|Home | About | Journals | Submit | Contact Us | Français|
Steroid receptor coactivator-3 (SRC-3) is a histone acetyltransferase and nuclear hormone receptor (NHR) co activator, located on 20q12, which is amplified in several epithelial cancers and well studied in breast cancer. However, its possible role in lung cancer pathogenesis is unknown. We found SRC-3 over-expressed in 27% of NSCLC patients (N=311) by immunohistochemistry, which correlated with poor disease-free (p=0.0015) and overall (p=0.0008) survival. Twenty-seven percent of NSCLCs exhibited SRC-3 gene amplification, and we found lung cancer cell lines expressed higher levels of SRC-3 than immortalized human bronchial epithelial cells (HBECs), which in turn expressed higher level of SRC-3 than cultured primary human HBECs. siRNA-mediated down-regulation of SRC-3 in high-expressing (but not low expressing) lung cancer cells significantly inhibited tumor cell growth and induced apoptosis. Finally, we found that SRC-3 expression is inversely correlated with gefitinib sensitivity and that SRC-3 knockdown results in EGFR-TKI-resistant lung cancers becoming more sensitive to gefitinib. Together these data suggest that SRC-3 may be an important oncogene and therapeutic target for lung cancer.
Steroid receptor coactivator-3 (SRC-3; AIB1/ACTR/RAC3/p/CIP) is a member of the p160 SRC family. SRC-3 has histone acetyltransferase activity and interacts with multiple nuclear receptors and transcription factors, to regulate the expression of their target genes, including estrogen receptor (ER), progesterone receptor (PR), E2F1, nuclear factor-kB (NF-kB) and activator protein-1 (AP-1) (1). SRC-3 also plays a role in EGFR signaling (2).
SRC-3 has been implicated in the development of many human cancers. Amplification and over-expression of SRC-3 is detected in 5%–10% of ovarian and 30%–60% of breast cancers (3). SRC-3 over-expression in breast cancer is associated with high levels of HER-2/neu and EGFR, tamoxifen resistance, and poor disease-free survival, suggesting that there may be cross-talk between the SRC-3, HER2/neu and ER signaling pathways in the genesis and progression of some breast tumors (4, 5). SRC-3 is also amplified and over-expressed in many other types of cancer, including prostate cancer (6), ER, PR-negative breast cancer (7), gastric cancer (GC) and colorectal carcinoma (CC) (1). Increased SRC-3 expression is also observed during pancreatic cancer (8) and esophageal tumor progression (9). However, a comprehensive profiling of SRC-3 gene and protein expression level in lung cancers is lacking, as is knowledge of the function of SRC-3 in lung cancer cell survival and proliferation.
In this study, we show that SRC-3 is over expressed in a subset of lung cancers which correlates with poor disease free and overall survival, and in some cases is associated with DNA amplification. Knockdown of SRC-3 in lung cancers leads to reduced cell growth, decreased anchorage-independent colony formation ability and increased apoptosis in non-small cell lung cancer (NSCLC) cell lines with high endogenous levels of SRC-3. In addition, we show that SRC-3 knockdown can potentiate the effect of gefitinib in EGFR-tyrosine kinase inhibitor (EGFR-TKI) resistant cells.
We obtained archived, formalin-fixed, paraffin-embedded (FFPE) tissue from surgically resected (with curative intent) NSCLC specimens (lobectomies and pneumonectomies) containing tumor and adjacent normal epithelium tissues from the Lung Cancer Specialized Program of Research Excellence (SPORE) Tissue Bank at The University of Texas M. D. Anderson Cancer Center (Houston, TX), which has been approved by the institutional review board. The tissue had been collected from 1997 to 2001. The tissue specimens were histologically examined and classified using the 2004 World Health Organization classification system(10). We selected 311 NSCLC tissue samples (188 adenocarcinomas and 123 squamous cell carcinomas) for our TMAs. TMAs were constructed using triplicate 1-mm diameter cores per tumor, and each core included central, intermediate, and peripheral tumor tissue. Detailed clinical and pathologic information, including demographics, smoking history (never- and ever-smokers), and smoking status (never, former, and current), clinical and pathologic tumor-node-metastasis (TNM) stage, overall survival (OS) duration, and time to recurrence was available for most cases. Patients who had smoked at least 100 cigarettes in their lifetime were defined as smokers, and smokers who quit smoking at least 12 months before their lung cancer diagnosis were defined as former smokers. Tumors were pathologic TNM stages I–IV according to the revised International System for Staging Lung Cancer.
