrogen receptor messenger RNA in human meningioma tissue and MCF7 breast cancer cells

J. Steroid Biochem. Molec. Biol. Vol. 45, No. 4, pp. 22%233, 1993

0960-0760/93 $6.00 + 0.00 Copyright © 1993 Pergamon Press Ltd

Printed in Great Britain. All fights reserved

WILD

TYPE

RECEPTOR

AND

ALTERNATIVELY

MESSENGER

TISSUE

AND

MCF7

RNA

SPLICED

IN HUMAN

BREAST

CANCER

ESTROGEN

MENINGIOMA CELLS*

S. G. A. KOEHORST,~"H. M. JACOBS,J. H. H. THIJSSENand M. A. BLANKENSTEIN Department of Endocrinology,AcademicHospital Utrecht, Utrecht, The Netherlands (Received 15 September 1992; accepted 23 November 1992)

Sunnnary--Human meningiomas are rich in progesterone receptors (PR), which appear to be expressed autonomously. To investigate whether estrogen receptor (ER) variants which do not bind the ligand, but may constitutively induce PR expression, prevail in meningioma, we amplified cDNA by PCR in order to detect mRNA coding for the ER in meningioma which were ER-negative/PR-positive at the protein level. We screened for a portion of the ER which includes the DNA binding domain, the hinge region and the ligand binding domain. For this part of the ER we found a wild type mRNA in all 8 meningiomas tested. No mutations were detected. Apart from this transcript we found two alternatively spliced products missing exons 4 and 7, respectively in 8/8 meningioma specimens. These two products were not exclusive for meningioma, since they were also detected in the MCF7 breast cancer cell line which was used as control. ER deletion mutants missing exon 7 have already been reported [Ref. 1; Molec. Endocr. 5 (1991) 1571-1577]. These are dominant negative. To our knowledge, this is the first report on ER mutants missing exon 4. The presence of ER variants missing exon 4, which is probably not able to bind heat shock protein 90 and therefore may be constitutively active, might explain the autonomous expression of PR in meningioma.

INTRODUCTION

Meningiomas are benign tumors arising from the leptomeninges. They account for approx. 18% of intracranial neoplasms and are found most commonly in middle aged women [2]. Epidemiological findings suggest that female sex hormones are involved in the etiology of human meningiomas [3-5]. Several groups [6-10] have investigated the presence of sex hormone receptors in meningioma. The presence of high concentrations of progesterone receptors (PR) in most human meningioma has been well established. This was done by ligand binding assay [11] as well as by enzyme immunoassay [12]. Considerable controversy existed about the presence of the estrogen receptor (ER). Blaauw et al. [11] tested 67 meningioma for ER by *Preliminary results of this work have been reported at the I V International Congress on Hormones and Cancer,

Amsterdam, The Netherlands, September 15-19, 1991. fTo whom correspondence should be addressed at: Department of Endocrinology G02.625, Academic Hospital Utrecht, P.O. Box 85500, NL-3508 GA Utrecht, The Netherlands. Abbreviations: aa, amino acid; cDNA, complementary DNA; DBD, DNA binding domain; ER, estrogen receptor; LBD, ligand binding domain; PCR, polymerase chain reaction; PR, progesterone receptor; wt, wild type. 227

ligand binding assay and found only 6 positive for ER. In contrast 60 of these samples were positive for PR [1 l]. Other groups [6, 7, 10] reported a much higher prevalence of ER in meningioma. The controversy disappeared, however, when it became clear that methodological differences caused the diverging results. When proper receptor assays, i.e. Scatchard plot analyses, were used, ER were seldom detected in meningioma [5, 11]. The ER could also not be detected in meningioma by immunohistochemical staining [13] and enzyme immunoassay [12]. Thus, in contrast to breast tumor[14] and uterine tissue[15], the expression of PR in meningioma is not regulated and conforms to the classical model, in which the synthesis of PR is stimulated by estrogens, through the available ER. In this respect, meningioma tissue shows a striking resemblance to ER-negative/PRpositive breast cancers, which account for 5% of all breast cancers. Fuqua et al. [16] detected, in ER-negative/PR-positive breast tumors, apart from the wild-type (wt) ER, m R N A coding for a variant ER. This variant was missing exon 5 and was shown to activate transcription of an estrogen-dependent-gene construct in yeast cells in a constitutive way [16].

