ABSTRACT
Previous studies defined pUL84 of human cytomegalovirus as an essential regulatory protein with nuclear localization that was proposed to act during initiation of viral-DNA synthesis. Recently, we demonstrated that a complex domain of 282 amino acids within pUL84 functions as a nonconventional nuclear localization signal. Sequence inspection of this domain revealed the presence of motifs with homology to leucine-rich nuclear export signals. Here, we report the identification of two functional, autonomous nuclear export signals and show that pUL84 acts as a CRM-1-dependent nucleocytoplasmic shuttling protein. This suggests an unexpected cytoplasmic role for this essential viral regulatory protein.
The open reading frame UL84 of human cytomegalovirus (HCMV) encodes an essential regulatory protein that appears to be localized within the cell nucleus (20, 33, 35). Initially, pUL84 was identified as a direct interaction partner of the nuclear HCMV regulatory protein IE2-p86 (28), which is the major transcription-activating protein of HCMV (22). Studies concerning the functional consequence of the pUL84-IE2 interaction revealed, on one hand, that this interaction down-regulates the transactivation activity of IE2 on some early promoters (13). On the other hand, it has been reported that this pUL84-IE2 complex is required for the activation of a bidirectional promoter located within the origin of lytic DNA replication (ori-Lyt) (34). Since pUL84 had additionally been shown (i) to be essential for virus replication (33, 35) and (ii) to be the only noncore protein required for origin-dependent DNA replication in a transient replication assay (24, 27), pUL84 was proposed to act as an initiator protein for viral-DNA synthesis of HCMV (34). The initiator proteins of some other herpesviruses were demonstrated to exert an inherent catalytic activity that may unwind a specific region of DNA within ori-Lyt, thus allowing the assembly of the DNA replication machinery (1, 2, 5, 9). In line with this, pUL84 has been shown to display UTPase activity and to exhibit some homology to the DExD/H box family of helicases (3). It is noteworthy that bioinformatic studies of this protein did not support the classification of pUL84 as a DExD/H box helicase but detected a homology to dUTPases (4). Taken together, pUL84 encodes an essential viral regulatory protein that is supposed to be active in the cell nucleus.
In this regard, we previously unraveled the molecular mechanism of pUL84 nuclear trafficking and could demonstrate that a nonconventional nuclear targeting domain comprising 282 amino acids within pUL84 mediates its interaction with importin alpha proteins (20). Interestingly, sequence inspection of the amino acid sequence of this nonconventional nuclear localization signal (NLS) revealed the presence of two small, leucine-rich regions that exactly match the consensus sequence of a classical nuclear export signal (NES) (20) (Fig. 1A). Both motifs could be aligned to confirmed NESs of other proteins (Fig. 1B), and one of the two motifs, UL84-NES2, yielded a positive NES motif score when we analyzed the HCMV UL84 primary sequence using an NES prediction server (16). In order to determine whether one of the putative pUL84 NESs is able to direct nuclear export of a heterologous protein, we performed microinjection experiments exactly as described previously (19). For this, we fused amino acids 228 to 237 and 359 to 366 of pUL84 to the C terminus of glutathione S-transferase (GST) (Fig. 1C), which served as a carrier protein. GST was used, since previous experiments showed that this protein on its own does not relocalize into the cytoplasm upon microinjection into the nucleus due to its capacity to form high-molecular-weight dimers (6, 25). The resulting recombinant fusion proteins were purified from Escherichia coli (Fig. 1D) and microinjected into the nuclei of HeLa cells in combination with rabbit immunoglobulin G (IgG) as a marker for the injection site (Fig. 1E, a to d). One hour after injection, the cells were fixed and double immunostained for GST and IgG (19). As shown in Fig. 1E, a to d, the coinjected rabbit IgG was detected exclusively within the nucleus, whereas GST-UL84-NES1 or GST-UL84-NES2 was translocated from the nuclear injection site to the cytoplasm in significant amounts. To exclude passive diffusion, we also injected the fusion proteins into the cytoplasm of HeLa cells. No nuclear accumulation of the respective proteins could be detected (Fig. 1E, e to h). This indicates that each transport signal is sufficient to target the GST protein for nuclear export and thus functions as an autonomous NES. Furthermore, it suggests that pUL84 may be capable of nucleocytoplasmic shuttling.
