The precise manner in which physical changes to the breast cancer susceptibility protein (BRCA1) affect its role in DNA repair events remain unclear. Indeed, cancer cells harboring mutations in BRCA1 suffer from genomic instability and increased DNA lesions. Here, we used a combination of molecular imaging and biochemical tools to study the properties of the BRCA1 in human cancer cells. Our results reveal new information for the manner in which full-length BRCA1 engages its binding partner, the BRCA1-associated Ring Domain protein (BARD1) under oxidative stress conditions. We also show how physical differences between wild type and mutated BRCA15382insC impact the cell’s response to oxidative damage. Overall, we demonstrate how clinically relevant changes to BRCA1 affect its structure-function relationship in hereditary breast cancer.
Germline mutations in the breast cancer susceptibility gene (BRCA1) are heavily linked to familial breast and ovarian cancers1,2,3,4. Women who inherit these mutations are ~60% more likely to develop the disease5,6. During normal female development, critical windows of vulnerability correlate with the early onset of breast cancer7. These events coincide with a buildup of DNA lesions in mammary tissue as reactive oxygen species are formed from the metabolic processing of estrogen8. Damaged DNA, if left unrepaired, can perpetuate errors in the genome9. Reduced expression levels of the BRCA1 protein or BRCA1 mutations, compounded with insufficient lesion repair, provide a tipping point toward cancer induction10. At the molecular level, the intricate details of these mechanisms are poorly understood.
In the nucleus, BRCA1 associates with its binding partner the BRCA1-associated Ring Domain protein (BARD1) to help coordinate the repair of DNA modifications. The BRCA1-BARD1 heterodimer performs these operations by interacting with other repair proteins, such as BRCA2, at damaged sites on DNA. In this context, BRCA1 acts as a tumor suppressor to ensure fidelity in the genome11,12. Inherited mutations in BRCA1 can cause functional deficiencies in the protein that affect BRCA1’s role in tumor suppression13,14. As decades of scientific research demonstrate BRCA1’s multifaceted role in DNA repair, information about the physical properties of BRCA1 are just coming to light.
While the molecular architecture of the entire BRCA1 protein has not been determined, structural information is available for the BRCA1 N-terminal RING domain15 and the C-terminal (BRCT) region16, where many clinical mutations reside. Equally important, the manner in which BRCA1 changes in response to cellular stress or inherited mutations remains unclear. Improving our knowledge of BRCA1’s structure-function relationship can provide new insights for therapeutic discovery. Here, we focused on defining differences between wild type and mutated BRCA1 under oxidative stress conditions that contribute to genomic instability. Our results reveal how modifications to BRCA1 influence nuclear events that can weaken DNA repair response.
How does oxidative stress effect wild type BRCA1 in breast cancer cells?
To investigate the manner in which BRCA1 responds to oxidative stress in breast cancer cells, we employed a combination of molecular imaging and biochemical tools. We first used primary ductal carcinoma cells (HCC70 line17; ATCC) known to express wild type BRCA1 and measured their susceptibility to reactive oxygen species (ROS). For these experiments, cells were incubated with culture media containing 1 mM hydrogen peroxide (H2O2) for up to 1 hour. ROS detection was determined by fluorescent measurements in whole cells using the Cellular Reactive Oxygen Species Detection kit (Deep Red Fluorescence; Abcam). Untreated cells showed minimal background signal throughout the experiments. Differences between H2O2-treated and untreated cells were detected within 30 minutes (Fig. 1a). These differences were quantified using fluorescence spectroscopy in duplicate experiments, each having four replicates. Greater ROS levels were detected in H2O2-treated cells compared with untreated cells (Fig. 1b).