S.G.A. KOErlORSTet al.

228

The purpose of the present investigation was to evaluate whether ER variants are also present in meningioma. To this end m R N A was prepared from meningioma tissue. PCR amplification of cDNA with primers spanning the DNA binding domain (DBD), the hinge region and the ligand binding domain (LBD) were used to detect low amounts of ER transcript [17]. The present paper describes the identification of the transcripts found.

EXPERIMENTAL

used for the preparation of cDNA [17]. Reactions were carried out in 20#1 10mM T r i s HCI (pH 8.3) containing 50mM KCI, 5 mM MgCI2, 10 ng random hexamer primer, 0.5 mM dNTP, 2 0 U RNAse inhibitor (Boehringer, Mannheim, Germany) and 40 U of M-MLV H RT superscript (BRL, Bethesda, MD, U.S.A.). Synthesis of cDNA was performed in a PCR apparatus (Perkin Elmer, Norwak, CT, U.S.A.). The synthesis consisted of one cycle: 30min incubation (42°C), 5min enzyme inactivation (95°C), 5min at 5°C. 10#1 was used for the PCR.

Tissue collection and cell lines

PCR assay

The E R + MCF7 breast cancer cell line was routinely grown with 10% fetal calf serum (Gibco BRL, Paisley, Scotland) with phenol red. Human meningioma tissue was placed on ice immediately after removal from the patient. Representative specimens were frozen at - 80°C until they were used for RNA extractions or receptor assay.

PCR reactions were carried out in 50~tl 10 mM Tris-HCl (pH 8.3) containing: 50 mM KC1, 1.5 mM MgC12, 100 ng primers and 1 U ampli Taq (Perkin). To amplify exons 2--6 primers 1 and 2 and for exons 4-8 primers 3 and 4 were used. Each cycle of amplification consisted of 30 sec denaturation (95"~'C), and I min annealing (53°C), followed by 90 sec extension (72°C). The ramp time between annealing and extension was 90sec. Each PCR reaction consisted of 35 cycles.

RNA extraction Total RNA was extracted from 250 mg tissue or 1 x 106 cells, with an acid guanidinium thiocyanate-phenol-chloroform mixture as described by Chomczynski et al. [18].

Hybridization in agarose gels

Oligonucleotides corresponding to the human ER cDNA were synthetized as described previously [17]: amplification exons 2-6: primer 1; sense 5 ' - G G A A G T A T G G C T A T G G A A TCT-3' (aa 171-178); primer 2; anti-sense 5' - G A T C T T C G A C A T G C T G C T G G - 3' (aa 423-430). Hybridization probe for PCR products from exons 2-6: probe 5; sense 5'CATAACGACTATATGTGTCCAGCCACC3' (aa 216-227). Amplification exons 4-8: primer 3; sense 5 ' - G G A G A C A T G A G A G C T GCCAAC-3' (aa 283-290); primer 4; anti-sense 5 ' - G A G A C G G A C C A A A G C C A C T T G - 3 ' (aa 562-568). Hybridization probe for PCR products from exons 4-8: probe 6; sense 5'A C C A A C C T G G C A G A C A G G G A G C T G - 3' (aa 347-354). Probes were synthesized using an Applied Biosystems D N A synthesizer (San Jose, CA, U.S.A.).

Half of the DNA of the PCR reactions was separated on 1% agarose gels. Gels were denatured for 30 min in a 0.5 M N a O H solution containing 0.15 M NaC1 and subsequently neutralized for 30 min in 0.5 M Tris-HCI buffer (pH 7.0) containing 0.15 M NaC! followed by a 10min wash in water. Gels were dried under vacuum for 30min at room temperature followed by 30 min at 60°C. The gels were prehybridized for 2 h in: 0.9 M NaCI; 50 mM sodium phosphate, pH 7.7; 5 mM Na2EDTA; 0.1% w/v bovine serum albumin; 0.1% w/v Ficoll; 0.1% w/v polyvinyl pyrollidone; 0.1% SDS and 100 #g/ml herring sperm DNA (Boehringer) at 42°C. Gels were hybridized overnight at 4 2 C with end-labeled oligonucleotides. After hybridization the gels were washed twice with: 0.75 M NaCI; 0.075 M Na3 citrate and 0.1% SDS at 4Z'C for 15 min and exposed to X-ray films for 12 h. For the detection of exons 2-6 of ER, probe 5 and for the detection of exons 4-8 of" ER, probe 6 was used as end-labeled oligonucleotides.