Using anti-SRC-3/AIB-1 mouse monoclonal antibody from BD Transduction Laboratories, CA, USA (Cat #: 611105) (6), immunohistochemical staining was performed as follows: 5-uM FFPE tissue sections were deparaffinized, hydrated, heated in a steamer for 10 minutes with 10 mM sodium citrate (pH 6.0) for antigen retrieval, and washed in Tris buffer. Peroxide blocking was done with 3% H2O2 in methanol at room temperature for 15 min, followed by 10% fetal bovine serum in tris-buffered saline-t for 30 min. The slides were incubated with primary antibody at 4 °C for 90 minutes, washed with phosphate-buffered saline, and incubated with biotin-labeled secondary antibody (Envision Dual Link +, DAKO, Carpinteria, CA) for 30 min. Staining for the slides was developed with 0.05% 3', 3-diaminobenzidine tetra hydrochloride, which had been freshly prepared in 0.05 mol/L Tris buffer at pH 7.6 containing 0.024% H2O2, and then the slides were counterstained with hematoxylin, dehydrated, and mounted. FFPE A549 cells were used as the positive control. For the negative control, we used the same specimens used for the positive controls but replaced the primary antibody with phosphate-buffered saline. For this antibody, we performed titration experiments using a relatively wide range of antibody concentration (1:50, 1:100, 1:200, and 1:500), including the concentration suggested by the manufacturer. One observer (M.G.R.) quantified the immunohistochemical expression using light microscopy (magnification 20×). Both nuclear and cytoplasmic expressions were quantified using a four-value intensity score (0, 1+, 2+, and 3+) and the percentage (0% to 100%) of reactivity. We defined the intensity categories as follows: 0 = no appreciable staining; 1+ = barely detectable staining in epithelial cells compared with the stromal cells; 2+ = readily appreciable staining; and, 3+ = dark brown staining of cells. Next, an expression score was obtained by multiplying the intensity and reactivity % values (theoretical range, 0–300). There were 3 cores of tissue per case in the tissue microarray. On each core, the score was obtained as above described. The average of the 3 cores was used as the score for this case. Progression was defined as either disease recurrence or death; survival time was defined as the time from diagnosis to death (or disease progression), which was censored by the last known number of follow up days. Lung cancer patients were dichotomized into high SRC-3 expression group (>1.67) and low SRC-3 expression group (≤1.67), and the cut-off value is the median SRC-3 expression among all lung cancer samples.
Comparative genomic hybridization (CGH) on cDNA microarrays was carried out as previously described (11). Briefly, 4 ug tumor and normal sex-matched reference genomic DNA were random-primer labeled with Cy5 and Cy3 respectively, then hybridized to a cDNA microarray (Stanford microarray core) containing ~39,000 cDNAs representing ~26,000 mapped genes/ESTs. Hybridized arrays were scanned on a GenePix scanner (Axon Instruments), and fluorescence ratios extracted using SpotReader software (Niles). Normalized log2 ratios were then mapped onto the genome using the NCBI genome assembly (Build 36). A preliminary report of these data (focused on TITF1 amplification) was recently published (12). Copy number at the SRC-3 locus was determined with the cghFLasso algorithm using a false discovery rate of 0.001 (13).
All cells were maintained in RPMI-1640 medium supplemented with 5% fetal bovine serum (FBS, Invitrogen). Most of these cell lines were established by John D. Minna and Adi Gazdar at the National Cancer Institute and the Hamon Cancer Center for Therapeutic Oncology at UT Southwestern Medical Center (14, 15). The other cell lines were obtained from the American Type Culture Collection. All of the cell lines have been DNA fingerprinted for provenance using the PowerPlex 1.2 kit (Promega). The DNA fingerprint are all confirmed to be same as DNA fingerprint library maintained either by ATCC or the Minna/Gazdar lab (which is the primary source of the lines). The lines were also tested to be free of mycoplasma by e-Myco kit (Boca Scientific).
Total RNAs were isolated using Trizol (Invitrogen) reagent following manufacturer's instruction. First-strand cDNA was reverse transcribed with 3ug of mRNA using Superscript II first-strand synthesis system (Invitrogen). The final volume was 20ul.
TaqMan Gene Expression Assays for SRC-3/NCoA3 gene and internal control GAPDH gene are purchased from Applied Biosystems. To generate the standard curve, serial dilution of MCF-7 cDNAs was run in triplicate for both SRC-3 and GAPDH. Efficiency for both primer probe is almost as high as 2 [E=10(−1/slope) =10(−1/−3.3) ≈2(data not shown)], appropriate to use ΔΔCt Method for data analysis.
