help button home button Endocrine Society JCEM
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a related Letter to the Editor
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Balleine, R. L.
Right arrow Articles by Clarke, C. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Balleine, R. L.
Right arrow Articles by Clarke, C. L.
The Journal of Clinical Endocrinology & Metabolism Vol. 84, No. 4 1370-1377
Copyright © 1999 by The Endocrine Society

Original Studies

Coexpression of Alternatively Spliced Estrogen and Progesterone Receptor Transcripts in Human Breast Cancer1

Rosemary L. Balleine2, Sybille M. N. Hunt and Christine L. Clarke

Westmead Institute for Cancer Research, University of Sydney, Westmead Hospital, Westmead, New South Wales 2145, Australia

Address all correspondence and requests for reprints to: Dr. Rosemary Balleine, Westmead Institute for Cancer Research, University of Sydney, Westmead Hospital, Westmead, New South Wales 2145, Australia. E-mail: rosemary{at}hemonc.wh.su.edu.au


   Abstract
Top
 Abstract
Introduction
 Subjects and Methods
 Results
 Discussion
 References
 
Primary transcripts of the human estrogen receptor (ER) and progesterone receptor (PR) are subject to a number of alternative splicing events resulting in a range of variant messenger ribonucleic acid species in receptor-positive tissues. Despite in vitro demonstrations of a possible role for some of these variants in hormonal sensitivity, the clinical significance of this process is uncertain. In this study the coexpression of variant ER and PR transcripts has been documented by RT-PCR and Southern blot analysis in a series of receptor-positive breast tumors. In 35 ER-positive tumors, a common profile of variant ER transcripts was present, with all tumors containing the {Delta}2ER and {Delta}7ER, 94% containing the {Delta}4ER, and 83% containing the {Delta}5ER. In 25 of these cases, which were also PR positive, the most highly expressed PR variants, the {Delta}4PR, {Delta}6PR, and {Delta}4/2PR, a transcript from which a 126-bp portion of PR exon 4 was deleted, were detected in over 90% of the cases. The alternatively spliced ER variants were expressed at higher relative levels than the PR species, which had mean levels of expression less than 10% that of wild-type PR. The most abundant species was the {Delta}7ER, which was present at levels ranging from 29–83% of the wild type. There was no relationship between the level of {Delta}7ER in individual tumors and the pattern of expression of the estrogen-responsive proteins PR and pS2. The common profile of alternatively spliced ER and PR transcripts in breast tumors means that this feature cannot be used as a discriminator of hormone responsiveness or other clinical end points. Further, the low level of expression of the majority of variant species calls into question their potential for impacting significantly on receptor function.


   Introduction
Top
 Abstract
 Introduction
Subjects and Methods
 Results
 Discussion
 References
 
THE OVARIAN steroid hormones estrogen and progesterone influence the development and function of the normal human breast and are implicated in the inception and progression of breast cancer. This hormone-responsive character of breast cancer not only gives insight into the essential nature of the disease, but allows a therapeutic opportunity, as it has been known for over a century that manipulation of the endocrine environment of a tumor can induce regression in some circumstances (1, 2). The effects of estrogen and progesterone in their target tissues are mediated by specific nuclear hormone receptors [estrogen receptor (ER) and progesterone receptor (PR)]. These receptors also mediate the effects of therapeutic endocrine agents in breast cancer, and their presence indicates a greater likelihood of response to this form of treatment (3, 4). However, the presence of ER and PR is not always accompanied by a response to endocrine agents (3, 4), and receptor expression may be retained after resistance to this form of treatment is acquired (5, 6).

The discrepancy between the presence of receptors and the likelihood of a response to endocrine agents in breast cancer has led to speculation that breast tumors may contain forms of ER or PR with aberrant function. To investigate this possibility, the ER has been extensively examined at the messenger ribonucleic acid (RNA) level, and it has emerged that the primary ER transcript in both normal and malignant tissue is commonly subjected to a number of alternative splicing events that give rise to a range of truncated transcripts (7). Since their discovery, there has been speculation that these transcripts may encode ER variants that are clinically important in breast cancer, and functional correlates of some of these have been examined in vitro. One species in particular, lacking exon 5 of the ER transcript, ({Delta}5ER), has generated considerable interest, as it has been found to have low level constitutive transcriptional activity (8), and relative to wild-type ER function, both dominant positive (9) and dominant negative (10) activities have been ascribed to the {Delta}5ER in specific experimental circumstances. Dominant negative activity has also been attributed to the exon 7-deleted ({Delta}7ER) (11), exon 3-deleted ({Delta}3ER) (12, 13), and exon 4-deleted ({Delta}4ER) (14) ER variants in some expression systems. Analogous to ER, the existence of alternatively spliced transcripts of PR has been demonstrated recently in breast tumors (15, 16), and in vitro functional studies have provided evidence that PR variants lacking exon 6 ({Delta}6PR) and exons 5 and 6 ({Delta}5+6) of the PR sequence may have dominant negative activity relative to that of the wild-type receptor (17).

Despite intriguing functional data from in vitro experiments, attempts to correlate the expression of alternatively spliced ER transcripts with clinical features of breast tumors have not clarified the role variant receptors may play in hormone resistance or disease progression. In a study by Zhang et al. the presence of multiple, alternatively spliced ER transcripts was documented in a series of 109 primary breast tumors, and no correlation between the presence of these variants and clinicopathological variables or disease recurrence was found (18). In two small studies, modest increases in the level of the {Delta}5ER variant have been reported in breast tumors compared with that in normal breast (19) and in recurrent breast cancer compared with that in primary tumors (20). The level of this variant was not elevated in tamoxifen-resistant tumors, however (21), nor was it related to disease-free survival in patients treated with adjuvant tamoxifen (18). Erenburg et al. have reported that breast tumors express significantly lower levels of the {Delta}3ER variant transcript than normal breast specimens and have proposed on the basis of the dominant negative activity of this receptor in vitro that this difference may play a role in the malignant phenotype (13).

