Subunits of the Heterotrimeric Transcription Factor NF-Y Are Imported into the Nucleus by Distinct Pathways Involving Importin β and Importin 13

Cell culture.HeLa cells were obtained from the German Collection of Microorganisms and Cell Cultures (DSMZ no. ACC57). The cells were grown in modified Eagle's medium (MEM; Biochrom) containing 10% (vol/vol) fetal calf serum (FCS; Biochrom) and maintained in a humidified incubator with 5% CO₂ atmosphere at 37°C.

Expression constructs.The coding regions of the respective proteins were first amplified from plasmid DNA (the cDNA of the NF-Y subunits was a kind gift from Roberto Mantovani, Dipartimento di Biologia Animale, Università di Modena e Reggio Emilia, Modena, Italy) using specific primer pairs with appropriate restriction sites.

The various coding regions for the glutathione S-transferase (GST) fusion proteins were cloned as follows: murine wild-type NF-YA and mutant NF-YA (developed by site-directed mutagenesis and characterized in Fig. 4A) were inserted as EcoRI-XhoI fragments, and human NF-YB was inserted as a BamHI-SmaI fragment into the respective sites of pGEX4T-1 (Amersham; N-GST-tagged); human NF-YC was inserted as a SpeI-SacI fragment into the SpeI-SacI sites of pET41a (Novagen; N-GST tagged).

The expression constructs for the zz/6z fusion proteins were generated as follows: murine NF-YA was inserted as an EcoRV-SpeI fragment and human NF-YB and human NF-YC were inserted as SmaI-NheI fragments from the corresponding pKS vector of the transfection system (described in reference 66) into the EcoRV-SpeI sites of p6z70. p6z70 is a modified pQE70 (QIAGEN) vector in which 6 z-tags (immunoglobulin G-binding domain from Staphylococcus aureus protein A) were inserted at the N terminus in addition to the C-terminal His tag. Human NF-YB was inserted as a SmaI-NheI fragment from the corresponding pKS vector of the transfection system (described in reference 66) into the EcoRV-SpeI sites of p2z70 (modified pQE70 in which 2 z-tags were inserted at the N terminus in addition to the C-terminal His tag).

The expression constructs for 6myc fusion proteins were generated as follows: murine NF-YA, human NF-YB, and human NF-YC were inserted as EcoRI-StuI fragments into the EcoRI-StuI sites of pCS2+MT (65, 75) containing 6 myc tags at the N terminus.

The expression constructs for β-galactosidase (β-gal) fusion proteins were generated as follows: the corresponding coding regions or gene fragments of the different NF-Y subunits were cloned via PCR into the modified pSVβ (Clontech) mammalian expression vector pKS10 (4, 66) using the unique restriction sites BglII and NheI. Deletion constructs were generated by combined insertion of two PCR-amplified gene fragments (PCR products) cloned into the BglII-NheI and NheI-NruI restrictions sites.

Hemagglutinin (HA)-NF-Y fusion protein expression constructs were generated as follows: murine NF-YA was inserted as an EcoRI-XhoI fragment and human NF-YB and human NF-YC were inserted as BamHI-XhoI fragments into the respective sites of pcHA (modified pcDNA3 [Invitrogen] in which the N-terminal HA tag was inserted as a KpnI-BamHI fragment).

Enhanced green fluorescent protein (EGFP)-NF-Y fusion protein expression constructs were generated as follows: murine NF-YA, human NF-YB, and human NF-YC were inserted as XhoI-ApaI fragments into the XhoI-ApaI sites of pcEGFP (modified pcDNA3 in which the N-terminal EGFP was inserted as an EcoRV-XhoI fragment).

The expression construct for the NF-YA-His fusion protein was generated as follows: murine NF-YA was inserted as an NdeI-SacI fragment into the NdeI-SacI sites of pJK45 (C-His-tagged construct, pJK45 is modified pET21b in which the PstI-NruI fragment of pACYC177 was inserted into the PstI-PshAI-digested pET21b).

The expression constructs for the GST-EGFP-Pax6 fusion protein was generated as follows: Xenopus Pax6 was inserted as an EcoRI-XhoI fragment into the EcoRI-XhoI sites of pGEX-EGFP (modified pGEX4T-1 in which EGFP was inserted as a BamHI-EcoRI fragment).

All constructs were verified by DNA sequencing (Andreas Nolte and Marco Winkler, Department of Developmental Biochemistry, University of Göttingen).

Site-directed mutagenesis.Nucleotide exchanges in pJK103, pJK104, pJK105, pJK106, pJK130, and pJK131 were done by site-directed mutagenesis as described by Ho et al. (31). The following primers were used to generate the various mutations: 5′-CTGTATGTGAATGCCGCACAGTATCACCGCATC-3′ (sense) and 5′-GATGCGGTGATACTGTGCGGCATTCACATACAG-3′ (antisense) for pJK103, 5′-AAACAGTATCACCGCATCCTTAATCTGGGACAAGCACGGGCTAAGCTAGA G-3′ (sense) and 5′-CTCTAGCTTAGCCCGTGCTTGTCCCAGATTAAGGATGCGGTGATACTGTTT-3′ (antisense) for pJK104, 5′-GAAGGGAAAATCCCAAAGGAACTAGGGCAATACCTCCATGAGTCTCGGCATCGG-3′ (sense) and 5′-CCGATGCCGAGACTCATGGAGGTATTGCCCTAGTTCCTTTGGGATTTTCCCTTC-3′ (antisense) for pJK105, 5′-CGGCATCGGCACGCCATGGCACTGGCGGGTGGGGAAGGGGGCCGCTTCTTC-3′ (sense) and 5′-GAAGAA GCGGCCCCCTTCCCCACCCGCCAGTGCCATGGCGTGCCGATGCCG-3′ (antisense) for pJK106. To generate pJK130 the same primers were used as for pJK104 and pJK105. In the case of pJK131 the same primers were used as for pJK104 and pJK106.

