HCF-dependent nuclear import of VP16
1999; Springer Nature; Volume: 18; Issue: 2 Linguagem: Inglês
10.1093/emboj/18.2.480
ISSN1460-2075
Autores Tópico(s)Genomics and Chromatin Dynamics
ResumoArticle15 January 1999free access HCF-dependent nuclear import of VP16 Sylvie La Boissière Sylvie La Boissière Marie Curie Research Institute, The Chart, Oxted, Surrey, RH8 OTL UK Search for more papers by this author Thomas Hughes Thomas Hughes Marie Curie Research Institute, The Chart, Oxted, Surrey, RH8 OTL UK Search for more papers by this author Peter O'Hare Corresponding Author Peter O'Hare Marie Curie Research Institute, The Chart, Oxted, Surrey, RH8 OTL UK Search for more papers by this author Sylvie La Boissière Sylvie La Boissière Marie Curie Research Institute, The Chart, Oxted, Surrey, RH8 OTL UK Search for more papers by this author Thomas Hughes Thomas Hughes Marie Curie Research Institute, The Chart, Oxted, Surrey, RH8 OTL UK Search for more papers by this author Peter O'Hare Corresponding Author Peter O'Hare Marie Curie Research Institute, The Chart, Oxted, Surrey, RH8 OTL UK Search for more papers by this author Author Information Sylvie La Boissière1, Thomas Hughes1 and Peter O'Hare 1 1Marie Curie Research Institute, The Chart, Oxted, Surrey, RH8 OTL UK *Corresponding author. E-mail: [email protected] The EMBO Journal (1999)18:480-489https://doi.org/10.1093/emboj/18.2.480 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Transactivation by VP16 requires the formation of a multicomponent complex, the TAATGAAAT recognition factor complex (TRF.C), that contains in addition to VP16, two cellular proteins, Oct-1 and HCF. HCF binds directly to VP16 and this promotes subsequent interaction of the VP16–HCF complex with the POU DNA-binding domain of Oct-1 and selective assembly onto target sites. Here we demonstrate a novel role of HCF in the intracellular compartmentalization of VP16. We show that while VP16 does not contain a consensus nuclear localization signal (NLS) and is largely cytoplasmic, co-expression with HCF resulted in VP16 nuclear accumulation. A candidate NLS within the C-terminus of HCF was identified and insertion of this motif into green fluorescent protein (GFP) promoted nuclear accumulation. Conversely, removal of this signal from HCF (HCFΔNLS) resulted in its cytoplasmic accumulation. Co-expression of HCFΔNLS with wild-type (wt) VP16, or of wt HCF with VP16 mutants lacking HCF-binding activity failed to promote the nuclear enrichment of VP16. These results indicate that in addition to its role in stabilizing TRF.C, HCF acts as a nuclear import factor for VP16. Introduction VP16, a structural component of herpes simplex virus (HSV) induces expression of the viral immediate-early (IE) genes by promoting the assembly of a multicomponent complex containing the cellular factors Oct-1 and HCF onto regulatory motifs present upstream of each of the IE genes (for reviews see O'Hare, 1993; Wilson et al., 1993a). Oct-1 is a member of the POU domain protein family (Herr et al., 1988; Sturm et al., 1988) whose members play a fundamental role in the regulation of transcription of distinct groups of genes in a wide array of cellular processes (Verrijzer and Van der Vliet, 1993). Analysis of the involvement of one of the members of this family of proteins in HSV IE gene expression has illustrated how selective utilization of such factors can control the coordinate induction of a distinct set of genes. The motifs responsible for VP16 induction encompass specialized recognition sites for Oct-1, and VP16 is recruited onto these sites by interactions with Oct-1 and HCF to form the TRF.C complex (Kristie and Roizman, 1987; Gerster and Roeder, 1988; O'Hare and Goding, 1988; Preston et al., 1988; apRhys et al., 1989; Stern et al., 1989; Wilson et al., 1993a). Selectivity in the assembly of TRF.C operates at the levels of both the protein–protein and protein–DNA interactions. Differences between the POU domains of Oct-1 and Oct-2 at particular surface-exposed residues of the homeodomain, notably a critical glutamic acid residue within helix 2, dictate selective recognition by VP16 (Stern and Herr, 1991; Lai et al., 1992; Pomerantz et al., 1992). With regard to DNA-binding site