To begin to understand this issue, the broader significance of splicing of receptor transcripts needs to be considered. The question of whether alternative splicing is a general phenomenon in breast tumors and whether tumors that contain alternatively spliced ER transcripts also contain alternatively spliced PR transcripts in particular needs to be addressed. The aim of this study was to document alternatively spliced ER and PR transcripts in a series of breast tumors and to determine whether the coexpression of alternatively spliced ER and PR transcripts was related to receptor function and, hence, hormone sensitivity in breast cancer.


   Subjects and Methods
Top
 Abstract
 Introduction
 Subjects and Methods
Results
 Discussion
 References
 
Patient material

Tumor samples were taken from 35 women with primary carcinoma of the breast treated at Westmead Hospital (Sydney, Australia). The patients ranged in age from 42–83 yr (mean, 60.8).

Detection of ER and PR

ER and PR measurements were made as part of the routine pathological assessment of tumors by enzyme immunoassay (EIA) using the Abbott ER EIA and PgR EIA kits (Abbott Laboratories, Diagnostics Division, North Chicago, IL), as previously described (22). ER or PR values of 15 fmol/mg protein or greater by EIA were deemed positive. The precise quantity of receptors in samples whose ER or PR content was higher than the upper limit of the standard curve was not determined; instead, these were reported as greater than the highest measurable value only. For the purposes of this analysis, this highest value has been used to approximate ER and PR content in such cases. The ER status of tumors was confirmed by immunohistochemical staining of paraffin-embedded tumor sections using a mouse monoclonal anti-human ER antibody Dako-ER 1D5 (Dako Corp., Carpinteria, CA) after microwave antigen retrieval.

Detection of pS2

The estrogen-responsive protein pS2 was detected by immunohistochemistry in paraffin-embedded tumor sections. Sections were incubated for 10 min at 37 C in a 0.05% (wt/vol) solution of protease, (Sigma Pronase Protease Type E, Sigma-Aldrich Co., Castle Hill, Australia) before incubation with a polyclonal anti-pS2 antibody, NCL-pS2 (Novocastra Laboratories Ltd., Newcastle-Upon-Tyne, UK), overnight. After blocking with 3% hydrogen peroxide, a secondary biotin-conjugated antirabbit Ig was applied to the sections for a minimum of 30 min, followed by streptavidin-horseradish peroxidase for an additional period of at least 30 min. Diaminobenzidine was used as the chromogen, and the sections were lightly counterstained with hematoxylin before coverslipping. Negative control slides, consisting of sections incubated overnight with diluent only or normal rabbit serum at 1:200 in diluent, rather than the primary antibody were performed, and a case known to give a positive result was included in each staining run.

Semiquantitative analysis of each section was carried out. The number of tumor cells stained positively was counted, and the percentage of cells stained was calculated for each case. Tumor cell counting was facilitated by viewing sections and counting with a video image analysis system (Optimas Corp., Seattle, WA). In accordance with some other reports (21, 23), tumors with 10% or more of cells staining positively were designated positive.

Extraction of RNA

Frozen tumor tissue stored at -70 C for up to 6 yr was pulverized in a supercooled vessel, suspended in a guanidinium isothiocyanate solution, and homogenized using a Dounce homogenizer (Kontes Co., Vineland, NJ) and Teflon pestle. Total RNA was extracted by the guanidinium isothiocyanate-cesium chloride method as previously described (24).

RT-PCR amplification of ER and PR transcripts

Fragments of the ER transcript were amplified by the RT-PCR using the primers indicated in Fig. 1Go. Oligonucleotide sequences are given below. Details of the ER sequence are taken from the reports of Greene et al. (25) and Ponglikitmongkol et al. (26). An EcoRI restriction site was added to the 5'-end of some sense primers, and this is indicated in bold type. Sequences for primers ER 1, ER 2 and ER D were taken from the report by Fuqua et al. (27): ER A, GGAATTC AGC CCG CTG ATG CTA CTG (ER 316–333); ER B, TCA TCA TTC CCA CTT CGT AGC (ER 734–754); ER C, GGAATTC TGC TTC AGG CTA CCA TTA TGG (ER 573–593); ER D, TGA ACC AGC TCC CTG TCT GCC AGG TTG GT (ER 1039–1067); ER E, GGAATTC AAA AAC AGG AGG AAG AGC TGC (ER 691–711); ER F, GCA AAC AGT AGC TTC CCT GGG (ER 1194–1214); ER F2, CAG GAT CTC TAG CCA GGC AC (ER 1142–1161); ER 1, GGA GAC ATG AGA GCT GCC AAC (ER 850–870); ER 2, CCA GCA GCA TGT CGA AGA TC (ER 1269–1288); ER G, GGAATTC CCT TCT AGA ATG TGC CTG GC (ER 1131–1150); ER H, TTC TCT TCC AGA GAC TTC AGG G (ER 1394–1415); ER 6, TAG AGG GCA TGG TGG AGA TC (ER 1253–1272); ER 8, CTT CAT GCT GTA CAG ATG CTC C (ER 1566–1587); and ER 8.2, GTA ACA AAG GCA T (ER 1553–1565).