Transfection and immunofluorescence.Transfection experiments and immunofluorescence were done as previously described (4, 66). Briefly, 1 × 10⁵ HeLa cells/ml MEM were seeded on 10-mm glass coverslips 24 h prior to transfection. Chemical transfection was done with 400 ng DNA of each construct using the Effectene Transfection Reagent (QIAGEN) according to the manufacturer's instructions. Either after 24 h or 48 h of incubation at 37°C, the cells were fixed with 3% paraformaldehyde in phosphate-buffered saline (PBS) for 15 min, permeabilized with 0.5% Triton X-100 in PBS for 10 min, and blocked with 3% bovine serum albumin (BSA) in PBS. The nuclei were visualized with 4′6′-diamidino-2-phenylindole (DAPI) (Hoechst), and the subcellular localization of the gene products, endogenous NF-Y subunits, and nonfluorescent transport substrates was detected by indirect immunofluorescence. For that purpose, monoclonal mouse anti-NF-YA (BD Pharmingen), polyclonal rabbit anti-CBF-A (Santa Cruz), polyclonal rabbit anti-CBF-C (Santa Cruz), monoclonal mouse anti-β-galactosidase (Sigma), monoclonal mouse anti-HA (Santa Cruz), monoclonal mouse anti-Myc (Invitrogen), monoclonal mouse anti-GST (Santa Cruz) antibodies (primary), and Cy3-labeled goat anti-mouse (Dianova), AlexaFluor488-labeled goat anti-mouse (Molecular Probes), AlexaFluor555-labeled goat anti-mouse (Molecular Probes), and AlexaFluor488-labeled goat anti-rabbit (Molecular Probes) antibodies (secondary) were used.

Recombinant protein expression and purification.The NF-Y proteins were expressed in BL21(DE3) strains (Novagen) as follows: cultures were grown at 37°C to an optical density at 600 nm of ∼1.0 and then were shifted to 25°C. After the temperature was equilibrated, the cultures were induced with isopropyl-β-d-thiogalactopyranoside (IPTG) (0.2 mM for single NF-YA, NF-YB, and NF-YC cultures, 0.4 mM for the NF-YB/NF-YC coexpression culture) and were grown for 3 h with shaking at 230 rpm. Cells were then collected by low-speed centrifugation, resuspended in buffer A (50 mM Tris-HCl, pH 8.0, 400 mM NaCl, and 5 mM β-mercaptoethanol), and lysed by sonication. The recombinant proteins were purified on nickel nitrilotriacetic (NTA) agarose (QIAGEN) or glutathione-Sepharose 4B (Amersham) and GST-Bind resin (Novagen) according to the manufacturer's instructions.

To preassemble the trimeric NF-Y holocomplex, recombinant NF-YA protein was immobilized on nickel NTA-agarose (QIAGEN) and used as bait to fish out the NF-YB/NF-YC dimeric complex from the soluble fraction recovered by high-speed centrifugation of the coexpression lysate. After 3 h of incubation, the supernatant was removed and the resin was washed extensively with buffer A. The formed NF-Y trimer was eluted with imidazole and was applied onto a gel filtration column Hiload 16/60 Superdex 200 (Amersham) equilibrated with buffer A. The purified complex was concentrated on Vivaspin (Viva Science) to a final concentration of ∼2 mg/ml.

The following import factors were expressed in Escherichia coli JM109 as described in the literature indicated and were purified on nickel NTA-agarose, followed by chromatography on Superdex 200: Xenopus importin α (28), human importin β (40), transportin (33), Xenopus importin 7, importin 5 (34), and importin 13 (50). Expression and purification (including affinity and ion exchange chromatography) of the following proteins was performed as described previously: NTF2 (40, 61), M9-GST (57), Ran and RanQ69L(GTP) (61), and 4z-rpL23a (34). GST-nucleoplasmin was expressed in E. coli BL21(DE3) for 3 h at 25°C and purified on glutathione-Sepharose 4B (Amersham).

GST pulldown assays.Bacteria expressing the recombinant GST fusion proteins were lysed in buffer A (50 mM Tris-HCl, pH 8.0, 400 mM NaCl, 5 mM β-mercaptoethanol) and centrifuged, and subsequently the supernatant was incubated with glutathione-Sepharose 4B beads or GST-Bind resin, respectively. Bound proteins (∼0.5 mg/ml of resin) were washed three times in binding buffer B (50 mM Tris-HCl, pH 7.5, 300 mM NaCl, 5 mM MgCl, 5 mM β-mercaptoethanol) and used as affinity matrix for the binding experiments. For each binding experiment appropriate amounts of affinity matrix (0.2 μM final concentration of bound protein) were incubated with either precleared bacterial lysates containing expressed import receptors (∼200 μl) or the corresponding purified recombinant import factor in buffer B (0.2 μM final concentration) at 4°C for 3 h in the presence or absence of 2 μM RanQ69L(GTP). Supernatants were removed and the beads were washed four times with ice-cold buffer B. Proteins were eluted with sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) sample buffer, boiled for 10 min, and analyzed by SDS-PAGE.

