Abstract
Background/Aim: Numerous missense mutations have been determined in the BRCT domain of the BRCA1 gene, affecting localization and interaction of BRCA1 with other proteins. Materials and Methods: We examined whether the M1775K and V1809F mutations in the BRCT domain affect BRCA1 cellular localization. Cells were transfected with pEGFP-C3-BRCA1 and detected by fluorescence microscopy. Results: Following induction of DNA damage, cytoplasmic mislocalization was observed for both M1775K and V1809F mutants compared to EGFP-BRCA1wt and the less common variant M1652I. These results indicate that M1775K and V1809F mutations may change the function of the protein by affecting BRCA1 localization. Conclusion: There is a correlation between subcellular localization of BRCA1 and diminished DNA repair observed in breast cancer cells, which may be explained by structural variations and altered binding properties of phosphopeptides.
- BRCA1
- BRCT
- mutants
- cellular localization
BRCA1 is a nuclear-cytoplasmic protein consisting of 1,863 residues, with significant roles in several cell processes (1-3). BRCA1 protein has two different domains with important roles in protein interactions (PPI), which have been shown to be involved in human cancers (4, 5). The BRCT domain of the BRCA1 protein folds, forming a parallel four-stranded beta-sheet at the central part of the BRCT domain. Mutations occur most frequently in the BRCT domain, affecting BRCA1 nuclear functions such as DNA repair (6) and transcriptional activity (7). We have previously analyzed mutations in the BRCT domain and found that they (i) alter the folding of the domain in vitro and (ii) interrupt the binding affinity of the BRCT region to synthetic phosphopeptides such as pBACH1/BRIP1 and pCtIP (8, 9). Nuclear localization is regulated by a nuclear localization signal [NLS (10)] and a nuclear export signal [NES (11)]. In most cases, exogenously expressed BRCA1 proteins are found to be nuclear due to NLSs (12).
BRCA1 mutations in the BRCT domain result in the mislocalization of BRCA1 to the cytoplasm. Two of the C-terminal mutations (M1775R and Y1853X) have been shown to inhibit nuclear localization and disrupt BRCA1 C-terminal folding (8, 9, 13). Mutations in these residues affect nuclear export and reduce nuclear import (12). Chen et al. (14) revealed that BRCA1 was mislocalized almost exclusively to the cytoplasm in breast cancer samples.
In our laboratory, more than twenty point mutations in the BRCA1-BRCT domain have already been studied for their impact on the structure and function of the protein (8, 9, 15). Among these mutations, M1775K and V1809F were selected for further study at the cellular level and specifically for their impact in the protein’s subcellular localization.
The mutant variant M1775K is a rare mutation that has been identified in only two families of European ancestry (8). In vitro experiments have shown that the M1775K missense variant affects the interaction with the synthetic peptides BACH1 and CtIP (16, 17). Similarly to M1775R, the mutation of the conserved Val1809 to Phe (V1809F) (8, 13, 17) may destabilize the folding of the BRCT domain. On the other hand, the variant M1652I at the first tandem repeat of the BRCT domain, has been recognized as neutral regarding breast cancer pathogenesis (17). Structural analysis has indicated that this variant has no significant effect on the structure of the BRCT domain. Even if the M1652I mutation is characterized as low risk (18), we included it in our study for comparison with the destabilizing mutants V1809F and M1775K.
To study how the selected mutants, V1809F and M1775K, affect the cellular topology of BRCA1 protein, we expressed BRCA1-EGFP fusion proteins carrying the specific mutations at the BRCT domain. The fused peptides were transfected into MCF-7 cells and the subcellular localization was assessed by fluorescence microscopy. Destabilizing mutations in the BRCT domain of BRCA1 may modify BRCA1 localization and affect the pathways involved in cell cycle control and DNA repair.
Materials and Methods
In silico methods. The 1Y98, 1T29, 1JNX, and 1T15 structures of the studied mutants were analyzed by using the PYMOL software (DeLano Scientific LLC, Palo Alto, CA, USA), and intermolecular and intramolecular changes were analyzed using the WHAT IF biotool.
