Kaposi's Sarcoma-Associated Herpesvirus-Induced Angiogenin Plays Roles in Latency via the Phospholipase Cγ Pathway: Blocking Angiogenin Inhibits Latent Gene Expression and Induces the Lytic Cycle

Abstract

During de novo infection of human dermal microvascular endothelial cells (HMVEC-d), Kaposi's sarcoma-associated herpesvirus (KSHV) induced the multifunctional angiogenin (ANG) protein, which entered the nuclei and nucleoli of infected cells and stimulated 45S rRNA gene transcription, proliferation, and tube formation, which were inhibited by blocking ANG nuclear translocation with the antibiotic neomycin (S. Sadagopan et al., J. Virol. 83:3342-3364, 2009). ANG was induced by KSHV latency protein LANA-1 (open reading frame 73 [ORF73]). Here we examined the presence and functions of ANG in KSHV-positive (KSHV⁺) primary effusion lymphoma (PEL/BCBL) cells. Significant ANG gene expression and secretion were observed in KSHV⁺ (BCBL-1 and BC-3) and KSHV⁺ and Epstein-Barr virus-positive (KSHV⁺ EBV⁺) (JSC-1) PEL cells and in BJAB-KSHV cells but not in EBV⁻ KSHV⁻ lymphoma cells (Akata, Loukes, Ramos, and BJAB), EBV⁺ lymphoma cells (Akata-EBV and Raji), and cells from an EBV⁺ lymphoblastoid cell line, thus suggesting a specific association of ANG in KSHV biology. Inhibition of nuclear translocation of ANG resulted in reduced BCBL-1 and TIVE-LTC (latently infected endothelial) cell survival and proliferation, while EBV⁻ and EBV⁺ Akata cells were unaffected. Blocking nuclear transport of ANG inhibited latent ORF73 gene expression and increased lytic switch ORF50 gene expression, both during de novo infection and in latently infected cells. A greater quantity of infectious KSHV was detected in the supernatants of neomycin-treated BCBL-1 cells than 12-O-tetradecanoylphorbol-13-acetate (TPA)-treated cells. Neomycin treatment and ANG silencing inhibited phospholipase Cγ (PLC-γ) and AKT phosphorylation, and in contrast, ANG induced ORF73 expression and PLC-γ and AKT phosphorylation. Further studies provided evidence that blockage of PLC-γ activation by neomycin appears to be mediating the inhibition of latent gene expression, since treatment with the conventional PLC-γ inhibitor U73122 also showed similar results. Silencing of ANG also resulted in reduced cell survival, reduced ORF73 gene expression, and lytic gene activation in BCBL-1 and TIVE-LTC cells and during de novo infection. Taken together, these studies suggest that KSHV has evolved to exploit ANG for its advantage via a so-far-unexplored PLC-γ pathway for maintaining its latency.

Kaposi's sarcoma-associated herpesvirus (KSHV) is etiologically associated with Kaposi's sarcoma (KS), primary effusion lymphoma (PEL), and multicentric Castleman's disease (4, 6, 43). PEL is an aggressive B-cell lymphoma with poor prognosis and for which there is no specific treatment. PEL (B-cell) lines, such as BC-1, HBL-6, and JSC, carry multiple genome copies of KSHV and related Epstein-Barr virus (EBV) in a latent state, whereas BCBL-1 and BC-3 cells carry only the KSHV genome (12). KSHV latency-associated open reading frames (ORFs), such as ORF73 (LANA-1), ORF72, ORF71, K12, K10.5, K1, and K2, as well as 12 pre-microRNAs (miRNAs), are expressed in PEL cells (3, 5, 12). These latent genes mediate functions such as controlling the KSHV lytic cycle switch ORF50 gene and evading host responses, including apoptosis (15), autophagy (23), interferons (IFNs) (2), etc., which are essential for the maintenance of latent infection. The consequences of these functions also lead to PEL cell survival, antiapoptosis, and continuous growth (13).