Preparation of fluorescence-labeled import substrates.Cargo substrates that did not contain GST or EGFP were fluorescence labeled as previously described (5). Briefly, purified 6z fusion proteins expressed in E. coli were dialyzed against 20 mM HEPES-KOH, pH 7.4, 110 mM potassium acetate, 5 mM magnesium acetate, and incubated overnight at 4°C with a 10-fold molar excess of fluorescein isothiocyanate (Sigma) or alternatively with a 10-fold molar excess of FLUOS (Roche). Labeling of 4z-rpL23a with fluorescein 5′-maleimide (Pierce) through the engineered cysteine residue was performed in 50 mM HEPES-KOH, pH 7.5, 300 mM potassium acetate, 5 mM magnesium acetate. The labeled reaction mixtures were passed over NAP 5 columns (Pharmacia) equilibrated in 20 mM HEPES, pH 7.3, 110 mM potassium acetate, 2 mM magnesium acetate, 1 mM EGTA, 2 mM dithiothreitol, and 250 mM sucrose to remove unincorporated fluorochrome.

Import assays.Import reactions were performed at 37°C essentially as described previously (34) based on the method established by Adam et al. (1). This assay allows reconstitution of nuclear import events by using recombinant purified transport factors without the interference of endogenous importins. Briefly, HeLa cells were grown on 10-mm glass coverslips to 40 to 80% confluence. Permeabilization was done with 40 μg/ml digitonin (Calbiochem) for 10 min on ice. The permeabilized cells were incubated for 25 min with 20 μl of a transport reaction mix consisting of cargo (2 μM) and either recombinant importin (2 μM) or 10 μl of reticulocyte lysate (Promega) in transport buffer (20 mM HEPES-KOH, pH 7.4, 110 mM potassium acetate, 5 mM magnesium acetate, 0.5 mM EGTA, 2 mM dithiothreitol, and 250 mM sucrose) along with an energy-regenerating system (0.5 mM ATP, 0.5 mM GTP, 10 mM creatine phosphate, and 50 μg/ml creatine kinase). In case of reconstitution experiments with recombinant transport factors, 3 μM Ran(GDP) and 0.5 μM NTF2 (called Ran mix) were added to the transport reaction mix to create a RanGTP gradient across the nuclear membrane. Cells were fixed by using 3% paraformaldehyde for 15 min and were mounted by using Vectashield with DAPI (Vector Laboratories). For the negative controls the assay was done in the absence of any transport factors, i.e., without reticulocyte lysate and recombinant transport factors. For competition experiments, 10 times the concentration of the actual cargo of nonfluorescent competitor protein was added to the import reaction. Import reactions were visualized by fluorescence microscopy with a Zeiss microscope (Axioskop 20) using a 40× objective lens (Plan Neofluar).

In vitro transcription-translation.In vitro transcription-translation of NF-YA, NF-YB, NF-YC, transcription factor IIIA (TFIIIA), and GST was performed from the corresponding SP6 promoter constructs (pCS2+MT). Using the T_NT-coupled reticulocyte lysate system (Promega) according to the manufacturer's instructions, the proteins were labeled with [³⁵S]methionine (Amersham). Reactions were performed at 30°C for 2 h in a 25-μl volume, and the samples were then either frozen or directly used for oocyte microinjections.

Oocyte microinjection and analysis.Preparation of oocytes, microinjection, and immunoprecipitation was performed as described before (30, 79). Briefly, collagenase-treated Xenopus laevis oocytes (stage V/VI) were kept over night in MBSH (88 mM NaCl, 2.4 mM NaHCO₃, 1 mM KCl, 10 mM HEPES, pH 7.6, 0.8 mM MgSO₄, 2 mM CaCl₂, 3.3 mM KNO₃) for recovery. Samples of 10 to 20 nl of the in vitro translation reaction were injected into the cytoplasm of oocytes. At the indicated time points, 10 to 15 oocytes were dissected manually and homogenized in NET-2 buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0.05% NP-40). Proteins of the nuclear and cytoplasmic fractions were precipitated using antibodies against the myc tag (Santa Cruz) and finally analyzed by SDS-PAGE and phosphorimaging (Amersham).

NF-Y subunits are predominantly nuclear at steady state.The subcellular localization of endogenous NF-Y subunits and differently tagged, overexpressed versions of the three NF-Y subunits was examined in HeLa cells. In this assay, β-galactosidase-tagged fusion constructs were under the control of an SV40 promoter causing moderate overexpression of the β-gal fusion proteins. In contrast, EGFP-, HA-, and myc-tagged fusion proteins were strongly overexpressed in HeLa cells, because their genes were under the control of a cytomegalovirus (CMV) promoter. Dominant nuclear localization patterns were evident for the endogenous and β-galactosidase-tagged NF-YA, NF-YB, and NF-YC subunits. In contrast, strong overexpression influenced the distribution of NF-YB and, significantly, of NF-YC, while the dominant nuclear localization pattern of NF-YA remained unchanged (Fig. 1A, lower panels). The fluorescent staining of EGFP-tagged NF-YB (EGFP-NF-YC) and HA-tagged NF-YB (HA-NF-YC) became more cytoplasmic compared to the β-galactosidase fusion protein, and EGFP- and HA-NF-YC remained essentially in the cytoplasm (Fig. 1A). When EGFP-NF-YC was overexpressed together with myc-tagged NF-YB and vice versa the subcellular localization changed, becoming largely nuclear (Fig. 1B). This behavior could also be observed coexpressing HA-NF-YC and myc-tagged NF-YB (Fig. 1B lowest panel), which further indicates that the nuclear localization of NF-YC depends on the presence of its histone fold partner NF-YB, as previously shown by Frontini et al. (24). Taken together, these results point toward NF-YA being imported into the nucleus independently of NF-YB and NF-YC, whereas these HFM-containing subunits mediate the nuclear localization of the corresponding histone fold partner.

