FAAP20: a novel ubiquitin-binding FA nuclear core-complex protein required for functional integrity of the FA-BRCA DNA repair pathway

Abstract

Fanconi anemia (FA) nuclear core complex is a multiprotein complex required for the functional integrity of the FA-BRCA pathway regulating DNA repair. This pathway is inactivated in FA, a devastating genetic disease, which leads to hematologic defects and cancer in patients. Here we report the isolation and characterization of a novel 20-kDa FANCA-associated protein (FAAP20). We show that FAAP20 is an integral component of the FA nuclear core complex. We identify a region on FANCA that physically interacts with FAAP20, and show that FANCA regulates stability of this protein. FAAP20 contains a conserved ubiquitin-binding zinc-finger domain (UBZ), and binds K-63-linked ubiquitin chains in vitro. The FAAP20-UBZ domain is not required for interaction with FANCA, but is required for DNA-damage-induced chromatin loading of FANCA and the functional integrity of the FA pathway. These findings reveal critical roles for FAAP20 in the FA-BRCA pathway of DNA damage repair and genome maintenance.

Figure 1

FAAP20 is a novel component of the FA nuclear core complex. (A) Silver-stained gel showing polypeptide bands isolated by 2-step purification, first purification using anti-FLAG (α-FLAG) and second purification using talon beads (Co⁺⁺), from nuclear extracts of HeLaS3:vector or HeLaS3:_His6FANCA/_FLAG-FAAP100 cells. Polypeptides identified by MS analysis are indicated, including the novel 20 kDa polypeptide, FAAP20 (asterisk). (B) Silver-stained gel showing the polypeptide bands purified from nuclear extracts of HeLaS3:vector or HeLaS3:_His6-FLAGFAAP20 using 2-step purification as described in panel A. Polypeptides identified by MS analysis are indicated. (C) Immunoblot of immunoprecipitated HeLa cell extract showing core-complex proteins present in IPs with anti-FAAP20 (lane 3), but absent in IPs with pre-immune serum (lane 4). FT is the flow-thru fraction. (D) Immunoblot showing FAAP20 coprecipitated in FLAG-IPs from HeLaS3:_His6-FLAGFANCG, HeLaS3: _His6-FLAGFANCL, and HeLaS3:_His6-FLAGFAAP100 and not in HeLaS3:vector. (E) Immunoblot showing identical gel filtration profiles on cofractionation of FANCA, FANCG, and FAAP20 using a superose 6 gel filtration column.

Figure 2

FANCA is required for stability of FAAP20. (A) Immunoblot showing levels of FANCA, FAAP20, and FANCG in various FA-patient-derived cells defective in one of the complementation groups (indicated above each lane); actin serves as loading control. In the absence of FANCA, FAAP20 amounts were negligible. (B) Immunoblot showing levels of FANCA, FAAP20, and FANCG from HSC93 WT cells and various FA-A patient-derived cells obtained from IFAR. FAAP20 amounts were negligible in the absence of FANCA. (C) Immunoblot showing that FAAP20 and FANCG but not FANCM are stabilized in FANCA-overexpressing HSC72 cells. A nonspecific band is indicated by an asterisk. (D) Immunoblot showing levels of FANCA and FAAP20 in IPs of HSC72 and HSC72:FANCA_FLAG cell extracts. Negligible amount of FAAP20 was immunoprecipitated from HSC72 cells in the absence of FANCA. A nonspecific band is indicated by an asterisk. (E) Immunoblot showing levels of FANCA and FAAP20 in cytosol and nuclear extracts of HSC93:shControl and HSC93:shFANCA cells; actin and ATR serve as controls. FANCA and FAAP20 are predominantly in the nuclear fraction, and FANCA knockdown results in reduced levels of FAAP20. (F) Immunoblot showing levels of FANCA and FAAP20 in total lysates of HeLa:vector and HeLa:shFANCA cells. Knockdown of FANCA results in decreased FAAP20. (G) Immunoblot of HeLa cell lysates showing that FAAP20 and FANCG, but not FANCL or FANCM, are induced by overexpression of FANCA. (H) Immunoblot showing levels of FANCA and FAAP20 in total lysates of HeLa:shControl and HeLa:shFANCA cells cultured in the presence (+) or absence (−) of Mg132 protease inhibitor. Inhibition of the proteasome pathway results in increased FAAP20 levels, despite low levels of FANCA. (I) Immunoblot showing core-complex proteins in input and IP samples from HSC72 cells stably expressing _His6-FLAGFANCA or _His6-FLAGFAAP20. FAAP20 induced on FANCA expression was able to coprecipitate with FANCA and FAAP20 failed to interact with other core-complex proteins in the absence of FANCA.

Figure 3

FAAP20 interact with FANCA in vivo and in vitro. (A) Schematic of FANCA deletion constructs expressed in HSC72 cells. Nuclear localization signal (NLS), leucine-zipper (LZ) motif and C-terminal FLAG tag are shown. (B) Immunoblot of FLAG IP showing that full-length FANCA and the N-terminal deletion interact with FAAP20, but the C-terminal deletion shows weak or no interaction. (C) Immunoblot showing FANCD2 mono-ubiquitination in MMC treated (MMC) or untreated (UT) HSC72 cells expressing various FANCA deletion fragments. Although full-length FANCA corrected the FANCD2 monoubiquitination defect in HSC72 cells, all 3 deletion constructs failed to do so. (D) Schematic of FANCA deletion constructs expressed in E coli cells. MBP is fused to N-terminus. (E) Schematic of FAAP20 full-length construct expressed in E coli. (F) Top panel: Coomassie-stained gel showing FAAP20 and FANCA purified from E coli cell lysate expressing _His6-FLAGFAAP20 or _MBPFANCA fragment or both using M2 agarose (FAAP20 target) or amylose (FANCA target) beads. _MBPFANCA_1065-1101 construct with the leucine-zipper motif failed to interact with FAAP20. _MBPFANCA_1095-1455 and _MBPFANCA_1095-1200 lacking the motif were able to interact with FAAP20. Bottom panel: immunoblot analysis using anti-MBP or anti-FAAP20 confirmed the identity of the bands as FANCA or FAAP20, respectively.

