Human cytomegalovirus UL94 is a nucleocytoplasmic shuttling protein containing two NLSs and one NES

Virus Research 166 (2012) 31–42

Contents lists available at SciVerse ScienceDirect

Virus Research journal homepage: www.elsevier.com/locate/virusres

Human cytomegalovirus UL94 is a nucleocytoplasmic shuttling protein containing two NLSs and one NES Yalan Liu a,b , Zhiping Zhang a , Xing Zhao a,b , Hongping Wei a , Jiaoyu Deng a , Zongqiang Cui a,∗∗ , Xian-En Zhang a,∗ a b

State Key Laboratory of Virology, Wuhan Institute of Virology, Chinese Academy of Sciences, Wuhan 430071, China Graduate School, Chinese Academy of Sciences, Beijing 100039, China

a r t i c l e

i n f o

Article history: Received 8 December 2011 Received in revised form 22 February 2012 Accepted 22 February 2012 Available online 5 March 2012 Keywords: HCMV UL94 Shuttle NLS NES

a b s t r a c t The tegument protein UL94 is a human cytomegalovirus (HCMV) late protein and its function has yet to be determined. Using live cell fluorescence recovery after photobleaching (FRAP) and fluorescence loss in photobleaching (FLIP) imaging, we found that UL94 is able to shuttle between the nucleus and cytoplasm. Analysis of UL94 mutants fused to EGFP showed that two newly characterized nuclear localization sequences (NLSs) and amino acid 343 play key roles in UL94 nuclear localization. Mutation of these sequences can alter the intracellular distribution of UL94 and disrupt its nucleocytoplasmic shuttling. Amino acid 343 of UL94 was also found to be crucial for its interaction with another HCMV tegument protein pp28. Furthermore, one nuclear export sequence (NES) was identified within UL94. Mutation of the key amino acids in the NES can also alter the intracellular distribution of UL94 and disrupt its shuttling function. Like other proteins containing a leucine-rich export signal, nuclear export of the UL94 was affected by leptomycin B, indicating that it is exported via the Crm1-dependent pathway. Our data provide a basis for further understanding the character and function of HCMV UL94. © 2012 Elsevier B.V. All rights reserved.

1. Introduction Human cytomegalovirus (HCMV) is a member of the betaherpesvirus family and can cause severe diseases, particularly in newborns and immunocompromised individuals (Mocarski and Courcelle, 2001; Pass, 2001). An infectious HCMV particle comprises three structural elements: an icosahedral capsid containing the double-stranded DNA genome, a tegument, and an envelope. The tegument, consisting of over 30 proteins, is unique to the herpesvirus family (Baldick and Shenk, 1996; Gibson, 1996; Mocarski and Courcelle, 2001). However, the characteristics and functions of many tegument proteins have not yet been determined. The HCMV tegument protein UL94 is a 345-amino acid protein encoded by the ul94 open reading frame (ORF). This protein is a true late protein that is detected only during the late stages of a productive HCMV infection and is not synthesized in the presence of the viral DNA replication inhibitor ganciclovir (Wing and Huang, 1995; Wing et al., 1996). It is thought that HCMV UL94 is a homolog of the HSV-1 tegument protein UL16 (Chee, 1991; Chee et al., 1990; Higgins and Sharp, 1989). However, little is known about the characteristics and functions of HCMV UL94. We have recently shown

∗ Corresponding author. Tel.: +86 10 58881508; fax: +86 27 87199492. ∗∗ Corresponding author. Tel.: +86 27 87199115; fax: +86 27 87199492. E-mail addresses: [email protected] (Z. Cui), [email protected] (X.-E. Zhang). 0168-1702/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.virusres.2012.02.023

that UL94 is a specific binding partner for pp28 (Liu et al., 2009). UL94, predominantly localized to the nucleus, could be directed into the cell cytoplasm by pp28 to colocalize with pp28 in a punctuate, juxtanuclear compartment, designated as the virus assembly compartment. Pp28 is an essential and abundant tegument protein required for HCMV assembly (final envelopment) and is related to the cell secretory pathway in the cytoplasm (Britt et al., 2004; Jones and Lee, 2004; Silva et al., 2003). In addition, intracellular localization of UL94 and the interaction between pp28 and UL94 may serve as a link in the sequential processes of HCMV capsidation. UL94 has also been revealed to interact with other tegument proteins UL82, UL25, and US22 by using Yeast Two Hybrid Analysis (To et al., 2011). Recently, using a murine cytomegalovirus (MCMV) model, Manninger et al.’s study revealed that the UL94 homologue gene is essential for viral trafficking at the secondary envelopment stage of MCMV (Maninger et al., 2011). Subsequently, Phillips and Bresnahan (2011) constructed a HCMV UL94-null mutant and revealed that UL94 functions late in infection to direct UL99 to the assembly complex, thereby facilitating secondary envelopment of virions. Therefore, it is important to investigate the localization and trafficking of HCMV UL94 in order to understand its function in viral infection and morphogenesis. In the current study, we first studied the intracellular dynamic behavior of UL94 using fluorescence recovery after photobleaching (FRAP) and fluorescence loss in photobleaching (FLIP) technologies. FRAP has been used extensively to examine the dynamics and

32

Y. Liu et al. / Virus Research 166 (2012) 31–42

mobility of proteins (Lippincott-Schwartz et al., 2001) and FLIP is ideal for studying the exchange of molecules between two compartments (e.g. compartments separated by lipid bilayers) (Cole et al., 1996; Nehls et al., 2000; Phair and Misteli, 2000). FRAP and FLIP imaging showed that UL94 is a nucleocytoplasmic shuttling protein. We then studied the mechanism that regulates the localization and trafficking of UL94. Two nuclear localization signals (NLSs) and one nuclear export sequence (NES) were discovered within UL94, neither of which has been reported previously. Amino acid 343 of UL94 was also found to be important for its nuclear localization and to play a key role in its interaction with pp28. Mutation of these localization-related sequences altered the intracellular distribution of UL94 and disrupted its nucleocytoplasmic shuttling. Furthermore, drug inhibition experiments indicated that the trafficking of UL94 may be linked to the classical Crm1-dependent nuclear export pathway.