For the SRC-3 expression in 33 tested cell lines, all the cDNAs were run for both SRC-3 and GAPDH in triplicate. Threshold and baseline were setup the same for both relative standard curve generation and comparative Ct generation for the whole panel. In each group of cells (i.e., NSCLC vs. SCLC vs. HBEC), the cell line expresses lowest level of SRC-3 in the group is defined as 1, and the other lines are relative levels comparing to the lowest one, using [Relative Expression Level = 100 × 2^ (−ΔΔCt)] for calculation.
All cells were maintained in RPMI-1640 medium supplemented with 5% fetal bovine serum (FBS, Invitrogen), cells were grown to 80% confluency prior to lysis. Cell lysates were prepared by 1% SDS containing lysis buffer followed by boiling. Cell extracts with equal amount of proteins were analyzed by immunoblotting. The SRC-3/AIB1 antibody (BD Transduction Laboratories) was used at 1:2000 dilutions. Antibodies for cleaved caspase-7 (1:1000) and phospho-SRC-3 (Thr24) (1:1000) were from Cell Signaling Technology and antibody for GAPDH was from Santa Cruz. Western blot band intensity is quantified using ImageJ following the program instruction.
Protein lysates were prepared as previously described (16). RPPA was produced and analyzed as previously described, with slight modifications (17). RPPA data was quantified using a SuperCurve method which detects changes in protein level by MicroVigene software (VigeneTech, Carlisle, MA) and an R package developed in house (18).
Human SRC-3 Smart pool siRNA and nonspecific control siRNA were obtained from Dharmacon Research, Inc. siRNAs were transfected with Dharmafect 2 (Dharmacon) into H1299, Dharmafect 4 into H1819, Dharmafect 3 into H2073 and Dharmafect 1 into A549 at 50nmol/L following the reverse transfection protocol.
48h after siRNA treatment, cells were harvested, counted, and single cells seeded (500–2,000 cells/well) in triplicate into 6-well plates. Cells were grown for 11 days, 12 days or 14 days (for H1299, H1819 and H2073 respectively), colonies stained with methylene blue (0.5% in 70% isopropyl alcohol) or crystal violet for 1 hour at room temperature and counted.
Clinical grade gefitinib and cetuximab were obtained from the UTSW pharmacy. Gefitinib was dissolved in DMSO to prepare 10 mM solution as drug stock, stored at 4°C. NSCLC cell line H1819 was treated with SRC-3 specific siRNA, non-targeting siRNA pool or transfection reagent only, 24 hrs later 100 nM Gefitinib added to the medium, and 72 hrs later harvested for Annexin V staining. Treated H1819 cells were trypsinized and counted, resuspended in 1× binding buffer (BD Pharmingen) (1×106 cells/ml). Annexin V-FITC (5 ul) and Propidium Iodide (5 ul) (BD Pharmingen) added, incubated in the dark at room temperature for 15 minutes, then the staining quenched by adding 400ul of 1× binding buffer to 100 ul of cells, followed by flow cytometry analysis, with percentage of Annexin V positive cells normalized to transfection reagent only treatment.
Using a specific anti-SRC-3 antibody we performed semi-quantitative immunohistochemistry on a tissue microarray of 311 clinically annotated NSCLCs and found a wide range of tumor SRC-3 expression (nuclear staining) (Figure 1, and see Materials and Methods); 85 NSCLCs (27%) expressed SRC-3 (staining score ≥10), 144 (46%) showed no SRC-3 staining, while 84 (27%) exhibited a SRC-3 nuclear staining score between 0 and 10. Kaplan-Meier analysis was performed on the 215 patient subset for which clinical data including survival were available. As shown in Figure 1C, patients whose tumors have high SRC-3 levels by immunohistochemistry have significantly shorter overall (P = 0.0008, log-rank test) and progression-free (P = 0.0015, log-rank test) survival time than patients with low SRC-3 levels. The median overall survival time was 54 months for the high SRC-3 group and 98 months for the low SRC-3 expression group (Figure 1C). After adjusting for the effects of histology, gender, race, tobacco, stage and lymph node status in the multivariate proportional hazard survival model, high SRC-3 was still significantly associated with poor overall (hazard ratio = 2.02, P = 0.0007) and progression free survival (hazard ratio=1.92, P =0.0011) (Table 1). Importantly, the expression of SRC-3 did not correlate with EGFR mutation (p=0.365, t-test) or KRAS mutation (p=0.256, t-test) (Supplementary Table 1).