View larger version (13K):
[in this window]
[in a new window]
  		Figure 1. Position of ER oligonucleotides used as primers for RT-PCR and probes for Southern blots of RT-PCR products. Details of the ER sequence are taken from the reports by Greene et al. (25 ) and Ponglikitmongkol et al. (26 ). Primers and probes used to examine each of exons 2–7 are as follows: exon 2: primers ER A and ER B, probes ER E and ER C; exon 3: primers ER C and ER D, probe ER 1; exon 4: primers ER E and ER F, probes ER F2 and ER 1; exon 5: primers ER 1 and ER 2, probes ER D and ER F2; exon 6: primers ER G and ER H, probe ER F; and exon 7: primers ER 6 and ER 8, probes ER 2, ER H and ER 8.2.

 
For each reaction, 200–600 ng tumor RNA were reverse transcribed by 50 U Moloney murine leukemia virus reverse transcriptase at 37 C for 60 min in a 20-µL reaction using the specific antisense primer in the presence of ribonuclease inhibitor and 1 mmol/L deoxy-NTPs. Subsequent amplification of the complementary DNA (cDNA) produced was achieved in a 100-µL reaction by addition of the sense primer, MgCl2 to a final concentration of 1.5 mmol/L, 2.5 U Taq polymerase, and the recommended concentration of Taq reaction buffer supplied by the manufacturer. In each reaction a sample without RNA was included as a negative control. RNA from the breast cancer cell line T47D was included as a positive control. PCR conditions were optimized for each primer set, and 30 or 40 cycles of amplification were performed, except for the reaction examining exon 7 of ER (see below). Amplification of fragments of the PR transcript by RT-PCR from the same tumors was performed by a similar method, which has previously been described (16).

Southern blot analysis

Aliquots of each PCR product mix were electrophoresed through agarose gels containing ethidium bromide and were transferred to nylon membranes, (Hybond N+, Amersham, Castle Hill, Australia), using capillary action. Membranes were probed with exon-specific oligonucleotide sequences from ER and PR and PR cDNA, labeled with 32P. Radioactively labeled membranes were evaluated using a Molecular Dynamics, Inc., PhosphorImager and ImageQuant software (Sunny- vale, CA).

A semiquantitative estimate of alternatively spliced transcript expression was made by comparing the intensities of the variant and wild-type bands on Southern blots as previously described (21, 28). Levels of the variant transcripts were expressed as a percentage of the wild-type level. Levels of the {Delta}5ER reported in Table 1Go are derived from the mean results of 2 separate determinations for each tumor. Relative expression of the most abundant variant, the {Delta}7ER, was examined in detail. For this reaction, the kinetics of amplification were documented by comparing the intensity of the wild-type and {Delta}7ER signals on Southern blot over a range of PCR cycles. These results showed that both products were amplified with similar efficiency, and the ratio of the two products was comparable throughout the reaction (data not shown). Semiquantitative results of the relative expression of {Delta}7ER are the mean values from 3 separate RT-PCR reactions for 30 tumors and from 2 reactions for 5 cases. In each case, measurements were made in both the plateau (30 cycles) and linear (19 cycles) phases of amplification.


View this table:
[in this window]
[in a new window]
  		Table 1. Level of variant receptor transcripts relative to wild type

 
Subcloning and sequencing

Sequence verification of the {Delta}7ER, {Delta}4PR, and {Delta}6PR variants was performed by subcloning the RT-PCR products into the pGEM-T vector (Promega Corp., Sydney, Australia). The {Delta}4/2PR variant was sequenced directly from the RT-PCR product. Sequencing was performed by automated dideoxy chain termination sequencing (373A DNA Sequencing System, Applied Biosystems, Foster City, CA).

Statistical methods

Statistical tests were performed using Abacus Concepts StatView Student software (Abacus Concepts, Inc., Berkeley, CA).


   Results
Top
 Abstract
 Introduction
 Subjects and Methods
 Results
Discussion
 References
 
Expression of alternatively spliced variant ER transcripts

Segments of the ER transcript spanning each of exons 2–7 were separately amplified by RT-PCR. Examples of Southern blots from these experiments are shown in Fig. 2Go. A region spanning exon 2 was amplified, and a product corresponding in size to the expected 446-bp wild-type ER was seen (Fig. 2AGo). This band hybridized with an oligonucleotide positioned in exon 3 (ER E) on Southern blots, and in addition, a smaller product corresponding in size to an ER transcript lacking exon 2 ({Delta}2ER) was revealed in all cases. Further evidence of the identity of the variant was derived from an experiment in which a duplicate of one of the Southern blots was probed with an oligonucleotide from the exon 2 ER sequence (ER C), and the smaller band failed to hybridize (not shown).



View larger version (77K):
[in this window]
[in a new window]
  		Figure 2. Southern blots of ER RT-PCR products demonstrating the presence of variant ER transcripts. A, Fragment spanning exon 2 probed with ER E positioned in exon 3. The wild-type ER (wtER) product is 446 bp in length, and {Delta}2ER is 191 bp. B, Fragment spanning exon 3 probed with ER 1 in exon 4. The wtER product is 502 bp in length. C, Fragment spanning exon 4 probed with ER F2 in exon 5. The wtER product is 531 bp, and the {Delta}4ER is 195 bp. D, Fragment spanning exon 5 probed with ER D in exon 4. The wtER is 439 bp, and the {Delta}5ER is 300 bp. E, Fragment spanning exon 6 probed with ER F in exon 5. The wtER fragment is 292 bp in length. F, Fragment spanning exon 7 probed with ER 8.2 from exon 8. The wtER product is 335 bp in length, and {Delta}7ER is 151 bp.

