A role for nucleosome assembly protein 1 in the nuclear transport of histones H2A and H2B

Abstract

Import of core histones into the nucleus is a prerequisite for their deposition onto DNA and the assembly of chromatin. Here we demonstrate that nucleosome assembly protein 1 (Nap1p), a protein previously implicated in the deposition of histones H2A and H2B, is also involved in the transport of these two histones. We demonstrate that Nap1p can bind directly to Kap114p, the primary karyopherin/importin responsible for the nuclear import of H2A and H2B. Nap1p also serves as a bridge between Kap114p and the histone nuclear localization sequence (NLS). Nap1p acts cooperatively to increase the affinity of Kap114p for these NLSs. Nuclear accumulation of histone NLS-green fluorescent protein (GFP) reporters was decreased in deltanap1 cells. Furthermore, we demonstrate that Nap1p promotes the association of the H2A and H2B NLSs specifically with the karyopherin Kap114p. Localization studies demonstrate that Nap1p is a nucleocytoplasmic shuttling protein, and genetic experiments suggest that its shuttling is important for maintaining chromatin structure in vivo. We propose a model in which Nap1p links the nuclear transport of H2A and H2B to chromatin assembly.

Fig. 1. Kap114p binds to Nap1p both in vivo and in vitro. (A) Kap114-PrA and associated proteins were isolated from cytosol. Kap114-PrA was visualized by Coomassie Blue staining, while Nap1p was visualized by western blotting (WB) and probing with an anti-Nap1p antibody. W, final wash fraction. (B) GST (1.9 µM), GST–Nap1p (675 nM) or GST–H2A_1–46 (1.6 µM) were immobilized on glutathione–Sepharose, and binding of MBP–Kap114p, MBP–Kap121p or MBP–Kap123p (60 nM of each as input) was tested. Bound fractions were separated by SDS–PAGE, and the bottom portion of the gel was stained with Coomassie Blue (CBB). Bound MBP–Kaps were visualized by western blotting and probing with an anti-MBP antibody. (C) Different portions of Nap1p fused to GST (1 µM) were tested for binding to MBP–Kap114p as in (B).

Fig. 2. Kap114p contains a distinct, RanGTP-insensitive binding site for Nap1p. (A) GST, GST–Nap1p, GST–H2A_1–46 or GST–H2B_1–52 (0.5 µM of each) were immobilized on glutathione–Sepharose. Binding of Kap114p (120 nM) was then tested in the presence or absence of free Nap1p (4 µM) or an H2A_1–46 NLS peptide (15 µM) as competitor. Bound fractions were analyzed as described. (B) MBP–Kap114p (60 nM) was pre-incubated with or without human Ran Q69L (at 10, 20 or 50 µM), and binding was tested with immobilized GST (0.75 µM), GST–H2A_1–46 (0.75 µM) or GST–Nap1p (0.65 µM). Bound fractions were analyzed as described.

Fig. 3. Nap1p and Kap114p bind directly to the N-termini of H2A and H2B. (A) Immobilized GST (4 µM), GST–H2A (1 µM) or GST–H2B (4 µM) were incubated with recombinant Nap1p (500 nM). The purified GST–histones, especially GST–H2B, contained degradation products that retained the GST tag. Bound fractions were analyzed by SDS–PAGE and Coomassie staining. (B) Analysis of Nap1p binding to GST–histone N-terminal tails (1 µM of each) as in (A). (C) MBP–Kap114p (200 nM) binding to GST–histones as in (A). The asterisk indicates a bacterial protein, probably the product of the dnaK gene, which co-purified with GST–H2A. (D) MBP–Kap114p (200 nM) binding to GST–histone N-terminal tails as in (B).

