Resolution of sister centromeres requires RanBP2-mediated SUMOylation of topoisomerase IIalpha

Abstract

RanBP2 is a nucleoporin with SUMO E3 ligase activity that functions in both nucleocytoplasmic transport and mitosis. However, the biological relevance of RanBP2 and the in vivo targets of its E3 ligase activity are unknown. Here we show that animals with low amounts of RanBP2 develop severe aneuploidy in the absence of overt transport defects. The main chromosome segregation defect in cells from these mice is anaphase-bridge formation. Topoisomerase IIalpha (Topo IIalpha), which decatenates sister centromeres prior to anaphase onset to prevent bridges, fails to accumulate at inner centromeres when RanBP2 levels are low. We find that RanBP2 sumoylates Topo IIalpha in mitosis and that this modification is required for its proper localization to inner centromeres. Furthermore, mice with low amounts of RanBP2 are highly sensitive to tumor formation. Together, these data identify RanBP2 as a chromosomal instability gene that regulates Topo IIalpha by sumoylation and suppresses tumorigenesis.

Figure 1. Generation of Mice with Graded Reduction in RanBP2 Dosage

(A) Schematic representation of the RanBP2 gene targeting strategy. The relevant portion of the RanBP2 locus, the targeting vector with loxP (triangles), the hypomorphic and knockout alleles, and the EcoRV restriction sites (E) and the probe used for Southern blotting are indicated. (B) (left) Southern blot analysis of RanBP2^+/H and RanBP2^+/+ mice. (Right) PCR-based genotype analysis of RanBP2 mutant mice. (C) Western blot analysis of MEFs isolated from mice carrying the indicated RanBP2 alleles using anti-RanBP2 antibody. Actin was used as a loading control.

Figure 2. Cells with Low Levels of RanBP2 Form Chromosome Bridges in Anaphase

(A) Chromosome dynamics and segregation of RanBP2^+/+ and RanBP2^−/H MEFs expressing H2B-YFP were observed by timelapse microscopy. Representative images of each mitotic stage are shown. Arrowheads mark chromatin bridges in RanBP2^−/H MEFs. Bar = 10 μm. (B) High-magnification images of RanBP2^−/H MEFs with indicated chromosome segregation defects. Bar = 10 μm. (C) Quantification of the chromosome segregation defects observed by live-cell imaging of MEFs of the indicated genotypes. n = total number of mitotic cells analyzed from at least three independent MEF lines. Error bars indicate SEM, *p < 0.05 versus wild-type cells (Chi-square test).

Figure 3. RanBP2 Binds to Topo IIα in Mitosis and Is Essential for Its Accumulation at Inner Centromeres

(A) Western blots of asynchronous MEF lysates probed with antibody to Topo IIα. Actin served as a loading control. (B) Western blots of synchronized RanBP2^+/+ and RanBP2^−/H MEF lysates. MEFs were synchronized in G0 by serum starvation and then released for the indicated durations in serum-containing medium. Nocodazole was added 23 hr after cells were released. Blots were probed for Topo IIα. Actin was used as loading control. (C) Immunolocalization of Topo IIα in MEFs with various levels of RanBP2 during prometaphase. Centromeres were visualized with ACA antibody. DNA was stained with Hoechst. Magnified images of the centromeric and inner centromeric regions are shown in the insets. Bar = 10 μm. (D) Quantification of prometaphases with inner centromeric versus diffuse localization of Topo IIα. Seventy-five prometaphases were analyzed per genotype (three independent MEF lines were analyzed per genotype). Error bars indicate SEM, *p < 0.05 versus wild-type cells (Chi-square test). (E) Immunoblots of mitotic (M), G1, or G2 HeLa extracts subjected to immunoprecipitation with RanBP2 antibody and analyzed for coprecipitation of Topo IIα. (F) Immunoblots of mitotic extracts from RanBP2^+/+ and RanBP2^−/H MEFs subjected to immunoprecipitation with RanBP2 antibody and analyzed for coimmunoprecipitation of Topo IIα. Phosphohistone H3 (P-H3) signals indicate that similar amounts of mitotic cells were present in the lysates used for immunoprecipitation.

Figure 4. Expression of the RanBP2 SUMO E3 Ligase Domain in RanBP2-Insufficient Cells Prevents Anaphase-Bridge Formation and Topo IIα Mislocalization

(A) Overview of the HA-tagged mutant RanBP2 proteins used in correction experiments. (B) Incidence of chromatin bridges in RanBP2^−/H MEFs expressing the proteins indicated in (A) or HA-tagged PIASy. Three independent MEF lines were evaluated per genotype (~25–40 anaphases per line). Error bars indicate SEM, *p < 0.05 versus RanBP2^−/H MEFs carrying empty expression vector (Chi-square test). (C) Immunolocalization of Topo IIα (green) in monastrol-treated RanBP2^−/H MEFs expressing the indicated proteins. Centromeres were visualized with human ACA antibody (red). Bar = 10 μm. (D) Quantification of prometaphases with inner centromeric versus diffuse Topo IIα localization. Seventy-five prometaphases were analyzed per genotype (three independent MEF lines were analyzed per genotype). Error bars indicate SEM. Chart legend is as in (B). *p < 0.05 versus RanBP2^−/H cells carrying empty expression vectors (Chi-square test). (E) Western blot analysis of HA immunoprecipitates from mitotic extracts of MEF cells expressing the indicated HA-RanBP2 fragments. Blots were probed for Topo IIα.