 
When the region spanning exon 3 was amplified, a band corresponding to the wild-type ER sequence was seen on agarose gels and on Southern blots probed with an oligonucleotide that binds to exon 4 of ER (ER 1; Fig. 2BGo). A smaller band was only faintly seen in some cases. To visualize products more clearly, RT-PCR products from a subset of cases were run on 6% polyacrylamide gels and stained with silver. By this method, a band of approximately the correct size for the exon 3-deleted ER variant ({Delta}3ER) was seen, which may represent a variant ER transcript, however, the intensity of this band was very low compared with that of wild-type ER (data not shown). It was concluded that the {Delta}3ER was not confidently identified in this series, and if present, the level of expression was likely to be very low.

The predicted exon 4-deleted ER transcript ({Delta}4ER) was seen in addition to wild-type ER on Southern blots probed with an exon 5 ER sequence (ER F2; Fig. 2CGo) in 33 of the 35 cases, and when 1 of the blots was stripped and reprobed with an oligonucleotide in exon 4 (ER 1), the variant was not detected (not shown). Primers used in the original description of the exon 5-deleted ER variant ({Delta}5ER) were used to amplify the ER sequence spanning exon 5 (27) (Fig. 2DGo). The {Delta}5ER, identified on the basis of its size and differential hybridization to an oligonucleotide probe in exon 4 (ER D) but not one in exon 5 (ER F2; not shown), was detected in 29 of the 35 tumors. Amplification of exon 6 revealed no evidence of an exon 6-deleted variant ({Delta}6ER) in any of the cases on Southern blots (Fig. 2EGo) or when a subset was run on acrylamide gels and stained with silver (data not shown).

In all of the ER-positive cases, amplification across exon 7 revealed wild-type and the exon 7-deleted variant ({Delta}7ER). In contrast to the other variants, the {Delta}7ER could be discerned on agarose gels containing ethidium bromide; consistent with this, it was the most intense signal relative to that of the wild type on Southern blots (Fig. 2FGo). The variant hybridized with ER sequences from exons 6 (ER 2) and 8 (ER 8.2), but not a probe positioned in exon 7 (ER H) (not shown). The identity of the variant was further proven by subcloning and sequencing of the wild-type and {Delta}7ER RT PCR products from two tumors (not shown).

Expression of alternatively spliced PR transcripts

Of the 35 ER-positive tumors examined, 26 were PR positive by EIA. The PR transcript was examined in 25 of these by amplifying 2 fragments by RT-PCR: one spanning exons 2, 3, and 4 and the other spanning exons 5, 6, and 7, (16). Using this strategy, a number of variant PR transcripts were detected in addition to wild-type PR. Southern blots from these experiments are illustrated in Fig. 3Go.



View larger version (36K):
[in this window]
[in a new window]
  		Figure 3. Southern blots of PR RT-PCR products. A, Fragment spanning PR exons 2, 3, and 4 probed with PR cDNA. The wild-type product is 731 bp in length. In addition, products of, 605/614, 579, 462, and 426 bp were seen corresponding to the {Delta}4/2PR and {Delta}3PR, {Delta}2PR, {Delta}2+3PR, and {Delta}4PR variants, respectively. B, Fragment spanning PR exons 5, 6, and 7 probed with PR cDNA. Wild-type PR is 788 bp in length, and a smaller product of 657 bp, representing the {Delta}6PR, is also seen. In a proportion of tumors a very faint band corresponding to the 513-bp {Delta}5+6PR was seen with longer exposure of the phosphorscreen (not shown). C, Sequence spanning the site of deletion of 126 bp of the PR sequence in the {Delta}4/2PR variant. The sequence was derived using an antisense PR primer and therefore reads from 3' to 5'. The precise location of the splice junction in the variant is impossible to determine, as the first two nucleotides of PR exon 4 are GT, and this nucleotide combination is also present at position 2276 of the PR cDNA sequence. It is likely, however, that the variant is formed by the splicing together of the native junction at the end of exon 3 and a cryptic splice site at position 2776 (29 ).

 
Amplification across exons 2, 3, and 4 revealed smaller bands (Fig. 3AGo), which, on the basis of size and the pattern of hybridization to oligonucleotide probes positioned in exons 2 and 3, are likely to represent variant PR transcripts from which exon 2 ({Delta}2PR) and both exons 2 and 3 ({Delta}2+3PR) were deleted. On Southern blots, each of these variants produced only a faint signal, which was not clearly discerned in all cases. A PR variant from which exon 4 of the PR sequence had been deleted ({Delta}4PR) was identified in 24 of the 25 cases. The identity of this species was confirmed by the pattern of differential hybridization to oligonucleotide probes; in addition, the {Delta}4PR variant was subcloned and sequenced from 3 tumors (data not shown). A product of approximately 600 bp was also seen in these experiments. Differential hybridization with oligonucleotides in exons 3 and 4 of PR revealed that 2 separate species were contributing to this band: a product from which exon 3 of the PR sequence was deleted ({Delta}3PR) and a novel variant from which the first 126 bp of exon 4 were deleted ({Delta}4/2PR; Fig. 3CGo).

When a fragment of PR spanning exons 5, 6, and 7 was amplified (Fig. 3BGo), a product of 657 bp was identified. When this was subcloned and sequenced, it was proven to be a PR variant from which the exon 6 sequence was deleted ({Delta}6PR). In addition, a faint band corresponding in size to a transcript lacking both exon 5 and 6 sequences ({Delta}5+6PR) was noted. No transcripts lacking exon 7 of PR were detected.