selectivity, only a subset of Oct-1-binding sites is competent to promote VP16 complex assembly. The critical feature of the competent sites is the conservation of a 3′ flanking motif, the GARAT motif, which is not necessary for Oct-1 binding per se, but is necessary for the recruitment of VP16 into the TRF.C complex by a mechanism involving recognition of specific conformational forms of Oct-1 (O'Hare et al., 1988; Walker et al., 1994). The precise role of HCF is the least understood aspect of the mechanism of action of VP16, and in contrast to the information on Oct-1, there have been no data on any aspect of the normal role of HCF in the cell until recently. Human HCF is transcribed in a wide range of tissue types (Frattini et al., 1994; Wilson et al., 1995b; Lu et al., 1997). Interestingly the HCF protein is translated as a 2035 amino acid precursor protein (∼300 kDa) which is subsequently cleaved at specific repeated motifs, located between residues 1000 and 1440, giving rise to a family of polypeptides ranging from 110 to 150 kDa (Kristie and Sharp, 1993; Wilson et al., 1993b, 1995a; Kristie et al., 1995). The HCF repeats consist of six copies of a 26-amino acid motif which represent the cleavage site for an as yet unidentified cellular protease (Wilson et al., 1995a). However, after proteolytic cleavage, N- and C-terminal products remain closely associated (Wilson et al., 1993b, 1995a). Recent studies investigating HCF structure–functional relationships in relation to its interactions with VP16 and TRF.C formation have shown that the N-terminal region, comprising residues 1–380, contains critical determinants for association with VP16 and for complex formation with the TAATGARAT sequence (La Boissière et al., 1997; Simmen et al., 1997; Wilson et al., 1997). This region of HCF is composed of six reiterations of a different repeating motif of ∼50 amino acids, designated kelch repeats, and originally defined in the Drosophila protein kelch (Xue and Cooley, 1993). We have recently demonstrated that a fusion protein containing both this N-terminal repeat region and the C-terminal region, excluding the central region of ∼1000 residues of the protein (HCF.NC), efficiently promoted complex formation (La Boissière et al., 1997). Although a mutation conferring temperature sensitivity on division of a hamster cell line was shown to be due to the substitution of a residue within the N-terminus of the hamster HCF (Goto et al., 1997), the normal role of HCF remains unclear. To date, most of our understanding of the VP16 transcription pathway has been obtained from biochemical analysis of the assembly of the DNA-binding complex and there has been little information on the intracellular interaction of the constituent components. Recent analyses of several systems, e.g. signal transduction via NF-κB (Gilmore et al., 1996), Wnt-1/β-catenin (Miller and Moon, 1996) and Notch regulatory components (Struhl and Adachi, 1998), have highlighted the role of compartmentalization in transcriptional regulation. We wished to pursue aspects of the compartmentalization of proteins involved in the VP16 activation pathway in vivo. In particular, we noted that VP16 does not possess its own nuclear localization signal and since biochemical studies have indicated that HCF and VP16 interact with one another independently of other factors in the complex, we wished to investigate the possibility of an additional role for HCF in VP16 activity. We show that while VP16 is mainly found in the cytoplasm of transfected cells, co-expression with HCF resulted in the relocalization of VP16 to a predominantly nuclear pattern. The subcellular distribution of VP16 in HCF-expressing cells was not reorganized using VP16 mutants which were unable to bind directly to HCF. Furthermore, by deletion analysis and by transfer experiments, we locate the NLS of HCF at its extreme C-terminus between residues 2015 and 2031, and show that a mutant lacking this NLS was unable to promote enhanced nuclear import of VP16. Taken together, these results indicate that in addition to its role in promoting assembly of TRF.C, HCF is also involved in nuclear trafficking of VP16. Results Analysis of HCF subcellular localization HCF has been previously