techniques (J. Sambrook). All constructs were verified by DNA sequencing (Invitrogen Biotechnology Co., Ltd., Shanghai, China). 2.2. Cell culture, transfection and FRET assay Vero (African green monkey kidney) cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM, Gibco, USA) supplemented with 10% (v/v) fetal bovine serum (FBS, Gibco, USA), penicillin (100 U/ml) and streptomycin (100 ␮g/ml), at 37 ◦ C in 5% CO2 . The day before transfection, Vero cells were seeded onto 35 mm glass bottom culture dishes at 3 × 105 cells per dish. Cells were transfected with plasmids using lipofectamine 2000 reagent (Invitrogen), according to the manufacturer’s instructions. Transfected cells were incubated at 37 ◦ C (5% CO2 ) overnight (18–20 h) and the medium was replaced with fresh medium before examination by microscopy. FRET assay was carried out as described previously (Liu et al., 2009).

2. Materials and methods 2.1. Plasmid construction For FRAP and FLIP analysis, the UL94 gene (GenBank ID: 3077496) and its mutants were inserted into pEGFP-C1 via Bgl II and EcoRI sites to construct the plasmid pEGFPUL94 and a panel of UL94 mutants (Fig. 3A and Table S1). The construction, cloning, and propagation of plasmids were carried out using standard

2.3. Live cell imaging and fluorescence recovery after photobleaching (FRAP) and fluorescence loss in photobleaching (FLIP) analysis Fluorescence localization was imaged using an inverted widefield fluorescence microscope (Axiovert 200, Carl Zeiss, Germany). For live cell imaging, the prepared cell culture dishes were placed in a temperature-controlled incubator at 37 ◦ C and detection was

Fig. 1. Selective FRAP was used to demonstrate repeated nucleocytoplasmic shuttling of UL94 in multinuclear cells. (A) Serial images of EGFPUL94 in a selective FRAP experiment. The bleached nucleus is indicated by a white arrowhead. (B) Curves showing the fluorescence intensity of EGFPUL94 in different regions in a selective FRAP experiment (one region (ROI 1) in bleached nucleus; two regions in the unbleached nucleus (ROI 2 and ROI 3) and one region (ROI 4) in the cytoplasm). Bar = 10 ␮m. (C) Western blot of EGFP and EGFPUL94.

Y. Liu et al. / Virus Research 166 (2012) 31–42

performed using a 100× oil objective (NA 1.32) with a 488 nm excitation beam. All confocal-FRAP experiments were performed using a TCS SP2 Leica laser scanning spectral confocal microscope equipped with a cooled CCD camera. The 488 nm lasers were set at 20% for imaging and 100% for bleaching. The nucleus was bleached with a pinhole setting of 1.13 Airy units. Cells were scanned in 2D in real time. For FRAP assay, the bleach area in a cell could be precisely defined under the TCS SP2 confocal microscope. Vero cells expressing EGFPUL94 were pre-treated with 50% polyethylene glycol (PEG) to induce multinuclear cells. FRAP was carried out in the presence of cycloheximide, an inhibitor of translation, to prevent artifactual recovery as a result of new protein synthesis. For FLIP analysis, the defined areas of the nucleoplasm were bleached for approximately 30 s with maximum laser intensity. To correct the acquired fluorescence intensities to obtain a generalized bleaching effect, which results from the imaging scan, an unbleached neighboring

33

cell in the same window was monitored. The total fluorescence of the bleached cell and of a neighboring cell was monitored between the bleaching times. The bleached regions are indicated by rectangles in the first post-bleach image. Each image series shows the fluorescence intensity before bleaching (0 s) and after consecutive bleaching periods. 2.4. Western blotting Cells were transfected with a series of UL94 mutant DNA. Cell lysates and supernatants were harvested 48 h after transfection. Transfected cells were lysed in RIPA buffer (P0013C, Beyotime, China). The lysates were separated by SDS-PAGE and transferred to 0.4 ␮m PVDF membrane (Millipore). For detection, the membrane were incubated for 1 h at 37 ◦ C with primary antibody (rabbit antiGFP polyclonal antibody (sc-8334, Santa Cruz)) at a dilution of 1/500

Fig. 2. Nucleocytoplasmic shuttling of UL94 demonstrated by FLIP. (A) Serial FLIP images of EGFPUL94 in a nuclear FLIP experiment; (B) serial FLIP images of EGFPUL94 in a cytoplasmic FLIP experiment; (C) curves showing the fluorescence intensity of EGFPUL94 in different regions in a nuclear FLIP experiment (the bleached area (ROI 1), one region in the near cytoplasm (ROI 2), one region in the distant cytoplasm (ROI 3) and a non-bleached cell (ROI 4)); and (D) curves showing the fluorescence intensity of EGFPUL94 in different regions in a cytoplasmic FLIP experiment (the bleached area (ROI 1); a distant cytoplasmic region (ROI 2), the nuclear area (ROI 3) and a non-bleached cell (ROI 4)). The monitored regions are indicated by white arrowheads in the first image. Bar = 10 ␮m.

