Functional characterization of a cancer causing mutation in human replication protein A

Abstract

Replication protein A (RPA) is the primary ssDNA-binding protein in eukaryotes. RPA is essential for DNA replication, repair, and recombination. Mutation of a conserved leucine residue to proline in the high-affinity DNA binding site of RPA (residue L221 in human RPA) has been shown to have defects in DNA repair and a high rate of chromosomal rearrangements in yeast. The homologous mutation in mice was found to be lethal when homozygous and to cause high rates of cancer when heterozygous. To understand the molecular defect causing these phenotypes, we created the homologous mutation in the human RPA1 gene (L221P) and analyzed its properties in cells and in vitro. RPA1(L221P) does not support cell cycle progression when it is the only form of RPA1 in HeLa cells. This phenotype is caused by defects in DNA replication and repair. No phenotype is observed when cells contain both wild-type and L221P forms of RPA1, indicating that L221P is not dominant. Recombinant L221P polypeptide forms a stable complex with the other subunits of RPA, indicating that the mutation does not destabilize the protein; however, the resulting complex has dramatically reduced ssDNA binding activity and cannot support SV40 DNA replication in vitro. These findings indicate that in mammals, the L221P mutation causes a defect in ssDNA binding and a nonfunctional protein complex. This suggests that haploinsufficiency of RPA causes an increase in the levels of DNA damage and in the incidence of cancer.

Figure 1. Structures and modeling of wild-type and L221P

(A) Structures of DBD-A from wild-type human RPA1 (blue) bound to DNA (purple; 1JMC (28)) and yeast RPA1 (yellow; 1YNX (39)). Position of L221 side chain is shown extending toward DNA from left side of binding cleft. (B) β1 and β2 strands of the DNA binding site shown for (left) structures of wild-type human and yeast and (right) modeled structures of human and yeast L221P. Shown from outside looking toward DNA centered on position 221. β1 and β2 labeled and discontinuity indicated (*) in human structure. Models determined using Modeller (40) from structures shown in (A)). DSSP was used for secondary structure assignment (41); structures rendered in PyMOL.

Figure 2. Effect of L221P on cell cycle progression

(A) Cells transfected with RPA1 siRNA and GFP-tagged wild-type (WT) or L221P RPA1 vector where indicated as described in Methods. At 96 hours, DNA content was analyzed with flow cytometry. Top row shows GFP expression (with % of GFP-positive cells indicated in upper left corner); bottom row shows DNA content as a histogram. (B) and (C) show analysis of the combined data from the experiment described in (A) and three similar, independent experiments. (B) Average transfection efficiency (GFP-positive cells) from cells treated with RPA1 siRNA and transfected with GFP-tagged wild-type (WT) or L221P RPA1 with standard deviation between experiments shown (error bars). (C) Average GFP fluorescence intensity was determined for mock treated cells and all GFP-positive cells (e.g. upper box in (A)) for RPA1 siRNA treated (mock), GFP-wild-type (WT) or GFP-L221P transfected cells shown with standard deviation between experiments indicated (error bars). Because gain settings varied between experiments, values from each experiment were normalized to wild-type RPA1 (WT). (D) Cells were treated and analyzed as described in (A), at 96 hours after mock or siRNA-transfection, cells were synchronized with 5µg/ml aphidicolin for 24 hours, then released media. Flow cytometry was used to analyze DNA content at 0, 8, and 24 hours after release. For both (A) and (D) DNA content shown for GFP-negative cells for Mock samples and GFP-positive cells for WT and L221P.

Figure 3. Cellular response to DNA damage

(A) Activation of ChK2 in the absence of induced DNA damage. At 96 hours after mock or RPA1 siRNA-transfection (si1), cells were stained for phosphorylated ChK2 and analyzed via flow cytometry. (B) Localization of wild-type and L221P mutant RPA complexes to sites of DNA damage. Cells were grown on coverslips and treated with 20 µM camptothecin for 4 hours. Non-chromatin bound RPA was extracted prior to fixing slide. Row one-DAPI staining; Row two-GFP-tagged RPA1; Row three-RPA2 antibody staining; Row four-merge of rows two and three. (C) Ability to recognize sites of DNA damage. Cells treated as in (B). Row one-DAPI staining; Row two-GFP-tagged RPA1; Row three-stained with phosphorylated H2AX antibody; Row four-merge of rows two and three. (D) Coexpression of L221P mutant and endogenous wild-type RPA1. 48 hours after wells were seeded, the cells were mock-transfected or transfected with GFP-tagged wild-type or L221P mutant RPA1 vector. At 96 hours, DNA content of cells was stained with propidium iodide and analyzed via flow cytometry. GFP-negative cells were selected for Mock samples; GFP-positive cells were selected for WT and L221P samples.

Figure 4. Biochemical properties of RPA-L221P

(A) Gel analysis of purified L221P RPA complex. 1µg of wild type RPA or RPA containing the L221P RPA1 subunit were separated on an 8–14% SDS-PAGE gel and stained with silver nitrate. Lane 1-wild type RPA; lane 2-L221P complex. Non-RPA impurities indicated by *. (B) Ability to support SV40 DNA replication in vitro. Background synthesis was assessed in the absence of RPA (-RPA) or T antigen (-Tag). Synthesis of complete reactions containing 400ng either wild-type or L221P. A representative assay is shown; all reactions done in duplicate and normalized to background. (C) Single-stranded DNA binding. Binding isotherms of a representative gel mobility shift assay for wild-type and L221P complexes.