reported to be present in a diffuse nuclear pattern in several cell types, although a distinct punctate accumulation was noted in HeLa cells (Kristie et al., 1995). To pursue investigation of the subcellular distribution of HCF we constructed SV5-epitope-tagged versions of both the full-length protein (HCF.FL) and of a version consisting of a fusion between the N- and C-terminal regions (HCF.NC) which promotes efficient complex formation with VP16 in in vitro DNA-binding assays (La Boissière et al., 1997). Schematic illustration of the various HCF constructs are indicated in Figure 1B. COS cells were transfected with the expression vector for HCF.FL or mock transfected, the cells were fixed 40 h later and HCF detected with either an anti-HCF polyclonal antibody, raised against the C-terminal residues 1508–2024, or the anti-SV5 antibody to detect the N-terminus. Examination of mock-transfected cells with the anti-HCF antibody showed that endogenous HCF was diffusely localized in the nucleus and exhibited nucleolar exclusion (Figure 2a). Cells expressing the introduced HCF.FL in the transfected cell population could be readily identified with the anti-HCF antibody (Figure 2c). In approximately half of the HCF.FL-expressing cells, the protein was mainly localized in the nucleus (Figure 2c, arrowheads), whereas in the remainder of the cells, the protein was more or less equally distributed in the cytoplasm and nucleus (Figure 2c, arrow). To identify selectively the HCF expressed from the introduced gene we used the anti-SV5 epitope tag antibody (Figure 2d). Again we observed a population exhibiting a nuclear only distribution and a population where the protein was present in both cytoplasm and nucleus. From examination of numerous fields in independent experiments, the population of cells containing the protein in both cytoplasmic and nuclear compartments was reproducibly higher by detection with the anti-SV5 antibody, representing ∼75% of the total (Figure 2d, arrows), with the remainder of the cells exhibiting nuclear accumulation (Figure 2d, arrowheads). Figure 1.(A) Schematic representation of the 2035 residue human HCF open reading frame (Wilson et al., 1993b) as discussed in the text. (B) Schematic representations of the constructs used in this study. The HCF codons are shown by the shaded boxes while the black box at the N-terminus represents the anti-SV5 epitope tag. Download figure Download PowerPoint Figure 2.Intracellular localization of HCF. COS-1 cells grown on coverslips were mock-transfected (a and b) or transfected (100 ng) individually with expression vectors for HCF.FL (c and d), HCF.NC (e and f), HCF.FLΔNLS (g and h), and HCF.NCΔNLS (i and j). Cells were fixed and reacted with the anti-HCF polyclonal antibody (a, c, e, g and i) or the anti-SV5 monoclonal antibody (b, d, f, h and j), and examined by confocal microscopy. Download figure Download PowerPoint We next analysed the localization pattern of HCF.NC, a 1031 amino acid protein lacking the central region from residue 491 to 1494 and thus the HCF cleavage sites. Using the anti-HCF antibody, HCF.NC exhibited a similar distribution pattern similar to HCF.FL (compare Figure 2c with e). Staining of HCF.NC-expressing cells with the anti-SV5 epitope tag antibody again revealed a similar localization pattern with cells exhibiting either a total diffuse distribution or a nuclear enrichment (Figure 2f). However, with the HCF.NC construct there was a change in the ratio of the patterns compared with HCF.FL. Typical fields from analysis of several independent experiments are represented in Figure 2d and f. Thus for HCF.NC, ∼75% of expressing cells exhibited pronounced nuclear enrichment, with 25% exhibiting both cytoplasm and nucleus localization (Figure 2f). The change in ratio in nuclear enrichment compared with HCF.FL, detected with the anti-SV5 N-terminal antibody, was a real one and finds an explanation in further results below (see Discussion). Redistribution of VP16 in HCF co-transfected cells Inspection of the VP16 amino acid sequence indicates that it does not contain a consensus nuclear localization signal. Since biochemical analysis demonstrates that HCF and VP16 associate with one another independently of