Fig. 4. A co-complex consisting of Kap114p, Nap1p and histones. (A) MBP–Kap114p was pre-incubated with or without human Ran Q69L (20 µM), and binding was then tested with immobilized GST or GST–H2A_1–46 (0.75 µM) in the presence or absence of Nap1p (1.5 µM). Unless indicated by WB (western blot), samples were analyzed by Coomassie staining. (B) Analysis of MBP–Kap114p binding as in (A) except that GST–H2A (full-length; 1 µM) was substituted for GST–H2A_1–46, Ran Q69L was used at 10 µM and Nap1p was used at 1 µM. (C) Yeast cytosol was prepared from a Kap114-PrA/H2B-PrA strain and analyzed by FPLC using a Superdex 200 column. The positions of protein standards are indicated above. (D) Recombinant MBP–Kap114p (0.75 µM), Nap1p (0.75 µM) and H2A–H2B dimer (3.6 µM) were pre-incubated at 4°C for complex formation and analyzed by FPLC as in (C). (E) GST–H2A_1–46 (50 nM) or GST–H2B_1–52 (100 nM) were immobilized on glutathione–Sepharose, and binding to MBP–Kap114p (100 nM) was tested in the presence or absence of recombinant Nap1p (20, 100 and 500 nM, or 1.25 µM).

Fig. 5. In vivo analysis of histone NLS–GFP reporters. (A) Wild-type or Δnap1 cells containing the indicated reporter were analyzed by fluorescence microscopy. (B) Mean nuclear:cytoplasmic fluorescence intensity ratios ± SD were determined from digital images using the indicated combinations of strains and NLS–GFP reporters. Numbers in parentheses indicate the mean nuclear:cytoplasmic ratio relative to wild-type cells expressed as a percentage.

Fig. 6. Nap1p is a karyopherin specificity determinant in vitro. Binding of recombinant MBP–Kap114p (A), MBP–Kap121p (B) or MBP–Kap123p (C) (120 nM each) was tested on immobilized GST, GST–H2A_1–46 and GST–H2B_1–52 (1.6 µM) in the presence or absence of recombinant Nap1p (0.8 µM). The bound fractions were analyzed as described.

Fig. 7. Nap1p is a karyopherin specificity determinant in vivo. H2B-PrA/Δkap114 and H2B-PrA/Δkap114/Δnap1 strains were created. (A and B) H2B-PrA and associated proteins were isolated from each strain using equivalent amounts of cytosol (standardized for total protein), separated by SDS–PAGE and stained with Coomassie Blue or western blotted with an anti-Kap121p antibody. W, final wash fraction. Fractions in (A) and (B) were run on the same gel and blot and thus represent the same staining and western blot. (C) Western blot analysis of whole-cell extracts from these strains probed for Kap121p. Recombinant MBP–Kap121p (1.5, 3.0 or 6.0 nmol) was used as a reference.

Fig. 8. Distinct domains of Nap1p mediate cytoplasmic or nuclear localization of Nap1p. (A and B) Nap1p (Nap1_1–417) or various deletions of Nap1p (as indicated by amino acid numbers) were expressed as fusions to GFP₂ in wild-type yeast. The coincident Hoechst staining is shown, and the positions of nuclei are arrowed.

Fig. 9. Nucleocytoplasmic shuttling of Nap1p. (A) Two putative NES mutants (nap1-1 and nap1-2) were created in Nap1p and expressed in wild-type yeast as GFP₂ fusions. The underlined residues were changed to alanine in these two mutants. A schematic of yeast Nap1p is shown on the right. Kap114p BD = Kap114p-binding domain. (B) nap1-1–GFP₂ was expressed in the indicated strains. (C) Nap1–GFP₂ was expressed in crm1-3 grown at 24°C or shifted to 37°C for 1 h. (D) The NES of Nap1p (residues 87–114) fused to GFP₂ was analyzed in wild-type yeast. (E) The Nap1p NES or PKI NES was fused to an exogenous NLS and GFP₂, and analyzed in an LMB^s strain with or without LMB treatment (200 nM) for 30 min. The coincident Hoechst staining is shown, and positions of nuclei are arrowed.

Fig. 10. Nucleocytoplasmic shuttling of Nap1p is necessary for the proper transcription of the Ty1 element. (A) β-galactosidase activity, as a measurement of Ty1–LacZ transcription, was measured in a wild-type (WT) and an isogenic Δnap1 strain. (B) β-galactosidase assays were performed in a Δnap1 strain after expression of wild-type NAP1, NAP1 residues 200–417 and the NAP1 NES mutant, nap1-1 or vector alone. (C) Transcription assays were performed using a non-Ty1 (CYC1-lacZ) reporter system in a wild-type (WT) and an isogenic Δnap1 strain.

Fig. 11. Model for the function of Nap1p in the nuclear transport of histones H2A and H2B.