• Open in new tab
• Download powerpoint

FIG. 1.

Subcellular localization of NF-Y subunits in HeLa cells. (A) The endogenous NF-Y subunits stained with anti-NF-YA, anti-CBF-A (NF-YB), and anti-CBF-C (NF-YC) antibodies and moderately overexpressed (SV40 promoter) β-galactosidase-tagged NF-Y subunits are predominantly nuclear at steady state in HeLa cells. Subcellular localization of strongly overexpressed (CMV promoter) EGFP-tagged and HA-tagged NF-Y subunits differs in HeLa cells. (B) EGFP-NF-YB, EGFP-NF-YC, and HA-NF-YC were cotransfected with the corresponding histone fold partner (myc tagged) and costained as indicated. In all cases the subcellular localization of the tagged fusion proteins was examined 24 h after transient transfection, either by direct (EGFP) or indirect (β-gal, HA, myc) fluorescence microscopy. Nuclei were stained with DAPI.

Sequence motifs necessary for nuclear localization of NF-Y subunits correspond to their conserved domains.To identify the regions essential for the nuclear localization of NF-Y subunits, we constructed a series of proteins which contained fragments of the individual NF-Y subunits fused to bacterial β-galactosidase as reporter protein and expressed the gene constructs in HeLa cells. The results of these in vivo transfection experiments, summarized in Fig. 2 and Fig. 3, revealed that the nuclear targeting signal of each subunit corresponded to its conserved domain. In particular, the region required for nuclear localization of NF-YB and NF-YC consists of their known protein-protein interaction domain termed histone fold motif (HFM). Constructs containing the N or C terminus of the protein or only parts of the HFM were not imported into the nucleus (except for the partial nuclear staining observed for pMB117; Fig. 2B) but remained in the cytoplasm (Fig. 2A and B). In NF-YA two overlapping regions within the conserved part containing either the subunit interaction domain (labeled BC) or the DNA-binding domain (labeled DNA) together with the linker region confer nuclear import when fused to β-galactosidase (Fig. 3, lower panel, pJK68 and pJK69). Surprisingly, deletion constructs bearing either of these two minimal regions targeting β-galactosidase into the nucleus were deficient in mediating nuclear uptake (Fig. 3, upper panel, pJK84 and pJK100). These results showed that the two minimal regions were not sufficient to confer nuclear import in the context of (nearly) the entire protein. In conclusion, NF-YA contains a nonclassical NLS (ncNLS) in the 56 amino acid long evolutionarily conserved region at the C terminus of the protein (amino acids 262 to 317). This is not unexpected, since NLS- and DNA-binding domains overlap in 90% of the proteins for which both the NLS- and DNA-binding regions are known (16).

• Open in new tab
• Download powerpoint

FIG. 2.

Nuclear transport of NF-YB and NF-YC fragments fused to β-galactosidase. The regions or sequence motifs that are essential for the nuclear localization of the histone fold motif (HFM)-containing subunits were determined by overexpressing fragments of NF-YB (A) and NF-YC (B). HeLa cells were transiently transfected with gene fragments of the subunits fused to bacterial β-galactosidase as reporter gene, and the subcellular localization of the gene products was examined by indirect immunofluorescence 24 h later. Nuclei were stained with DAPI. On the left, the names of the expression constructs used in this study and the amino acids contained in the constructs are listed. Lines indicate the protein regions, and gray bars represent the histone fold motif (HFM) or parts of it. On the right, representative photographs of the subcellular localization of the constructs. The regions required for nuclear localization of NF-YB and NF-YC correspond to the evolutionarily conserved HFM. aa, amino acid.

• Open in new tab
• Download powerpoint

FIG. 3.

NF-YA contains an NLS in the evolutionarily conserved domain. The NLS was mapped essentially as described in the legend to Fig. 2. On the left, the names of the expression constructs used in this study and the amino acids contained (or deleted) in the constructs are listed. Lines indicate the protein regions, bent lines indicate deleted regions, and gray bars represent the conserved domain including subunit interaction domain (BC) and DNA binding domain (DNA) or parts of it. On the right are representative photographs of the subcellular localization of the constructs. The 56-amino-acid-long conserved region at the C terminus of NF-YA contains a nonclassical NLS. aa, amino acid.