Analysis of SRC-3 (at 20q12) DNA copy number alterations by array-based comparative genomic hybridization (aCGH) in a panel of lung and breast cancer cell lines revealed SRC-3 copy number gain in 14 of 55 (25%) NSCLC lines, 8 of 23 (35%) SCLC lines and 15 of 31 (48%) breast cancer cell lines, while 4 of 55 (7%) NSCLC lines, 1 of 23 (4%) of SCLC and 1 of 31 (3%) breast cancer cell lines show copy number loss at this locus (Supplementary Table 2). SRC-3 mRNA expression analysis by quantitative RT PCR (Q RTPCR) of 16 NSCLC, 11 human bronchial epithelial cell (HBEC) lines immortalized by expression of CDK4 and telomerase (HBEC-KT) and six primary, unimmortalized human bronchial epithelial cells (HBEC-UI) showed >30% of NSCLCs over-express SRC-3 at a level at least 3–4 fold higher than the average for the HBECs. Overall, there was a greater than 20-fold variation in the mRNA level of SRC-3 expression within NSCLCs and the average SRC-3 expression level of NSCLCs was greater than the immortalized HBECs, which in turn was greater than that of the non immortalized HBECs. The means and 95% confidence intervals of mRNA expression of SRC-3 for NSCLC, HBEC-KT and HBEC-UI are 12.6 [7.3–17.8], 9.0 [6.2, 11.8] and 3.6 [3.0, 4.3], respectively. The overall p-value of comparing three groups is 0.078 based on the analysis of variance test, therefore the differences seen between mRNA expression of SRC-3 in NSCLC, HBEC-KT and HBEC-UI are significant (Figure 2A). NSCLC lines (HCC44, HCC15, H28) with SRC-3 expression levels lower than the HBECs were those with SRC-3 DNA copy number loss. In addition, SRC-3 protein expression was highly variable in HBECs and NSCLC lines as analyzed by western blotting (Figure 2B).
The function of SRC-3 is regulated by protein phosphorylation and among the eight mapped phosphorylation sites in SRC-3, modification of Thr24 is crucial for the interaction with ER, AR and NF-kB to achieve maximum transcriptional co-activation activity (19). Phosphorylation at this site is also important for the transforming ability of SRC-3 (19). Using an anti-phospho-SRC-3 (Thr24) specific antibody, we found the phosphorylation level of SRC-3 to be highly variable amongst the lung cancer cell and HBEC lines (Figure 2B). For example, HBEC13KT and H2073 show no detectable pSRC-3, though they express modest amounts of total protein. On the other hand, H1299 has the highest pSRC-3/SRC-3 ratio, though it does not express the highest level of total SRC-3 protein.
To investigate whether the cells expressing high-levels of SRC-3 rely on it for survival and proliferation, a pool of 4 siRNA oligonucleotides targeting SRC-3 was transfected into H1299, which expresses significant levels of both total and phosphorylated (Thr24) SRC-3. The cell viability and liquid colony formation were found to be decreased by more than 70% when SRC-3 was knocked down (Figure 3). Cleaved-caspase 7 expression was detected 72 hours after siRNA transfection, and was sustained up to 96 hours, indicating that SRC-3 knockdown in H1299 cells is associated with activation of the apoptosis pathway (Figure 3A). The smartpool siRNAs targeting SRC-3 have been deconvoluted into the individual targeting sequences, and the degree of SRC-3 knockdown is correlated with the degree of effect observed (Supplementary Figure 1). Note that the control siRNA oligo itself is somewhat toxic, and rendered a modest effect (~20% reduction) on cell viability, yet showed no effect on the caspase activation. Contrary to the results in H1299 cells, H1819 and H2073 NSCLCs (expressing medium and low SRC-3 levels, respectively) showed only modest or no effect on growth and liquid colony formation after SRC-3 knockdown (Supplementary Figure 2). This result suggested that only NSCLCs expressing high levels of SRC-3 (or pSRC-3) show dependence on continued SRC-3 expression for survival and growth.