cDNA synthesis

Cloning and sequencing of specific PCR products

Following establishing the RNA integrity by gel electrophoresis, 1/tg of total RNA was

Specific PCR products were isolated from a 1% agarose gel using the Prep-A-Gene kit

Oligonucleotide primers and probes

Human meningioma ER variants

(Biorad, Richmond, CA, U.S.A.) and blunt-end cloned into pUc 19 (Boehringer). Double stranded DNA was used for dideoxy sequencing with M13 universal primers (Pharmacia, Uppsala, Sweden). Sequencing reactions were performed according to manufacturer's instructions, using [~t-35S]dATP, electrophoresed on 6% acrylamide gel containing 8 M urea, and visualized by autoradiography.

229 RESULTS

Expression of ER transcripts in meningioma We first examined the DBD in meningioma because of the fact that, if PR synthesis is induced by estrogen at least the DBD of ER has to be present. Therefore from four ERnegative/PR-positive meningioma (PR content: 309, 267, 179 and 135 fmol PR/mg protein) the

(o) Sample

I

1

I

2

I

3

I

4

I

,5

I 6

I 7

i 8

19

i

1766

1033 653

3ected size 770 bp

453 298

(b) Sample I

1 i 2

i 3

i 4

i 5

I 6

i 7

I 8

i 9

i

1766

1033 653

~ - - - - E x p e c t e d size 850 bp

453 2 9 8 ~

Fig. 1. Expression of ER transcripts in the E R + / P R + MCF7 breast cancer cell line (sample 1: ER; 150, PR; 220), a solid breast tumor (sample 2: ER; 188, PR; 874) and a meningioma (sample 3: ER; 7, PR; 92). Four ER-/PR+ meningiomas (samples 4-7: ER; <3, PR resp.; 309; 267; 179; 135) a ER-/PRmeningioma (sample 8: ER; <3, PR; <3) and ER-/PR- solid breast cancer tumor (sample 9: ER; <3, PR; < 3) (all ER and PR levels are in fmol/mg protein), cDNA was prepared and exons 2-6 (a) and 4-8 (b) were amplified by PCR and analyzed by gel electrophoresis and hybridization with ER-specific oligonucleotide probes.

S . G . A . KOEHORST et al.

230

breast tumor cell line MCF7 and an ER-/PRnegative meningioma and solid breast tumor, mRNA was isolated and reversely transcribed to cDNA. From this cDNA exons 2-6 was amplified by PCR. PCR controls, in which no cDNA was used, were all negative and are therefore not shown. The PCR products were separated on a 1% gel and the specificity of the products was tested by hybridization with an end-labeled internal probe. After hybridization two specific products were present in the meningioma as well as in the MCF7 cell line. One of the transcripts had the expected size of 770 basepairs (bp), corresponding to the wt receptor. One variant was detected with a size of 450 bp [Fig. l(a)]. Subsequently, we also decided to examine the LBD. Therefore cDNA exons 4-8 was amplified using primers 3 and 4 flanking the LBD. After separation on a 1% gel and hybridization with an internal probe, two specific products were detected. The MCF7 cell line as well as the

meningiomas expressed an approx. 500bp product in addition to the expected 850bp product [Fig. l(b)].