34

Y. Liu et al. / Virus Research 166 (2012) 31–42

Fig. 3. Two putative NLSs were identified within UL94. (A) Schematic representation of the UL94 deletion mutants; (B) EGFPUL94(1–345) in Vero cells; (C) EGFPUL94(1–340) in Vero cells; (D) EGFPUL94(1–330) in Vero cells; (E) EGFPUL94(1–320)M in Vero cells; (F) EGFPUL94(1–310) in Vero cells; (G) EGFPUL94(1–300) in Vero cells; (H) EGFPUL94(1–267) in Vero cells; (I) EGFPUL94(1–200) in Vero cells; (J) EGFPUL94(1–166) in Vero cells; (K) EGFPUL94(1–133) in Vero cells; (L) EGFPUL94(1–100) in Vero cells; (M) EGFPUL94(1–50) in Vero cells; (N) EGFPUL94(51–100) in Vero cells; (O) EGFPUL94(25–40) in Vero cells; (P) EGFPUL94(87–97) in Vero cells; (Q) EGFPUL94(25–40)M in Vero cells; and (R) EGFPUL94(87–97)M in Vero cells. Fluorescence of EGFP fusion proteins is green. Cell nuclei are pseudocolored blue, following staining with Hoechst 33258 dye. Bar = 10 ␮m. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Y. Liu et al. / Virus Research 166 (2012) 31–42

in PBS–2% (w/v) BSA, followed by incubation for 1 h at 37 ◦ C with peroxidase-goat anti-rabbit IgG antibody (BA1054, Boster, China) at a dilution of 1/10,000 in PBS-2% (w/v) BSA. 3. Results 3.1. Nucleocytoplasmic shuttling of UL94 revealed by FRAP and FLIP Localization of UL94 to the nucleus, but with some protein still detectable in the cytoplasm (Liu et al., 2009) implies that it may shuttle between the nucleus and cytoplasm. Here, whether UL94 is able to shuttle in uninfected cells was examined in the absence of additional viral factors. Nucleocytoplasmic shuttling includes nuclear import and subsequent export back to the cytoplasm. This

35

process was tested for UL94 by FRAP experiments in multinucleated cells using the method of previous research (Howell and Truant, 2002; Koster et al., 2005). As shown in Fig. 1A, after completely bleaching EGFPUL94 in one nucleus of a multinuclear cell (ROI 1), re-entry of EGFPUL94 from the non-bleached nucleus was observed for 60 min. Fig. 1B shows the quantitative analysis for the relative fluorescence intensity variation of EGFPUL94 in different regions in Fig. 1A. The increase in relative fluorescence intensity in the bleached nucleus (ROI 1) corresponded to a decrease of EGFPUL94 in the non-bleached nucleus (ROI 2 and ROI 3) and the cytoplasm (ROI 4). This result demonstrates that EGFPUL94 shuttles between the nucleus and cytoplasm. Additional Western blotting was performed to confirm that UL94-fused EGFP were correctly expressed in live cells and that there was no obvious degradation of the EGFP fusion protein (Fig. 1C).

Fig. 4. Amino acids 26–35 and 90–94 within UL94 were defined as NLSs. (A) EGFPUL94(25–40)/K26A in Vero cells; (B) EGFPUL94(25–40)/K30A in Vero cells; (C) EGFPUL94(25–40)/R32A in Vero cells; (D) EGFPUL94(25–40)/K33A in Vero cells; (E) EGFPUL94(25–40)/R35A in Vero cells; (F) EGFPUL94(26–35) in Vero cells; (G) EGFPUL94(87–97)/R90A in Vero cells; (H) EGFPUL94(87–97)/R91A in Vero cells; (I) EGFPUL94(87–97)/R92A in Vero cells; (J) EGFPUL94(87–97)/R93A in Vero cells; (K) EGFPUL94(87–97)/R94A in Vero cells; (L) EGFPUL94(90–94) in Vero cells; (M1) EGFPUL94M1M2 in Vero cells (63%); and (M2) EGFPUL94M1M2 in Vero cells (37%). Fluorescence of EGFP fusion proteins is green. Cell nuclei are pseudocolored blue, following staining with Hoechst 33258 dye. Bar = 10 ␮m. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

36

Y. Liu et al. / Virus Research 166 (2012) 31–42

The FLIP assay was also used to study the exchange of EGFPUL94 between the nuclear and cytoplasmic compartments. As shown in Fig. 2A and C, when nuclear EGFPUL94 was repeatedly bleached, the fluorescence in different regions of the cell was recorded. The result showed that bleaching of EGFPUL94 in a particular nuclear region (ROI 1) resulted in the complete loss of detectable fluorescence in the whole nucleus. Fluorescence in the cytoplasm (ROI 2 and ROI 3) decreased with the fluorescence decrease in the nucleus because of the exchange of molecules between bleached (ROI 1) and non-bleached regions (ROI 2 and ROI 3). The control (ROI 4) for the fluorescence intensity (in a neighboring cell) remained constant during the bleaching procedure showing that loss of fluorescence in the cytoplasm was not caused by cytoplasmic photo-bleaching. The repeated bleaching of EGFPUL94 in a small cytoplasmic region was also carried out and the fluorescence intensity of the bleached cell and that of a neighboring cell control was monitored during the bleaching intervals. As shown in Fig. 2B and D, repeated bleaching of the cytoplasmic ROI resulted in the complete loss of detectable fluorescence throughout the cytoplasm and also caused a decrease in fluorescence intensity within the cell nucleus. These results confirm that the EGFPUL94 protein is able to shuttle between the nucleus and cytoplasm. 3.2. Amino acids 26–35 and 90–94 of UL94 function as NLSs Since UL94 is able to shuttle between the cytoplasm and nucleus, there should be particular amino acids within UL94 that are responsible for its shuttling. To map the key amino acid sequences within UL94 responsible for nuclear import and export, we constructed a panel of UL94 deletion mutants fused to EGFP (Fig. 3A) and analyzed their localization in living cells. The deletion fragments UL94(1–166), UL94(1–200), UL94(1–267), UL94(1–300), UL94(1–310), UL94(1–320), UL94(1–330), UL94(1–340) all localized to the juxtanuclear cytoplasm (Fig. 3C–J), while UL94(1–133) and UL94(1–100) localized to the nucleus (Fig. 3K and L). This result suggests that there should be NLSs within amino acids 1–100 of UL94. Then, the intracellular localization of EGFPUL94(1–50) and EGFPUL94(51–100) were analyzed. Both of them localized mainly to the nucleus, although some cytoplasmic fluorescence was also evident (Fig. 3M and N). Analysis shows that there are two basic amino acid clusters within sequences 25–40 and 87–97 which might be candidates for NLSs. Therefore, EGFPUL94(25–40) and EGFPUL94(87–97) were constructed and expressed in Vero cells. As shown in Fig. 3O and P, EGFPUL94(87–97) was detected primarily in the nucleus. EGFPUL94(25–40) also showed mainly nuclear localization, but had more cytoplasmic fluorescence than EGFPUL94(87–97). The basic amino acids within UL94(25–40) and UL94(87–97) were all mutated to alanine to construct UL94(25–40)M and UL94(87–97)M. When expressed in Vero cells, both were distributed diffusely throughout the cells (Fig. 3Q and R). This localization pattern was similar to the EGFP-only control (Fig. 8D). These results strongly suggest the presence of both NLSs within amino acids 25–40 and 87–97. Point mutation of the basic amino acid within amino acids 25–40 and 87–97 was then performed to evaluate the role of each basic amino acid in the nuclear localization of UL94(25–40) and UL94(87–97). Fluorescence imaging showed that amino acids 26, 30, 32, 33, and 35 were critical for the nuclear localization of EGFPUL94(25–40), amino acids 90, 91, 92, 93 and 94 were crucial for the nuclear localization of EGFPUL94(87–97) (Fig. 4). Then, we further restricted the two NLSs to UL94(26–35) and UL94(90–94). Fluorescence imaging showed that EGFPUL94(26–35) and EGFPUL94(90–94) localized mainly to the nucleus (Fig. 4F and L). Therefore, the NLS1 of UL94 was defined as UL94(26–35) and the NLS2 of UL94 was defined as UL94(90–94). These two NLSs may function separately and neither