Positively charged amino acids are required for nuclear targeting of NF-YA.The ncNLS of NF-YA contains 11 arginine and 7 lysine residues within the conserved region of 56 amino acids. The number of positively charged amino acids is very similar to the BIB domain of rpL23a that contains nine arginine and eight lysine residues (43 amino acids long) (34). To analyze the function of the basic amino acid residues in the NLS sequence of NF-YA, we generated mutants substituting some of the positively charged amino acid residues (Fig. 4A). These NF-YA mutants, along with wild-type NF-YA, all of them fused to β-galactosidase, were expressed in HeLa cells, and the subcellular localization was detected by indirect immunofluorescence. Based on the fluorescence intensity the nucleocytoplasmic distribution of the overexpressed NF-YA mutants was quantified, and the results of the in vivo transfection studies are summarized in Fig. 4B. Nuclear localization of NF-YA was strongly reduced when the basic cluster (KRR) in the subunit interaction domain (BC) was mutated (pJK104, 46% nuclear uptake corresponding to an equal distribution of fluorescence in nucleus and cytoplasm). In contrast, the mutation of other positively charged amino acid residues and basic clusters did not (pJK103, pJK106) or just moderately (pJK105) reduce the nuclear localization of (mutant) NF-YA (Fig. 4B). However, mutation of the three basic amino acid residues (KRR) in the subunit interaction domain in addition to the mutations as described in pJK105 or pJK106, respectively (6 out of the 18 positively charged amino acid residues mutated in total), led to a significant loss of nuclear accumulation (Fig. 4B; pJK130, 21% nuclear localization; pJK131, 17% nuclear localization). Two conclusions can be drawn from these observations. Firstly, a certain number of positively charged amino acid residues in the NLS sequence of NF-YA is required for nuclear uptake. Secondly, some of the basic amino acid residues in the region essential for subunit interactions (46) are also crucial for nuclear localization (for example, amino acids mutated in pJK104; Fig. 4A). Additionally, the mutant NF-YA constructs were expressed as GST fusion proteins in E. coli and tested for binding to importin β (Fig. 4C). The results of these in vitro binding studies are described and discussed below in detail (in the context of the identification of importin β as import receptor for NF-YA).

• Open in new tab
• Download powerpoint

FIG. 4.

Characterization of the ncNLS of NF-YA by mutagenesis of basic clusters; positive charge is required for nuclear targeting. (A) Wild-type and mutant NF-YA were fused to either β-galactosidase or GST (construct names are listed on the left). The amino acids that were substituted for the basic amino acid residues are indicated by a black background. (B) The β-galactosidase fusion constructs were overexpressed in HeLa cells, and the subcellular localization of the NLS mutants was detected by indirect immunofluorescence as described in the legend to Fig. 2. The nucleocytoplasmic distribution of the overexpressed NF-YA mutants was quantified with the program ImageJ (nuclear uptake in percentages, written below the representative photographs; RSD, relative standard deviation). The results indicate that nuclear import of NF-YA was reduced when the basic cluster (KRR) in the subunit interaction domain (BC) was mutated (pJK104). Mutation of 6 out of 18 basic amino acids inside the conserved domain (ncNLS) almost completely blocked nuclear uptake (pJK130, pJK131). (C) In addition, binding studies with importin β were done using GST fusion constructs carrying the same mutations in the ncNLS of NF-YA as used in the in vivo import experiments described in panel B. The names of these constructs (pJK197 to pJK202) are included in panel A. The GST fusion constructs were expressed in E. coli, immobilized on glutathione-Sepharose, and incubated with importin β from bacterial lysate. The bound fractions were analyzed by SDS-PAGE followed by Coomassie staining. The basic cluster (KRR) in the subunit interaction domain proved to be crucial for importin β binding, and the additional mutation of three more positively charged amino acid residues caused a complete loss of importin β binding (pJK201, pJK202). The gene product pJK203 (1 to 269 amino acids), which resembles a C-terminal NF-YA deletion construct lacking nearly the entire conserved domain (harboring the ncNLS), was used as negative binding control. DNA, DNA binding domain; Q, glutamine-rich trans-activation domain; aa, amino acid; M, molecular mass in kilodaltons.

NF-Y subunits are actively transported in Xenopus oocytes.To characterize and compare the nuclear transport of the different NF-Y subunits, in vitro transcribed and translated, ³⁵S-labeled, myc-tagged fusion proteins were injected into the cytoplasm of Xenopus oocytes. Nuclear and cytoplasmic fractions were separated manually at the indicated time points, homogenized, immunoprecipitated using antibodies against the myc tag, and analyzed by SDS-PAGE with subsequent phosphorimaging. As shown in Fig. 5A, 24 h after microinjection the majority of the cytoplasmically injected NF-Y subunits were detected in the nuclear fractions. However, the kinetics of nuclear translocation differed among the three NF-Y subunits. While NF-YA was imported rapidly into the nucleus of Xenopus oocytes, NF-YB and NF-YC were imported much more slowly (for better illustration, see the graphic in Fig. 5B). One explanation for the different import kinetics would be that the interaction with the endogenous histone fold partner, which is abundant in Xenopus oocytes (42), is a prerequisite for nuclear uptake. When the temperature was lowered to 4°C the nuclear import of the three subunits was nearly blocked (Fig. 5C). This strong temperature dependence points to an active nuclear transport of the NF-Y subunits rather than passive diffusion. Furthermore, the relative amount of nuclear import of NF-YB and NF-YC depended on the quantity of injected protein. The more protein that was injected, the lower the percentage of nuclear uptake (data not shown). These results are in accordance with the differences observed between strong and moderate overexpression of NF-YB and NF-YC in HeLa cells and further indicate that nuclear accumulation presumably relies on interaction/dimerization of the two HFM-containing subunits.

• Open in new tab
• Download powerpoint

FIG. 5.