Generating protein-peptide complexes. We collected available information on the molecular structures of native protein-peptide complexes of BRCA1 and their corresponding mutants investigated in this study. The structures of wtBRCA1-BRCT (PDB ID: 1N5O), M1775K-BRCA1-BRCT (PDB ID: 2ING), V1809F-BRCA1-BRCT (PDB ID: 1T2U) on the unbound state were retrieved. These structures were all similar (Cα RMSD <1.0 Å). Additionally, wtBRCA1-BRCT binds both BACH1 and CtIP peptides in a near-rigid manner, as highlighted by the low Cα RMSD when its bound and unbound state are compared [Cα RMSD <1.0 Å when either the BACH1 (PDB ID: 1T29) or the CtlP peptide (PDB ID: 1Y98) is bound]. Consequently, we assumed that since binding occurs in a near-rigid manner, modeling of the interfaces between mutant structures and peptides is amenable by simple structural superposition. All-atom structural alignment of the free states of M1775K-BRCA1-BRCT (PDB ID: 2ING), V1809F-BRCA1-BRCT (PDB ID: 1T2U) was performed against the corresponding bound complexes. After superposition, coordinates of the mutant BRCA1-BRCT structures were stored along with the bound peptides. For the M1652I-BRCA1-BRCT mutant for which no crystal structure is available, in silico mutagenesis was performed and modeling of the bound complexes was followed as described previously (15). Coordinate files generated are available upon request.
Energy calculations of BRCA1-BRCT-peptide complexes and corresponding mutants. Energy calculations of BRCA1-BRCT-peptide complexes were performed as previously described (19, 20), by using the HADDOCK server (21, 22). Electrostatic energies (Eelec) were calculated by using a shift function. Switching function (between 6.5 and 8.5 Å) calculates van der Waals interactions (Evdw). Non-bonded connections are determined with the OPLS force field (21, 22) (cut-off=8.5 Å). The salvation energy was calculated by empirical interatomic solvation parameters. We demonstrated 50 models for each case, starting with distinct random velocities. By using the HADDOCK program, the average score of 4 main models is estimated.
Docking calculations of BRCA1-BRCT-peptide complexes and corresponding mutants. Additional docking calculations were performed to calculate the distribution of binding modes on the target interface and their underlying energies. Active residues, as defined in HADDOCK, were assigned after calculating residues with an inter-molecular distance of 5 Å in all 8 bound assemblies. These residues served as seeds to target interaction interfaces. Docking predictions were performed using the HADDOCK webserver as described by De Vries et al. (22). We did not cluster at 5 Å during docking as recommended by the HADDOCK group, because we deal with interfaces that bury similar accessible surface areas (in Å2) as ordinary protein-protein interfaces (23) and not significantly smaller. Therefore, clustering at 7.5 Å was preferred.
Cell culture and transfection conditions. MCF-7 cancer cells obtained from Prof. H.-W. Stürzbecher, Lübeck University, were grown in Dulbecco’s media (DMEM) with 10% fetal calf serum (FCS) (Sigma-Aldrich, St Louis, MO, USA) and cultured at 37°C in 5% CO2 atmosphere.
The transfection was performed using the eukaryotic expression vector pEGFP-C3 with BRCA1-BRCT cDNAs inserted (mutants and wt) and fused to EGFP. Transfection was performed at about 75% cell confluence with 2 μg of plasmid DNA per well of a 6-well plate, using Lipofectamine Reagent (Life Technologie (Sigma-Aldrich). Transfected cells were grown for 18 h and then protein expression was examined by immunofluorescence.
Mutagenesis and plasmid construction. The EGFP-BRCA1 construct included the full length BRCA1 cDNA (24, 25) in-frame with the EGFP-containing vector pEGFP-C3 (Clontech Laboratories, Inc. Takara Bio, Shiga, Japan), which was kindly provided by Dr. Ody Sibon (24). PCR-based mutagenesis was used to create a series of point mutations in the BRCA1 gene.
Mutagenesis was performed using QuickChange site-directed mutagenesis kit (Stratagene, Foster City, CA, USA), according to the manufacturer’s instructions. Mutations were confirmed by DNA sequencing.
Immunofluorescence microscopy. Cells expressing EGFP-tagged proteins were fixed in 4% Paraformaldehyde in PBS for 20 min at room temperature, washed three times with PBS, and nuclei were stained with Hoechst 33285 (Sigma-Aldrich). Finally, localization of BRCA1 proteins was followed by EGFP fluorescence. The endogenous RAD51 and p53 proteins (Sigma-Aldrich) were detected by immunofluorescence using monoclonal antibodies (25, 26). A secondary goat anti-mouse antibody conjugated with fluorescein isothiocyanate (Sigma-Aldrich) was used for the detection of the primary antibodies used.