The mechanisms by which KSHV controls the lytic cycle, as well as the switch from latency to lytic replication, are areas of intense study and incompletely understood. ORF50 is not only sufficient but required for KSHV lytic induction and increases the expression of early and late lytic genes, resulting in the production of viral progeny. About 1 to 3% of PEL cells spontaneously enter the lytic cycle, and it can be induced by chemical agents such as 12-O-tetradecanoyl-phorbol-13-acetate (TPA) and n-butyrate; however, the mechanism by which chemical agents induce the lytic cycle is not fully clear (13).

In vitro KSHV infection of human dermal microvascular endothelial cells (HMVEC-d) is characterized by the concurrent expression of latent genes and a limited number of lytic genes, subsequent decline and/or absence of lytic gene expression, and persistent latent gene expression (21). Gene array studies demonstrated that KSHV reprogrammed HMVEC-d cell transcriptional machinery regulating apoptosis, cell cycle regulation, signaling, inflammatory response, and angiogenesis (21). Our subsequent cytokine array analysis showed that KSHV infection induced a significant increase in the secretion of several endothelial cell angiogenic molecules (vascular endothelial growth factor [VEGF], angiopoietin, angiogenin, and SDF-1), growth factors (platelet-derived growth factor [PDGF], fibroblast growth factor [FGF], granulocyte-macrophage colony-stimulating factor [GM-CSF], and insulin-like growth factor 1 [IGF-1]), chemokines (monocyte chemoattractant protein 2 [MCP-2], macrophage inflammatory protein [MIP], monocyte induced by IFN-γ [MIG], and eotaxin), and proinflammatory (interleukin-2 [IL-2], IL-3, IL-8, GRO, and IL-16) and anti-inflammatory (IL-4, IL-5, and IL-15) cytokines, and many of these factors were induced in an NF-κB-dependent manner (38). The angiogenic factor angiogenin was among the most highly upregulated cytokines during de novo KSHV infection (38).

Angiogenin, a multifunctional 14-kDa angiogenic protein, was first isolated from HT-29 human colon adenocarcinoma cell conditioned media based on its angiogenic activity (16). Angiogenin has been shown to play a role in tumor angiogenesis, and its expression is upregulated in several types of cancer, including pancreatic, breast, prostate, cervical, ovarian, colon, colorectal, gastric, urothelial, and endometrial cancers (16). Antiangiogenin monoclonal antibodies used as antagonists inhibited the establishment, progression, and metastasis of human cancer cells inoculated in athymic mice (35).

The half-lives of angiogenin mRNA and protein are estimated to be about 9 to 12 h and 12 to 24 h, respectively (15a, 44a). When human angiogenin is added to nonpermeabilized human umbilical vein endothelial cells, it translocates into the nucleus in a time-dependent manner (17a). Exogenous angiogenin appears in the nucleus in 2 min; reaches saturation in 15 min, when 85% of the internalized angiogenin is in the nuclei; and remains associated with the nucleus for at least 4 h. Exogenous angiogenin could also be detected in the cell extracts even after 48 h following addition (15a).

Angiogenin exerts multiple effects at the endothelial cell surface, nucleus, and nucleoli (46). In semi- and subconfluent cells, angiogenin binds to a 170-kDa protein and activates phospholipase C (PLC), phospholipase A2 (PLA2) (31), protein kinase B (PKB/AKT) (19), extracellular signal-regulated kinase 1/2 (ERK1/2) (25), and stress-associated protein kinase/c-Jun N-terminal kinase (SAPK/JNK) (25). Angiogenin nuclear translocation is strictly dependent on cell type and cell density (32). It remains cytoplasmic in fibroblast cells, suggesting that the nuclear function of angiogenin is limited to endothelial cells. Nuclear translocation does not occur when the cells are confluent but occurs in cells with about 30% confluence (subconfluent) and cells with <60% confluence (semiconfluent). Angiogenin translocates through a microtubule-independent pathway into the nucleus of sub- and semiconfluent cells via an unknown mechanism (32). Angiogenin moves into the nucleolus of subconfluent cells and mediates rRNA transcription by binding to CT repeats in the promoter region of the rRNA gene (48). Angiogenin also exerts its ribonucleolytic activity by catalyzing the generation of 18S and 28S rRNA. Nuclear translocation of angiogenin in endothelial cells has been shown to be necessary for the angiogenic potentials of not only angiogenin but also VEGF and basic fibroblast growth factor (bFGF) (20).