In order to prove this, we next performed an interspecies heterokaryon analysis using HCMV-infected primary human foreskin fibroblasts (HFF) (19). For this, HCMV (strain AD169)-infected cells were fused with nonpermissive murine NIH 3T3 cells in order to produce heterokaryons (Fig. 2A). After fusion, the cells were fixed and the localization of viral nucleoproteins was assessed by indirect immunofluorescence. Murine nuclei were identified by counterstaining them with Hoechst 33258 dye, yielding a characteristic punctate heterochromatin staining (Fig. 2A, a). If pUL84 constitutes a bona fide nucleocytoplasmic shuttling protein, it would be expected to move from the infected HFF nuclei into the cytoplasm and subsequently to enter those murine nuclei that are part of the heterokaryon. As shown in Fig. 2A, b, by 3.5 h postfusion, pUL84 had migrated into the murine nuclei. In contrast, IE1, which is an immediate-early regulatory protein of HCMV, was not detected in murine nuclei of interspecies heterokaryons (Fig. 2A, c). Thus, we conclude that in infected cells, pUL84 continuously shuttles between the nucleus and the cytoplasm.
Next, we asked whether nucleocytoplasmic shuttling of pUL84 also occurs in the absence of additional viral factors (Fig. 2B). To investigate this, HeLa cells were cotransfected with a pUL84 expression plasmid (20) and either the internal control plasmid CFN-βGAL or CFNrev-βGAL (19, 26). CFNrev-βGAL encodes β-galactosidase (β-Gal) fused to an NLS and an NES; thus, this fusion protein shuttles between the nucleus and the cytoplasm (Fig. 2B, c). CFN-βGAL expresses only an NLS-β-Gal fusion protein and is therefore restricted to nuclei (Fig. 2B, g). Transfected cells were allowed to synthesize pUL84, together with one of the control proteins, and were subsequently subjected to heterokaryon formation. After fixation, the cells were costained for pUL84 and β-Gal (19). In interspecies heterokaryons that coexpressed pUL84 and β-Gal-NLS/NES, both proteins were observed in murine and human nuclei (Fig. 2B, a to d). In contrast, when the localization of pUL84 and β-Gal-NLS was assessed, pUL84 alone was found to be present in the human and murine nuclei, whereas β-Gal-NLS was detected exclusively in the human nuclei (Fig. 2B, e to h). This suggests that the pUL84 nucleocytoplasmic shuttling activity is not dependent on either viral infection or the expression of HCMV-encoded cofactors.
The nuclear export of proteins bearing a leucine-rich NES is usually mediated by the export receptor CRM1/exportin1 (18). The antibiotic leptomycin B (LMB) specifically blocks CRM1-mediated nuclear export by disrupting the interaction of NESs with the receptor (10, 14, 15, 32). To investigate the contribution of the CRM1 pathway to the nuclear export of pUL84, HeLa cells were cotransfected with a pUL84 expression plasmid and a control plasmid encoding the CRM1-independent nucleocytoplasmic shuttling protein pUL69 (19). Subsequently, heterokaryon assays were performed as described above, with the exception that 3 h prior to and following heterokaryon formation, the cells were treated with 2.5 ng/ml LMB to inhibit CRM1 function. In the absence of LMB, both proteins, pUL84 and pUL69, could be detected in murine nuclei after fusion with transfected HeLa cells (Fig. 2C, a to d). In contrast, treatment with LMB completely blocked shuttling of pUL84 but did not prevent the translocation of pUL69 in the same heterokaryon (Fig. 2C, e to h), indicating that pUL84 uses the CRM1-mediated nuclear export pathway.
