Article Figures & Data
Figures
- Figure 1.
AI resistance is characterized by hormone receptor switching and an aggressive phenotype. A, Kaplan–Meier estimates of disease-free survival in tamoxifen-treated (n = 77) and AI-treated (n = 89) patients according to PR expression. PR-positive patients treated with an AI did significantly better than PR-negative patients during the first 2 years of follow-up (P = 0.0206). B, table showing hormone receptor status of matched primary and resistant tumors for 6 AI-treated patients. Changes in receptor status are highlighted in pink. C, Western blot analysis shows slightly increased expression of ERα in LetR cells compared with Aro cells. D, AI-resistant cell model (LetR cells, black bars) shows reduced proliferative response to steroids and increased growth factor response compared with sensitive cells (Aro cells, gray bars). Results are mean ± SEM (n = 3). *, P < 0.01. E, LetR cells are more motile than Aro cells (P < 0.0001). Histogram shows the mean migratory area per cell (μm2) ± SEM (n = 3). The metastatic MDA-MB231 cells are shown for comparison. (Scale bars, 200 μm). F, higher levels of both MMP9 mRNA by RT-PCR and active MMP9 by gelatin zymography in LetR cells compared with Aro cells. G, LetR cells do not form organized acini in 3D culture. Aro cells, similar to the highly polarized MCF10A cells, form 3D organized structures with hollow lumen. LetR cells fail to hollow out a lumen and remain disorganized, more comparable with SKBR3 cells. Cells are stained with DAPI (blue) and phalloidin (red), and images are representative of 3 separate experiments. (Scale bars, 20 μm).
- Figure 2.
Response of SRC-1, Ets2, and Myc to steroid treatments. A, location of SRC-1–binding peak within the proximal promoter region of Myc gene as detected by ChIP sequencing analysis in endocrine-resistant LY2 cells. RNA sequencing confirms expression of Myc mRNA in these cells. X marks the location of an E2F-binding site within the Myc promoter. B, protein levels of SRC-1, Ets2, phospho-Ets2 (pEts2), and Myc are higher in LetR than in Aro and MCF7 cells. Western blot images are representative, and densitometry graphs represent relative mean normalized expression (n = 3). Error bars represent SEM. C, SRC-1 and Myc protein expression is sensitive to letrozole treatment in Aro cells but insensitive to letrozole in LetR cells. Ets2 expression is not regulated by steroid treatments in either cell line. Cells were treated with vehicle (V), estrogen (E), androstenedione (A), letrozole (L), or a combination (A+L). Western blot images are representative, and densitometry graphs represent relative mean normalized expression (n = 3). Error bars represent SEM. D, confocal images of SRC-1 localization in Aro and LetR cells in the presence and absence of androstenedione and letrozole alone and in combination. Nuclear localization of SRC-1 increases in Aro cells in response to androstenedione and is reduced when letrozole is added. By contrast, nuclear intensity of SRC-1 is strong in LetR cells independent of treatments. Images are taken at ×40 magnification with a confocal fluorescent microscope.
- Figure 3.
SRC-1 and Ets2 regulate expression of target genes Myc and MMP9. A, SRC-1 and Ets2 coimmunoprecipitate with strongest interaction after 45-minute steroid treatment. Aro cells were treated with androstenedione for 0 to 4 hours. Protein was immunoprecipitated (IP) with an anti-Ets2 antibody and immunoblotted (IB) for SRC-1 and Ets2. B, SRC-1 and Ets2 are recruited to the Myc and MMP9 promoters. ChIP analysis in Aro and LetR cells with the same treatments as in Fig. 2C. Recruitment to both promoters was letrozole sensitive in Aro cells and letrozole insensitive in LetR cells. Graphs show real-time PCR relative quantification of ChIP results. Anti-H4 antibody was used as a positive control and IgG as a negative ChIP control. Genomic DNA (+ve) and water (−ve) were used as PCR controls. A distal promoter region was used to confirm specificity of recruitment to the promoter region. C, overexpression of SRC-1 resulted in increased transcript levels of both Myc and MMP9 (RT-PCR analysis) in Aro cells. Increased Myc expression was also seen at the protein level (Western blot) but no change in secreted levels of MMP9 protein was observed (zymography). D, overexpression of Ets2 resulted in increased transcript levels of both Myc and MMP9 (RT-PCR analysis) in Aro cells. Increased Myc expression was also seen at the protein level (Western blot), but no change in secreted levels of MMP9 protein was observed (zymography). E, overexpression of Ets2 resulted in increased expression of Myc and MMP9 in an SRC-1–dependent manner. Myc and MMP9 mRNA expression was increased in response to Ets2 overexpression in LetR cells, but this increase was inhibited when the cells were concomitantly transfected with SRC-1 siRNA.