Sequence analysis of ER transcript The PCR products from 3 ER-negative/PRpositive meningioma (PR content: 309, 386 and 123 fmol PR/mg protein) MCF7 were blunt-end cloned in pUc 19 for dideoxy sequence analysis. The transcripts generated from cDNA of MCF7 were used as reference sequence and positive control. The sequence was compared to published data [20]. The sequence of the 770 bp ER transcript generated by amplification of exons 2-6 from meningioma was identical to that of the transcript generated from the MCF7 cell line control and in full agreement with the published sequence for the MCF7 pOR8 cDNA clone[19,20]. The specific 450bp fragment which was generated in the same PCR reaction for the meningioma samples as well as by the MCF7 control contain the wt sequence for

Exon 5 411

Exon 4

Exon 5 D,,

Exon3

Exon 4 " ' E xon 3

Fig. 2(a)--legend opposite

Human meningioma ER variants

G

(b)

A

T

231

,c /G

i

C

E xon 8 ,-,=11

Exon 7

,! Exon 8 ib.

Exon 6

Exon 7 Exon 6

Fig. 2(b) Fig. 2. Dideoxy sequence analysis of specific PCR products from human meningioma tissue. 3/5 exon boundary from the variant missing exon 4 and 3/4, 4/5 exon boundaries from the wt ER (a). 6/8 exon boundary from the variant missing exon 7 and 6/7, 7/8 exon boundaries from the wt ER (b).

exons 3, 5 and 6. Exon 4, however, was missing [see Fig. 2(a)]. Amplification of the LBD generated the expected 850bp transcript. This sequence was identical to the control MCF7 transcript. However comparison of this sequence with the published MCF7 pOR8 cDNA clone[19,20] revealed a T - - G residue change at nucleotide 1491 (results not shown). This base change has also been observed in human placental DNA [21]. Therefore the pOR8 sequence probably contains a cloning artifact. The specific 500 bp variant also contained the wt sequence for exons 4, 5, 6 and 8 but was missing exon 7 [see Fig. 2(b)]. This variant was also generated in the MCF7 control. DISCUSSION

Since by ligand binding assays[10] or immunohistochemistry[13] no ER could be de-

tected in meningioma, it was a surprise that we found mRNA coding for the ER in meningioma. This apparent paradox might be caused by the presence of a receptor which does not bind its ligand but has retained the DBD and hence the potential to induce transcription. Such a truncated receptor might be able to act as an oncogene and stimulate growth. So far one variant has been described by Fuqua et al. [16] in breast cancers. They studied ER-negative/PR-positive breast cancer and detected, apart from mRNA for the wt receptor, an exon 5 deletion variant that results in a truncation at aa 370. This cloned 40 kDa variant constitutively stimulated estrogen-responsive genes in a yeast reporter system. Because of the resemblance, at least at the protein level of the receptor content in ERnegative/PR-positive breast tumors and meningioma we searched for ER variants in meningioma and we amplified the DBD, the

232

S . G . A . KOEHORST et al.

hinge region and the LBD by PCR using cDNA from meningioma, mRNA coding for the DBD, the hinge region and the LBD of ER, with no mutations or minor deletions, were found in meningioma. This ER could not be detected by the routinely used ligand binding assay. This can be caused by mutations or deletions outside the DBD and LBD. Another explanation could be the sensitivity of the PCR technique. Henry et al. [22] found that the ER mRNA assay is more sensitive than the ligand binding assay. The wt ER content of meningioma is probably less than 3 fmol/mg protein and therefore in the ligand binding assay meningiomas are classified as ER-negative. In contrast, however, in 8 out of 8 meningiomas we were able to identify wt and variant ER mRNA by PCR of cDNA. This is similar to the findings of Fuqua et al. [16], who found mRNA for wt ER in breast tumors which were ER-negative/PR-positive by ligand binding assay. Besides the wt ER, we found two variant ER in meningioma as well in the MCF7 control. One variant with a major deletion in the LBD was an alternatively spliced product missing exon 7. This variant has been described previously [1] and is common in all breast tumors. This variant is not hormone independent or transcriptionally active as proven by McGuire et al. [1]. They also showed that this variant was dominant negative and interferes with the wt receptor's ability to induce transcription in a yeast expression vector system. Therefore this variant can probably not account for the apparently autonomous PR synthesis in meningioma. We used the same primers which were used for the detection of the ER variant missing exon 5 in breast tumors [16] but we could not detect this variant in meningioma tissue or the MCF7 breast tumor cell line. We detected a variant missing exon 4 not only in meningioma, but also in the MCF7 control. The sequence of this variant stays in frame and codes for a ER missing aa 254-365. To our knowledge such a deletion mutant has not been described previously. It is not inconceivable that this variant is responsible for the synthesis of PR in meningioma in a constitutive manner. Exon 4 includes the last part of the DBD, the hinge region and the first hundred amino acids of the LBD. Heat shock protein 90 (hsp 90) plays a role in the activation of the ER after binding of estrogen. A highly positively charged region situated be-

