A Potential α-Helix Motif in the Amino Terminus of LANA Encoded by Kaposi's Sarcoma-Associated Herpesvirus Is Critical for Nuclear Accumulation of HIF-1α in Normoxia^▿

DNA constructs.

Plasmids pDSRed/LANA (LANA-RFP), pA3M/LANA1-329^925-1162 (LN-1), pA3M/LANA1-340 (LN-2), and pA3M were described previously (3, 37). Plasmids pA3M/LANA1-233 (LN-3), pA3M/LANA46-145 (LN-4), pA3M/LANA89-233 (LN-5), pA3M/LANA1-46^<942>129-233 (LN-6), and pA3M/LANA1-233H47AC57A (LN-7) were generated by PCR amplicons subcloned into the pA3M vector by KpnI/EcoRV digestion. Construct pEGFP/HIF-1α was prepared by PCR amplification of the ORF from pCEP4/HIF-1α (26). The amplicon was digested with BamHI/KpnI and cloned into the BglII and KpnI sites of pEGFP-C1 (Clontech, Inc., Mountain View, CA). pEGFP/HIF-1α 1-330(N) was generated by pEGFP/HIF-1α digestion with EcoRI and the self-ligation of the large fragment. pEGFP/HIF-1α 1-3005̂30-826(NC) was obtained from the large fragment of the EcoRI-digested pEGFP/HIF-1α, and its ends were filled by Klenow enzyme followed by blunt-end ligation. pEGFP/HIF-1α 300-530(M) was generated by PCR from pCEP4/HIF-1α, and then the amplicon was digested with BamHI/EcoRI cloned into the BglII and EcoRI sites of pEGFP-C1. All clones were confirmed by DNA sequencing. pCEP4/HIF-1α was provided by Gregg L. Semenza of Johns Hopkins University (26). The wild-type hypoxia response element (wHRE) consisted of a trimerized 24-mer containing 18 bp of sequence from the PGK promoter, including the HRE (5-TGTCACGTCCTGCACGACTCTAGT [HRE is underlined]), and its mutant (mHRE) had the ACG of the HIF-1 binding site mutated to CAT, abolishing binding, in the pGL2-Basic vector provided by Craig B. Thompson (University of Pennsylvania School of Medicine). pGL3-1.5VEGF plasmid containing the 47-bp HRE and pGL3-1.2VEGF with HRE deletion were provided by Amit Maity (University of Pennsylvania School of Medicine, Philadelphia) (22).

Cell cultures and transfection.

KSHV-positive type cells (BCBL-1 and BC-3) and KSHV-negative type cells (BJAB and DG75) were maintained in RPMI 1640 medium with 7% fetal bovine serum (FBS), 4 μM l-glutamine, penicillin, and streptomycin. The renal carcinoma VHL-null cell line 786-O (provided by Volker H. Haase from the University of Pennsylvania School of Medicine), the human osteosarcoma p53-null cell line Saos-2 (provided by Jon Aster from Brigham and Women's Hospital, Boston, MA), and human embryonic kidney 293 (HEK293) and U2OS cells (from the American Type Culture Collection) were maintained in Dulbecco modified Eagle medium (DMEM) supplemented with 5% FBS, 4 μM l-glutamine, penicillin, and streptomycin. All cells were incubated at 37°C in a humidified environmental incubator supplemented with 5% CO₂. Ten million cells with 400 μl of medium were transfected by electroporation with a Bio-Rad Gene Pulser in 0.4-cm-gap cuvettes at 220 V and 975 μF. To create a hypoxic environment, cells were allowed to grow for 24 h with 100 μM CoCl₂ treatment before use (31).

Immunofluorescence assays.