Activation of PLC-γ is required for nuclear translocation of angiogenin (17). Neomycin, an aminoglycoside antibiotic, inhibits this nuclear translocation by inhibiting PLC-γ activation without affecting the viability of the cells or ERK1/2 phosphorylation (25). Neomycin-mediated inhibition is due to cleavage of phosphatidylinositol (PtdIns), phosphatidylinositol-4-phosphate (PtdIns4P), and phosphatidylinositol 4,5-biphosphate [PtdIns(4,5)P2] by phospholipase C and the release of two intracellular second messengers, inositol 1,4,5-triphosphate [Ins(1,4,5)P3] and diacylglycerol (DAG; acyl2Gro), by directly binding to PtdIns4P and to PtdIns(4,5)P2 (26). In contrast paromomycin, an analogue aminoglycoside, does not inhibit nuclear translocation of angiogenin. Neomycin, an FDA-approved antibiotic, inhibits xenographic human tumor growth in athymic mice (16) and lung adenocarcinoma cell proliferation (49), while neamine, a nonototoxic derivative of neomycin, prevents and reverses prostate intraepithelial neoplasia (18).

In our earlier study, we examined the role of angiogenin in KSHV infection of endothelial cells (37). Angiogenin was expressed in KS tissues, and KSHV induced a time- and dose-dependent increase in angiogenin secretion during de novo KSHV infection. KSHV LANA-1 expressed from lentivirus or plasmids induced angiogenin gene expression and angiogenin secretion (37). Angiogenin was translocated into the nucleoli and nuclei of subconfluent and semiconfluent endothelial (HMVEC-d) cells (37). Neomycin treatment blocked nuclear translocation of angiogenin in KSHV-infected HMVEC-d cells and inhibited angiogenin function, including cell proliferation, antiapoptosis, 45S rRNA synthesis, cell migration, and angiogenesis (37).

Since we also detected angiogenin secretion in BCBL-1 cells (37), here we analyzed the significance of angiogenin in BCBL-1 cells. The studies presented here suggest that KSHV utilizes angiogenin to maintain its latency, probably by activating the PLC-γ pathway, and show for the first time the importance of PLC-γ activation for the maintenance of latency.

MATERIALS AND METHODS

Cells.

Primary human dermal microvascular endothelial (HMVEC-d) cells (CC-2543; Clonetics Walkersville, MD) and telomerase-immortalized human umbilical vein (TIVE)-LTC cells were grown in endothelial basal medium (EBM-2) with growth factors (Clonetics). Human foreskin fibroblast (HFF), HEK 293T (human embryonic kidney), and HeLa cells were cultured in Dulbecco's modified Eagle's medium (DMEM; Gibco BRL, Grand Island, NY) supplemented with 10% heat-inactivated fetal bovine serum (FBS), 2 mM l-glutamine, and antibiotics. EBV-negative and KSHV-negative (EBV⁻ KSHV⁻) lymphoma cells (Akata, Loukes, Ramos, and BJAB), EBV⁺ lymphoma cells (Akata-EBV and Raji), and EBV⁺ lymphoblastoid cell line (LCL) cells (primary B cells transformed by B95-8 EBV), KSHV⁺ BC-3 and BCBL-1 PEL cells, KSHV⁺ EBV⁺ JSC-1 PEL cells, and BJAB-KSHV (34) cells were cultured in RPMI 1640 (Gibco BRL) medium with 10% fetal bovine serum (FBS) (HyClone, Logan, UT), 2 mM l-glutamine, and antibiotics.