Active nuclear transport of NF-Y subunits in Xenopus laevis oocytes. ³⁵S-labeled NF-YA, NF-YB, NF-YC myc-tagged fusion proteins were injected into the cytoplasm of Xenopus oocytes; nuclear (N) and cytoplasmic (C) fractions were separated manually, either immediately (0 h) or at the indicated time points (up to 24 h of incubation), and were analyzed for the nucleocytoplasmic distribution of the different subunits by SDS-PAGE. To control the injection/separation of nuclear and cytoplasmic fractions, coinjection of either transcription factor IIIA (TFIIIA) or GST protein was used. The nucleocytoplasmic distribution of the injected proteins was quantified with the program ImageQuant 5.2 after phosphorimaging (in percentages, written below the gels). (A) Twenty-four hours after microinjection the majority of the cytoplasmically injected NF-Y subunits were detected in the nuclear fractions. (B) The kinetics of nuclear translocation differ among the three subunits. NF-YA was imported rapidly, while the histone fold motif subunits NF-YB and NF-YC were imported much slower. (C) When the temperature was lowered to 4°C the nuclear import of the subunits was blocked. This strong temperature dependence points to an active nuclear transport of the NF-Y subunits rather than passive diffusion.

NF-Y subunits interact with different import receptors.Assuming that the NF-Y subunits were actively imported into the nucleus, their nuclear transport would have been mediated most likely by transport receptors of the importin β superfamily. The NLS of NF-YA resembles a nonclassical (importin α-independent) arginine-rich type similar to the BIB domain of rpL23a that confers import by importin β, importin 5, importin 7, and transportin. Thus, it was obvious to investigate the role of these transport factors with regard to the nuclear import of NF-YA. In addition, importin 13 was included, since binding of the NF-YB-like protein YBL1 (11) to immobilized importin 13 had been reported (50). YBL1 shows a strong NF-YB homology covering the histone fold motif, but despite their similarity YBL1 differs from NF-YB in respect to specific functions (11). In order to identify potential import receptors, GST pulldown assays were performed with NF-Y subunits. The three NF-Y subunits were expressed as GST fusions in E. coli and immobilized on glutathione-Sepharose beads. In addition, the HFM-containing subunits were coexpressed in E. coli as either GST-NF-YB/GST-NF-YC or zz-NF-YB/GST-NF-YC complex and also used as bait after immobilization on glutathione-Sepharose beads. The beads were incubated with an importin α/β heterodimer, importin β, transportin, importin 5, and importin 13, all from bacterial lysates, and were subsequently washed. The retained proteins were analyzed by SDS-PAGE followed by Coomassie staining. Importin β, transportin, and importin 5 showed binding to NF-YA, whereas under the same conditions none of the import factors bound efficiently to the monomeric NF-YB and NF-YC subunits. Surprisingly, importin 13 only bound to the NF-YB/NF-YC dimer, while binding of other transport receptors to the preformed complex was insignificant (Fig. 6).

• Open in new tab
• Download powerpoint

FIG. 6.

Interaction of NF-Y subunits with different import receptors. The three NF-Y subunits were expressed as GST fusions in E. coli and immobilized on glutathione-Sepharose. In addition, the HFM-containing subunits were coexpressed in E. coli as GST-NF-YB/GST-NF-YC complex and zz-NF-YB/GST-NF-YC complex and used as bait after immobilization on glutathione-Sepharose. The immobilized fusion proteins were incubated with either importin α/β, importin β, transportin, importin 5, or importin 13, all from bacterial lysates. The bound fractions were analyzed by SDS-PAGE followed by Coomassie staining. Note that importin β, transportin, and importin 5 were bound to NF-YA, whereas under the same conditions none of the used import factors bound efficiently to single HFM-containing subunits NF-YB and NF-YC. Surprisingly, importin 13 only bound the dimerized NF-YB/NF-YC complex while binding of other transport receptors to the preformed HMF-dimer was insignificant. Further, the importin α/β heterodimer just bound efficiently to the control cargo nucleoplasmin and importins did not bind to GST alone. M, molecular mass in kilodaltons.

Importins bind specifically to NF-Y subunits.In order to verify that binding of the different members of the importin β family to NF-YA and binding of importin 13 to the NF-YB/NF-YC complex is specific, GST pulldown assays were performed in the presence and absence of RanGTP, which was used to simulate nuclear Ran conditions. Immobilized GST-NF-YA was incubated with purified recombinant importin β, importin 7, transportin, and importin 5 (Fig. 7A). All these import receptors were bound specifically to NF-YA, and this binding was reduced by the GTP-bound form of RanQ69L. Among the four importins that bind specifically to NF-YA, importin β showed the highest binding competence (Fig. 7B). On the other hand, the immobilized NF-YB/NF-YC complex showed binding of recombinant importin 13 from a bacterial lysate. The retained proteins were analyzed by SDS-PAGE, and indeed importin 13 was bound specifically to the histone fold dimer and this binding was abolished by RanGTP, as expected for an import cargo (Fig. 7C). Since importin β had shown the highest binding competence to GST-NF-YA, these in vitro binding studies were extended to the different NLS mutants characterized in Fig. 4A and B. The mutant NF-YA GST-fusion constructs were expressed in E. coli, immobilized on glutathione-Sepharose, and incubated with importin β (from bacterial lysate), and bound fractions were analyzed by SDS-PAGE followed by Coomassie staining. As shown in Fig. 4C, in vitro binding assays indicate that the basic clusters in the subunit interaction domain (KRR, pJK198) and linker region (RRK, pJK199) are crucial for importin β binding. The additional mutation of three more positively charged amino acid residues caused a complete loss of importin β binding (pJK201, pJK202). These binding data fit well with the results from the in vivo transfection studies (Fig. 4B) and strongly support our conclusion that a certain number of positively charged amino acid residues in the NLS sequence of NF-YA are required for nuclear targeting.

• Open in new tab
• Download powerpoint

FIG. 7.