Fig. 5. Amino acid 343 of UL94 plays an important role in its nuclear localization. (A) EGFPUL94(1–344) in Vero cells; (B) EGFPUL94(1–343) in Vero cells; (C) EGFPUL94(1–342) in Vero cells; (D) EGFPUL94(1–341) in Vero cells; and (E) EGFPUL94/L343A in Vero cells. Fluorescence of EGFP fusion proteins is green. Cell nuclei are pseudocolored blue, following staining with Hoechst 33258 dye. Bar = 10 ␮m. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

has been reported previously. Comparison of the distribution of EGFPUL94(26–35) with that of EGFPUL94(90–94) in live cells shows that NLS2 is inherently stronger than NLS1. Also, EGFPUL94(1–100) showed more nuclear localization than EGFPUL94(26–35) and EGFPUL94(90–94), indicating that the two signals work additively. 3.3. Amino acid 343 of UL94 plays an important role in its nuclear localization Considering the obvious localization difference between UL94(1–340) and full-length UL94(1–345) (Fig. 3B and C), we also investigated the roles of C-terminal amino acids 341–345. A series of the last five amino acids within the C-terminal deletion mutants of UL94[UL94(1–341), UL94(1–342), UL94(1–343),UL94(1–344)] were fused to EGFP and expressed in living cells. Fluorescence imaging showed that deletion fragments UL94(1–344), UL94(1–343) localized to the nucleus, with the same level of cytoplasmic fluorescence as the full length UL94 (Fig. 5A and B), while UL94(1–342) and UL94(1–341) localized to the juxtanuclear cytoplasm (Fig. 5C and D). We then constructed a mutant of UL94 containing single-point mutation (UL94/L343A), fused it to EGFP (EGFPUL94/L343A) and expressed it in Vero cells. As shown in Fig. 5E, EGFPUL94/L343A localized to the juxtanuclear cytoplasm. These results show that

Y. Liu et al. / Virus Research 166 (2012) 31–42

37

Fig. 6. Mutation of nuclear localization-related sequences of UL94 disrupts its shuttling function. (A) Serial FLIP images of EGFPUL94M1M2/L343A in a nuclear FLIP experiment and (B) curves showing the fluorescence intensity of EGFPUL94M1M2/L343A in different regions in a nuclear FLIP experiment (the bleached area (ROI 1), one region in the near cytoplasm (ROI 2), one region in the distant cytoplasm (ROI 3) and a non-bleached cell (ROI 4)). Fluorescence of EGFP fusion proteins is green. The monitored regions are indicated by white arrowheads in the first image. Bar = 10 ␮m. (C) Western blot of EGFPUL94M1M2/L343A and EGFPUL94. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

amino acid 343 of UL94 plays an important role in its nuclear localization. 3.4. Mutation of nuclear localization-related sequences alters the intracellular distribution of UL94 and disrupts its nucleocytoplasmic shuttling To elucidate the function of NLSs within the full-length UL94 protein, some UL94 mutants including EGFPUL94M1M2 (all basic amino acids within fragments 26–35 and 90–94 were mutated to alanine) and EGFPUL94M1M2/L343A (all basic amino acids within fragment 26–35, 90–94 and the leucine 343 were mutated to alanine) were constructed and transiently expressed in Vero cells. EGFPUl94M1M2 completely localized to the perinuclear cytoplasm in 63% of the cells (108/172) (Fig. 4M1) and localized to both the nucleus and the cytoplasm in 37% of cells (Fig. 4M2). No fluorescence was observed in the nucleolus. EGFPUl94M1M2/L343A completely localized in the cytoplasm (Fig. 6A). These results demonstrate that mutation of NLS1 and NLS2 abrogated the nucleolus localization of UL94 and decreased its nuclear distribution and mutation of NLS1 and NLS2 plus amino acid 343 can completely change the intracellular distribution of UL94 from the nucleus to cytoplasm. However, whether these mutations impair the shuttling function of UL94 required investigation. FLIP assay was carried out to study the exchange of EGFPUL94M1M2/L343A between the nuclear cytoplasmic compartments. As shown in Fig. 6, repeated