- Figure 4.
SRC-1 has a functional role in migration and inhibits acini formation in LetR cells. A, SRC-1 knockdown in LetR cells results in decreased migration. Histogram shows the mean migratory area per cell (μm2) ± SEM and was significantly less for SRC-1 knockdown than for nontargeting (nt) control (P = 0.0007). Aro cells are shown for comparison. (Scale bars, 200 μm). Western blot confirms SRC-1 protein knockdown. B, SRC-1 knockdown in LetR cells results in increased ability to form organized 3D acini. Cells from 3D assay are stained with DAPI (blue) and phalloidin (red), and results are representative of 3 separate experiments. Aro cells are shown for comparison. (Scale bars, 20 μm). C, functional migratory role of SRC-1 in AI resistance is not dependent on ERα. Western blot confirms successful ERα knockdown with siRNA. Histogram shows only a marginal decrease in the mean migratory area per cell in LetR cells following ERα knockdown. These cells migrate significantly more than LetR cells with SRC-1 knockdown (P = 0.0377).
- Figure 5.
SRC-1 significantly associates with disease recurrence in AI-treated patients. A, immunohistochemical staining of SRC-1 in tissue microarray cores, counterstained with hematoxylin. Examples of SRC-1–positive and–negative primary tumors are shown. B, Kaplan–Meier estimates of disease-free survival in the tamoxifen- (n = 75) and AI- (n = 84) treated populations. SRC-1–positive primary tumors (red line) were significantly associated with reduced disease-free survival in the tamoxifen-treated population (P = 0.0326) but not significantly in the AI-treated population (P = 0.6894). C, expression of SRC-1 was increased and more nuclear in AI-resistant metastatic tumors than in matched primary tumors (n = 3). Representative images are shown of a matched primary breast tumor and a metastatic lung tumor from 1 AI-treated patient. IgG was used as a negative control. Kaplan–Meier estimates represent disease-free survival in the AI-treated population (n = 84) according to SRC-1 staining in the primary or resistant metastatic tumor. SRC-1 is significantly associated with reduced disease-free survival (P = 0.0106). (continued on following page) D, immunofluorescent analysis of SRC-1 (green) and phospho-Ets2 (red) expression in matched primary breast and AI-resistant tumor samples (n = 3), counterstained with DAPI (blue). Expression of both proteins was stronger and more nuclear in the resistant samples. Representative images are shown (n = 3). Merged image shows that both proteins colocalize in the nucleus of these metastatic cells (white arrows). The extent of coassociation was measured by Pearson correlation, R(r), and is significantly higher in the resistant tissue than in the primary breast tissue (P = 0.0004, n = 3). E, immunohistochemical analysis of Myc and MMP9 in matched primary breast and AI-resistant tumor samples (n = 3), counterstained with hematoxylin. Representative images are shown. Myc was strongly expressed in resistant tumors. MMP9 was weakly expressed in resistant tumors and absent from primary tumors.
Tables
- Table 1.
Associations of SRC-1 with clinicopathologic variables using Fisher exact test in endocrine, AI-, and tamoxifen-treated patient populations
Parameter Endocrine-treated population AI (N = 84) Tamoxifen (N = 75) N = 141 SRC-1% P SRC-1% P SRC-1% P PR status positive 106 61.3 57.4 62.5 negative 35 62.9 1.000 43.5 0.328 78.9 0.263 Her2 status positive 26 65.4 46.7 70.6 negative 115 60.9 0.834 55.1 0.659 65.5 1.000 Recurrence positive 23 87.0 66.7 93.3 negative 118 56.8 0.009 52.0 0.494 60.0 0.015 Nodal status positive 71 67.6 79.5 70.7 negative 68 55.9 0.167 50.0 0.658 61.8 0.466 Tumor grade I 21 57.1 50.0 66.7 II 79 63.3 56.5 65.1 III 41 61.0 0.861 50.0 0.818 70.0 0.943 Tumor stage I 56 53.6 41.9 62.1 II 64 57.8 50.0 63.9 III 18 94.4 100.0 85.7 IV 3 100.0 0.003 — 0.001 100.0 0.513 NOTE: The percentage of SRC-1% patients within each parameter is listed. Parameters include recurrence (positive) or no recurrence (negative); node positive (1 or more nodes positive) or negative (no positive nodes).
Supplementary Data
Files in this Data Supplement:
- Supplementary Figure 1 - PDF file - 175K
- Supplementary Figure 2 - PDF file - 154K
- Supplementary Table 1 - PDF file - 28K
Volume 72, Issue 2