We then examined the biochemical nature of BRCA1 and its interactions with BARD1 in HCC70 cells. Cells were collected and lysed, and the cytoplasmic and nuclear material of the lysed cells was separated using the NE-PER kit (Thermo Scientific). Nuclear proteins were enriched by incubating with Nickel-Nitrilotriacetic acid (Ni-NTA)-coated agarose beads for 60 minutes at 4 °C. BRCA1-BARD1 naturally bound to the beads according to western blot analysis. The Ni-NTA elution profile indicated an enrichment of BRCA1 (~220 kDa) and BARD1 (~87 kDa) in fraction E2 along with low levels of K48-linked ubiquitin migrating at ~220 kDa (Fig. 1c; Supplementary Fig. 1). To detect biochemical interactions between these components, we performed co-immunoprecipitation (co-IP) experiments using magnetic beads decorated with IgG antibodies against BRCA1. Western blot detection on the IP material identified interactions between BRCA1 and BARD1 (Fig. 1c; Supplementary Fig. 2). Also, K48-linked ubiquitin moieties co-migrated with BRCA1. The low quantities of K48-linked ubiquitin were a potential indicator of BRCA1 degradation during normal cellular processing events.
Nuclear BRCA1-BARD1 is relatively stable under oxidative conditions
Next, we examined the effects of oxidative damage in the cells’ nucleus. Cells were treated with 1 mM H2O2 and fluorescence microscopy was used to detect antibodies against the oxidated DNA base, 8-Oxo-guanine (8-OxoG) (Santa Cruz Biotechnology). Within 40 minutes of H2O2 treatment, the signal for 8-OxoG (red fluorescence) increased in and around the nucleus (blue fluorescence) (Fig. 2a, top panels). We also examined 8-OxoG formation in cells that experienced mild thermal stress (HCC70stress) prior to H2O2 treatment. As mild thermal treatment can prime cells to better manage stressful conditions18,19, HCC70stress cells provided a comparison for oxidative resistance.
Following a 10-minute incubation with H2O2, 8-OxoG levels increased in all cells. After 20 minutes of treatment, 8-OxoG levels increased in HCC70 cells and this progression continued for up to 40 minutes. Following 60 minutes of treatment, cellular viability limited measurements. In parallel, the HCC70stress cells showed a decrease in 8-OxoG after 20 minutes of treatment and this trend persisted up to 40-minutes (Fig. 2a, bottom panels). Untreated cells showed no signal for 8-OxoG within a 40-minute incubation. Overall, the HCC70 cells were more susceptible to oxidative damage while the pre-stressed HCC70stress cells were more resistant to H2O2 treatment.
For HCC70stress cells and control cells, BRCA1 and BARD1 were primarily found in the nuclear fractions and in similar quantities. Upon treatment with 1 mM H2O2, all cells showed a modest reduction in BRCA1 (~18%) and BARD1 (~27%) levels according to densitometry measurements. Unexpectedly, we detected a notable decrease in BRCA2 (~66%), a known DNA damage response protein (Fig. 2b).
We further evaluated BRCA1’s ability to interact with BARD1 under oxidative stress conditions. Co-IP results showed that BRCA1 associated with BARD1 in high quantities (Fig. 2c; Supplementary Fig. 2). Control experiments using species-specific IgG antibodies or probing for phosphorylated (pSer2) RNA polymerase II (RNAP II), showed little to no protein association (Fig. 2c, bottom row; Supplementary Fig. 2). Experiments performed on control cells showed approximately the same level of interaction as the stressed cells (similar to Fig. 1c). These results indicated that wild type BRCA1 was relatively stable under oxidative conditions in the nucleus, and that its association with BARD1 was strongly maintained.
The BRCA15382insC mutation influences cellular susceptibility to oxidative damage
Next, we examined cells that produced the prevalent BRCA15382insC cancer-related mutation. We hypothesized that mutated BRCA15382insC produced in the HCC1937 breast cancer line20 (ATCC) may affect the physical properties and functional capacity of the protein. We assessed global ROS values in HCC1937 cells using fluorescence imaging experiments performed on whole cells. We detected similarly low starting amounts of ROS in HCC1937 and HCC70 cells expressing wild type BRCA1. After treating cells with 1 mM H2O2 for 40 minutes, ROS levels increased in treated cells compared to control cells evaluated across duplicate experiments, each having four replicates (Fig. 3a). Quantifying these results, we found HCC1937 cells were susceptible to oxidative damage and ROS formation similar to HCC70 cells expressing wild type BRCA1 (Fig. 3b).