bleaching of EGFPUL94M1M2/L343A in a particular nuclear region (Fig. 6A) was carried out and the fluorescence intensity of the bleached cells and that of a neighboring cell control were monitored during the bleaching intervals. Bleaching of a particular nuclear region (ROI 1) resulted in the complete loss of detectable fluorescence in the whole nucleus while fluorescence in the cytoplasm (ROI 2 and ROI 3) and the control (ROI 4) remained constant during the bleaching procedure, showing that UL94M1M2/L343A could not import to the nucleus (Fig. 6A and B) anymore. Additional Western blotting was performed to confirm that EGFPUL94M1M2/L343A correctly expressed in live cells and that there was no obvious degradation of the EGFP fusion protein (Fig. 6C). Therefore, the mutation of two NLSs and amino acid 343 alters the intracellular distribution of UL94 and disrupts its nucleocytoplasmic shuttling. 3.5. Amino acid 343 of UL94 is found to be related to its interaction with pp28 Our previous research showed that UL94 was a specific binding partner for pp28. Here we further investigated whether the new found localization signal sequences are related to UL94 interaction with pp28. UL94 mutants with the 343 point mutation were found to no longer colocalize specifically with pp28 in their co-expressed cells. As shown in Fig. 7, UL94/L343A-EYFP and pp28-ECFP both localized in the cell cytoplasm but had no obvious colocalization signal (Fig. 7A) compared with UL94-EYFP and pp28-ECFP

38

Y. Liu et al. / Virus Research 166 (2012) 31–42

Fig. 7. Amino acid 343 of UL94 plays a key role in its interaction with pp28. (A) EGFPUL94/L343A and pp28 coexpressed in Vero cells and (B) EGFPUL94 and pp28 coexpressed in Vero cells. Pp28-ECFP fluorescence is cyan and EYFP fusion protein fluorescence is yellow. Cell nuclei are pseudocolored blue, following staining with Hoechst 33258 dye. Bar = 10 ␮m. (C) A typical FRET result of the interaction between UL94/L343A and pp28; (D) a typical FRET result of the interaction between UL94 and pp28; and (E) a typical FRET result of the interaction between EYFP and ECFP (negative control). ROI 1 is the area acceptor photobleached. Dpre and Apre are fluorescence images of donor and acceptor before acceptor photobleaching, respectively. Dpost and Apost are fluorescence images of donor and acceptor after acceptor photobleaching, respectively. The FRET efficiency (FRETeff ) of ROI 1 was calculated as FRETeff = [Dpost − Dpre ]/Dpost . (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

(Fig. 7B). The acceptor photobleaching protocol for FRET analysis was also performed to detect interactions between pp28 and some UL94 mutants in living cells (Karpova et al., 2003). Fig. 7C showed a typical FRET assay between pp28-ECFP and UL94/L343AEYFP. The images and quantification analysis showed that an increase in the fluorescence intensity in the CFP channel could not be detected concomitant with a decrease in YFP fluorescence. The significant mean FRETeff value between UL94/L343A-EYFP and pp28-ECFP was calculated as 0.24 ± 0.13% (n = 30), which was close to the negative control FRETeff value (0.33 ± 0.09%; n = 25) (Fig. 7E) and significantly lower than the value between UL94EYFP and pp28-ECFP (13.43 ± 0.88%; n = 23) (Fig. 7D). No interaction between UL94 mutant and pp28 was detected, demonstrating that amino acid 343 of UL94 plays an important role in its interaction with pp28. 3.6. Amino acids 55–64 of UL94 function as a nuclear export sequence From the intracellular localization results obtained for the deletion mutants, two clues about UL94 nuclear export were found. First, deletion mutant UL94(1–166) was only detected within the cytoplasm of transfected cells, while UL94(1–133) showed strict nuclear localization. This suggested that the region of UL94 required for nuclear export can be mapped to amino acids 134–166. Second, UL94(1–50) localized to the nucleolus, while UL94(51–100) did not. There are several leucine residues within amino acids 55–64 of UL94, implying that there should also be a nucleolar export signal (NoES) or NES within amino acids 51–100 of UL94. Thus, two potential nuclear export sequences were suggested within the amino end of pUL94 (residues 1–166), and might be in sequences 51–100 and 134–166, respectively. To investigate this, EGFPUL94(134–166) and EGFPUL94(55–64) were expressed in Vero cells. As shown in Fig. 8, EGFPUL94(55–64) localized to the cytoplasm (Fig. 8B), while EGFPUL94(134–166) was

Fig. 8. Amino acids 55–64 of UL94 function as a nuclear export sequence. (A) EGFPUL94(134–166) in Vero cells; (B) EGFPUL94(55–64) in Vero cells; (C) EGFPUL94(55–64)M in Vero cells; and (D) EGFP-only control in Vero cells. Fluorescence of EGFP fusion proteins is green. Cell nuclei are pseudocolored blue, following staining with Hoechst 33258 dye. Bar = 10 ␮m. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Y. Liu et al. / Virus Research 166 (2012) 31–42

39

Fig. 9. Amino acids 55–64 within UL94 are defined as NES. (A) EGFPUL94(55–64)/I56A in Vero cells; (B) EGFPUL94(55–64)/L57A in Vero cells; (C) EGFPUL94(55–64)/L60A in Vero cells; (D) EGFPUL94(55–64)/L61A in Vero cells; (E) EGFPUL94(55–64)/L62A in Vero cells; (F) EGFPUL94(55–64)/L63A in Vero cells; (G) EGFPUL94(55–64)/I56A+L57A in Vero cells; (H) EGFPUL94(55–64)/L60A+L61A in Vero cells; (I) EGFPUL94(55–64)/L61A+L62A in Vero cells; (J) EGFPUL94(55–64)/L62A+L63A in Vero cells; (K) EGFPUL94(55–64)/L61A+L63A in Vero cells; and (L) EGFPUL94/L61A+L63A in Vero cells. Fluorescence of EGFP fusion proteins is green. Cell nuclei are pseudocolored blue, following staining with Hoechst 33258 dye. Bar = 10 ␮m. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