Relative expression of ER and PR alternatively spliced transcripts

When the level of expression of variant transcripts was compared semiquantitatively with that of the wild-type ER or PR transcript, it was found that for each variant there was a characteristic level of expression, and the range for the tumors examined was quite small (Table 1Go). In general, ER variant transcripts were more abundant than PR variants.

The most highly expressed species was the {Delta}7ER, which was detected at levels ranging from 29–83% of wild-type ER (mean, 51%). Mean levels of expression of the other ER variants were less than 25% that of wild type. Of the PR variant transcripts, the most abundant species was the {Delta}6PR, which was detected in 23 of 25 tumors and had a mean level of expression of 8% that of wild-type PR. The other 2 relatively abundant PR variant species were {Delta}4PR and {Delta}4/2PR, which had mean levels of expression of 2% and 4% that of the wild type, respectively.

For each alternatively spliced transcript species detected, there were individual tumors that contained high levels compared with the overall population; however, a general propensity to form alternatively spliced transcripts was not seen as these cases did not necessarily express high levels of other variant species (data not shown).

Coexpression of alternatively spliced ER and PR transcripts and expression of the estrogen-responsive proteins PR and pS2

To determine whether the expression of variant transcripts was related to markers of hormonal sensitivity in breast cancers, the profile of expression of the estrogen-responsive proteins, PR and pS2, was examined in the 35 primary breast cancer specimens (Table 2Go). All of the tumors were ER positive by EIA, except for a single case with ER of 4 fmol/mg protein but ER positive by immunohistochemistry which was therefore included. In this small cohort there was considerable heterogeneity in the expression of the two proteins, with 51% being PR+pS2+, 14% being PR-pS2+, 23% being PR+pS2-, and 11% being PR-pS2-.


View this table:
[in this window]
[in a new window]
  		Table 2. Presence of alternatively spliced variant ER and PR transcripts in 35 primary breast tumors

 
Despite the heterogeneity in putative ER function evidenced by the combined PR/pS2 phenotype, a common profile of variant ER transcripts was seen in this cohort. All of the tumors contained the {Delta}2ER and {Delta}7ER, and none contained the {Delta}3ER and {Delta}6ER variants. The {Delta}4ER was absent from 2 cases only. The {Delta}5ER was present in 29 of the 35 cases. Three of the 6 cases in which the variant was not identified contained both PR and pS2; one was PR-pS2+, one was PR+pS2-, and the remaining case expressed neither protein. The distribution with respect to PR and pS2 was similar for cases that contained the {Delta}5ER as for those that did not. It was clear, therefore, that there was no relationship between the profile of ER variants and receptor function.

The level of expression of the most abundant variant, {Delta}7ER, was compared with the expression of PR and pS2. When the cases were subdivided according to the pattern of expression of PR and pS2, no relationship with the {Delta}7ER level was seen (by one-factor ANOVA, P = 0.19; Fig. 4Go).



View larger version (15K):
[in this window]
[in a new window]
  		Figure 4. Relative level of the {Delta}7ER variant transcript in 35 ER-positive breast tumors, divided according to the pattern of expression of the estrogen-responsive proteins PR and pS2. The mean level of {Delta}7ER compared with that of wild-type ER was 56% for tumors that were PR+pS2+, 42% for tumors that were PR-pS2+, 50% for tumors that were PR+pS2-, and 45% for tumors that were PR-pS2-. On the basis of {Delta}7ER expression there was no significant difference among the four groups (by one-factor ANOVA, P = 0.19).

 
In the subset of tumors that were PR+, there was also a common profile of variant transcripts, as {Delta}3PR, {Delta}4PR, {Delta}4/2PR, and {Delta}6PR species were detected in the majority of tumors. The {Delta}2PR and transcripts lacking two exons ({Delta}2+3PR and {Delta}5+6PR) were detected infrequently (Table 2Go); however, it was notable that these latter transcripts were present in very low abundance, so it cannot be excluded that failure to detect these in some tumors may be related to the limit of detection of the method.

When expression of the most abundant alternatively spliced PR transcripts ({Delta}4PR, {Delta}4/2PR, and {Delta}6PR) was compared on the basis of pS2 positivity, no difference was noted between pS2+ and pS2- tumors (Table 2Go).


   Discussion
Top
 Abstract
 Introduction
 Subjects and Methods
 Results
 Discussion
References
 
In this series of 35 ER-positive primary breast tumors, the pattern of expression of the estrogen-responsive proteins PR and pS2 revealed considerable variability in putative ER function; however, the pattern of alternative splicing of the ER transcript was uniform. All of the tumors contained the {Delta}2ER and {Delta}7ER variant transcripts, whereas {Delta}3ER and {Delta}6ER variants were not seen. The {Delta}4ER was absent from two cases only. The {Delta}5ER was detected in the majority of tumors, but there was no apparent relationship between expression of the {Delta}5ER and the pattern of expression of PR and pS2. There was no concordance between the small groups of cases in which {Delta}4ER and {Delta}5ER were not found. It is apparent, therefore, that there is a common profile of variant ER transcripts in breast cancer, and this feature is not related to receptor function, as revealed by the presence of estrogen-dependent end points. These data are consistent with results reported by Zhang et al. (18).

A subset of 25 tumors from this cohort that were both ER and PR positive were examined using a similar strategy for the presence of variant PR transcripts. Multiple truncated PR transcripts were detected in all cases. The {Delta}3PR, {Delta}4PR, {Delta}6PR, and {Delta}4/2PR species were detected in almost all tumors. There was a single PR+pS2+ case in which {Delta}4PR, {Delta}6PR, and {Delta}4/2PR were not seen and two other cases that failed to express either {Delta}3PR or {Delta}6PR. The level of expression of the other less commonly detected variants, {Delta}2PR, {Delta}2+3PR, and {Delta}5+6PR, was very low, and it is likely therefore that their apparent absence from some cases is due to their being below the level of detection.