Cells transfected on coverslips were fixed in 3% paraformaldehyde for 10 min at 4°C at 24 h posttransfection. Antibody working dilutions were as follows: mouse anti-HIF-1α (BD Transduction Laboratory, San Jose, CA) and anti-myc (9E10) or rabbit anti-LANA polyclonal antibodies (kindly provided by Bala Chandran of Rosalind Franklin University of Medicine and Science, North Chicago, IL) at 1:50. Cells were permeabilized in immunofluorescence buffer (0.2% fish skin gelatin and 0.2% Triton X-100 in phosphate-buffered saline [PBS]) for 5 min, followed by 1 h of incubation in immunofluorescence buffer with primary antibodies. Cells were washed three times in PBS. Fluorescence-labeled Alexa Fluor 488 or 594 anti-mouse or anti-rabbit antibodies (Molecular Probes, Inc., Eugene, OR) were used as secondary antibodies for 30 min. DAPI (4′,6′-diamidino-2-phenylindole at 0.5 μM was used for nuclear staining (Pierce, Inc., Rockford, IL), and coverslips were washed in PBS and mounted on glass slides using Prolong antifade mounting medium (Molecular Probes). All steps were performed at room temperature. Fluorescence confocal microscopy was performed with an Olympus microscope using FluoView FV300 software (Olympus, Inc., Melville, NY).

Fractionation of nuclear or cytoplasm proteins.

Twenty million transfected cells were harvested and washed twice with ice-cold PBS, followed by resuspension of the cell pellet in hypotonic buffer A (10 mM HEPES-K⁺ [pH 7.5], 10 mM KCl, 1.5 mM MgCl₂, 0.5 dithiothreitol) in the presence of protease inhibitor cocktail (PIC; 1 mM phenylmethylsulfonyl fluoride, 10 μg of aprotinin/ml, 10 μg of leupeptin/ml, 10 μg of pepstatin A/ml, 10 μg of phenanthroline/ml, 16 μg of benzamidine/ml). Cells were pelleted by spinning them at 1,000 rpm 5 min. The cells were lysed in ice-cold 0.5% NP-40 containing buffer A with PIC on ice for 10 min. The nuclei were pelleted by centrifugation at 3,000 rpm for 2 min at 4°C. The supernatant (cytoplasm protein) was harvested and frozen at −80°C for use. The nuclear pellets were washed with buffer A (without NP-40), followed by resuspension in buffer C (20 mM HEPES-K⁺ [pH 7.9], 420 mM NaCl, 0.2 mM EDTA, 1.5 mM MgCl₂, 0.5 dithiothreitol, 25% glycerol) with PIC. Nuclei were incubated on ice for 30 min and vortex mixed periodically. Supernatants containing nuclear protein were collected by spinning at 14,500 rpm for 10 min at 4°C and then snap-frozen for further use. Antibodies to nuclear protein Sp1 (1C6; Santa Cruz, Inc., Santa Cruz, CA) and cytoplasm protein Hsp70 (BD Transduction Laboratory) were used as markers.

Immunoprecipitation and immunoblotting.

For immunoprecipitation, B-cell lysates were extracted by using radioimmunoprecipitation assay buffer (50 mM Tris [pH 7.6], 150 mM NaCl, 2 mM EDTA, 1% Nonidet P-40, 1 mM phenylmethylsulfonyl fluoride, aprotinin [1 μg/ml], pepstatin [1 μg/ml], 25 mM N-ethylmaleimide). The protein concentration was determined by using a Bio-Rad protein assay. After one-step preclearing (rotation 1 h, 4°C) with protein A/G (50/50) Sepharose Fast-Flow (Amersham Biosciences, Inc., Piscataway, NJ), the antibodies were added to the cell protein extract in a binding buffer adjusted to 20 mM Tris (pH 7.5)-200 mM NaCl-0.1% Nonidet P-40. After overnight incubation, immunocomplexes were recovered with protein A/G (50/50) Sepharose after 1 h of incubation. After three washes with binding buffer, the proteins were eluted in sample buffer.

For immunoblotting, proteins were resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to polyvinylidene fluoride membrane. After a blocking step with 5% nonfat dry milk in PBS (2 mM KCl, 120 mM NaCl, 1.5 mM KH₂PO₄, 8 mM Na₂HPO₄) containing 0.1% Tween 20, the membrane was incubated with primary antibodies overnight at 4°C. The next day, the appropriate infrared-conjugated secondary antibodies (Alexa Fluor 800 or 680) were incubated for 1 h at room temperature, followed by scanning with a Li-Cor Odyssey scanner (Li-Cor, Inc., Lincoln, NE). The blot was washed three times with TBST (25 mM Tris-HCl [pH 8.0], 125 mm NaCl, 0.1% Tween 20), and each wash was for an interval of 5 min.