Antibodies, reagents, and growth factors.

Rabbit polyclonal antibodies against human angiogenin were from Santa Cruz Biotechnology, Inc., Santa Cruz, CA. Anti-goat, anti-rabbit, and anti-mouse antibodies linked to Alexa Fluor 488 or Alexa Fluor 594 and antifade DAPI (4′,6-diamidino-2-phenylindole) were purchased from Molecular Probes, Eugene, OR. Recombinant human angiogenin, bovine serum albumin (BSA), 12-O-tetradecanoylphorbol-13-acetate (TPA), Triton X-100, neomycin, paromomycin, U73122, paraformaldehyde, and antibodies against lamin B and tubulin were from Sigma, St. Louis, MO. Rabbit antibodies detecting phospho-PLC-γ, total PLC-γ, total AKT, and total ERK2 and mouse antibodies detecting phospho-ERK1/2 and phospho-AKT were from Cell Signaling Technology, Beverly, MA.

Virus.

The induction of the KSHV lytic cycle in BCBL-1 cells, supernatant collection, and virus purification procedures were described previously (21). The purity of KSHV was assessed according to general guidelines established in our laboratory (21, 33). For virus purification after neomycin treatment, BCBL-1 cells were treated with 200 μM neomycin for 3 days before supernatant was harvested and concentrated by ultracentrifugation. KSHV DNA was extracted from the purified virus, and copy numbers were quantitated by real-time DNA PCR using primers amplifying the KSHV ORF73 gene (21). For TIVE-LTC cell culture supernatant studies, supernatant was harvested 3 days after 200 μM neomycin treatment, spun at 5,000 rpm for 10 min to remove cell debris, and concentrated, and the DNA was extracted and subjected to real-time DNA PCR for the ORF73 gene.

Angiogenin protein quantitation.

The levels of angiogenin in the culture supernatant of Akata, Akata-EBV, LCL, Loukes, Raji, Ramos, BJAB, BJAB-KSHV, JSC-1, BCBL-1, and BC-3 cells treated with increasing concentrations of neomycin were quantitated with a human angiogenin enzyme-linked immunosorbent assay (ang-ELISA) (R&D Systems, Minneapolis, MN). Briefly, supernatants were collected from the cells, spun at 1,000 rpm for 10 min at 4°C to remove the particulates, and stored at −80°C until use. Total soluble protein was quantitated by bicinchoninic acid (BCA) protein assay (Pierce, Rockford, IL). Since the concentration of protein present in the supernatant was the only variable between various cell types, an equal concentration of protein made up to 200 μl was loaded onto each well, pretreated with antiangiogenin antibody followed by secondary antibody, developed with the substrate, and read at 450 nm. For estimation of angiogenin protein concentration in studies using small interfering RNA (siRNA) coding for angiogenin (si-angiogenin), HEK 293T cells were transfected with plasmids containing siRNAs encoding angiogenin (si-ang1 and si-ang2), Renilla luciferase (si-RL), or lamin (si-lamin). Twenty-four hours posttransfection, medium was changed to complete Dulbecco's modified Eagle's medium (DMEM) containing serum. Cells were allowed to grow for 24 h, followed by serum starvation for 24 h, and the harvested supernatant was used for ang-ELISA.

IFA.