tween aa 251 and 271 is very important for the formation of the non-DNA binding 8-9 S ER complexes, which bind hsp 90 [23]. A variant missing the aa 251-271, such as the variant missing exon 4, probably cannot bind hsp 90 and therefore is always in the 4-5 S DNA binding form [23]. The ability of hormone independent transcription of this variant remains to be proven. In the C-terminal side of the DBD in several steroid receptors, amino acids are found that bear strong homology to the nuclear translocation signal of SV40 T antigen [24]. Nearly identical sequences are found in the glucocorticoid, progesterone, androgen and mineralcorticoid receptors from various species. In contrast, sequences in the region of the estrogen, vitamin D, thyroid, and retinoic acid receptors do not exhibit strong homology to the T antigen nuclear localization signal [25]. Although all the constitutive nuclear localization signals, as they were recently characterized, are located in exon 4 [26], the exon 4 variant would be cytoplasmic. But, taking into account that the estimated size of the exon 4 variant would be 53-54 kDa and therefore small enough to allow passive diffusion through nuclear pores [27] the location of the exon 4 variant could also be nuclear. Meningiomas thus appear to contain several ER variants; i.e. the wt ER, which according to our present understanding of the mechanism of action of steroid hormones, should be able to induce PR synthesis after binding of the ligand; a variant which lacks exon 7, and a variant missing exon 4. The ER variant missing exon 7 will probably not induce PR synthesis, on the contrary it will interfere with transcriptionally active ER [1]. The variant missing exon 4, which is reported here for the first time may not bind hsp 90 and may therefore be responsible for PR expression in meningioma.

REFERENCES 1. McGuire W. L., C h a m n e s G. C. and Fuqua S. A. W.: Estrogen receptor variants in clinical breast cancer. Molec. Endocr. 5 (1991) 1571-1577. 2. Rausing A., Ybo W. and Stenflo J.: lntracranial meningioma, a population study of ten years. Aeta NeuroL Scand. 46 (1970) 102-110. 3. Bickerstaff E. R., Small J. M. and Guest I. A.: The relapsing course of certain meningiomas in relation to pregnancy and menstruation. J. Neurol, Neurosurg. Psyehiat. 21 (1985) 89-91. 4. Schoenberg B. S., Christine B. W. and Whisnant J. P.: Nervous system neoplasms and primary malignancies of other sites. Neurology 25 (1975) 705 712.