Several studies have revealed a relationship between SRC-3 expression level and EGFR family pathway signaling in breast cancer (4, 5). Thus, we wished to know if SRC-3 expression in NSCLC correlates with the response to EGFR-targeted chemotherapy. We examined the expression of 160 proteins potentially involved in signaling and oncogenesis by reverse phase protein arrays (RPPAs) using validated antibodies and correlated protein levels with in vitro sensitivity and resistance to erlotinib, gefitinib and cetuximab in 48 NSCLC lines using a MTS proliferation assay (20). High EGFR protein expression (often associated with the presence of EGFR oncogenic mutations and amplification) was correlated with cetuximab, erlotinib and gefitinib sensitivity (Supplementary Table 3), while high SRC-3 protein expression correlated with the resistance to these drugs (Figure 4A). This result led us to wonder if SRC-3 knockdown would sensitize NSCLCs with wild-type EGFR to EGFR tyrosine kinase inhibitors (TKIs). H1819 is a NSCLC cell line expressing moderate levels of SRC-3 and with wild-type EGFR but demonstrates phosphorylation of EGFR, HER2, Erb-B3 and downstream effectors such as AKT and P44/42 ERK indicating the EGFR pathway is active (Supplementary Figure 3) (20). H1819 cells are resistant to gefitinib (IC50=15 μM), and knocking down SRC-3 via siRNA in these cells only led to a modest growth reduction (Supplementary Figure 2A). While neither of these treatments alone elicited apoptosis (with gefitinib given at 0.1μM, a concentration that is within the range of achievable serum levels in patients), combined siRNA SRC-3 knockdown and 0.1 μM gefitinib led to induction of apoptosis detected by Annexin V staining (Figure 4B). Also, cells treated with the SRC-3-targeted siRNA and gefitinib were less well attached, rounded up, and showed a “spider web”-like appearance, when compared to the cells treated with SRC-3 siRNA or gefitinib alone (not shown).
Studies of SRC-3 in lung cancer have revealed DNA copy number increases and great variation in expression levels with ~25% of NSCLCs dramatically over expressing this gene compared to other NSCLCs, normal lung epithelium, and immortalized human bronchial epithelial cells. Patients whose tumor showed SRC-3 over expression also had inferior survival compared to those with low or no SRC-3 expression. Increased expression was also associated with SRC-3 DNA copy number increase. This frequency of high SRC-3 expression is comparable to that found in breast, ovarian and prostate cancer, all of which can become dependent on SRC-3 for proliferation (3, 6, 7). In addition, NSCLCs varied in their levels of pSRC-3 expression and the variations in SRC-3 DNA copy number, mRNA, protein and phosphoprotein levels suggests that there is regulation of SRC-3 at the transcriptional, translational and post-translational levels. For example, a recent study identified an ubiquitin ligase, CHIP, that can directly target SRC-3 for ubiquitinylation and degradation and these changes inhibit the anchorage-independent cell growth and the metastatic potential of cancer cells (21). This is consistent with our general finding that siRNA-mediated SRC-3 knockdown inhibited cell growth and colony formation in NSCLCs expressing high levels of SRC-3 and pSRC-3, while little or no effects of SRC-3 knockdown was seen in NSCLCs with low endogenous SRC-3 levels. It is true that in the A549 cell line, which expresses the highest level of the protein, knockdown of SRC-3 resulted in only a 20% reduction in cell viability. However, we were unsuccessful in achieving efficient knockdown in this line, the maximum knockdown efficiency is 50% using up to 100nM siRNAs, which still left a significant amount of SRC-3 protein in the cells.
Finally, we discovered a correlation between SRC-3 expression levels and resistance to EGFR-TKIs gefitinib and erlotinib. SRC-3 knockdown led to dramatic sensitization of a NSCLC with wild-type EGFR to EGFR TKI targeted therapy. The mechanism by which SRC-3 knockdown leads to gefitinib sensitization remains to be elucidated. One possibility is that SRC-3 might be crucial for the activation of other receptor tyrosine kinases (such as IGF1R) in lung cancers that can bypass EGFR to activate critical downstream signaling pathways. For example, SRC-3 is known to mediate IGF-1-induced phenotypic changes in human breast cancer cells (22) and SRC-3 deficiency affects breast cancer initiation and progression in mice (23). After this manuscript was submitted for review, a paper was published that showed that overexpression and amplification of SRC-3 was found in 48.3% and 8.2% of NSCLCs, respectively, and that the overexpression of SRC-3 negatively affects survival of surgically resected NSCLC patients(24). This paper provides independent validation of our findings. The difference in over-expression and amplification frequencies is most likely due to the fact that an Asian population was the subject of their experiments while a Caucasian population was the subject of ours.
Taken together, our results suggest that SRC-3 is an important new oncogene and potential therapeutic target for lung cancer.