Specific binding of importins to NF-Y subunits. GST pulldown assays were performed as described in the legend to Fig. 6. (A) Immobilized GST-NF-YA was incubated with purified recombinant importin β, importin 7, transportin (trn), and importin 5. Starting material (10% of the importins that were used) and bound fractions were analyzed by SDS-PAGE followed by Coomassie staining. As shown in panel A, the import receptors were bound specifically to immobilized NF-YA, and this binding was reduced by the GTP-bound form of RanQ69L (2 μM) which was used to simulate nuclear Ran conditions. (B) Among the transport factors binding to NF-YA, importin β showed the highest binding competence. (C) Immobilized zz-NF-YB/GST-NF-YC complex and GST-NF-YB/GST-NF-YC complex (see also arrows on the right) were used to bind recombinant importin 13 from a bacterial lysate. Importin 13 was bound specifically to the immobilized NF-YB/NF-YC dimer, and this binding was abolished by RanGTP. Hence, the NF-YB/NF-YC dimer behaves as an importin 13-specific import substrate. M, molecular mass in kilodaltons; rec. imp, recombinant importin.

Binding sites in the NF-YB/NF-YC complex for NF-YA and importin 13 overlap.Since the minimal regions essential for nuclear localization of NF-YB and NF-YC correspond to their histone fold motifs (as shown by the in vivo transfection experiments in Fig. 2) and only the preformed NF-YB/NF-YC dimer was able to bind to importin 13, it is reasonable to assume that the binding site in the NF-YB/NF-YC complex for importin 13 consists of the dimerized histone fold motifs. To further test this assumption, binding of recombinant importin 13 to an immobilized zz-NF-YB/GST-NF-YC complex was performed in the presence of an excess of NF-YA, which is thought to bind to a groove formed by the dimerized histone fold motifs (63). This approach showed that importin 13 interacts with the NF-YB/NF-YC complex in an NF-YA sensitive manner. Importin 13 was bound specifically to the histone fold dimer, and this binding was abolished by excess of NF-YA (Fig. 8A, note the binding of NF-YA instead of importin 13). To test the specificity of the competition, GST-NF-YB/GST-NF-YC complex on beads was exposed to purified recombinant importin 13 and increasing amounts of NF-YA. Again, the results show that importin 13 competes with NF-YA for binding to the NF-YB/NF-YC dimer (Fig. 8B). Addition of equimolar concentrations of importin 13 and NF-YA to the NF-YB/NF-YC complex results in weaker binding of importin 13 (Fig. 8B, lane 7), indicating a higher affinity of NF-YA than importin 13 for the NF-YB/NF-YC dimer.

• Open in new tab
• Download powerpoint

FIG. 8.

Binding sites in the NF-YB/NF-YC complex for NF-YA and importin 13 overlap. Importin 13 interacts with the NF-YB/NF-YC complex in a NF-YA sensitive manner. (A) Importin 13 was bound specifically to the immobilized zz-NF-YB/GST-NF-YC complex, and this binding was abolished by a 10-fold molar excess of NF-YA. As can be seen in the last lane, NF-YA was bound to the histone fold dimer instead of importin 13. (B) GST-NF-YB/GST-NF-YC complex on beads (also labeled B/C) was exposed to purified recombinant importin 13 (1 μM, lane 3 to 10) and increasing amounts of His-tagged NF-YA (lane 4, 0.2 μM; lane 5, 0.4 μM; lane 6, 0.8 μM; lane 7, 1 μM; lane 8, 1.25 μM; lane 9, 2.5 μM; lane 10, 5 μM). The results show that importin 13 competes with NF-YA for binding to the NF-YB/NF-YC dimer. Addition of equimolar concentrations of importin 13 and NF-YA to the NF-YB/NF-YC complex (lane 7) results in weaker binding of importin 13. M, molecular mass in kilodaltons; rec., recombinant; imp, importin.

NLS in NF-YA is not recognized by importins once the subunit is assembled into the NF-Y complex.Since the identified NLS of NF-YA also contains the known subunit interaction domain, we wished to determine if NF-YA can bind import receptors once it is assembled into the trimeric NF-Y complex. To address this question, two GST pulldown experiments were performed. Firstly, a GST-NF-YB/GST-NF-YC complex was used as bait and exposed to recombinant importin β and/or NF-YA (Fig. 9A). NF-YA and importin β were added either at the same time (lane 7) or stepwise, NF-YA first and importin β 1 h later (Fig. 9A, lane 8). Again, bound fractions were analyzed by SDS-PAGE followed by Coomassie staining. When NF-YA was prebound to the NF-YB/NF-YC complex, no binding of importin β could be observed, showing that NF-YA was unable to mediate the binding of the entire heterotrimeric NF-Y complex to importin β. In the presence of an importin β excess, NF-YA showed reduced binding to the histone fold dimer (data not shown), also indicating that the importin β/NF-YA complex cannot bind to the NF-YB/NF-YC dimer. Secondly, a preassembled NF-Y holocomplex (see Materials and Methods regarding recombinant protein expression and purification) was immobilized on glutathione-Sepharose beads and incubated with recombinant importin β, transportin, and importin 5 (Fig. 9B). None of the import factors bound efficiently to the preassembled NF-Y holocomplex. In conclusion, the NLS of NF-YA cannot be recognized by import receptors once the trimeric NF-Y complex is formed, which is probably due to the fact that the subunit interaction domain and NLS overlap.

• Open in new tab
• Download powerpoint

FIG. 9.