Next, we examined the biochemical nature of the mutated BRCA15382insC protein produced in HCC1937 cells. Treated cells were collected and lysed following a 40-minute incubation with H2O2. Control cells were incubated with media lacking H2O2 and processed in the same manner. Nuclear proteins were separated and enriched using the same procedure implemented for wild type BRCA1. Western blot analysis revealed that BRCA15382insC, which is truncated by ~10 kDa, migrated the same as wild type BRCA1 in untreated cells (Fig. 3c). Physical changes or modifications to BRCA15382insC may account for this slower than expected migration. K48-linked ubiquitin moieties co-migrated with BRCA15382insC at ~220 kDa (Fig. 3c; Supplementary Fig. 3a,b).
Western blot analysis of the H2O2-treated cells identified BRCA15382insC migrated at a much slower rate (~270 kDa) than wild type BRCA1 (~220 kDa) (Fig. 3d; Supplementary Fig. 3a,b). This shift in mobility also corresponded with a shift in K48-linked ubiquitin moieties (~270 kDa). In both treated and untreated cells intact BARD1 migrated at ~87 kDa although its detection was more complex in the treated cells (Supplementary Fig. 3c). Taken together, these results suggested that oxidative damage can enhance K48-linked ubiquitin attachments to mutated BRCA15382insC.
How does the BRCA15382insC mutation affect protein stability in the nucleus?
To gain insight on the functional nature of BRCA15382insC during DNA damage response, we treated HCC1937 cells with 1 mM H2O2 for up to 40 minutes and used fluorescence microscopy to detect 8-OxoG accumulation. Treated HCC1937 cells showed the same trend as treated HCC70 cells. We detected 8-OxoG in and around the nucleus within 20 minutes and levels continued to accumulate for up to 40 minutes (Fig. 4a). Untreated cells showed evidence of 8-OxoG at the end of the 40-minute incubation period and some signal was also detected at the start of the experiment (Fig. 4a, white arrows). These results indicated that cells expressing BRCA15382insC were not well-equipped to manage oxidative conditions in the nucleus.
We then investigated how BRCA15382insC and its binding partners were biochemically effected by oxidative effects. We evaluated nuclear levels of BRCA1, BARD1, and BRCA2 in different cell lines using western blot analysis and densitometry measurements performed on three replicate experiments (Fig. 4b). Treated HCC1937 cells exhibited more stable levels of BRCA2 than the other cells. By comparison, BRCA1 levels were reduced in treated HCC1937 cells by ~50% relative to the other cells. BARD1 levels in the HCC1937 cells were lowered by ~80%. Consistent with western blot results in Fig. 3, BRCA15382InsC in H2O2-treated cells also migrated slower than wild type BRCA1. We posited these changes were due to modifications to BRCA15382InsC that occurred under oxidative conditions. Specifically, we found an increase in K48-linked ubiquitination in the nuclear extract of treated HCC1937 cells. The enhanced K48-ubiquitin signal corresponded with a mobility shift in BRCA15382InsC in treated cells (Figs 3d and 4c). An accumulation of K48-linked ubiquitin adducts on BRCA15382insC can cause the mutated protein to be degraded at higher levels. This mechanism would account for the reduced amounts of BRCA1 found in the treated HCC1937 cells. While other forms of ubiquitination or phosphorylation were also feasible they were not detected (Fig. 4c; Supplementary Fig. 4).
The BRCA15382insC structure is modified under oxidative conditions
To understand the physical consequences of oxidative damage to BRCA1, we examined documented sites for post-translational modifications using the online bioinformatics tool, PhosphoSitePlus21 (http://www.phosphosite.org/). Identified modifications to human BRCA1 include phosphorylation, ubiquitination, sumoylation, and acetylation (Fig. 5a). Although there are many lysine residues dispersed throughout the BRCA1 sequence, specific lysines have been documented through mass spectrometry analysis as ubiquitination sites21,22. These residues include K970 and K1667. There is currently no structural information available for the region of BRCA1 that encompasses K970. Residue K1667 is located in the BRCA1 C-terminal region (pdb code, 3K0H)23 known as the BRCT1 domain (Fig. 5b). This location is ~200 residues prior to the end of the BRCA1 sequence and is close to the phosphopeptide binding site.