Figure 4

FAAP20 binds ubiquitin in vitro. (A) Immunoblot showing mono-ubiquitin (Ub), K-48 and K-63–linked ubiquitin chains. Interaction of M2-agarose bound _His6-FLAGFAAP20 was tested with mono-ubiquitin, K-48 and K-63–linked ubiquitin chains. FAAP20 binds K-63 linked ubiquitin chains (lane 8), but not mono-ubiquitin (lane 6) or K-48 linked (lane 7). Mutation in the UBZ domain abolished binding (lane 9). RAD18 served as a positive control (lanes 3-5). (B) Immunoblot showing K-63- linked ubiquitin chains. Interaction of M2-agarose bound _His6-FLAGFAAP20 WT (Wt) and various point mutants was tested with K-63–linked ubiquitin chains. FAAP20 constructs with single residue mutations in the UBZ domain, D164A mutation (lane 3) abolishes binding of FAAP20, whereas D162A (lane 2) and E173A (lane 4) bind K-63–linked chains. (C) Immunoblot showing no apparent difference between WT FAAP20 and UBZ mutant (C170A or D164A) binding to core-complex proteins. (D) Interaction of M2-agarose bound _His6-FLAGFAAP20 alone (−) or in complex with _MBPFANCA_1065-1455 (+) was tested with mono-ubiquitin, K-48 and K-63–linked ubiquitin chains. No apparent difference was found in binding of FAAP20 with K-63–linked chains in the absence or presence of FANCA.

Figure 5

FAAP20 is required for activation of the FA pathway. (A) Immunoblot showing levels of FANCD2 mono-ubiquitination and FAAP20 in HeLa cells stably expressing shControl or shFAAP20#1 siRNAs. Knockdown of FAAP20 expression reduced levels of mono-ubiquitinated FANCD2 in cells treated with MMC or HU compared with UT cells. (B) Immunofluorescence analysis of FANCD2 nuclear foci. HeLa cells stably expressing shControl showed an induction of FANCD2 foci on MMC and HU treatment; knockdown of FAAP20 by shFAAP20#1 resulted in decreased foci formation. Expression of WT _His6-FLAGFAAP20_shRES/Wt rescues capacity to form foci but mutant _His6-FLAGFAAP20_shRES/D164A does not. The percentage of cells with 5 or more foci was determined by examining at least 150 cells. Data are presented as the average of 3 independent experiments with standard deviations. (C) Bar diagram showing chromosome aberrations analysis data. Human HSC93 lymphoblast cells stably expressing control shRNA (shControl), shFAAP20 or shFAAP20, and _His6-FLAGFAAP20_shRES together were analyzed for diepoxybutane-induced chromosomal aberrations like breaks, gaps, and radials. Compared with shControl cells, shFAAP20 cells showed higher number of aberrations per cell and this phenotype was rescued by expressing wildtype FAAP20 resistant to shRNA. Fifty metaphase spreads were prepared and scored for chromosomal aberrations as described in supplemental Methods. (D) Cell-cycle analysis of PI and RNase-stained HeLa:shControl and HeLa:shFAAP20#1 cells that were left UT or treated with MMC for 24 hours. HeLa:shFAAP20#1 MMC treated cells showed an increased number of cells in G₂ phase compared with HeLa:shControl MMC treated cells. (E) MMC survival curve showing that reduced FAAP20 expression results in increased sensitivity to MMC; control levels of MMC sensitivity are restored by expressing FAAP20_shRES/wt, but not FAAP20_shRES/C170A or FAAP20_shRES/D164A mutants. Data represent percent survival, compared with untreated, MMC-naive cells. Each experiment was performed in triplicate, and mean values are shown with standard deviations, derived by comparing each dose to no MMC (0 value on the x-axis). (F) Immunoblot showing that FANCA is reduced when FAAP20 is knocked down using either of 2 shRNAs. (G) Immunoblot showing FANCA is reduced when FAAP20 is knockdown and the reduced levels can be rescued by expressing either WT (WT) or 1 of 2 FAAP20 mutants (D164A or C170A). (H) Immunoblot showing chromatin association of FANCA, FANCG, and FAAP20. HeLa:shControl cells treated with MMC or HU exhibited increased association of both FANCA and FANCG with chromatin, compared with UT cells (lanes 1-3). In contrast, HeLa cells depleted of FAAP20 showed reduced FANCA-chromatin association (lanes 4-6). Induction of chromatin association of FANCA in FAAP20 depleted cells can be rescued by ectopic expression of WT FAAP20 (lanes 7-9), but not by the FAAP20_D164A mutant (lanes 10-12) suggesting FAAP20-ubiquitin binding activity is required for chromatin association of FANCA and FANCG on DNA damage.