assembly of the DNA-binding complex with Oct-1, we wished to determine if HCF was involved in VP16 subcellular localization. We first analysed the distribution pattern of VP16 in cells containing endogenous HCF only. Cells were transfected with a VP16 expression vector (pRG50), fixed 40 h post-transfection and VP16 detected with the monoclonal antibody LP1 (Figure 3A, left panel). The majority of VP16 was diffusely localized in the cytoplasm of cells, with a relatively minor amount in the nucleus. Similar results were observed in Vero and BHK cells (data not shown). This cytoplasmic localization of VP16 was also observed when VP16 was expressed from a weaker version of the CMV enhancer/promoter, pRG36 (Figure 3A, middle panel). Figure 3.Intracellular localization of VP16 expressed individually or coexpressed with HCF. (A) COS-1 (left and middle panels) or Vero cells (right panel) were independently transfected with expression vectors for VP16 (pcVP16, 50 ng, left panel; pRG36, 50 ng, middle panel; pcVP16, 100 ng, right panel). Cells were fixed and reacted with LP1. (B) COS-1 cells were cotransfected with expression vectors for HCF.FL (100 ng) and VP16 (pRG36, 50 ng) and processed for detection with LP1 (top panels) and anti-HCF antibody (bottom panels). Two independent fields scanned for VP16 (top panel) and HCF (bottom panel) are shown. (C) Exactly as for (B) except that HCF.NC was used. (D) As for (B) except using pcVP16 (100 ng) and HCF.NC (200 ng) in Vero cells. Download figure Download PowerPoint To analyse the effect of HCF on subcellular localization of VP16, cells were co-transfected with expression vectors for VP16 (pRG36) and HCF.FL, and the proteins detected simultaneously with anti-VP16 or anti-HCF antibodies. Strikingly, in those cells expressing HCF.FL there was a significant redistribution of VP16, and the protein was now seen only or mainly in the nucleus. This alteration of VP16 subcellular localization was observed in virtually all cells expressing exogenous HCF and was particularly pronounced in cells expressing relatively higher levels of HCF (Figure 3B). A separate field showing nuclear accumulation of VP16 in several cells co-expressing exogenous HCF is shown in Figure 3B (right panel). This pattern of nuclear accumulation of VP16 was only observed in the presence of exogenous HCF (contrast cell with VP16 only, arrowhead left panel) and indicated that HCF may be a limiting factor in nuclear entry of VP16. Although all previous data indicate that there is no direct complex between Oct-1 and VP16, in the absence of HCF, we performed similar co-transfection experiments with Oct-1 and VP16. In contrast to the results with HCF, overexpression of Oct-1 did not influence the distribution of VP16 (data not shown). We wished to determine whether VP16 relocalization was also observed with HCF.NC since its distribution pattern was similar to HCF.FL, but HCF.NC is not processed by the repeat-specific pathway. As shown in Figure 3C, the VP16 subcellular localization was also altered in HCF.NC-expressing cells, resulting in similar nuclear accumulation (Figure 3C, c.f. VP16 alone, Figure 3A). Identical results were obtained in Vero cells, where VP16 alone was present throughout the cell with significant cytoplasmic distribution (Figure 3A, right panel) while in HCF.NC expressing Vero cells, VP16 showed nuclear enrichment (Figure 3D, arrowed cell). Deletion of the NLS of HCF resulted in the accumulation of the protein in the cytoplasm Analysis of the primary amino acid sequence of HCF revealed the presence of a putative bipartite consensus NLS between residues 2015 and 2031. This type of NLS motif is defined as two basic residues, a spacer region of any 10 amino acids and a basic cluster in which three out of the next five residues must be basic (Nigg, 1997). To investigate whether this region was responsible for the nuclear localization of HCF, a deletion mutant of the full-length protein was constructed in which the last 33 residues were removed (HCF.FLΔNLS). Its subcellular pattern was analysed using either the anti-HCF (Figure 2g) or the anti-SV5 antibody (Figure 2h). Examination with the anti-HCF antibody showed that HCF.FLΔNLS was localized in the cytoplasm (Figure 