distributed diffusely throughout the cells (Fig. 8A). The isoleucine or leucine residues within UL94(55–64) were all mutated to alanine to construct UL94(55–64)M and fused to EGFP. When EGFPUL94(55–64)M expressed in Vero cells, it was distributed diffusely throughout the cells (Fig. 8C), suggesting that UL94(55–64) is a leucine-rich NES and those leucine residues play a critical role in its nuclear export. Point mutation of the isoleucine or leucine within amino acids 55–64 was then performed to further evaluate the role of these residues in the nuclear export of UL94(55–64). Fluorescence imaging showed that acids 61 and 63 are critical for nuclear export of EGFPUL94(55–64) (Fig. 9). 3.7. Mutation of NES within UL94 also alters its intracellular distribution and shuttling function The function of newly characterized NES within the fulllength UL94 protein was also investigated, the UL94 mutant EGFPUL94/L61A+L63A (amino acid 61 and 63 were mutated to alanine), were also constructed and transiently expressed in Vero cells. The level of cytoplasmic fluorescence of EGFPUL94/L61A+L63A was obviously decreased comparing with that of the EGFPUL94

(Fig. 9L). The exchange of EGFPUL94/L61A+L63A between the nuclear and cytoplasmic compartments was also analyzed by FLIP assay. As shown in Fig. 10, EGFPUL94/L61A+L63A could import to the nucleus (Fig. 10A and C) but could not export from the nucleus (Fig. 10B and D) any more. Additional Western blotting was performed to confirm that EGFPUL94/L61A+L63A correctly expressed in live cells and that there was no obvious degradation of the EGFP fusion protein (Fig. 10E). Therefore, the mutation of NES within UL94 disrupted its intracellular distribution and shuttling function. 3.8. The NES of UL94 exports via the classical Crm1-dependent pathway Previous studies have shown that Crm1 is the export receptor for the majority of leucine-rich NES proteins (Fornerod et al., 1997; Fukuda et al., 1997; Stade et al., 1997), such as the NES of the HIV-1 Rev protein, and that Leptomycin B (LMB) prevents its association with the NES substrate without affecting other known nuclear transport pathways, thereby inhibiting the nuclear export of leucine-rich NES-containing proteins (Kudo et al., 1999; Wolff

40

Y. Liu et al. / Virus Research 166 (2012) 31–42

Fig. 10. Mutation of NES of UL94 disrupts its shuttling function. (A) Serial FLIP images of EGFPUL94/L61A+L63A in a nuclear FLIP experiment; (B) serial FLIP images of EGFPUL94/L61A+L63A in a cytoplasmic FLIP experiment; (C) curves showing the fluorescence intensity of EGFPUL94/L61A+L63A in different regions in a nuclear FLIP experiment (the bleached area (ROI 1), one region in the near cytoplasm (ROI 2), one region in the distant cytoplasm (ROI 3) and a non-bleached cell (ROI 4)); and (D) curves showing the fluorescence intensity of EGFPUL94/L61A+L63A in different regions in a cytoplasmic FLIP experiment (the bleached area (ROI 1), a distant cytoplasmic region (ROI 2), the nuclear area (ROI 3) and a non-bleached cell (ROI 4)). Fluorescence of EGFP fusion proteins is green. The monitored regions are indicated by white arrowheads in the first image. Bar = 10 ␮m. (E) Western blot of EGFPUL94/L61A+L63A and EGFPUL94. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

et al., 1997). To address the question of whether the nuclear export sequence of UL94 is dependent on Crm1, Vero cells expressing EGFPUL94(55–64), or EYFPRevNES (control) were both treated with LMB at 10 ng/ml. As shown in Fig. 11A, prior to LMB treatment, EYFPRevNES localized only to the cytoplasm. However, after treatment with LMB, EYFPRevNES was distributed throughout the cell (Fig. 11B). Similar to EYFPRevNES, EGFPUL94(55–64) localized only to cytoplasm before treatment with LMB (Fig. 11C); however, after LMB treatment, EGFPUL94(55–64) was detected primarily in the nucleus (Fig. 11D). To investigate if the NES within the context of full-length UL94 can respond to LMB, Vero cells expressing EGFPUL94 were treated with LMB too. Prior to LMB treatment, EGFPUL94 distributed mainly in the nucleus, with obvious fluorescence in the cytoplasm (Fig. 11E). The level of cytoplasmic fluorescence of EGFPUL94 was apparently decreased after treatment with LMB (Fig. 11F). So, like many other proteins containing a leucine-rich export signal, UL94 should be exported out of the nucleus in the Crm1-dependent pathway. 4. Discussion The function of human cytomegalovirus tegument protein UL94 remains to be determined. Our previous study revealed that UL94, predominantly localized to the nucleus, could be directed into the cellular cytoplasm by pp28 to co-localize with pp28, suggesting

that the cellular localization of UL94 likely plays a role in HCMV viral assembly. In the current study, using FRAP and FLIP imaging, we demonstrated that UL94 can shuttle between the nucleus and the cytoplasm. As schematically represented in Fig. 12, two newly characterized NLSs both containing basic amino acid clusters, and amino acid 343 were found to be responsible for the nuclear localization of UL94. Point mutation analysis was also performed to evaluate the role of basic amino acids within the two NLSs in nuclear localization of UL94, showing that residues 26, 30, 32, 33 and 35 were the key amino acids responsible for the localization of EGFPUL94(25–40) to the nucleus, and that amino acids 90, 91, 92, 93 and 94 were crucial for nuclear localization of EGFPUL94(87–97). Therefore, the NLS1 of UL94 was defined as UL94(26–35) and the NLS2 of UL94 was defined as UL94(90–94). The residues responsible for NLSs found in UL94 are in accordance with the analysis from PSORTII program (University of Tokyo, Japan). In fact, comparing the distribution of these point mutations with EGFPUL94(25–40) and EGFPUL94(87–97) (Fig. 4), these residues are not only responsible for the localization of EGFPUL94(25–40) and EGFPUL94(87–97) to the nucleus but also to the nucleolus. Mutation of NLS1 and NLS2 resulted in the export of UL94 from the nucleolus in all cells and from the nucleus to the cytoplasm in approximately 60% of cells. These results demonstrate that mutation of NLS1 and NLS2 abrogated the nucleolus localization of UL94 and decreased its nuclear