RNA interference.

The small hairpin RNAs complementary to the C-terminal (GCTAGGCCACAACACATCT) fragment of LANA as described previously (11) were cloned into the pSIREN vector according to the instructions of manufacture (BD Clontech) to generate sh-LANA (2). pSIREN vector with luciferase target sequence (sh-Luc) was used as a control. Ten million BCBL-1 cells were transfected by electroporation with 5 μg of sh-LANA or sh-Luc. BCBL-1 stable cells knockdown for LANA were selected and maintained in 4 μg of puromycin/ml. Cells were fixed in 3% paraformaldehyde, and an immunofluorescence assay was performed as described previously (3).

Infection of HEK293 cells with KSHV virus and induction for virus production.

HEK293 cells grown at 60 to 80% confluence in 100-mm-diameter tissue culture dishes were infected with the same amount of concentrated virus from an identical number of BCBL-1 cells. Virus produced from approximately 20 million BCBL-1 cells was used for infecting a single 100-mm culture dish in the presence of Polybrene (Sigma, St. Louis, MO) at different time points.

For virus induction, five hundred million exponentially growing BCBL-1 cells were induced with 20 ng of tetradecanoyl phorbol acetate/ml and 1.5 mM sodium butyrate (Sigma-Aldrich, St. Louis, MO) for 5 days at 37°C with 5% CO₂. The supernatant was collected and filtered through a 0.45-μm-pore-size filter, and viral particles were spun down at 25,000 rpm for 2 h at 4°C. The concentrated virus was collected and used for infection experiments.

Determination of HIF-1α stability.

HEK293 cells were cotransfected pCEP4/HIF-1α with or without pA3M/LANA as described above. After 24 h of transfection, the cells were incubated with 100 μM cycloheximide (C4859; Sigma) for 0 to 4 h and then harvested. Equal amounts of total proteins from each treatment were taken to perform Western blotting or immunoprecipitation.

VEGF expression.

VEGF expression in the cell supernatant was monitored by using a Quantikine human VEGF immunoassay kit (R&D Systems, Minneapolis, MN). Briefly, when cells were 70 to 80% confluent (2 × 10⁶ cells/well), they were washed twice in DMEM and further incubated in phenol red-free DMEM supplemented with 5% FBS at 37°C. After 24 h of incubation, supernatants were collected in 1.5-ml vials, spun at 4°C to remove the particulates, and then stored at −80°C. The supernatant (200 μl) was thawed and used to test VEGF expression by performing an enzyme-linked immunosorbent assay (ELISA) according to the manufacturer's recommendations.

Luciferase assay.

Ten million cells were transfected with luciferase reporter (5 μg) combined with specific expression constructs. The transfected cells were harvested at 48 h posttransfection and subsequently washed once with PBS (HyClone, Inc., Logan, UT), followed by lysis with 200 μl of reporter lysis buffer (Promega, Inc., Madison, WI). Then, 40 μg of total protein of the lysate was mixed with 100 μl of the luciferase assay reagent. Luminescence was measured for 10 s with an OptiComp luminometer (MGM Instruments, Inc., Hamden, CT). The luciferase activity was normalized with cotransfected β-galactosidase activity (i.e., the optical density at 450 nm), which was driven by the cytomegalovirus promoter. The relative luciferase activity was expressed as the fold activation relative to the reporter construct alone. Assays were performed in triplicate.

HIF-1α is accumulated in KSHV-infected cells.