DNA damage conditions. After transfection with the fused EGFP-BRCA1 (BRCA1wt and BRCA1 mutants) plasmids, the MCF-7 cells were cultured for 18 h, irradiated, cultured for an additional 4 h, and fixed as described. Cell cultures were exposed to UV irradiation for 15 min at 20 J/m2 energy (26, 27) and then, their viability was estimated by staining with trypan blue, which stains only dead cells. EGFP-BRCA1 expression was detected by fluorescence microscopy (Figure 1).
Results
Structural analysis of the BRCA1-BRCT domain and its role in the subcellular localization of BRCA1. The BRCA1-BRCT mutations M1775K and V1809F influence the structure of the BRCT domain and the interaction with BACH1 and CtIP. Furthermore, localization of the mutant proteins to the nucleus was reduced.
The CtIP peptide binds with BRCA1 in a CDK-dependent manner and forms the complex MRN-CtIP-BRCA1, which mediates Double Strand Brake (DSB) repair (28, 29). CtIP binds to the BRCT domain of BRCA1 after CtIP’s phosphorylation by CDKs (30, 31). Furthermore, BRCA1 mutations disrupt targeting of the protein into the nucleus.
The mutations affected the ability of the protein to interact with BARD1 and BACH1. These interacting BRCA1 proteins are crucial for the integration of BRCA1 to the sites of DNA damage and for DNA repair (32, 33).
The main functionally sensitive regions of BRCA1, such as the RING and BRCT domains, seem to be crucial for the subcellular localization of the BRCA1 protein and mutations in these regions have been shown to block nuclear localization (12, 34).
Subcellular localization of fusion EGFP-BRCA1-BRCT mutants. In order to assess the effect of the point mutations on the localization of proteins, the selected BRCT point mutants were fused to GFP (GFP-BRCA1-BRCT-x, x refers to the mutation used) and transfected into MCF-7 cells. The expressed proteins were studied by fluorescence microscopy and the transfection rate was approximately 35%.
The wild type fusion protein EGFP-BRCA1wt and the fusion protein EGFP-BRCA1-M1652I (characterized as a variation) were detected both in the nucleus and the cytoplasm (Figure 1), suggesting that the change in the structure caused by replacement of M1652 to Ile does not seem to affect the nuclear transport of BRCA1. However, the localization of the mutated proteins was different. Both EGFP-BRCA1 mutant proteins (V1809F and M1775K) were localized in the cytoplasm (Figure 2).
Following UV irradiation, the wild type and the M1652I were localized mostly in the nucleus (85%), whereas before UV exposure, they were mostly in the cytoplasm (80%). Irradiation did not change the cytoplasmic expression of the EGFP-BRCA1 mutant proteins (V1809F and M1775K).
Co-localization of EGFP-BRCA1-BRCT mutants with endogenous p53 protein. To comparatively analyze the localization of GFP-BRCA1-BRCT mutants and that of endogenous proteins, p53 localization was studied by immunofluorescence using specific antibodies against p53. p53 is a DNA repair and cell cycle control protein with nuclear and cytoplasmic localization, depending on the cell cycle stage and the DNA repair status (35). The endogenous p53 protein was localized in the nucleus and the variants V1809F and M1775K in both radiated and non-radiated MCF-7 cells were localized in the cytoplasm (Figure 3).
Discussion
BRCA1 translocates to the nucleus in order to regulate the DNA damage response and homologous recombination. Specifically, BRCA1 becomes phosphorylated and translocate into the nucleus during the beginning of the S-phase and then, becomes ubiquitinated and degraded by the proteasome in the G1 phase (36) (Table I).
During the G1 phase, BRCA1 localizes in the cytoplasm until the S-phase. In the cytoplasm, BRCA1 is detected at the centrosomes (37). BRCA1 may form complex with Bcl-2 at the endoplasmic reticulum (38, 39) and these endoplasmic reticulum cytoplasmic interactions may affect the regulation of apoptosis. The BRCA1 domains targeted by gene mutations are at the amino and carboxy terminal ends and mutations in the carboxy-terminal BRCT domain influence nuclear import of BRCA1 (Figure 4). These data agree with the results of Elstrodt et al. (40) in different breast cancer cell types.
The studied mutations are located at crucial regions of the BRCT domain. The M1775K has been shown not to affect the binding of the synthetic phosphopeptides pBACH1/BRIP1 or pCtIP (13, 17). According to kinetic analysis of the interaction between the BRCA1 C-terminal domains and phosphorylated peptides, two distinct binding conformations exist (41, 42). Moreover, energy calculations using HADDOCK (24, 27) demonstrated the importance of side-chain conformation of Lys1775 for peptide binding. Mutation of this residue to Met has been previously performed and was shown to significantly decrease the affinity of the complex (13, 17). The positive HADDOCK score, affected by van der Waals energies, clearly demonstrates that clashes in the interface that are introduced by mutation will subsequently result in a different binding mode of the protein to the peptide (Table II).