The immunofluorescence assay (IFA) was done as described previously (40). Briefly 1 × 10⁶ BCBL-1, BC-3, Akata, Akata-EBV, and JSC-1 cells were harvested and resuspended in 250 μl 1× phosphate-buffered saline (PBS). Fifteen microliters of this suspension was laid on microscopy slides, air dried under laminar air flow, fixed in acetone for 10 min, blocked with 5% BSA in 1× PBS, and stained with primary antibody for 1 h at 37°C. Cells were washed and incubated with an appropriate dilution of secondary antibodies conjugated with Alexa Fluor 488 for 1 h at room temperature. Stained cells were washed and viewed under a fluorescence microscope equipped with Nikon Metamorph digital imaging system. Nuclei were visualized by using DAPI as a counterstain. For calculation of the percentage of cells expressing viral latent and lytic genes, the numbers of LANA-1-, ORF59-, or K8.1A-positive cells were counted in three different fields and normalized to the number of live cells by considering DAPI-positive cells as 100%. Only live cells stained positive for DAPI.

Measurement of KSHV DNA internalization.

Untreated HMVEC-d cells or HMVEC-d cells incubated with various concentrations of neomycin or 10 μM LY294002 for 1 h were infected with KSHV (10 DNA copies per cell). After different time points of infection with virus, cells were washed twice with Hanks’ balanced salt solution (HBSS) to remove unbound virus. Cells were treated with 0.25% trypsin-EDTA for 5 min at 37°C to remove the bound, noninternalized virus. Detached cells were washed twice to remove the virus and harvested. Total DNA isolated from cells was tested by real-time DNA PCR for ORF73 as described previously (21). For BCBL-1 cell culture supernatant studies, 10⁶ BCBL-1 cells were harvested 3 days and 6 days after the indicated treatments and spun at 5,000 rpm for 10 min to remove cell debris, and the supernatant was used to infect HMVEC-d cells. DNA was extracted 2 h postinfection (p.i.) and subjected to real-time DNA PCR for ORF73.

Real-time qRT-PCR.

Angiogenin expression was detected by real-time quantitative reverse transcription-PCR (qRT-PCR) using SYBR green chemistry. ORF73 and ORF50 expression was detected by real-time RT-PCR using TaqMan chemistry. For viral gene expression in BCBL-1 cells with neomycin, 5 × 10⁶ BCBL-1 cells were treated with 25, 50, 100, 200, 500, and 1,000 μM neomycin for 3 days or 6 days, and total RNA was extracted. Two micrograms of total RNA was treated with DNase for 1 h and then reverse transcribed into cDNA by using a random hexamer with a high-capacity cDNA reverse transcription mix for RT-PCR (Applied Biosystems, Foster City, CA). cDNA was used as a template with primers specific for angiogenin, ORF73, and ORF50 (21, 37). Tubulin was used as an internal control (37). PCR was performed with the Applied Biosystems 7500 real-time PCR system with SYBR green or TaqMan PCR master mix (ABI; Applied Biosystems). The standard amplification program included 40 cycles of two steps each comprised of heating to 95 and 60°C. Fluorescent product was detected at the last step of each cycle. For SYBR green reactions, the final mRNA levels of the genes studied were normalized by the comparative threshold cycle (C_T) method.

Lentivirus production and infection of BCBL-1 and TIVE-LTC cells.

Lentiviral infection was done as described by Vart et al. (44). Vesicular stomatitis virus G envelope-pseudotyped lentivirus was produced with a four-plasmid transfection system, as previously described (8, 30). Briefly, HEK 293T cells were transfected with lentiviral constructs expressing short hairpin RNA (shRNA) to angiogenin, si-angiogenins 1, 2, 3, 4, and 5 (si-ang1 to -5, respectively) (Open Biosystems, Huntsville, AL), si-Renilla luciferase (si-RL), si-lamin, and packaging plasmids by the calcium phosphate precipitation method. The cells were incubated overnight at 3% CO₂, and then at 48 and 72 h p.i., culture supernatant containing the lentiviral virions was harvested, and cell debris was cleared by passing through a 0.45-μm-pore filter and either aliquoted directly or concentrated and stored at −80°C. Infections were carried out by incubating the desired amount of virus preparation with BCBL-1 or TIVE-LTC cells in culture for 8 h, and the medium was replaced with complete growth medium and incubated for 48 h. Total RNA was extracted from lentivirus-transduced BCBL-1 or TIVE-LTC cells by using TRIzol reagent (Invitrogen) and subjected to DNase I digestion (Invitrogen). This RNA was subsequently used for cDNA synthesis using a high-capacity cDNA reverse transcription kit (ABI). Real-time quantitative RT-PCR (qRT-PCR) was done for angiogenin to confirm the knockdown and then used for quantitation of ORF73 and ORF50 gene expression.