These data suggest that primary breast tumors are not distinguished by the profile of expression of ER or PR variant transcripts and that this feature cannot, therefore, be indicative of hormone responsiveness. Further, a general propensity for alternative splicing of ER and PR transcripts was not seen, as cases that expressed high levels of a specific variant relative to other members of the cohort did not necessarily contain high levels of others.

Alternative splicing of the primary transcript is the most likely mechanism for the formation of variant transcripts from which entire exon sequences have been deleted. With respect to both ER and PR, the multiplicity of variants detected and the distinct profiles of these indicate that the primary transcripts are subject to a number of alternative splicing events, and there are specific splice junctions that are vulnerable to this process. It is noted, for example that although exons 2, 4, 5, and 7 were commonly excluded from ER, a variant ER transcript lacking exon 6 was not detected in this series and has never been identified. Similarly for PR, deletion of exons 2, 3, 4, and 6 were observed, whereas an exon 7-deleted variant was not seen, and exon 5 was deleted only in combination with exon 6. Insight into the selection of splice sites may come from the novel PR variant {Delta}4/2PR that was detected in this series. This transcript is the result of deletion of 126 bp from exon 4 of the PR sequence and is likely to be formed by the splicing together of the native donor splice site from the 3'-end of PR exon 3 with a cryptic acceptor site within exon 4 at position 2776 (29). It is noted that the bases bounding this cryptic splice junction conform to the acceptor splice site consensus sequence (Table 3Go), suggesting that the formation of the {Delta}4/2PR variant may be due to the cryptic splice site successfully competing with the native sequence for recognition in the splicing reaction in a small proportion of transcripts. In the formation of others of the alternatively spliced ER and PR variants, however, the degree of conformation to the consensus sequence of splice sites that are never excluded in the formation of variant transcripts is not greater than those that are, suggesting that factors other than splice site strength are likely to determine the pattern of alternative splicing.


View this table:
[in this window]
[in a new window]
  		Table 3. Comparison of the consensus splice site sequences with the intron/exon boundary sequences of ER and PR

 
One of the outstanding questions pertaining to the alternatively spliced ER transcripts is to what extent these transcripts are translated into variant proteins. Despite the availability of monoclonal anti-ER antibodies directed against the N-terminus of the protein, which should be intact in the variant, and the demonstration of antibody binding to variant proteins in vitro (11, 30, 31), studies of ER at the protein level do not readily correspond to the multitude of variant ER transcripts that have been described. There is some evidence that at least some of the proteins corresponding to alternatively spliced ER variants exist, however. Castles et al. detected a protein corresponding in size to the {Delta}5ER by immunoprecipitation followed by immunoblotting in the BT20 breast cancer cell line (32), and Desai et al. using an anti-{Delta}5ER monoclonal antibody reported specific staining in frozen sections from two breast tumors (10). In addition, Park et al. demonstrated the {Delta}4ER protein on Western blots of ovarian carcinoma and other normal tissues (14), and Erenburg et al. reported low level expression of a protein corresponding to the {Delta}3ER in the MCF-7 breast cancer cell line (13). Candidate variant PR proteins that may be encoded by the alternatively spliced PR transcripts have not been specifically demonstrated. A truncated PR protein seen in 25% of PR-positive breast tumors on Western blot is unlikely to correspond to one of these variants (16); however, other small immunoreactive bands seen in some tumors on these blots are potential candidates (22).

In vitro studies of variant receptor function have supported the view that these variants may impact on hormone sensitivity; however, this has not been confirmed to date by studies of clinical material. One reason for this discrepancy might be that the effects demonstrated in vitro have been in the context of much higher relative expression of the variants than are present in vivo. Wang et al. reported that cotransfection of a 20-fold excess of {Delta}2ER to wild-type ER in HeLa cells resulted in mild inhibition of ER function, and when {Delta}3ER was expressed at an equivalent level as wild-type ER, a 30% reduction in ER activity was noted (12). With respect to PR variants, Richer et al. reported that a 10-fold excess of {Delta}6PR reduced the activity of PR B by 60–70% and equimolar {Delta}5+6PR reduced the activity of PR B by 20% and that of PR A by 65% (17). In this study, however, the mean level of expression of {Delta}2ER transcript was 18% that of wild-type ER, and the mean level of expression of {Delta}6PR was 8% that of wild-type PR. The {Delta}3ER and {Delta}5+6PR variant transcripts were barely detectable. Although results derived from RNA extracted from a tissue homogenate are likely to be a crude reflection only of events at a molecular level, the low level of relative expression of the alternatively spliced transcripts in breast tumors imply that corresponding variants would require more powerful dominant activity than has been reported to date to significantly impact on hormone sensitivity.

The most abundant variant species detected in this series was the {Delta}7ER, which was present at levels ranging from 29–83% of wild-type levels. In this range of expression, an effect on ER function might be expected, as Fuqua et al. reported a 60% reduction in ER function in a yeast expression system when the {Delta}7ER was expressed at an equivalent level as wild-type ER (11). In contrast to reports that the {Delta}7ER was more highly expressed in ER+PR- tumors than ER+PR+ cases (11, 18), in this study the relative level of the {Delta}7ER could not be related to the expression of the estrogen-responsive proteins PR and pS2, and there was no evidence therefore to implicate the {Delta}7ER in aberrant ER function.