Based on these results, we further explored a potential effect of these mutants on the localization of the protein by expressing wilt type and mutant BRCA1 fused to EGFP in MCF-7 breast cancer cells.
As expected, both M1775K and V1809F mutants, regardless of the cell cycle phase (dividing cells), were sequestered only in the cytoplasm in all cells examined, whereas wtBRCA1 and the M411652I mutant were detected both in the cytoplasm and the nucleus.
Immunohistochemical microscopy studies have revealed that in human breast tumors and breast cancer cell lines expressing BRCA1 mutants, there is a shift in the localization of BRCA1 to the cytoplasm (43, 44). According to Solyom et al., abraxas interacts with the BRCA1 BRCT repeats and contributes to the response of BRCA1 to DNA damage (45, 46).
Interaction of the BRCA1 BRCT domain with BACH1 (46, 47) and the regulator of transcription CtIP is crucial during G2/M phase checkpoint (32). BRCT domains of the yeast DNA repair protein Rad9 have been shown to interact with phosphopeptides, suggesting that the BRCT domains are conserved (48-51). Moreover, BRCA1 deficient cells exhibit loss of homologous recombination repair of DSBs (52-54).
The residue 1775 of the BRCT domain seems to be very important for the functional integrity of both the domain and the whole molecule (9, 17). According to our results, the M1775K missense variant leads to the expression of the fusion protein GFP-BRCA1-M1775K mostly to the cytoplasm, presumably by blocking the nuclear entrance of BRCA1.
The comparison of the localization of the BRCA1 mutant proteins with that of endogenous proteins implicated in DNA repair, such as p53, was used to assess the impact of the mutations in protein function. UV radiation resulted in the nuclear translocation of p53 proteins, while the BRCA1-M1775K mutant remained cytoplasmic. Additionally, the variant V1809F even if it is distant from the Phe13 and the related binding pocket, influences intra-molecular hydrophobic interactions and interrupts the protein’s interactions (8, 54). This may explain the loss of interactions with pBACH1/BRIP1 and pCtIP and the mislocalization of BRCA1–BRCT mutants. The mutations M1775K and V1809F change the interaction site as well as the stability of the molecule. Furthermore, the cytoplasmic localization of the fusion GFP-BRCA1-V1809F protein may support the destabilizing hypothesis.
To provide further evidence in support of the hypothesis of direct impact of structural changes on the nuclear translocation of BRCA1, we studied the effect of M1652I mutation on the protein’s function. This variant has been classified as neutral by previous studies (18, 55). The GFP-BRCA1-M1652I protein showed a cytoplasmic-nuclear profile in normal conditions with an increased nuclear detection after UV radiation, similar to the GFP-BRCA1-wt. Since the M1652I mutation is at the surface of the protein, it has no effect on its integrity and stability. Thus, the binding interaction should be examined in future studies.
Conclusion
The M1775K and V1809F mutations resulted in cytoplasmic mislocalization of the mutant proteins and inhibition of their translocation to the nucleus, suggesting that the integrity of the BRCA1-BRCT domain is very important for the function and structure of the protein.
These mutations may regulate subcellular compartmentalization and function of BRCA1-related pathways and procedures. Furthermore, our data suggest that analysis of subcellular localization of other, not studied so far, BRCA1 mutants present in breast cancer patients, may provide valuable information regarding the mutant’s mechanism of function and may facilitate the design of specifically targeted drugs towards personalized treatment schemes.
Acknowledgements
The Authors would like to thank Dr. Ody Sibon, Department of Radiation and Stress Cell Biology, University of Groningen, Netherlands for kindly providing us the pEGFP-C3-BRCA1 vector and Prof. Dr. H.-W. Stürzbecher for provision of MCF-7 cell line.
Footnotes
Authors’ Contributions
Drikos Ioannis, Boutou Effrosyni, Kastritis Panagiotis, Vorgias Constantinos carried out and participated in the experimental procedures and manuscript demonstrations. Drikos Ioannis, Boutou Effrosyni, Vorgias Constantinos participated in the design of the study and helped drafting the manuscript. All Authors read and approved the final manuscript.
Conflicts of Interest
The Authors declare that they have no competing interests in relation to this study.
- Received January 10, 2021.
- Revision received April 24, 2021.
- Accepted April 26, 2021.
- Copyright © 2021 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.