The mutated BRCA15382insC protein sequence (Fig. 5a) can be translated and modeled up to residue G176324,25. After this glycine residue, mutated BRCA1 shows no secondary structure and is truncated in the BRCT2 domain (Fig. 5b). The truncated protein is also thought to be misfolded to some extent, which can trigger K48-linked ubiquitination and proteasomal degradation26. The proximity of K1667 to the mutation site (Fig. 5b, red star) makes it a prime candidate for ubiquitination in response to errors in protein folding. Similarly, residue K1759 is located in the connecting loop between BRCT1 and BRCT2, and it is more accessible in the mutated BRCA15382insC structure. This gain in accessibility also makes K1759 a strong candidate for ubiquitination in mutated BRCA15382insC (Fig. 5b).
We biochemically tested for other modifications to BRCA15382insC including phosphorylation and sumoylation. No phosphorylation signal was detected on western blot analysis when probing for these modifications, especially S1524 (Fig. 4c). This finding is consistent with the work of Okada and Ouchi27. Sumoylation was also not noted in our analysis, which is not surprising considering the transient nature of these modifications and the fact that H2O2 can prevent these attachments from occurring28.
In summary, we present biochemical and molecular imaging information to explain differences in wild type and mutated BRCA1 under oxidative conditions. We found notable changes in the physical properties of mutated BRCA1 related to ubiquitination patterns and the molecular architecture of the BRCT region. Under oxidative conditions, mutated BRCA15382insC exhibited a slower than expected mobility according to western blot analysis. A shift in BRCA15382insC migration corresponded with changes in K48-linked ubiquitin migration. These observations support the idea that modifications to BRCA15382insC were enhanced in cells experiencing oxidative damage.
To better understand the manner in which modifications to BRCA1 can affect its function, we assessed ROS accumulation in whole cells and in the nucleus. Cells expressing wild type or mutated BRCA1 were susceptible to oxidative stress. However, cells producing BRCA15382insC showed a major decline in proteins levels and inherent amounts of oxidated DNA bases. These results indicate long-term genomic instability and insufficient lesion repair in cells harboring the BRCA15382insC mutation. Our findings are consistent with other reports for BRCA15382insC and may account for its documented deficiencies in protein-protein interactions29,30. Cells expressing wild type BRCA1 showed relatively stable protein levels and lacked inherent oxidated DNA bases. These observations underscore the functional importance of BRCA1-BARD1 in responding to oxidative conditions in the nucleus.
To place our results in the context of other studies, we looked to researchers who have tested the ubiquitin ligase activity of BRCA1. Studies performed on cells expressing different BRCA1 constructs demonstrated the importance of the RING domain in mediating ubiquitin transfer events. Specifically, the I26A mutation located in the RING domain tends to limit BRCA1’s association with E2 binding partners and reduce its ubiquitination capacity29. Other studies demonstrated that the I26A mutant expressed in HCC1937 cells, although deficient in ligase activity, exhibited the same physical properties and mobility as wild type BRCA1 in rescue experiments30. Still, others have shown that BRCA15382insC has natural deficiencies in ubiquitin ligase activity due to inadequacies in the mutated BRCT domain31,32. There is no compelling evidence in the literature to indicate BRCA1 transfers K48-type ubiquitin moieties to protein substrates, rather K6-type ubiquitin moieties are typically transferred33,34,35,36,37. Therefore, the enhanced modifications to BRCA15382insC formed in HCC1937 cells under oxidative conditions are unique and not likely due to ubiquitin intermediates associated with its E3-ligase function.
Another explanation for the changes to BRCA15382insC may be related to auto-ubiquitination. Extensive work by other researchers demonstrates enhanced BRCA1 activity through auto-ubiquitination mechanisms also involving K6-type ubiquitin linkages33,35,36,38. According to our oxidative damage response assays, reduced BRCA1 activity was observed as protein levels declined. In the case of the BRCA15382insC mutation, protein levels were further diminished upon modification. Therefore, our results do not support the idea of auto-ubiquitination to mutated BRCA15382insC. Rather, they point to differences in protein stability levels of BRCA15382insC in response to ROS accumulation. Our working hypothesis is that protein misfolding in BRCA15382insC triggers an increase in K48-linked ubiquitination, possibly within the BRCT domain, and that this effect is enhanced by oxidative stress.