In summary, we have found in a series of 35 primary breast tumors that multiple alternatively spliced ER and PR transcripts coexist and that there is a common profile of these variants. This uniformity is in contrast to considerable variation in receptor levels and expression hormone-responsive markers and implies that alternative splicing of the primary transcript may not be a determinant of receptor function.


   Acknowledgments
 
The support of the Departments of Tissue Pathology and Radiation Oncology at Westmead Hospital is gratefully acknowledged.


   Footnotes
 
1 This work was supported by the National Health and Medical Research Council of Australia, the New South Wales Cancer Council, the Leo and Jenny Leukemia and Cancer Foundation, and the University of Sydney Cancer Research Fund. Back

2 R.L.B. was a National Health and Medical Research Council Medical Postgraduate Scholar. Back

Received October 27, 1998.

Revised January 12, 1999.

Accepted January 20, 1999.


   References
Top
 Abstract
 Introduction
 Subjects and Methods
 Results
 Discussion
 References
 

  1. Santen RJ, Manni A, Harvey H, Redmond C. 1990 Endocrine treatment of breast cancer in women. Endocr Rev. 11:221–265.[Medline]
  2. Wyld DK, Chester JD, Perren TJ. 1998 Endocrine aspects of the clinical management of breast cancer–current issues. Endocr Relat Cancer. 5:97–110.[Abstract]
  3. McGuire WL, Chamness GC, Fuqua SAW. 1991 Estrogen receptor variants in clinical breast cancer. Mol Endocrinol. 5:1571–1577.[Abstract]
  4. Horwitz KB. 1981 Is a functional estrogen receptor always required for progesterone receptor induction in breast cancer? J Steroid Biochem. 15:209–217.[CrossRef][Medline]
  5. Encarnacion CA, Ciocca DR, McGuire WL, et al. 1993 Measurement of steroid hormone receptors in breast cancer patients on tamoxifen. Breast Cancer Res Treat. 26:237–246.[CrossRef][Medline]
  6. Johnston SRD, Saccani-Jotti G, Smith IE, et al. 1995 Changes in estrogen receptor, progesterone receptor, and pS2 expression in tamoxifen resistant human breast cancer. Cancer Res. 55:3331–3338.[Abstract/Free Full Text]
  7. Murphy LC, Dotzlaw H, Leygue E, et al. 1997 Estrogen receptor variants and mutations. J Steroid Biochem Mol Biol. 62:363–372.[CrossRef][Medline]
  8. Fuqua SAW, Fitzgerald SD, Chamness GC, et al. 1991 Variant human breast tumor estrogen receptor with constitutive transcriptional activity. Cancer Res. 51:105–109.[Abstract/Free Full Text]
  9. Chaidarun SS, Alexander JM. 1998 A tumor-specific truncated estrogen receptor splice variant enhances estrogen-stimulated gene expression. Mol Endocrinol. 12:1355–1366.[Abstract/Free Full Text]
  10. Desai AJ, Luqmani YA, Walters JE, et al. 1997 Presence of exon 5-deleted oestrogen receptor in human breast cancer: functional analysis and clinical significance. Br J Cancer. 75:1173–1184.[Medline]
  11. Fuqua SAW, Fitzgerald SD, Allred DC, et al. 1992 Inhibition of estrogen receptor action by a naturally occurring variant in human breast tumors. Cancer Res. 52:483–486.[Abstract/Free Full Text]
  12. Wang Y, Miksicek RJ. 1991 Identification of a dominant negative form of the human estrogen receptor. Mol Endocrinol. 5:1707–1715.[Abstract]
  13. Erenburg I, Schachter B, Mira y Lopez R, Ossowski L. 1997 Loss of an estrogen receptor isoform (ER{alpha}{Delta}3) in breast cancer and the consequences of its reexpression: interference with estrogen-stimulated properties of malignant transformation. Mol Endocrinol. 11:2004–2015.[Abstract/Free Full Text]
  14. Park W, Choi J-J, Hwang E-S, Lee J-H. 1996 Identification of a variant estrogen receptor lacking exon 4 and its coexpression with wild-type estrogen receptor in ovarian carcinomas. Clin Cancer Res. 2:2029–2035.[Abstract]
  15. Leygue E, Dotzlaw H, Watson PH, Murphy LC. 1996 Identification of novel exon-deleted progesterone receptor variant mRNAs in human breast tissue. Biochem Biophys Res Comun. 228:63–68.[CrossRef][Medline]
  16. Yeates C, Hunt SMN, Balleine RL, Clarke CL. 1998 Characterization of a truncated progesterone receptor protein in breast tumors. J Clin Endocrinol Metab. 83:460–467.[Abstract/Free Full Text]
  17. Richer JK, Lange CA, Wierman AM, et al. 1998 Progesterone receptor variants found in breast cells repress transcription by wild-type receptors. Breast Cancer Res Treat. 48:231–241.[CrossRef][Medline]
  18. Zhang Q-X, Hilsenbeck SG, Fuqua SAW, Borg A. 1996 Multiple splicing variants of the estrogen receptor are present in individual human breast tumors. J Steroid Biochem Mol Biol. 59:251–260.[CrossRef][Medline]
  19. Leygue ER, Watson PH, Murphy LC. 1996 Estrogen receptor variants in normal human mammary tissue. J Natl Cancer Inst. 88:284–290.[Abstract/Free Full Text]
  20. Gallacchi P, Schoumacher F, Eppenberger-Castori S, et al. 1998 Increased expression of estrogen-receptor exon-5-deletion variant in relapse tissues of human breast cancer. Int J Cancer. 79:44–48.[CrossRef][Medline]
  21. Daffada AAI, Johnston SRD, Smith IE, et al. 1995 Exon 5 deletion variant estrogen receptor messenger RNA expression in relation to tamoxifen resistance and progesterone receptor/pS2 status in human breast cancer. Cancer Res. 55:288–293.[Abstract/Free Full Text]
  22. Graham JD, Yeates C, Balleine RL, et al. 1995 Characterization of progesterone receptor A and B expression in human breast cancer. Cancer Res. 55:5063–5068.[Abstract/Free Full Text]
  23. Detre S, King N, Salter J, et al. 1994 Immunohistochemical and biochemical analysis of the oestrogen regulated protein pS2, and its relation with oestrogen receptor and progesterone receptor in breast cancer. J Clin Pathol. 47:240–244.[Abstract/Free Full Text]
  24. Clarke CL, Roman SD, Graham J, Koga M, Sutherland RL. 1990 Progesterone receptor regulation by retinoic acid in the human breast cancer cell line T47D. J Biol Chem. 265:12694–12700.[Abstract/Free Full Text]
  25. Greene GL, Gilna P, Waterfield M, et al. 1986 Sequence and expression of human estrogen receptor complementary DNA. Science. 231:1150–1154.[Abstract/Free Full Text]
  26. Ponglikitmongkol M, Green S, Chambon P. 1988 Genomic organization of the human oestrogen receptor gene. EMBO J. 7:3385–3388.[Medline]
  27. Fuqua SAW, Falette NF, McGuire WL. 1990 Sensitive detection of estrogen receptor RNA by polymerase chain reaction assay. J Natl Cancer Inst. 82:858–861.[Abstract/Free Full Text]
  28. Daffada AAI, Johnston SRD, Nicholls J, Dowsett M. 1994 Detection of wild-type and exon 5-deleted splice variant oestrogen receptor (ER) mRNA in ER-positive and -negative breast cancer cell lines by reverse transcription/polymerase chain reaction. J Mol Endocrinol. 13:265–273.[Abstract]
  29. Kastner P, Krust A, Turcotte B, et al. 1990 Two distinct estrogen-regulated promoters generate transcripts encoding the two functionally different human progesterone receptor forms A and B. EMBO J. 9:1603–1614.[Medline]
  30. Koehorst SGA, Cox JJ, Donker GH, et al. 1994 Functional analysis of an alternatively spliced estrogen receptor lacking exon 4 isolated from MCF-7 breast cancer cells and meningioma tissue. Mol Cell Endocrinol. 101:237–245.[CrossRef][Medline]
  31. Rea D, Parker MG. 1996 Effects of an exon 5 variant of the estrogen receptor in MCF-7 breast cancer cells. Cancer Res. 56:1556–1563.[Abstract/Free Full Text]
  32. Castles CG, Fuqua SAW, Klotz DM, Hill SM. 1993 Expression of a constitutively active estrogen receptor variant in the estrogen receptor negative BT-20 human breast cancer cell line. Cancer Res. 53:5934–5939.[Abstract/Free Full Text]
  33. Smith CW, Patton JG, Nadal-Ginard B. 1989 Alternative splicing in the control of gene expression. Annu Rev Genet. 23:527–577.[CrossRef][Medline]
  34. Lewin B. 1994 The apparatus for nuclear splicing. In: Genes, V. Oxford, New York, Tokyo: Oxford University Press; 911–940.
  35. Misrahi M, Venencie P-Y, Saugier-Veber P, et al. 1993 Structure of the human progesterone receptor gene. Biochim Biophys Acta. 1216:289–292.[Medline]