Figure 5. RanBP2 Sumoylates Topo IIα in Mitosis

(A) In vitro sumoylation of Topo IIα by the RanBP2 E3 ligase domain. Twenty nanograms of recombinant human Topo IIα was tested for SUMO1 modification in the presence or absence of 100 ng wild-type (His-HA-RanBP2[2553–2838]) or mutant (His-HA-RanBP2[2553–2838]^{L2651A,L2653A}) RanBP2 SUMO E3 ligase domain, 110 ng E1, 50 ng Ubc9, 5 μg SUMO1, and 1 mM ATP. Samples were incubated at 37°C for 1 hr and then analyzed by western blotting with antibody to Topo IIα. (B) In vivo sumoylation of Topo IIα by RanBP2. Mitotic shake-off lysates from RanBP2^+/+ and RanBP2^−/H MEFs containing the indicated expression vectors were prepared and analyzed by western blotting for Topo IIα and HA. P-H3 and actin antibodies were used to verify that similar amounts of mitotic cells and protein were present in lysates used. (C) RanBP2 sumoylation of Topo IIα under physiological conditions. Mitotic shake-off lysates were prepared from RanBP2^+/+ and RanBP2^−/H MEFs and analyzed by western blotting for Topo IIα. P-H3 and actin antibodies were used to ensure that lysates contained similar amounts of mitotic cells and protein. (D) Localization of SUMO1-Topo IIα-EGFP in RanBP2^−/H MEFs during prometaphase. (Left) Confocal image of a prometaphase expressing SUMO1-Topo IIα-EGFP. Centromeres were visualized by immunostaining with ACA antibody. Bar = 5 μm. (Right) Quantification of cells with inner centromeric versus diffuse localization of SUMO1-Topo IIα-EGFP during prometaphase. Two independent RanBP2^−/H MEF lines were analyzed (50 prometaphases per line). Error bars indicate SEM, *p < 0.05 versus untransfected RanBP2^−/H MEFs (paired t test). Localization of Topo IIα in untransfected RanBP2^−/H MEFs was determined by immunostaining for Topo IIα and centromeres as in Figure 3C. (E) Incidence of chromatin bridges in RanBP2^−/H MEFs in the absence or presence of SUMO1-Topo IIα-EGFP. Two independent RanBP2^−/H MEF lines were analyzed (50 prometaphases per line). Error bars indicate SEM, *p < 0.05 versus RanBP2^−/H untransfected MEFs (paired t test). (F) Immunolocalization of Topo IIα in PIASy^−/− MEFs during prometaphase. (Left) Confocal image of a PIASy^−/− MEF immunostained for Topo IIα and centromeres. Bar = 5 μm. (Right) Quantification of PIASy^+/+ and PIASy^−/− prometaphases with inner centromeric versus diffuse localization of Topo IIα. Three independent MEF lines were analyzed (~25–50 prometaphases per line). Error bars indicate SEM. (G) Incidence of chromatin bridges in PIASy^−/− MEFs examined by live-cell imaging. Three independent PIASy^+/+ and PIASy^−/− MEF lines were analyzed (~25–40 anaphases per line). Error bars indicate SEM. (H) Sumoylation of endogenous Topo IIα in PIASy^−/− MEFs. Experimental details were as in (C). (I) Proposed model for the role of RanBP2 in sister-chromosome separation. In mitosis, RanBP2 (in conjunction with Ubc9) binds to and sumoylates Topo IIα in the mitotic cytosol. This modification serves as a signal for Topo IIα targeting to inner centromeres, where it functions to decatenate sister centromeres, thus allowing for proper separation of sister chromosomes in anaphase. As SUMO-modified Topo IIα represents only a small fraction of the total Topo IIα pool (Figure 5C), SUMO conjugation may only be required for initial targeting of Topo IIα to centromeres but not for maintenance of centromeric localization.

Figure 6. RanBP2 Suppresses Spontaneous and DMBA-Induced Tumorigenesis

(A) Gross appearance (top) and histology (bottom) of skin tumor of a DMBA-treated RanBP2^−/H mouse. (B) Skin tumor incidence of DMBA-treated mice of the indicated genotypes. *p < 0.05 versus wild-type mice (Log rank test). (C) Skin tumor burden of DMBA-treated mice of the indicated genotypes. Error bars indicate SEM, *p < 0.05 versus wild-type mice (Mann Whitney test). (D) Gross appearance (top) and histology (bottom) of lung tumors of a DMBA-treated RanBP2^−/H mouse. (E) Lung tumor incidence of DMBA-treated mice of the indicated genotypes. Error bars indicate SEM, *p < 0.05 versus wild-type mice (Chi-square test). (F) Lung tumor burden of DMBA-treated mice of the indicated genotypes. Error bars indicate SEM, *p < 0.05 versus wild-type mice (Mann Whitney test). (G) Tumor-free survival of RanBP2^+/+, RanBP2^H/H, and RanBP2^−/H mice. *p < 0.01 versus wild-type mice (Log rank test). (H) Tumor spectrum of mice with various RanBP2 genotypes. (I) Gross images and histological analysis of representative spontaneous tumors from RanBP2^−/H mice. (J) Chromosome counts on primary tumors of RanBP2^H/H and RanBP2^−/H mice.