Collectively, we demonstrate that protein levels, genetic mutations, and post-translational modifications – particularly ubiquitination – can affect BRCA1’s functional state in breast cancer cells. Our present work puts forth the idea of testing mechanistic-based therapies such as deubiquitinating enzymes to restore the physical properties of mutated BRCA15382insC in cells. Future structural studies on full-length BRCA1 can complement our findings to better inform us of the physical variations in BRCA1 related to clinically harmful mutations.
Authentication and preparation of cells and enriched nuclear fractions
Breast cancer cells (HCC70 and HCC1937 lines) were were purchased from the American Type Culture Collection (ATCC) and independently characterized by ATCC as being primary ductal carcinoma cells and triple negative in nature (i.e., lacking expression of estrogen receptor, progesterone receptor, and Her2). Cells were used within 6 months of resuscitation. Cells were grown in complete growth medium (RPMI-1640; Mediatech) supplemented with 10% fetal bovine serum (ATCC) and 0.5X penicillin-streptomycin (Thermo Fisher) at 37 °C and 5% CO2 until confluent. Matched HCC70 cells were stressed by the addition of complete media warmed to 24 °C, then returned to 5% CO2 and 37 °C for ~5–7 cycles. Control cells having a similar passage number were cultured with media warmed to 37 °C. For each experiment, we collected ~1,000,000 cells using Trypsin-EDTA (Thermo Fisher) followed by centrifugation (500× g, 5 minutes). The resulting cell pellet was washed with PBS followed by re-pelleting. The NE-PER kit (Thermo Scientific) was used to separate cytoplasmic and nuclear fractions. The nuclear fractions were diluted (~1 mg/ml) in 20 mM HEPES buffer (pH 7.2) containing 2 mM MgCl2, 2 mM CaCl2, 5 mM imidazole and Complete protease inhibitor cocktail (EDTA-free, Roche). BRCA1-BARD1 heterodimers were enriched from the soluble nuclear fraction by incubating with pre-equilibrated Nickel-Nitrilotriacetic acid (Ni-NTA) agarose beads (Qiagen) for 1 hour at 4 °C with rotation. The Ni-NTA beads were washed with 20 mM HEPES buffer (pH 7.2) containing 140 mM NaCl2, 5 mM imidazole and proteins were eluted in the same buffer having 150 mM imidazole. Protein concentrations were estimated using the Pierce Bradford assay (Thermo Scientific).
Fluorescence imaging of 8-OxoG DNA lesions
Cells (~50,000 per chamber) were seeded in an eight-chamber slide and incubated at 37 °C and 5% CO2 overnight. To induce oxidative damage in cells, 1 mM hydrogen peroxide (H2O2) (Sigma) was added to culture media and cells were examined at different time points up to 1-hour post-treatment. Cells were washed with standard PBS solution then fixed with 4% paraformaldehyde (Electron Microscopy Sciences) for 15 minutes. Cells were permeabilized by incubating with PBS solution containing 0.5% Triton X-100 (Sigma) for 10 minutes followed by a 1-hour blocking step with PBS supplemented with 10% normal goat serum (Jackson Immuno Research, #005–000–121) and 0.2% Triton X-100. Control cells were treated with complete culture media lacking H2O2. Cells were incubated with anti-8-oxoG DNA Lesion (483.15) (Santa Cruz, #sc-130914) at 4 °C overnight. Following a wash step with PBS, antibodies were detected with goat anti-mouse IgM-TR (Santa Cruz, #sc-2983). Nuclei were stained with Hoechst 33342 and cells were imaged using an inverted fluorescence microscope (Zeiss Axio Vert.A1; Carl Zeiss Microscopy).