This article has been cited by other articles:


Home page
 Endocr. Rev.Home page
M. Schumacher, R. Guennoun, A. Ghoumari, C. Massaad, F. Robert, M. El-Etr, Y. Akwa, K. Rajkowski, and E.-E. Baulieu
Novel Perspectives for Progesterone in Hormone Replacement Therapy, with Special Reference to the Nervous System
Endocr. Rev., June 1, 2007; 28(4): 387 - 439.
[Abstract] [Full Text] [PDF]

Home page
 J. Clin. Endocrinol. Metab.Home page
Y. Guo, H. Shen, P. Xiao, D.-H. Xiong, T.-L. Yang, Y.-F. Guo, J.-R. Long, R. R. Recker, and H.-W. Deng
Genomewide Linkage Scan for Quantitative Trait Loci Underlying Variation in Age at Menarche
J. Clin. Endocrinol. Metab., March 1, 2006; 91(3): 1009 - 1014.
[Abstract] [Full Text] [PDF]

Home page
 Mol Hum ReprodHome page
P.B. Marshburn, J. Zhang, Z.B. Mostafavi, M.L. Matthews, J. White, and B.S. Hurst
Variant progesterone receptor mRNAs are co-expressed with the wild-type progesterone receptor mRNA in human endometrium during all phases of the menstrual cycle
Mol. Hum. Reprod., November 1, 2005; 11(11): 809 - 815.
[Abstract] [Full Text] [PDF]

Home page
 J. Clin. Pathol.Home page
E Nardon, I Buda, G Stanta, E Buratti, M Fonda, and L Cattin
Insulin-like growth factor system gene expression in women with type 2 diabetes and breast cancer
J. Clin. Pathol., August 1, 2003; 56(8): 599 - 604.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a related Letter to the Editor
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Balleine, R. L.
Right arrow Articles by Clarke, C. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Balleine, R. L.
Right arrow Articles by Clarke, C. L.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Endocrinology Endocrine Reviews J. Clin. End. & Metab.
Molecular Endocrinology Recent Prog. Horm. Res. All Endocrine Journals