Human meningioma ER variants 5. Lodrini S. and Savoiardo G. O.: Metastases of carcinoma to intracranial meningioma: report of two cases and review of the literature. Cancer 48 (1981) 2668-2673. 6. Blankenstein M. A., Blaauw G., Lamberts S. W. J. and Mulder E.: Presence of progesterone receptors and absence of oestrogen receptors in human intracranial meningioma cytosols. Eur. J. Cancer. Clin. Oncol. 19 (1983) 365-370. 7. Martuza R. L., Miller D. C. and MacLaughlin D. T.: Estrogen and progestin binding by cytosolic and nuclear fractions of human meningiomas. J. Neurosurg. 62 (1985) 750-756. 8. Lesch K. P., Scott W. and Gross S.: Estrophilin immunoreactivity versus estrogen receptor binding activity in meningiomas: evidence for multiple estrogen binding sites. Surg. Neurol. 28 (1987) 181-188. 9. Blankenstein M. A., Blaauw G., van 't Verlaat J. W., van der Meulen-Dijk C. and Thijssen J. H. H.: Steroids receptors in cerebral tumours, possible consequence for endocrine treatment? In Hormonal Manipulation of Cancer: Peptides, Growth factors and New (Anti) Steroidal agents (Edited by J. G. M. Klijn, R. Paridaens and J. A. Foekens). Raven Press, New York (1987) pp. 6 !-70. 10. Markwalder T. M., Zava D. T., Goidhirsch A. and Markwalder R. V.: Estrogen and Progesterone receptors in meningiomas in relation to clinical and pathological features. Surg. Neurol. 20 (1983) 42-47. 11. Blaauw G., Blankenstein M. A. and Lamberts S. W. J.: Sex steroid receptors in human meningioma. Acta Neurochirurgica 79 (1986) 42-46. 12. Blankenstein M. A., van der Meulen-Dijk C. and Thijssen J. H. H.: Assay of oestrogen and progestin receptors in human meningioma cytosols using immunological methods. Clin. Chim. Acta 165 (1987) 189-195. 13. Blankenstein M. A., Berns P. M. J. J., Blaauw G., Mulder E. and Thijssen J. H. H.: Search for estrogen receptors in human meningioma tissue sections with a monoclonal antibody against the human estrogen receptor. Cancer Res. 46 (Suppl.) (1986) 4268s-4270s. 14. Horwitz K. B., Koseki Y. and MeGuire W. L.: Estrogen control of progesterone receptor in human breast cancer: Role of estradiol and antiestrogen. Endocrinology 103 (1978) 1742-1751. 15. Gorski J., Welshons W. and Saki D.: Remodeling the estrogen receptor model. Molec. Cell. Endocr. 36 (1984) 11-15.

233

16. Fuqua S. A., Fitzgerald S. D., Chamness G. C., Tandon A. K., McDonell D. P., Nawaz Z., O'Malley B. W. and McGuire W. L.: Variant human breast tumor estrogen receptor with constitutive transcriptional activity. Cancer Res. 51 (1991) 105-109. 17. Koehorst S. G. A., Jacobs H. M., Tilanus M. G. J., Bouwens A. G. M., Thijssen J. H. H. and Blankenstein M. A.: Aberrant oestrogen receptor species in human meningioma tissue. J. Steroid Biochem. Molec. Biol. 43 (1992) 57-61. 18. Chomczynski P. and Sacchi N.: Single step method of RNA isolation by acid guanidinium thiocyanate phenol chloroform mixture. Analyt. Biochem. 162 (1987) 156-159. 19. Greene G. L., Gilna P., Waterfield M., Baker A., Hort Y. and Shine J.: Sequence and expression of human estrogen receptor complementary DNA. Science 231 (1986) 1150-1154. 20. Green S., Walter P., Kumar V., Krust A., Bornert J. M., Argos P. and Chambon P.: Human oestrogen receptor cDNA: sequence expression and homology to v-erb-A. Nature 320 (1986) 134-139. 21. Tora L., Mulliek A., Metzger D., Ponglikitmongkol M., Park I. and Chambon P.: The cloned human oestrogen receptor contains a mutation which alters its hormone binding properties. EMBO Jl 8 (1989) 1981-1986. 22. Henry J. A., Nicholson S., Farndon J. R., Westley B. R. and May F. E. B.: Measurement of oestrogen receptor mRNA levels in human breast turnouts. Br. J. Cancer 58 (1988) 600--605. 23. Chambraud B., Berry M., Redeuilh G., Chambon P. and Baulieu E. E.: Several regions of human estrogen receptor are involved in the formation of receptor-heat shock protein 90 complex. J. Biol. Chem. 265 (1990) 20686-20691. 24. Kalderon D., Roberts B. L., Richardson W. D. and Smith A. E.: A short amino acid sequence able to specify nuclear location. Cell 39 (1984) 499-509. 25. Carson-Jurica M. A., Sehrader W. T. and O'Malley B. W.: Steroid receptor family: Structure and functions. Endocrine Rev. 11 (1990) 201-220. 26. Ylikomi T., Bocquel M. T., Berry M, Gronemeyer H. and Chambon P.: Cooperation of proto-signals for nuclear accumulation of estrogen and progesterone receptors. EMBO JI 11 (1992) 3681-3694. 27. Lang I., Scholz M. and Peters R.: Molecular mobility and nucleocytoplasmic flux in hepatoma cells. J. Cell Biol. 102 (1986) 1183-1190.