Previously, we observed an interaction between the key KSHV latent antigen LANA and HIF-1α, which led to virus lytic replication under hypoxic conditions (3). More recently, we saw that both the HIF-1α suppressors VHL and p53 were downregulated by LANA under normoxic conditions (2). To explore this potential effect of LANA on HIF-1α accumulation in KSHV latent infection, we compared the expression levels of HIF-1α in two KSHV-positive (BCBL-1 and BC-3) PEL cells and two KSHV-negative (BJAB and DG75) B-lymphocyte cell lines by immunoprecipitation and immunoblot assays targeting HIF-1α. The results clearly showed that HIF-1α levels were dramatically elevated in the KSHV positive BCBL-1 cells and slightly increased in the BC-3 cells (Fig. 1A). To demonstrate the role of LANA as an inducer of HIF-1α expression during the KSHV latent infection, we performed KSHV latent infection in vitro using the susceptible cell line HEK293. The data from Western blot analysis of infected 293 cells at 12, 24, and 36 h postinfection by KSHV showed that increased HIF-1α expression was consistent with increased expression of LANA compared to cells uninfected at the zero time point (Fig. 1B). Unexpectedly, we also observed an increase in modification of HIF-1α at the 24- and 36-h time points when LANA was expressed at relatively higher levels (Fig. 1B). Previous studies have shown that except for latent genes there are many other KSHV immediate-early- and early-stage genes expressed (17). To verify the role of LANA on HIF-1α accumulation, we tested the HIF-1α levels in the BCBL-1 cells with specific small interfering RNA targeting LANA mRNA. The data from both immunofluorescence and immunoprecipitation assays showed that the increased expression of HIF-1α in the PEL cell lines was significantly diminished once the endogenous viral LANA was knockdown (Fig. 1C and D). This indicates that LANA contributes to HIF-1α accumulation in KSHV latent infections.

LANA enhances the stability of HIF-1α.

To determine whether the effect of LANA on HIF-1α results in an increase in HIF-1α stability, HEK293 cells expressing HIF-1α alone or both HIF-1α and LANA were treated with cycloheximide for 1 to 4 h after 24-h posttransfection. Immunoblotting against total HIF-1α showed that the stability of HIF-1α protein was significantly enhanced by LANA coexpressed compared to HIF-1α alone (Fig. 2, left panel). Determination of the relative percentage of HIF-1α remaining showed that the half-life HIF-1α was increased ∼2-fold with LANA coexpression (Fig. 2, right panel). These data suggest that LANA also contributes to stabilization of HIF-1α.

LANA promotes HIF-1α nuclear accumulation.

To further investigate the role of LANA in HIF-1α nuclear accumulation in KSHV latently infected cells, we wanted to look specifically at changes in HIF-1α localization in LANA-expressing cells. First, we constructed a vector carrying an in-frame GFP fusion of full-length HIF-1α (schematically shown in Fig. 3A). U2OS cells with clear cytoplasmic and nuclear compartments were then transiently transfected with either the parental GFP construct or the chimeric GFP-HIF-1α expression construct in the presence or absence of LANA-red fluorescence protein (RFP). Fluorescence was observed in ca. 20% of the cells, which reflected the transfection efficiency. As in previous studies (14, 20), the fluorescence of GFP or GFP-HIF-1α alone was uniformly distributed throughout the cell (Fig. 3B, left panels), and hypoxia dramatically increased the nuclear intensity of GFP-HIF-1α signals but had no effect on GFP alone (Fig. 3B, middle panels). However, the transient expression of GFP-HIF-1α with LANA-RFP constructs in normoxia resulted in clearly nuclear localization of HIF-1α, which also occurs by hypoxic treatment, even though the nuclear accumulation patterns of HIF-1α induced by LANA and hypoxia were distinctly different (Fig. 3B, right panels). To quantitatively compare the signals of HIF-1α in the nuclear and cytoplasmic compartments, the nuclear and cytoplasmic fractions from U2OS cells cotransfected with GFP-HIF-1α in the presence or absence of LANA-RFP were subjected to immunoblotting analysis to detect HIF-1α (GFP) and LANA. The nuclear protein Sp1 and cytoplasm protein Hsp70 immunoblot analysis were performed to determine the efficiency of nuclear and cytoplasm fractionation. The results showed that the signal for HIF-1α in the nucleus was significantly higher than in the cytoplasm when LANA was coexpressed, but similar signals were seen in both the cytoplasm and nucleus when GFP-HIF-1α was expressed alone (Fig. 3C). Therefore, this suggests a new role for LANA in inducing HIF-1α nuclear accumulation under normoxic conditions.

LANA enhances the activities of a HIF-1α responsive promoter.