Y. Liu et al. / Virus Research 166 (2012) 31–42

41

Fig. 11. NES of UL94 is sensitive to LMB. (A) Representative fluorescence images of living cells expressing RevNES-EYFP; (B) representative fluorescence images of living cells expressing RevNES-EYFP with LMB (10 ng/ml) processing; (C) representative fluorescence images of living cells expressing EGFPUL94NES; (D) representative fluorescence images of living cells expressing EGFPUL94NES with LMB (10 ng/ml) processing; (E) representative fluorescence images of living cells expressing EGFPUL94; and (F) representative fluorescence images of living cells expressing EGFPUL94 with LMB (10 ng/ml) processing. EGFPUL94NES and EGFPUL94 fluorescence is green and RevNES-EYFP fluorescence is yellow. Cell nuclei are pseudocolored blue, following staining with Hoechst 33258 dye. Bar = 10 ␮m. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 12. Mapping of the localization-related sequences within UL94. Schematic diagram of the amino acid sequences of NLS1, NLS2, NES, and Leu343 within UL94.

distribution, suggesting that the two NLSs may also function in its nucleolus localization. Interestingly, mutation of amino acid 343 can completely change the intracellular distribution of UL94 from the nucleus to cytoplasm. UL94 mutants with a 343 point mutation were also unable to co-localize with pp28 and had no interaction with pp28. It seems that amino acid 343 is more important than the two NLSs for nuclear localization of UL94. It is possible that the efficiency of the two NLSs within full-length UL94 may decrease because of inter/intramolecular masking. Another possible explanation may be that the localization of UL94 may have something to do with the integrity of its 3D structure in which amino acid 343 is highly involved. Amino acid sequence 55–64 functions as a NES in UL94 and translocates EGFP from the nucleus to cytoplasm. Its nuclear export function can be inhibited by LMB, suggesting that the UL94 protein is exported out of the nucleus via the classical Crm1-dependent pathway. EGFPUL94(1–133) completely localized to the nucleus while EGFPUL94(1–166) localized to the juxtanuclear cytoplasm. However, UL94(134–166) did not translocate EGFP from the nucleus to cytoplasm. Therefore, it is probable that UL94(134–166) functions to regulate the NLS or NES. UL94 may have a similar function to UL16 of HSV-1, playing a role in capsid maturation (Ward et al., 1996). Our previous results showed that UL94 and pp28 interacts in the viral assembly compartment (AC) and this interaction may serve as a bridge connecting capsid and envelope. It is possible that UL94 is first packaged to the viral capsid in the nucleus and then exported out of the nucleus along with the partially tegumented virus particles, which may then acquire other tegument proteins in the AC in the cytoplasm. The shuttling of UL94 likely plays a key role during this process. The

NLS(s) and amino acid 343 function in the nuclear entry of UL94 and give UL94 the chance to be packaged to the viral capsid. The NES directs UL94 nuclear export to enable interaction with other tegument proteins such as pp28 and participates in viral assembly. While beyond the scope of this study, future studies will be conducted to verify and clarify these sequential events. 5. Conclusion Using live cell FRAP and FLIP assays, we have discovered the nucleocytoplasmic shuttling of the HCMV UL94 protein. Two NLSs and one NES were identified within UL94 to be responsible for its nuclear localization and nuclear export respectively. Amino acid 343 of UL94 also played an important role for the UL94 nuclear localization and its interaction with HCMV pp28. Mutation of these localization-related sequences within UL94 altered its intracellular distribution and shuttling function. In addition, it was found that the nuclear export of the UL94 is mediated by the Crm1 pathway. These results provide new insights into HCMV UL94 which may play important roles in the virus assembly. Acknowledgments Z.Q. Cui and Z.P. Zhang were supported by the National Basic Research Program of China 2011 CB933600 and the Knowledge Innovation Program of the Chinese Academy of Sciences KSCX2EW-Q-15. The others were supported by Chinese Academy of Sciences (KJCX2-YW-M15).

