ESR1 and PGR polymorphisms are associated with estrogen and progesterone receptor expression in breast tumors

Abstract

Hormone receptor-positive (HR+) breast cancers express the estrogen (ERα) and/or progesterone (PgR) receptors. Inherited single nucleotide polymorphisms (SNPs) in ESR1, the gene encoding ERα, have been reported to predict tamoxifen effectiveness. We hypothesized that these associations could be attributed to altered tumor gene/protein expression of ESR1/ERα and that SNPs in the PGR gene predict tumor PGR/PgR expression. Formalin-fixed paraffin-embedded breast cancer tumor specimens were analyzed for ESR1 and PGR gene transcript expression by the reverse transcription polymerase chain reaction based Oncotype DX assay and for ERα and PgR protein expression by immunohistochemistry (IHC) and an automated quantitative immunofluorescence assay (AQUA). Germline genotypes for SNPs in ESR1 (n = 41) and PGR (n = 8) were determined by allele-specific TaqMan assays. One SNP in ESR1 (rs9322336) was significantly associated with ESR1 gene transcript expression (P = 0.006) but not ERα protein expression (P > 0.05). A PGR SNP (rs518162) was associated with decreased PGR gene transcript expression (P = 0.003) and PgR protein expression measured by IHC (P = 0.016), but not AQUA (P = 0.054). There were modest, but statistically significant correlations between gene and protein expression for ESR1/ERα and PGR/PgR and for protein expression measured by IHC and AQUA (Pearson correlation = 0.32-0.64, all P < 0.001). Inherited ESR1 and PGR genotypes may affect tumor ESR1/ERα and PGR/PgR expression, respectively, which are moderately correlated. This work supports further research into germline predictors of tumor characteristics and treatment effectiveness, which may someday inform selection of hormonal treatments for patients with HR+ breast cancer.

Keywords: AQUA; RT-PCR; estrogen receptor; genotype; immunohistochemistry; progesterone receptor.

Fig. 1.

Immunohistochemistry (IHC) staining for estrogen receptor (ERα) and progesterone receptor (PgR). Representative images of tumors with high and low staining for ERα (high staining, A; low staining, B) and PgR (high staining, C; low staining, D).

Fig. 2.

Consort diagram depicting patient flow from the observational clinical study into this secondary analysis. DCIS, ductal carcinoma in situ.

Fig. 3.

Association between the ESR1 single nucleotide polymorphism (SNP) rs9322336 and ESR1 gene and ERα protein expression. There was a statistically significant association between rs9322336 in the ESR1 gene and normalized ESR1 gene expression (A); however, this SNP was not associated with ERα protein expression measured by IHC (B) or automated quantitative immunofluorescence assay (AQUA, C).

Fig. 4.

Association between the PGR SNP rs518162 and PGR gene and PgR protein expression. There was a statistically significant association between rs518162 in the PGR gene and normalized PGR gene expression (A) and PgR protein expression measured by IHC (B) but not AQUA (C).

Fig. 5.

Correlation of gene and protein expression. Moderate, but statistically significant, correlations were found for ESR1/ERα and PGR/PgR gene expression measured by RT-PCR and protein expression measured by IHC (A and B) and for protein expression measured by IHC and AQUA (C and D). p = P value, r = Pearson correlation coefficient. The dotted lines in A and B signify the cut-offs for ER and PgR positivity of the RT-PCR assay.