To investigate whether the effect of LANA enhancing HIF-1α accumulation would lead to an increase in transcriptional activities of HIF-1α under normoxic conditions, we analyzed the specific effects of LANA on the transcriptional activities of native HIF-1α targeting its HRE and compared this to HIF-1α expression alone. The promoter elements of a trimerized wild-type HRE (wHRE) and its corresponding mutant (mHRE) responsive elements fused upstream to the luciferase reporter individually were tested in these reporter assays using BJAB and HEK293T cell lines. The data showed that the wHRE promoter was highly responsive to LANA expression in both HEK293T and BJAB cell lines (Fig. 4A). However, little or no response was seen in the mutant HRE reporter (Fig. 4A). Specifically, using the luciferase reporters individually containing the well-known HIF-1α downstream genes VEGF promoter and its relative HRE deleted mutant, the reporter assays data showed that LANA dramatically enhance VEGF promoter activity and results in a drop in activity of ca. 30% once the HREs in the VEGF promoter were deleted (Fig. 4B). This suggests that the induction of VEGF by LANA is at least in part as a result of HIF-1α accumulation and its responsive element within the VEGF promoter.

To support these transcriptional studies, ELISAs were performed to determine the levels of VEGF, a downstream target of HIF-1α, which was previously demonstrated to be induced by KSHV infection (21, 47). The data obtained from cells transiently expressing LANA showed that the levels of VEGF produced were also significantly increased in HEK293T cells transfected with LANA compared to only HIF-1α expression (Fig. 4C). More importantly, increasing amounts of LANA showed a dramatic increase in VEGF levels as seen by ELISA (Fig. 4C).

HIF-1α nuclear accumulation mediated by LANA is dependent on its ODD domain.

In order to identify which structural motif involved in nuclear import of HIF-1α mediated by LANA, we generated four expression vectors containing GFP fused to wild type and three different domains of HIF-1α (schematically represented in Fig. 5A). After transient expression of these constructs with or without LANA-RFP in U2OS cells, the cells were analyzed for localization of the fusion proteins. As shown in Fig. 5B, the fusion protein containing the most central region of HIF-1α which spans the oxygen-dependent degradation (ODD) domain resulted in a predominantly nuclear accumulation when coexpressed with LANA-RFP (Fig. 5B). Moreover, equal cytoplasm and nuclear distribution of HIF-1α was seen when the ODD domain was deleted, indicating that its localization was not altered when LANA-RFP was coexpressed (Fig. 5B). This strongly supports the conclusion that the ODD domain of HIF-1α is critical for LANA-mediated HIF-1α nuclear translocation. Interestingly, unlike other truncated and full-length HIF-1α, the GFP fusion with HIF-1α amino acids 1 to 300 showed significant cytoplasmic fluorescence activity whether or not LANA-RFP was coexpressed in normoxic condition (Fig. 5B). To establish a quantitative set of analyses for our data, 50 fluorescent cells were routinely analyzed for compartmentalization of HIF-1α and subdivided into four categories according to previous studies (14, 45): N, exclusively nuclear fluorescence; N>C, nuclear fluorescence more than cytoplasmic fluorescence; N=C, equal distribution of nuclear and cytoplasmic fluorescence; and N<C, exclusively cytoplasmic fluorescence. The percentage of cells identified in each category is shown in Table 1. With LANA-RFP coexpression, 72% of cells transiently expressing the wild-type GFP-HIF-1α construct showed exclusively nuclear compartmentalization of HIF-1α (category N), whereas only a limited number of cells (6%) had similar signal with GFP-HIF-1α expression alone (Table 1). The GFP fused with truncated HIF-1α polypeptides showed that only the construct transiently expressing GFP-HIF-1α/330-530 had 10% of cells belonging to category N when LANA-RFP was coexpressed. No other truncated GFP-HIF-1α-expressing cells were seen in this category. This indicated that the ODD domain of HIF-1α is critical for LANA-mediated HIF-1α nuclear accumulation.

The amino-terminal residues 46 to 89 are required for LANA to mediate HIF-1α nuclear accumulation.

