Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 24 (3), 538-545

RADX Modulates RAD51 Activity to Control Replication Fork Protection


RADX Modulates RAD51 Activity to Control Replication Fork Protection

Kamakoti P Bhat et al. Cell Rep.


RAD51 promotes homologous recombination repair (HR) of double-strand breaks and acts during DNA replication to facilitate fork reversal and protect nascent DNA strands from nuclease digestion. Several additional HR proteins regulate fork protection by promoting RAD51 filament formation. Here, we show that RADX modulates stalled fork protection by antagonizing RAD51. Consequently, silencing RADX restores fork protection in cells deficient for BRCA1, BRCA2, FANCA, FANCD2, or BOD1L. Inactivating RADX prevents both MRE11- and DNA2-dependent fork degradation. Furthermore, RADX overexpression causes fork degradation that is dependent on these nucleases and fork reversal. The amount of RAD51 determines the fate of stalled replication forks, with more RAD51 required for fork protection than fork reversal. Finally, we find that RADX effectively competes with RAD51 for binding to single-stranded DNA, supporting a model in which RADX buffers RAD51 to ensure the right amount of reversal and protection to maintain genome stability.

Keywords: BRCA1; Fanconi anemia; MRE11; RAD51; RADX; fork protection; fork reversal; replication stress.

Conflict of interest statement


The authors declare no competing interests.


Figure 1
Figure 1. RADX silencing rescues the MRE11-dependent fork protection defects caused by loss of RAD51 stability
(A) Graphical depiction of the fork protection assay with representative images. (A, B, C, E, F) U2OS or RADXΔ U2OS, (D) RPE-hTERT, (G) FANCD1/BRCA2-mutant fibroblasts, or (H) fibroblasts expressing the T131P RAD51 mutant were transfected with the indicated siRNAs then labeled sequentially with CldU and IdU before treatment with 3mM HU for 5 hours. The lengths of DNA fibers were measured and mean+/−SEM of the IdU/CldU ratio is depicted. P values were derived from Kruskal-Wallis ANOVA with a Dunn’s post-test. Each experiment was repeated at least twice and a representative result is depicted. (siNT = non-targeting siRNA). (I and J) siRNA transfected parental or RADXΔ U2OS were plated for clonogenic survival assays in the absence (I) or presence (J) of drug. P values were calculated from a two-way ANOVA with Tukey’s post-test. Mean+/−SEM from n=3 is depicted. See also Supplemental Figures S1 and S2.
Figure 2
Figure 2. RADX silencing rescues DNA2 dependent fork degradation and RADX overexpression causes degradation of reversed forks
(A–F) Fork protection assays were completed in U2OS cells or RADX overexpressing (OE RADX) U2OS cells transfected with the indicated siRNAs or treated with Mirin. All cells were treated with HU for 5 hours except for an 8-hour treatment in (B). P values were derived from Kruskal-Wallis ANOVA with a Dunn’s post-test. See also Supplemental Figures S1 and S3.
Figure 3
Figure 3. More RAD51 is required for fork protection than fork reversal
(A) Fork protection assays in U2OS cells transfected with seven different RAD51 siRNAs. (B–D) U2OS or RADX overexpressing (OE RADX) cells were transfected with the “J11” siRNA (B and D) or “J12” siRNA (C) to RAD51 at the indicated amounts prior to performing the fork protection assay. P values were derived from Kruskal-Wallis ANOVA with a Dunn’s post-test. Immunoblots from transfected cells corresponding to the same samples are shown below the graphs. (E) Model illustrating differential RAD51 requirements.
Figure 4
Figure 4. RADX can outcompete RAD51 for binding to ssDNA
(A–C) Electrophoretic mobility shift assays for RADX and RAD51 binding to a dT-50 oligonucleotide. The experiment was performed by incubating 0.5 nM ssDNA with RADX, RAD51, or both proteins in the indicated order and concentrations in the presence of ATP. The Ob-m is a DNA binding mutant of RADX. (D) ssDNA pre-bound with RAD51 in the presence of ATP was incubated with increasing amounts of RADX prior to capture with magnetic beads and analysis of captured and supernatant proteins by immunoblotting. (E) Quantitation of RAD51 and RADX levels in 293T and U2OS cells (see also Supplemental Figure S4). The median is depicted. All experiments are representative of two repeats.

Similar articles

See all similar articles

Cited by 12 articles

See all "Cited by" articles

Publication types

MeSH terms