HeLa cell transfection studies.

HeLa cells were transfected with si-ang1, si-ang2, si-RL, or si-lamin plasmids. Twenty-four hours posttransfection, medium was changed to complete DMEM containing serum. Cells were allowed to grow for 24 h and infected with KSHV (10 DNA copies per cell) for 48 h, and then total RNA was extracted with TRIzol reagent (Invitrogen, Carlsbad, CA) and subjected to DNase I digestion (Invitrogen). This RNA was used for cDNA synthesis using a high-capacity cDNA reverse transcription kit (ABI) and real-time qRT-PCR for the ORF73 and ORF50 genes.

Nuclear extract preparation.

BCBL-1 cells were treated with the indicated concentrations of neomycin, U73122, or TPA for 3 days, cells were harvested, and nuclear extracts were prepared with a nuclear extract kit (Active Motif Corp., Carlsbad, CA) as per the manufacturer's instructions. After measurement of protein concentrations with BCA protein assay reagent, nuclear and cytoplasmic extracts were subjected to Western blot analysis with antiangiogenin antibody. The purity of the nuclear extracts and cytoskeletal contamination were assessed by immunoblotting with anti-lamin B and anti-β-tubulin antibodies, respectively.

Western blotting.

Lysates or nuclear extracts (10 μg) were resolved on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gels, transferred to nitrocellulose membranes, blocked with 5% skim milk, and immunoblotted with rabbit polyclonal antiangiogenin (Santa Cruz), antitubulin, and anti-lamin B antibodies for nuclear extracts and with anti-phospho-PLC-γ, anti-ERK1/2, anti-AKT, anti-total PLC-γ, anti-protein kinase Cζ (anti-PKCζ), anti-ERK2, and anti-AKT antibodies for whole-cell lysates. Immunoreactive bands were developed by enhanced chemiluminescence reaction (Pierce) and quantified following standard protocols (40).

Cell survival (MTT) assay.

KSHV⁺ and KSHV⁻ B cells and TIVE-LTC cells were treated with increasing concentrations of neomycin. BCBL-1 cells were treated with increasing concentrations of U73122. Three days posttreatment, 50 μl of MTT [3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide] reagent (Promega, Madison, WI) was added to each well, the mixture was incubated for 4 h, and the absorbance at 570 nm was read. For cell survival studies with si-angiogenin, 10⁴ BCBL-1 cells transduced with lentivirus-expressed si-ang1, si-ang2, si-RL, and si-lamin were seeded on a 96-well plate for 24 h, treated with 50 μl MTT reagent, and incubated for 4 h at 37°C, and the absorbance at 570 nm was read.

Trypan blue cell viability assay.

The in vitro effects of neomycin treatment on TIVE, TIVE-LTC, BCBL, BC-3, JSC-1, Akata, and Akata-EBV cells were determined by trypan blue assay in quadruplicate. The cells were treated with 200 μM neomycin. At 0, 1, 2, and 3 days posttreatment, cells were mixed with trypan blue dye and the numbers of viable cells were counted. The number of cells in untreated cultures on each day was considered as 100% for comparison of cell survival with neomycin treatment.

Statistical analysis.

