Loss of the functional NLSs in hKid reduces its interaction with mitotic chromosomes in living cells. (A) Representative images of Venus–wt hKid and Venus–ΔNLSs hKid in stably transformed live HeLa cells. Cells in metaphase (left) and anaphase (right) are shown. To compare the localization signals in the cells, the fluorescent intensities were digitally compensated. The actual fluorescent intensity of Venus–ΔNLSs hKid was much higher than that of Venus–wt hKid. Bar, 10 μm. (B) FRAP analysis of Venus–wt hKid or Venus–ΔNLSs hKid on chromosomes. The mean intensity of a bleached chromosomal spot (a), a nonbleached chromosomal spot (b, control), and a nonchromosomal spot (c, background) were measured over time (micrographs). All FRAP measurements were plotted as a function of time (graphs). The mean values and their standard deviations from six to nine cells are plotted. The diagram below depicts our experimental design. All possibilities for the replacement and turnover of hKid between each compartment are indicated by arrows. Bar, 10 μm. (C) FLIP analysis of Venus–wt hKid or Venus–ΔNLSs hKid in the presence (−nocodazole) or absence (+nocodazole) of a mitotic spindle. A spot outside the chromosomes and mitotic spindle (a) was irradiated, and the mean intensities of a chromosome spot (b) and a spot outside the cells (c, background) were measured before and during irradiation (micrographs). All FLIP measurements were plotted as a function of time (graphs). The mean values and their standard deviations from 37–42 cells are plotted. The diagrams below depict our experimental design. Arrows indicate putative replacements and the turnover of hKid between each compartment. Each t_1/2 value was calculated from the best-fit curve as described in Materials and methods. Bar, 10 μm.

Importin-β–mediated chromosome loading of hKid is promoted by the production of Ran-GTP. (A) Digitonin-permeabilized mitotic HeLa cells were incubated with 0.4 μM FLAG-hKid in the presence or absence of 0.8 μM importin-α, 0.8 μM importin-β, and 4 μM Ran-GDP with or without an energy source. (B) Digitonin-permeabilized mitotic HeLa cells were incubated with 0.4 μM FLAG-hKid in the presence or absence of 0.8 μM importin-α, 0.8 μM importin-β, and 4 μM RanGDP or 4 μM T24N Ran with an energy source. After incubation, the cells were fixed and subjected to indirect immunofluorescence staining and the images were processed as described in Fig 3. Bars, 10 μm. Quantification of the fluorescent intensities of the images taken under the same conditions showed that the chromosomal accumulation of FLAG-hKid was reduced to 57% when Ran-GDP was replaced with T24N Ran.

Ran-GTP production induces local cargo release at chromosomes and supports the chromosomal accumulation of hKid. Digitonin-permeabilized mitotic HeLa cells were incubated with 0.4 μM FLAG-hKid or 0.4 μM Cy3-BSA-NLS in the presence of 0.4 μM CFP–importin-α and 0.4 μM YFP–importin-β, with or without 4 μM Ran-GDP, an energy source, and 1.7 μM CAS. After incubation, the cells were fixed and subjected to indirect immunofluorescence staining. The images were processed as described in Fig 3. Bar, 10 μm.

Model of the importin-β–mediated chromosome loading of hKid. The association of hKid with importin-β via importin-α triggers the initial targeting of hKid to mitotic chromosomes, whereas cargo release, mediated by the action of Ran-GTP, which is generated locally at the chromosomes, promotes the accumulation of hKid on the chromosomes. Cargo release is likely coupled with the deposition of hKid onto the chromosomes.

hKid possesses two functional NLSs that bind importin-β via importin-α. (A) 30 pmol of recombinant GST–importin-β, 60 pmol FLAG-hKid, and 30 pmol importin-α were incubated in 50 μl of solution and then pulled down with glutathione-Sepharose beads. The bound proteins were analyzed by SDS-PAGE followed by Coomassie brilliant blue staining. hKid bound importin-β in the presence of importin-α. (B) 20 pmol of recombinant GST–importin-α or its ED mutant (defective in basic NLS binding) was incubated with 40 pmol FLAG-hKid, 40 pmol importin-β, or both in 50 μl of solution and then pulled down with glutathione-Sepharose beads. Coomassie brilliant blue staining of the bound proteins analyzed by SDS-PAGE is shown. Importin-α bound hKid directly, whereas the mutant version of importin-α did not bind hKid. Importin-β binding was unaffected by the mutations introduced in importin-α. (C, top) Nuclear migration of 0.4 μM of recombinant FLAG-hKid was examined in digitonin-permeabilized HeLa cells in the presence of 0.4 μM importin-α, 0.4 μM importin-β, 4 μM Ran-GDP, and an energy source. (C, bottom) Localization of EGFP-hKid in HeLa cells 3 h after nuclear injection of the expression plasmid. Mutations in both NLSs abolished the nuclear localization of hKid. Bars, 10 μm. (D) Schematic diagram of hKid. The positions of the two NLSs where mutations were introduced are indicated by arrows.

Importin-α/β inhibits spindle binding but facilitates chromosome loading of hKid in digitonin-permeabilized cells. (A) Digitonin-permeabilized mitotic HeLa cells were incubated with 0.4 μM FLAG-hKid or FLAG–ΔNLSs hKid in the absence or presence of 0.8 μM each importin-α and -β, with or without 4 μM Q69L Ran-GTP. Note that the recombinant proteins were present in excess compared with the mitotic chromosomes and mitotic spindles contained in the permeabilized cells, indicating that the reduction observed in the spindle signal of hKid upon addition of importin-α/β was caused by the inhibition of spindle binding. (B) Digitonin-permeabilized mitotic HeLa cells were incubated with 0.4 μM FLAG-hKid in the presence of either 0.8 μM importin-α or -β. All reactions were performed in the presence of apyrase, which depleted the energy source. The same results were obtained without the addition of the energy source to the reaction mixture (not depicted). After incubation, the cells were fixed and processed for indirect immunofluorescence staining to detect FLAG-hKids and the spindle microtubules, as described in Materials and methods. DNA was counterstained with DAPI. The images are projections of five deconvolved sections taken at 0.2-μm focus intervals. Bars, 10 μm.

The effect of importin-α/β and Ran-GTP production on hKid chromosomal loading is independent of the mitotic spindle. HeLa cells treated with 2 μM nocodazole for 2 h were permeabilized with digitonin and then incubated with 0.4 μM FLAG-hKid or 0.4 μM FLAG–ΔNLSs hKid in the presence or absence of 0.8 μM importin-α, 0.8 μM importin-β, and 4 μM Ran-GDP with or without an energy source. After incubation, the cells were fixed and subjected to indirect immunofluorescence staining. The images are projections of 46 deconvolved sections taken at 0.2-μm focus intervals. Bars, 10 μm.

Expression of mutant Ran defective in the GTPase cycle affects the mitotic localization of endogenous hKid in living cells. (A) Synchronized HeLa cells were transfected with expression vectors encoding wt Ran, Q69L Ran, or T24N Ran as described in Materials and methods. 10 h after their release from a second thymidine block, the cells were fixed and stained for endogenous hKid and Ran. The images are projections of five deconvolved sections taken at 0.2-μm focus intervals. Bars, 10 μm. (B) Quantification of the results in A. Cells expressing the indicated Ran proteins were fixed and stained. The fluorescent intensity of hKid on the chromosomes (1) and centrosomes (2) was measured using SoftWorx software and their ratios were plotted. The values are expressed as the mean ± standard deviation. Significant differences were detected using t test. *, P < 0.05. Bar, 10 μm.