2g, arrow). The weaker nuclear staining (Figure 2g, arrowhead) was no more intense than that observed in untransfected cells (Figure 2a) and represented endogenous HCF. The absence of nuclear accumulation of the HCF.FLΔNLS was clearly observed using the anti-SV5 antibody (Figure 2h). Similar analyses were performed on HCF.NC lacking the NLS motif (HCF.NCΔNLS). Again, with either the anti-HCF or anti-SV5 antibodies we observed a lack of nuclear accumulation of HCF.NCΔNLS (compare Figure 2e, i and f, j). Of the two versions lacking the NLS, HCF.FLΔNLS and HCF.NCΔNLS, we observed slightly higher amounts of the smaller HCF.NC version in the nucleus (compare Figure 2h and j). Nevertheless, there was a striking difference in the ratio of cells showing nuclear accumulation between HCF.NC and HCF.NCΔNLS. To confirm the proposal that the candidate sequence acted as a NLS we asked whether it could confer nuclear localization onto a reporter protein. We constructed a vector encoding the green fluorescent protein (GFP) fused at its C-terminus to the HCF NLS and examined localization in live cells (Figure 4). The control GFP protein was present in both the cytoplasm and the nucleus, while the addition of the NLS of HCF clearly induced nuclear accumulation, indicating that this sequence is a NLS and is transferable to heterologous proteins. Altogether these results provide compelling evidence that the putative NLS located at the extreme C-terminus of HCF is functional and is likely to be the only NLS in HCF. Figure 4.The NLS of HCF directs GFP into the nucleus. COS-1 cells were transfected with 100 ng of either GFP (left panel) or GFP.HCF.NLS (right panel) expression vectors. At 24 h post-transfection, live cells were examined for the distribution of GFP and GFP.HCF.NLS. Download figure Download PowerPoint Deletion of the NLS in HCF abrogates VP16 relocation We next tested whether the HCF constructs in which the NLS had been deleted were still able to induce VP16 nuclear accumulation. Cells were transfected with expression vectors for VP16 together for those for HCF.FLΔNLS or HCF.NCΔNLS, and processed for VP16 and HCF localization as above (Figure 5A). In contrast to the effect of the NLS-containing versions, VP16 underwent no reorganization in either HCF.FLΔNLS or HCF.NCΔNLS-expressing cells. We note that in co-expressing cells, we frequently observed an accumulation of VP16 and HCF ΔNLS species around the perimeter of the nucleus. Figure 5.In vitro and in vivo interactions between VP16 and HCFΔNLS. (A) COS-1 cells were cotransfected with the expression vector for VP16 (pRG36, 50 ng) together with that for HCF.FL (100 ng) or HCF.FLΔNLS (100 ng). Cells were fixed and processed as in Figure 3. (B) Samples representing 10% of the in vitro translated inputs for VP16, HCF.NC and HCF.NCΔNLS used in each reaction are shown (lanes 1–3 respectively). Equal amounts of HCF.NC (lane 7) or HCF.NCΔNLS (lane 8) were incubated with VP16 and immunoprecipitated with LP1. Control immunoprecipitations were performed without VP16 (lanes 5 and 6) or without HCF (lane 4). (C) Doubling doses of VP16 starting at 0.25 μl of programmed lysate were incubated with the POU domain, TAAT24 probe and equal amounts of either in vitro-translated HCF.NC or HCF.NCΔNLS. Independent POU-binding and POU–VP16–HCF complexes are indicated by arrows. A non-specific DNA activity from the reticulocyte lysate is indicated by an asterisk. Download figure Download PowerPoint To rule out the possibility that the lack of nuclear enrichment of VP16 in HCFΔNLS-expressing cells was due to an absence of interaction between the two proteins, we first analysed direct binding between VP16 and HCF in co-immunoprecipitation assays. We used in vitro-translated VP16 and the anti-VP16 antibody LP1, in the absence and presence of either HCF.NC or HCF.NCΔNLS (Figure 5B). Controls showed that in the absence of VP16 (Figure 5B, lanes 5 and 6), only background levels of HCF were precipitated. In the presence of VP16, HCF.NC was co-precipitated (Figure 5B, lane 7), an interaction which we have previously shown to be specific and require key determinants in VP16 (La Boissière et al., 1997; see also Figure 6). Deletion of the NLS had no effect on this interaction (Figure 5B, compare lanes 7 and 8). Figure 6.In vitro and in vivo interactions between HCF and VP16 362in and 364YA variants. (A) Schematic representation of VP16 illustrating the positions of the mutations used here within the region from 355 to 390. (B) Samples representing 10% of the input wt VP16 (lane 1), variant 362in (lane 2), variant 364YA (lane 3) and HCF.NC (lane 4) used in each reaction are shown. HCF.NC was incubated with equal amounts of wt VP16 (lane 9), 362in (lane 10) or 364YA mutants (lane 11), and immunoprecipitated with LP1. Control immunoprecipitations were performed without HCF (lanes 5 to 7) or without VP16 (lane 8). (C) Equal amounts (in doubling doses) of in vitro-translated wt VP16 or VP16 variants were incubated with in vitro-translated HCF.NC (1 μl), the POU domain and the TAAT24 probe. Complexes are indicated as for Figure 5. (D) COS-1 cells were transfected with the expression vector for VP16 362in variant (50 ng), fixed and processed as Figure 3. (E) COS-1 cells were transfected (50 ng) with vectors for VP16 362in or 364YA variants together with the vectors (100 ng) for either HCF.FL or HCF.NC and processed as for Figure 3. Each pair of panels represents a field showing the VP16 localization (left hand side) or the HCF localization (right hand side) as indicated. Download figure Download PowerPoint We next compared the ability of both HCF constructs to participate in the DNA-binding complex with Oct-1 and VP16. In vitro-translated VP16 was incubated in increasing amounts with the purified POU domain of Oct-1, equal amounts of in vitro-translated HCF.NC or HCF.NCΔNLS, and a probe containing the octamer-GARAT sequence from the HSV IE110k promoter (Figure 5C). The results show the formation of the POU–VP16–HCF complex, that this was dependent on HCF (see lane with top dose of VP16 and no HCF), and that deletion of the NLS had no significant effect. Taken together, the results demonstrate that the absence of nuclear enrichment of VP16 in HCFΔNLS-cotransfected cells was not caused by the inability of the two proteins to interact with one another and reinforce the proposal that enhanced nuclear import of VP16 requires HCF protein competent for nuclear entry. VP16 mutants unable to interact with HCF do not exhibit enhanced nuclear import In parallel with the experiments above, we examined the effect of HCF containing the NLS but in this case with VP16 mutants which would be impaired in their ability to interact with HCF. The role of the region encompassing residues 360–390 in complex formation has been extensively studied and several critical residues involved in either step have been identified (Ace et al., 1988; Werstuck and Capone, 1989; Greaves and O'Hare, 1990; Lai and Herr, 1997). We therefore constructed VP16 variants incorporating a 3 residue linker insertion at position 362 (362in), or a single tyrosine to alanine substitution at position 364 (364YA). We first analysed the ability of these VP16 mutant proteins to associate with HCF in solution by co-immunoprecipitation of in vitro-translated HCF.NC and VP16 using LP1 (Figure 6B). In the presence of wild-type (wt) VP16, HCF was coprecipitated (lane 9), whereas no significant interaction with HCF was observed above background for either 362in (lane 10) or 364YA (lane 11). Controls showed that equal amounts of the VP16 wt and variants were used (lanes 1–3). We assessed whether these mutations would also have a deleterious effect on complex formation. Equal aliquots of in vitro-translated wt VP16 or variants were incubated in a dose–response experiment, with the purified POU domain, in vitro-translated HCF.NC, and the DNA probe (Figure 6C). The result shows that wt VP16 efficiently promoted the formation of a POU–VP16–HCF.NC complex whereas the two VP16 variants were defective in complex assembly. These results reinforce the critical requirement of this region of VP16 in complex assembly and show that the insertion of three residues in VP16 at position 362 or that a single point mutation at position 364 is sufficient to prevent direct association between VP16–HCF, in agreement with recent results (Lai and Herr, 1997). For the analysis of distribution patterns of these variants we first examined their