pUL84 encodes two NESs that are equally capable of mediating the nuclear export of a heterologous protein (Fig. 1). Although the physiological significance of two NESs in pUL84 is presently unclear, this finding is not without precedent, since multiple leucine-rich NESs have been uncovered in a series of proteins of cellular or viral origin (7, 23, 29). In order to analyze whether the two identified NESs represent the only export signals within pUL84, we next aimed at inactivating these signals in the context of the wild-type protein. It has been demonstrated that substitution of alanine for any leucine within the core motif of the NES abrogates export activity (12, 29). Thus, we generated eukaryotic expression plasmids that contained the mutant UL84 sequences indicated in Fig. 3. However, since both export signals are located within the pUL84 importin alpha binding domain (20), we were concerned that mutations at these sites might affect the nuclear import of the protein. Therefore, we initially determined the subcellular localization of each of the mutants in transfected HeLa cells. Figure 3A summarizes the results of immunolocalization experiments. It shows that all pUL84 mutants carrying mutations in UL84-NES1 showed nuclear localization, as did the wild-type protein (Fig. 3A, a to h). In contrast, when leucine residues at positions 359/361 or 364/366 within UL84-NES2 were replaced by alanines, the respective mutants displayed a partial cytoplasmic localization (Fig. 3A, k and l, and m and n). In light of these results, we generated plasmid LLL228/30/59AAA, encoding pUL84 with mutations at key residues in both UL84-NES1 and UL84-NES2 that were predicted to inhibit nuclear export but not nuclear import (Fig. 3A, o and p). When we subjected this mutant to the interspecies heterokaryon assay, we observed that the nuclear-export capability of the pUL84 mutant was completely abolished, excluding the existence of other, unrecognized NESs (Fig. 4, i to l). However, when heterokaryon assays were performed with mutants LL228/30AA and L359A, which inactivated only one of the two NESs, weak pUL84 staining was detected in murine nuclei (Fig. 4, a to d and e to h). This indicates that both signals contribute independently to the nucleocytoplasmic shuttling activity of the wild-type protein.
Taken together, these experiments describe a novel nuclear-export activity of pUL84 and thus define this protein as the second nucleocytoplasmic shuttling protein to be identified in HCMV. This suggests an unexpected cytoplasmic role for this essential regulatory protein. Although additional studies are needed to elucidate the relevance of pUL84 shuttling, the presently available evidence indicates that this capacity might be crucial for its biological function, as underlined by the presence of two functional NESs embedded in a highly complex bidirectional-transport domain. This conclusion leads us to speculate that pUL84 might carry out an unknown function in the cytoplasm, as has recently been shown for the herpes simplex virus type 1 shuttling protein ICP27 (17); alternatively, pUL84 might act in addition to pUL69 as a carrier that transports macromolecules between the two cellular compartments. Recently, pUL69 was defined as an HCMV-encoded RNA-binding protein with activities in the nuclear export of RNA (21, 30). Since RNA binding could also be observed for pUL84 (T. Stamminger, unpublished data), it is tempting to speculate that pUL84 may also function in the nuclear export of a subset of viral RNAs. However, since pUL84 has been implicated in viral-DNA replication, studies to elucidate the relevance of nucleocytoplasmic shuttling for the initiation of viral-DNA replication will be of particular importance.
pUL69 contains two active leucine-rich NESs. (A) Schematic representation of the UL84 protein showing the NLS/importin alpha binding domain (20). Two putative leucine-rich NES motifs located within the nonconventional NLS/importin alpha binding domain of pUL84 are highlighted. (B) Alignment and comparison of the two putative pUL84 NES motifs with known leucine-rich NESs and with the consensus sequence for leucine-rich NESs. The indicated pUL84 motifs are compared with the NES of the protein kinase inhibitor α (PKI) (31) or the fragile X mental retardation protein (FMRP) (11). Additionally, the NES of the human immunodeficiency virus type 1 Rev protein (HIV1-REV) (8) and a derived consensus NES are listed. Conserved residues in the NESs are shown in boldface. The numbers refer to the positions of the amino acid sequences within each protein. (C) Schematic representation of the UL84 coding sequence showing the two putative NESs fused to the C terminus of GST to produce GST-UL84-NES1 and GST-UL84-NES2. (D) Procaryotic expression and purification of GST-UL84-NES1 and GST-UL84-NES2. Shown is a Coomassie blue-stained gel: extracts from E. coli cells grown without isopropyl-β-d-thiogalactopyranoside (IPTG) (lanes 1 and 4) and grown in the presence of IPTG (lanes 2 and 5) are shown; lanes 3 and 6, purified GST fusion proteins. (E) GST fusion proteins were microinjected into the nuclei (a to d) or the cytoplasm (e to h) of HeLa cells, together with rabbit IgG as a marker for the injection site. At 1 h after injection, the cells were fixed and immunostained for GST-UL84-NES1 or GST-UL84-NES2 and the coinjected IgG control.