Previously, we demonstrated that LANA acts as an adapter of the EC₅S ubiquitin complex blocking the HIF-1α suppressors VHL and p53 through its unconventional suppressors of cytokine signaling (SOCS)-box motif (Fig. 6A) (2). To test whether LANA-mediated HIF-1α nuclear accumulation is dependent on this motif, we coexpressed the LANA truncated mutants containing the entire (LN-1) or partly (LN-2 and LN-3) SOCS-box motif with GFP-HIF-1α in U2OS cells (Fig. 6A). Immunofluorescence assay data showed that the LANA mutant truncated from 1 to 233 residues containing only the BC-box component of the SOCS-box motif is sufficient to mediate HIF-1α nuclear accumulation (Fig. 6B). This suggests that the amino-terminal domain of LANA can mediate HIF-1α accumulation that is independent of its SOCS-box motif. Importantly, in the VHL-deficient 786-O cells, HIF-1α was also induced to nuclear accumulation by LANA coexpression but not HIF-1α alone, indicating that this accumulation is not dependent on VHL (Fig. 6C, top panel). Interestingly, in the p53-deficient Saos-2 cells, whether HIF-1α was expressed with LANA or not, the majority of HIF-1α was localized in nucleus (Fig. 5C, bottom panel).

To further define which specific domain of LANA is necessary to mediate HIF-1α nuclear accumulation, we sought to identify the potential secondary structures that are located at the amino termini of LANA to associate with HIF-1α. The cysteine and histidine residues, within the region of LANA from amino acids 1 to 233 were then investigated. Figure 7A shows the positions of these residues. Based on the predicted potential functional RING domain formation via different disulfide bonds, three LANA mutants tagged with the myc epitope and mutated at these specific positions were generated (Fig. 7B). Immunofluorescence assays with these truncated LANA polypeptides coexpressed with GFP-HIF-1α constructs in U2OS cells were performed. The data showed that the LANA mutant with residues 47 to 129 deleted did diminish the level of HIF-1α nuclear accumulation (Fig. 7C, top panel). However, the LANA truncated mutant (LN5) containing only residues 46 to 145 was also not efficient in inducing HIF-1α nuclear localization (Fig. 7C). Interestingly, the LANA mutant with 1 to 89 residues deleted (LN6) not only lost the capability of inducing HIF-1α nuclear localization but was also dramatically reduced in its own ability to localize to the nucleus (Fig. 7C). To verify that the secondary structure containing histidine 47 and cysteine 57 is necessary for LANA to mediate HIF-1α nuclear accumulation, we generated double mutations of these two residues, which were converted to alanine and then performed similar immunofluorescence assays. Surprisingly, the data showed that mutation of these two residues is sufficient to destroy the effect of LANA on HIF-1α nuclear accumulation (Fig. 7C). In the cytoplasm and nuclear fractionation assays, the results also showed that the LANA mutants LN-5 with 46 to 129 residues deleted or LN-7 with only the H47 and C57 double mutations, the effect of LANA on HIF-1α nuclear accumulation was abolished (Fig. 7D). Expectedly, deletion of the nuclear localization sequence that existed in residues 1 to 23 of LANA resulted in LANA being mostly accumulated in the cytoplasmic compartment (Fig. 7D).

To determine whether the protein-protein interaction between LANA and HIF-1α is required for HIF-1α nuclear accumulation, we also performed immunoprecipitation assays in HEK293 cells using the LANA truncated mutants described above. As expected, the mutants with residues 46 to 129 deleted or with residues 1 to 89 deleted failed to interact with HIF-1α (Fig. 8A and B). However, compared to the truncated mutant of LANA 1 to 233, the mutant containing residues 46 to 145 associated with HIF-1α, even though it was not able to efficiently induce HIF-1α nuclear accumulation (Fig. 8A and B). This suggests that the secondary structure located in the region from residues 46 to 89 is necessary for LANA to interact with HIF-1α but not sufficient to induce HIF-1α nuclear accumulation. The construct with the histidine 47 and cysteine 57 both mutated also lost this association with HIF-1α, supporting this conclusion (Fig. 8A and B). Surprisingly, the secondary structure analysis of LANA reveals that residues H47 and C57 may be potentially located at separate α-helix motifs important for contact with HIF-1α (Fig. 8C).