ROS detection and quantitation
To detect and quantify levels of ROS accumulation in whole cells, we used the Cellular Reactive Oxygen Species Detection Assay Kit (Deep Red fluorescence; Abcam, #ab186029). Cells (~20,000 per well) were plated in a 96-well black wall/clear bottomed microplate and incubated at 37 °C and 5% CO2 overnight. Cells were treated with 1 mM hydrogen peroxide (H2O2) in standard PBS solution containing 10% Fetal Bovine Serum (FBS) for 40 minutes. In parallel, control cells were treated with the same PBS solution lacking H2O2 for 40 minutes. Treated and control cells were incubated in ROS Deep Read working solution for 30–60 minutes in the dark at 37 °C and 5% CO2. The deep red fluorescence was read using the SpectraMax i3 microplate reader (at excitation 650 nm, emission 675 nm; Molecular Devices). Cells were imaged using an inverted fluorescence microscope.
Co-IP experiments and immunoblot analysis
For co-immunoprecipitation (IP) experiments, antibodies against BRCA1 (5 μg Santa Cruz Biotechnology (SCBT); sc-642, C-20) or normal rabbit IgG (5 μg Cell Signaling Technology; 2729) were diluted in PBS-T (0.02% Tween-20) before incubating with 0.75 mg Dynabeads Protein G (Thermo Fisher). The mixtures were incubated with rotation for 30 minutes at 4 °C. Antibody-coated beads were washed in 20 mM HEPES buffer (pH 7.2) containing 140 mM NaCl2 prior to adding enriched nuclear fractions. Eluates from the Ni-NTA beads were pooled (200 μg per experiment) and supplemented with protease and phosphatase inhibitor cocktails (Thermo Fisher). This material was incubated with the antibody-coated beads overnight at 4 °C with gentle rotation. The beads were then washed followed by elution with NuPAGE LDS sample buffer (Thermo Fisher).
Western blot analysis was performed to identify proteins separated on 3–8% Tris-Acetate NuPAGE mini gels (Thermo Fisher) and transferred onto Immobilon-P membranes (Millipore) in a Mini-PROTEAN Tetra system (Bio-Rad). Blots were blocked with 1% non-fat dry milk (NFDM) in TBS-T (0.05%) for 1 hour with gentle rocking. Primary antibodies were diluted in blocking solution and incubated overnight at 4 °C. The following primary antibodies were used in our studies: BRCA1 (SCBT; sc-642, C-20), pBRCA1 (S1524-specific; Bethyl, A300–001A), BRCA2 (EMD Millipore; OP95), BARD1 (SCBT; sc-11438), RNA Polymerase II (Pser2-specific; Covance MMS-129, H5), ubiquitin (K48-linkage-specific, Abcam, ab140601), and β-Actin (Sigma Aldrich; A5441). Blots were washed three times with TBS-T (0.05%). Goat anti-rabbit, goat anti-mouse, mouse anti-rabbit (light chain-specific, IP), or goat anti-mouse (light chain-specific, IP) secondary antibodies conjugated to horseradish peroxidase (Jackson ImmunoResearch) were incubated for 1 hour followed by additional washing. ECL Prime western blotting reagent (GE Healthcare) or West Femto (Thermo Fisher) was used for detection. A ChemiDoc MP (Bio-Rad) was used for imaging and densitometry.
Proteomic and homology modeling analysis
To evaluate documented post-translational modifications to the BRCA1 primary sequence, we utilized the PhosphoSitePlus website, which is freely available at http://www.phosphosite.org/. The human isoform of BRCA1 (ACC_ID, P38398) was used in our analysis. Ubiquitination results were reported based on the number of records in which this modification was assigned using proteomic discovery-mode mass spectrometry. The homology model presented here was generated using the online SWISS-MODEL website (http://swissmodel.expasy.org) and was previously reported24,25.
How to cite this article: Gilmore, B. L. et al. Molecular Analysis of BRCA1 in Human Breast Cancer Cells Under Oxidative Stress. Sci. Rep. 7, 43435; doi: 10.1038/srep43435 (2017).
Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
The authors acknowledge Dr. Sarika Srivastava, Dr. Robert Gourdie, and Dr. Elizabeth Alli for useful discussions and assistance with fluorescence microscopy. This work was supported by funds from the Commonwealth Health Research Board , the Concern Foundation , NIH/NCI [1R01CA193578-01A1], and the UVA-VTC Neuroscience Seed Fund Award.