In Fig. 2 through 7 and 9, and in Fig. S1 in the supplemental material, statistical significance (t test) was determined with respect to untreated cells for each time point. In Fig. 3, 4, and 5, the fold change in gene expression was calculated by considering gene expression in untreated cells as 1. In Fig. 8, gene expression in si-RL was considered as 1. Each reaction was done in duplicate, and each bar represents the average + standard deviation (SD) from three independent experiments.

RESULTS

Increased angiogenin expression is associated with KSHV-infected B-lymphoma cells.

To understand the significance of angiogenin secretion in BCBL-1 cells, first we used quantitative real-time PCR assays to measure angiogenin gene expression in EBV⁻ KSHV⁻ lymphoma cells (Akata, Loukes, Ramos, and BJAB), EBV⁺ lymphoma cells (Akata-EBV and Raji) and an EBV⁺ lymphoblastoid cell line, KSHV⁺ BC-3 and BCBL-1 PEL cells, KSHV⁺ EBV⁺ JSC-1 PEL cells, and BJAB-KSHV cells. No significant angiogenin gene expression was observed in KSHV⁻, EBV⁻, and EBV⁺ cells (Fig. 1A). In contrast, we observed about 5.2-, 5.3-, 7.2-, and 6.7-fold increases in angiogenin expression in BJAB-KSHV, JSC-1, BCBL-1, and BC-3 cells, respectively, compared to the level in EBV⁻ Akata cells (Fig. 1A). These results were confirmed by the detection of significant amounts of angiogenin (250 to 400 pg/ml) in BCBL-1, BC-3, BJAB-KSHV, and JSC-1 cell supernatants and less than 30 pg/ml from EBV⁻ and EBV⁺ Akata cell supernatants (Fig. 1B). Quantitative real-time PCR confirmed the expression of KSHV ORF73 and EBV EBNA-1 in these cells (data not shown). By IFA, significant angiogenin expression was detected in more than 60% of BCBL-1, BC-3, and JSC-1 cells, while angiogenin was barely detectable in <1 to 2% of EBV⁻ and EBV⁺ Akata cells (Fig. 1C). These results suggested that KSHV but not EBV infection is associated with increased angiogenin expression in B-lymphoma cells.

FIG. 1.

Angiogenin expression and secretion in KSHV-infected and EBV-infected B-lymphoma cell lines. (A) Total RNA was extracted from the cells, converted to cDNA, used to estimate angiogenin gene expression by quantitative real-time PCR, normalized to tubulin, and expressed as fold change in angiogenin gene expression. (B) Supernatants harvested from the indicated cell lines were used to measure the levels of angiogenin secretion by ang-ELISA. Fold increase was calculated by considering angiogenin secretion in Akata cells as 1. (C) Immunofluorescence analysis demonstrating angiogenin expression. Cells were harvested, washed, fixed with acetone, and incubated overnight with antiangiogenin antibody in 2% BSA. Cells were washed with PBS, incubated with secondary antibody conjugated with Alexa Fluor 488 (green) for 1 h at room temperature, washed, mounted with antifade DAPI, and visualized under fluorescence microscope. Enlarged images show angiogenin and DAPI results merged in a single cell. **, P < 0.001; ***, P < 0.0001.

Nuclear translocation of angiogenin in BCBL-1 cells is blocked by neomycin.