42

Y. Liu et al. / Virus Research 166 (2012) 31–42

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.virusres.2012.02.023. References Baldick, C.J., Shenk, T., 1996. Proteins associated with purified human cytomegalovirus particles. J. Virol. 70 (9), 6097–6105. Britt, W.J., Jarvis, M., Seo, J.Y., Drummond, D., Nelson, J., 2004. Rapid genetic engineering of human cytomegalovirus by using a lambda phage linear recombination system: demonstration that pp28 (UL99) is essential for production of infectious virus. J. Virol. 78 (1), 539–543. Chee, M., 1991. The HCMV genome project: what has been learned and what can be expected in the future. Transplant. Proc. 23 (3 Suppl. 3), 174–180 (discussion 180). Chee, M.S., Bankier, A.T., Beck, S., Bohni, R., Brown, C.M., Cerny, R., Horsnell, T., Hutchison 3rd, C.A., Kouzarides, T., Martignetti, J.A., et al., 1990. Analysis of the protein-coding content of the sequence of human cytomegalovirus strain AD169. Curr. Top. Microbiol. Immunol. 154, 125–169. Cole, N.B., Smith, C.L., Sciaky, N., Terasaki, M., Edidin, M., Lippincott-Schwartz, J., 1996. Diffusional mobility of Golgi proteins in membranes of living cells. Science 273 (5276), 797–801. Fornerod, M., Ohno, M., Yoshida, M., Mattaj, I.W., 1997. Crm1 is an export receptor for leucine-rich nuclear export signals. Cell 90 (6), 1051–1060. Fukuda, M., Asano, S., Nakamura, T., Adachi, M., Yoshida, M., Yanagida, M., Nishida, E., 1997. Crm1 is responsible for intracellular transport mediated by the nuclear export signal. Nature 390 (6657), 308–311. Gibson, W., 1996. Structure and assembly of the virion. Intervirology 39 (5-6), 389–400. Higgins, D.G., Sharp, P.M., 1989. Fast and sensitive multiple sequence alignments on a microcomputer. Comput. Appl. Biosci. 5 (2), 151–153. Howell, J.L., Truant, R., 2002. Live-cell nucleocytoplasmic protein shuttle assay utilizing laser confocal microscopy and FRAP. Biotechniques 32 (1), 80–82, 84, 86-87. Sambrook, J., Russell, D.W., 2001. Molecular Cloning: A laboratory manual, 3rd ed. Cold Spring Harbor Laboratory Press, New York. Jones, T.R., Lee, S.W., 2004. An acidic cluster of human cytomegalovirus UL99 tegument protein is required for trafficking and function. J. Virol. 78 (3), 1488–1502. Karpova, T.S., Baumann, C.T., He, L., Wu, X., Grammer, A., Lipsky, P., Hager, G.L., McNally, J.G., 2003. Fluorescence resonance energy transfer from cyan to yellow fluorescent protein detected by acceptor photobleaching using confocal microscopy and a single laser. J. Microsc. Oxford 209, 56–70. Koster, M., Frahm, T., Hauser, H., 2005. Nucleocytoplasmic shuttling revealed by FRAP and FLIP technologies. Curr. Opin. Biotechnol. 16 (1), 28–34.

Kudo, N., Matsumori, N., Taoka, H., Fujiwara, D., Schreiner, E.P., Wolff, B., Yoshida, M., Horinouchi, S., 1999. Leptomycin B inactivates Crm1/exportin1 by covalent modification at a cysteine residue in the central conserved region. Proc. Natl. Acad. Sci. U.S.A. 96 (16), 9112–9117. Lippincott-Schwartz, J., Snapp, E., Kenworthy, A., 2001. Studying protein dynamics in living cells. Nat. Rev. Mol. Cell Biol. 2 (6), 444–456. Liu, Y.L., Cui, Z.Q., Zhang, Z.P., Wei, H.P., Zhou, Y.F., Wang, M.L., Zhang, X.E., 2009. The tegument protein UL94 of human cytomegalovirus as a binding partner for tegument protein pp28 identified by intracellular imaging. Virology 388 (1), 68–77. Maninger, S., Bosse, J.B., Lemnitzer, F., Pogoda, M., Mohr, C.A., von Einem, J., Walther, P., Koszinowski, U.H., Ruzsics, Z., 2011. M94 is essential for the secondary envelopment of murine cytomegalovirus. J. Virol. 85 (18), 9254–9267. Mocarski, E.S., Courcelle, C.T., 2001. Cytomegaloviruses and their replication. In: Knipe, D.M., Howley, P.M., Griffin, D.E., Lamp, R.A., Martin, M.A., Roizman, B., Straus, S.E. (Eds.), Fields Virology. , 4th ed. Lippincott-Raven, pp. 2629–2673. Nehls, S., Snapp, E.L., Cole, N.B., Zaal, K.J., Kenworthy, A.K., Roberts, T.H., Ellenberg, J., Presley, J.F., Siggia, E., Lippincott-Schwartz, J., 2000. Dynamics and retention of misfolded proteins in native ER membranes. Nat. Cell Biol. 2 (5), 288–295. Pass, R.F., 2001. Cytomegaloviruses. In: Knipe, D.M., Howley, P.M., Griffin, D.E., Lamp, R.A., Martin, M.A., Roizman, B., Straus, S.E. (Eds.), Fields Virology. , 4th ed. Lippincott-Raven, pp. 2675–2705. Phair, R.D., Misteli, T., 2000. High mobility of proteins in the mammalian cell nucleus. Nature 404 (6778), 604–609. Phillips, S.L., Bresnahan, W.A., 2011. The human cytomegalovirus tegument protein UL94 is essential for secondary envelopment of HCMV virions. J. Virol. Silva, M.C., Yu, Q.C., Enquist, L., Shenk, T., 2003. Human cytomegalovirus UL99-encoded pp28 is required for the cytoplasmic envelopment of tegumentassociated capsids. J. Virol. 77 (19), 10594–10605. Stade, K., Ford, C.S., Guthrie, C., Weis, K., 1997. Exportin 1 (Crm1p) is an essential nuclear export factor. Cell 90 (6), 1041–1050. To, A., Bai, Y., Shen, A., Gong, H., Umamoto, S., Lu, S., Liu, F., 2011. Yeast two hybrid analyses reveal novel binary interactions between human cytomegalovirusencoded virion proteins. PLoS One 6 (4), e17796. Ward, P.L., Ogle, W.O., Roizman, B., 1996. Assemblons: nuclear structures defined by aggregation of immature capsids and some tegument proteins of herpes simplex virus 1. J. Virol. 70 (7), 4623–4631. Wing, B.A., Huang, E.S., 1995. Analysis and mapping of a family of 3 -coterminal transcripts containing coding sequences for human cytomegalovirus open reading frames UL93 through UL99. J. Virol. 69 (3), 1521–1531. Wing, B.A., Lee, G.C.Y., Huang, E.S., 1996. The human cytomegalovirus UL94 open reading frame encodes a conserved herpesvirus capsid/tegument-associated virion protein that is expressed with true late kinetics. J. Virol. 70 (6), 3339–3345. Wolff, B., Sanglier, J.J., Wang, Y., 1997. Leptomycin B is an inhibitor of nuclear export: Inhibition of nucleo-cytoplasmic translocation of the human immunodeficiency virus type 1 (HIV-1) Rev protein and Rev-dependent mRNA. Chem. Biol. 4 (2), 139–147.