location in the absence of exogenous HCF and found no detectable difference compared with wt VP16 (Figure 6D, 362in). However, in cotransfection experiments with HCF, no nuclear accumulation of the VP16 variants was observed in those cells co-expressing HCF (Figure 6E). Similar results were obtained with 364YA and HCF.NC or 362in with HCF.FL (data not shown). In contrast to wt VP16, the VP16 variants did not accumulate in the nucleus in cells expressing the exogenous HCF, whether the HCF was nuclear or nuclear plus cytoplasmic (see arrowed cells, Figure 3B–D). The result indicates that the reorganization of VP16 pattern requires an association between VP16 and HCF. Discussion HCF is one of two cellular proteins required to assemble the transcription complex through which VP16 functions. Biochemical analyses from several laboratories have shown that VP16 directly binds to HCF, that HCF is required for normal complex assembly and that HCF is itself incorporated into the complex (Gerster and Roeder, 1988; Katan et al., 1990; Xiao and Capone, 1990; Kristie and Sharp, 1993; Wilson et al., 1993b). In this study, we have examined the cell biology of the interaction between VP16 with results which provide compelling evidence for a novel role of HCF in VP16 function. HCF is involved in nuclear entry of VP16 VP16 does not contain a consensus NLS and exhibited a predominantly cytoplasmic location. Co-expression with HCF resulted in VP16 nuclear accumulation. We considered it possible, though unlikely, that the nuclear accumulation of VP16 might be indirectly associated with HCF co-expression. For example it may have been that for some reason, expression of HCF resulted in lower levels of VP16, that some other factor limited VP16 entry and then that lower levels exhibited predominantly nuclear localization. In control experiments, expression of HCF did not promote nuclear accumulation of a VP16 mutant unable to bind to HCF, and conversely a HCF mutant lacking its NLS, while still able to bind VP16, did not promote VP16 nuclear entry. Altogether the results indicate that HCF is itself a limiting factor in VP16 nuclear import and acts as a chaperone for the nuclear entry of VP16 by direct interaction. Another possibility, but less likely in our view, is that rather than actively transport VP16, HCF acts as a sink, retaining VP16 in the nucleus. While there are previous examples of interactions between HSV viral regulatory proteins (Knipe and Smith, 1986; Mullen et al., 1995; Zhu et al., 1996), or between viral structural proteins influencing nuclear localization (Nicholson et al., 1994; Rixon et al., 1996), to our knowledge, outside of the importin α/β protein mechanism of nuclear entry upon which many classes of proteins depend, this is the first report of a specific cellular NLS-bearing protein chaperoning a viral non-NLS protein into the nucleus. The results have implications for understanding the normal cellular role of HCF and the VP16 pathway. HCF compartmentalization To date the cellular function of HCF is not known, although its strong conservation in for example rodents (Frattini et al., 1996), its presence in Caenorhabditis elegans (Wilson et al., 1993b) and the identification of a mutation in HCF conferring temperature sensitivity on cell proliferation (Goto et al., 1997) all indicate that it plays important role(s) in cell metabolism. The protein contains a kelch repeat motif at its N-terminus, cleavage sites between residues 1000 and 1440 for an unidentified protease, and the C-terminal NLS identified in this work. The kelch repeat domain is the main site of interaction with VP16 (La Boissière et al., 1997; Simmen et al., 1997; Wilson et al., 1997), and also contains the mutation conferring temperature sensitivity on cell proliferation (Goto et al., 1997). In those proteins with a conserved kelch domain, the main role which at least some have in common that would underpin the conservation of the domain, is in cytoskeletal organization, particularly actin filament binding or cross linking (e.g. Eichinger et al., 1996; Hernandez et al., 1997). For HCF specifically, the only currently identified interacting cellular component is a transcription factor of the basic-leuci
Referência(s)