**2.1 ERα splice variants**

The two most referenced ERα isoforms that seem to be of particular significance are **ERα46** and **ERα36** as they were reported to oppose genomic actions of full length **ERα66** (figure 4).

The **ERα46** isoform has been identified in the MCF7 breast cancer cell line (Penot *et al.*, 2005) in which it is coexpressed with full length ERα66. The presence of ERα46 has also been confirmed in osteoblasts (Wang *et al.*, 2005) and endothelial cells (Figtree *et al.*, 2003). This isoform is formed by skipping exon 1 encoding the N-terminus (A/B) and it is devoid of AF-1 activity. In contrast with full length ERα66, the truncated isoform ERα46 does not mediate E2 dependent cell proliferation and high levels of this isoform have been shown to be associated with cell cycle arrest in the G0/G1 phase and a state of refraction to E2 stimulated growth, which is normally reached at hyperconfluency of the cells (Penot *et al.*, 2005). Similarly to ERβ, ERα46 is a potent ligand-dependent transcription factor containing AF-2 and a powerful inhibitor of ERα AF-1 dependent transcription (Figtree *et al.*, 2003). By inhibition of ERα66 dependent gene transcription, ERα46 isoform inhibits estrogenic induction of c-Fos and Cyclin D1 promoters, which are involved in cell cycle control. Coexpression of ERα46 with ERα66 in an SaOs osteoblast cell line results in concentration dependent inhibition of E2 stimulated cell proliferation (Ogawa *et al.*, 1998b), an effect similar to the consequence observed with coexpression of ERα with ERβ (Sotoca *et al.*, 2008; Ström *et al.*, 2004).

The second truncated ERα isoform ERα36 was first described recently (Wang *et al.*, 2005), and it has been shown to lack both the AF-1 and AF-2 transactivation functions of full length ERα. However it has functional DBD, partial dimerization and LBD domains. ERα36 contains an exon coding for myristoylation sites, hence predicting an interaction with the plasma membrane. Transcription of this ERα36 isoform is initiated from a previously unidentified promoter in the first intron of the ERα gene and the unique 27 amino acid Cterminal sequence is encoded by a novel ERα exon, localized downstream of exon 8 to replace the last 138 amino acids encoded by exon 7-8 (Wang *et al.*, 2005).

Fig. 4. Schematic comparison between full length ERα and its most referenced truncated isoforms.

This novel isoform has been cloned from a human placenta cDNA library, which indicates that it is a naturally occurring isoform of ERα. With no functional AF-1 and AF-2 ERα36 does not have any direct transcriptional activity. However, it is a robust inhibitor of full length ERα and ERβ dependent transactivation (ZhaoYi Wang *et al.*, 2006). It is mainly localized in the plasma membrane and works in a different way than full length protein. Even though it lacks transcriptional activity it can activate non genomic ER pathways such as MAPK/ERK signaling in response to E2 which is of particular significance in response to antiestrogens such as tamoxifen, 4OH-tamoxifen and ICI-182.780 (ZhaoYi Wang *et al.*, 2006). As a result of MAPK/ERK pathway activation by E2 and these antiestrogens a signal is transduced to the nucleus and consequently Elk1 transcription factor is activated. The effect of MAPK/ERK activation mediated by ERα36 is increased cell proliferation in response to E2 as well as antiestrogens in doses that shut down transcriptional activity of full length ERα and ERβ proteins (ZhaoYi Wang *et al.*, 2006).

The **ERα80** isoform was detected in the MCF7:2A cell line, which is a subclone MCF7 cell line derived from long term growth in the absence of E2. This ERα80 isoform was produced by duplication of exons 6 and 7 (Pink *et al.*, 1996). No evident function has been described so far.

proliferation (Ogawa *et al.*, 1998b), an effect similar to the consequence observed with

The second truncated ERα isoform ERα36 was first described recently (Wang *et al.*, 2005), and it has been shown to lack both the AF-1 and AF-2 transactivation functions of full length ERα. However it has functional DBD, partial dimerization and LBD domains. ERα36 contains an exon coding for myristoylation sites, hence predicting an interaction with the plasma membrane. Transcription of this ERα36 isoform is initiated from a previously unidentified promoter in the first intron of the ERα gene and the unique 27 amino acid Cterminal sequence is encoded by a novel ERα exon, localized downstream of exon 8 to

Fig. 4. Schematic comparison between full length ERα and its most referenced truncated

ERα and ERβ proteins (ZhaoYi Wang *et al.*, 2006).

This novel isoform has been cloned from a human placenta cDNA library, which indicates that it is a naturally occurring isoform of ERα. With no functional AF-1 and AF-2 ERα36 does not have any direct transcriptional activity. However, it is a robust inhibitor of full length ERα and ERβ dependent transactivation (ZhaoYi Wang *et al.*, 2006). It is mainly localized in the plasma membrane and works in a different way than full length protein. Even though it lacks transcriptional activity it can activate non genomic ER pathways such as MAPK/ERK signaling in response to E2 which is of particular significance in response to antiestrogens such as tamoxifen, 4OH-tamoxifen and ICI-182.780 (ZhaoYi Wang *et al.*, 2006). As a result of MAPK/ERK pathway activation by E2 and these antiestrogens a signal is transduced to the nucleus and consequently Elk1 transcription factor is activated. The effect of MAPK/ERK activation mediated by ERα36 is increased cell proliferation in response to E2 as well as antiestrogens in doses that shut down transcriptional activity of full length

The **ERα80** isoform was detected in the MCF7:2A cell line, which is a subclone MCF7 cell line derived from long term growth in the absence of E2. This ERα80 isoform was produced by duplication of exons 6 and 7 (Pink *et al.*, 1996). No evident function has been

isoforms.

described so far.

coexpression of ERα with ERβ (Sotoca *et al.*, 2008; Ström *et al.*, 2004).

replace the last 138 amino acids encoded by exon 7-8 (Wang *et al.*, 2005).

Several other multiple splice variants (**ERαΔE2, ERαΔE3, ERαΔE4, ERαΔE5, ERαΔE6, ERαEΔ5,7, ERαEΔ7**…) as a result of exon splicing deletions have been confirmed in human (Poola *et al.*, 2000; Zhang *et al.*, 1996) showing a dominant inhibitory effect in normal ER function. A list of selected ERα splice variants and their expression in various breast tissues (normal and tumor) and breast cancer cell lines is given in Table 2.


Table 2. List of selected ERα splice variants and their expression in various breast tissues (normal and tumour) and breast cancer cell lines.