NLS in NF-YA is not recognized by importins once the subunit is assembled into the NF-Y complex. (A) GST-NF-YB/GST-NF-YC (labeled B/C) on beads was exposed to recombinant importin β with and without addition of His-tagged NF-YA to determine whether NF-YA can bridge binding of importin β to the trimeric NF-Y complex. NF-YA and importin β were added either at the same time (lane 7) or stepwise, NF-YA first and importin β 1 h later (lane 8). The last lane (10) shows the control binding of importin β to bead-bound GST-NF-YA (labeled A) alone. (B) Preassembled NF-Y holocomplex (see Materials and Methods) consisting of NF-YA-His, GST-NF-YB, and GST-NF-YC was immobilized on glutathione-Sepharose beads and incubated with recombinant importin β, transportin (trn), and importin 5. None of the import receptors that bound to GST-NF-YA alone (see Fig. 7B) were bound to the preassembled NF-Y holocomplex. Starting material (20% of the recombinant importin β and NF-YA that were used; panel A) and bound fractions were analyzed by SDS-PAGE followed by Coomassie staining. M, molecular mass in kilodaltons; imp, importin.

NF-Y subunits are imported into nuclei of permeabilized cells by distinct pathways.Finally, we wanted to examine whether the differentially interacting importins represent functional import receptors for NF-YA or for the NF-YB/NF-YC complex. For that purpose, HeLa cells were grown on glass coverslips and the plasma membranes were selectively permeabilized with digitonin to reconstitute nuclear import by using recombinant purified import factors (1). The use of this system revealed that the nuclear import of fluorescently labeled 6z-NF-YA was dependent on the presence of rabbit reticulocyte lysate as a source of import factors (Fig. 10A). To analyze the transport factors involved in the nuclear import of NF-YA in vitro, import assays were performed in the presence of nonfluorescent core histones as potential import competitors. The addition of individual nonfluorescent core histones to the import mix completely blocked the nuclear import of NF-YA (shown for H2A and H2B in Fig. 10A). The competition by core histones indicated that the nuclear import of NF-YA depends entirely on at least one of the import receptors that mediate nuclear transport of the core histones, namely importin β, importin 5, importin 7, importin 9, or transportin (3, 53). Despite the fact that importin β, importin 5, importin 7, and transportin bound specifically to GST-NF-YA (see Fig. 7A), it was only imported into the nuclei of permeabilized cells when importin β together with Ran mix and an energy-regenerating system was present (Fig. 10C). While the monomeric HFM-containing subunits NF-YB and NF-YC (as 6z fusion protein) were not imported into nuclei of permeabilized cells (Fig. 10B), nuclear accumulation of the GST-NF-YB/GST-NF-YC complex was strictly importin 13 dependent and stimulated by Ran mix and an energy regenerating system (Fig. 10E). Other importins, such as importin β, had no significant effect. Addition of an NF-YA excess to the import mix led to a reduced nuclear import of the preformed NF-YB/NF-YC complex in the presence of importin 13 (data not shown), indicating that only the histone fold dimer can be recognized by importin 13 and that NF-YA competes with importin 13 for binding to the NF-YB/NF-YC dimer (see Fig. 8). Taken together, we can conclude that importin β functions as import receptor for NF-YA, whereas nuclear transport of the NF-YB/NF-YC complex indeed can be mediated by importin 13.

• Open in new tab
• Download powerpoint

FIG. 10.

NF-Y subunits are imported into nuclei of permeabilized cells by distinct pathways. Importin β mediates import of NF-YA, and importin 13 confers nuclear uptake of a preformed NF-YB/NF-YC dimer. Digitonin-permeabilized cells were incubated with 1 μM of each substrate, rabbit reticulocyte lysate (retic) or the indicated combinations of nuclear transport receptors (2 μM each), a Ran mix (including RanGDP and NTF2), and an energy regenerating system (energy [E]) for 25 min at 37°C. For a negative control, retic or the recombinant import receptor was replaced by transport buffer (buffer). Panels show the cargo distribution after import and fixation. Substrates carrying GST (except GST-EGFP-Pax6) were detected by indirect immunofluorescence, whereas fluorescently labeled cargoes (6z-NF-YA, 6z-NF-YB, 6z-NF-YC, and 4z-rpL23a) were detected by direct immunofluorescence. (A) 6z-NF-YA was imported into nuclei of permeabilized cells in the presence of retic and energy; however, 6z-NF-YA tended to slightly diffuse into nuclei in the absence of import receptors (buffer). Nuclear import reactions were also performed in the presence of retic, energy, and either 10 μM H2A or H2B (nonfluorescent) as indicated. Both core histones completely inhibited the nuclear uptake of 6z-NF-YA. (B) 6z-NF-YB and 6z-NF-YC were not imported into nuclei of permeabilized cells in the presence of retic and energy but remained in the cytoplasm in vicinity to the nuclear membrane. (C) Nuclear accumulation of GST-NF-YA was dependent on the presence of importin β, Ran mix, and an energy-regenerating system. However, GST-NF-YA remained in the cytoplasm in the presence of other importins such as transportin and importin 5. (E) The GST-NF-YB/GST-NF-YC complex was imported into nuclei of permeabilized cells in the presence of importin 13, Ran mix, and an energy regenerating system but was not imported in the absence of importin 13 or in the absence of Ran and the energy regenerating system. (D and F) In vitro nuclear import assay of the reference substrates 4z-rpL23a, GST-M9, and GST-EGFP-Pax6 in the presence and absence of appropriate transport receptors. imp, importin.