Nucleocytoplasmic shuttling of pUL84 in infected and transfected cells. (A) Heterokaryons were generated by fusion of HCMV-infected primary HFF and murine NIH 3T3 cells. Prior to and following heterokaryon formation, de novo protein synthesis was inhibited using cycloheximide. At 3.5 h after fusion, the cells were fixed, and a double-immunofluorescence analysis was performed with a polyclonal antiserum directed against pUL84 (b) and a monoclonal antibody against the IE1 protein (c). Staining with Hoechst 33258 (a) was used to differentiate between human and murine nuclei within the heterokaryon. Murine nuclei display a characteristic punctate pattern, whereas human nuclei are diffusely stained with the reagent; murine nuclei are indicated by arrows. Panel d shows the phase-contrast image of the heterokaryons; the cytoplasmic edge is highlighted by a broken line. (B) HeLa cells were cotransfected with expression plasmids for pUL84 and one of the internal control plasmids, β-Gal-NLS/NES or β-Gal-NLS, as indicated (a to d, UL84 and β-Gal-NLS/NES; e to h, UL84 and β-Gal-NLS). The transfected cells were subsequently analyzed in heterokaryon assays as described in the legend to panel A. Double-immunofluorescence analysis with polyclonal anti-pUL84 serum and a monoclonal antibody against β-Gal was performed in order to detect the expressed proteins. (C) pUL84 shuttles between the nucleus and the cytoplasm in a CRM1-dependent manner. Expression constructs for pUL84 and pUL69 were cotransfected into HeLa cells, and the transfected cells were subjected to heterokaryon assays. Three hours prior to fusion and throughout the experiment, the cells were incubated in the absence (−) (a to d) or presence (+) (e to h) of LMB. The indicated proteins were detected by double-label immunofluorescence using a polyclonal pUL84 antiserum and a monoclonal antibody directed against pUL69 (19).
Subcellular localization of UL84 mutants carrying mutated nuclear export signals. (A) A series of pUL84 mutants carrying alanine replacement mutations either in NES1 or NES2 was generated, and the subcellular localization of the resulting mutants was analyzed via indirect immunofluorescence analysis. The mutants indicated on the left were transiently expressed in HeLa cells, which were subsequently fixed and immunostained (right) using an anti-pUL84 antiserum (αUL84); DAPI, DNA staining of the transfected HeLa cells. (B) Western blot analysis of expression levels of the indicated UL84 mutants in transfected HeLa cell cultures.
Nucleocytoplasmic shuttling activities of pUL84 mutants carrying alanine substitutions within NES1, NES2, or NES1 and NES2. (Left) Schematic representation of the respective pUL84 mutants and β-Gal-NLS/NES, which was used as an internal shuttling control in the interspecies heterokaryon analysis. (Right) Heterokaryon experiments (as described in the legend to Fig. 2) were performed to visualize the nuclear-export activities of the indicated proteins.
ACKNOWLEDGMENTS
We thank Matthias Dobbelstein (Göttingen, Germany) for providing plasmids.
This work was supported by the Wilhelm Sander Stiftung, the DFG (SFB473), and the IZKF Erlangen.
FOOTNOTES
- Received 15 May 2006.
- Accepted 20 July 2006.
- Copyright © 2006 American Society for Microbiology