Angiogenin movement inside endothelial cells is cell density dependent; it is cytoplasmic in confluent cells, nuclear in semiconfluent cells, and nucleolar in subconfluent cells (32). Nuclear translocation of angiogenin was critical for its function, which was blocked by neomycin treatment (20). To test the importance of nuclear angiogenin in BCBL-1 cells, cells were treated with 10 μM and 200 μM neomycin for 3 days, fixed, permeabilized, and stained with an antiangiogenin antibody. A distinct nuclear localization of angiogenin was observed in untreated BCBL-1 cells (Fig. 2A). A 10 μM concentration of neomycin had no impact on nuclear translocation of angiogenin, whereas 200 μM neomycin significantly impaired angiogenin's movement into the nucleus, which resulted in scattered angiogenin staining in the cytoplasm (Fig. 2A). To confirm IFA findings, Western blot analysis was done with nuclear extracts from untreated and neomycin-treated BCBL-1 cells with an antiangiogenin antibody. We observed a nearly 80% reduction in nuclear translocation of angiogenin in BCBL-1 cells treated with 200 μM neomycin (Fig. 2B, lane 3) but no significant effect with 10 μM neomycin (Fig. 2B, lane 2). Paromomycin (100 μM), an analog of neomycin, did not inhibit nuclear translocation of angiogenin (Fig. 2B, lane 4). When cytoplasmic extracts from BCBL-1 cells were Western blotted for angiogenin, we observed an increase in the cytoplasmic retention of angiogenin with 200 μM neomycin and not with 10 μM neomycin or 100 μM paromomycin treatments (Fig. 2B). The purity of the nuclear fraction was confirmed with anti-lamin B antibody (Fig. 2B, bottom panel) and by the absence of reactivity with antitubulin antibody (Fig. 2B, bottom panel).

FIG. 2.

Immunofluorescence analysis showing a block in nuclear translocation of angiogenin by neomycin treatment and its effect on KSHV⁺, KSHV⁻, EBV⁺, and EBV⁻ B-lymphoma cell, PEL cell, and LCL survival. (A) BCBL-1 cells were treated with the indicated concentrations of neomycin (Neo) for 3 days, fixed, permeabilized, stained for angiogenin (green), and counterstained with DAPI (blue). Arrows indicate angiogenin nuclear staining in BCBL-1 cells. Magnification, ×40. The enlarged images show a single cell displaying nuclear localization of angiogenin. Magnification, ×80. (B) Nuclear and cytoplasmic extracts were prepared from BCBL-1 cells either untreated (lane 1) or treated with 10 μM neomycin (lane 2), 200 μM neomycin (lane 3), and 100 μM paromomycin (Paro; lane 4) and Western blotted with antiangiogenin antibody (top two panels). The nuclear extract blot was stripped and probed with anti-lamin B and antitubulin antibodies (bottom two panels). (C) Supernatants harvested from BCBL-1 cells treated with the indicated concentrations of neomycin were used to measure the levels of angiogenin secretion by ang-ELISA. (D, E, and F) Cell survival assay upon neomycin treatment. The MTT assay was done at 3 days posttreatment for the indicated cell lines with increasing concentrations of neomycin. **, P < 0.001; ***, P < 0.0001.

Our previous studies with HMVEC-d cells showed that even though neomycin treatment blocked the movement of angiogenin into the nucleus, it did not inhibit angiogenin secretion significantly (37). Similarly, when angiogenin ELISA was done with supernatants harvested from BCBL-1 cells treated for 3 days with increasing neomycin concentrations, we did not observe any significant reduction in angiogenin secretion (Fig. 2C). These results indicated that neomycin treatment blocked nuclear translocation of angiogenin without any significant effect on the cytoplasmic level of angiogenin and secretion of angiogenin.

Nuclear angiogenin is essential for PEL cell survival.

We next carried out an MTT cell survival assay to analyze the effect of blocking nuclear translocation of angiogenin. Active mitochondria of living cells are required to cleave MTT, yielding a dark blue formazan, and even freshly dead cells do not cleave significant amounts of MTT. Neomycin treatment reduced BCBL-1 cell survival in a dose-dependent manner (Fig. 2D). A 50% reduction in BCBL-1 cell survival was detected at 3 days posttreatment with 200 μM neomycin (Fig. 2D). To confirm our finding, the cell survival assay was performed with multiple KSHV⁺, EBV⁺, KSHV⁻, and EBV⁻ B-cell lines. We observed reduced cell survival by MTT assay when BJAB-KSHV, BC-3, and JSC-1 cells were treated with increasing concentrations of neomycin (Fig. 2E). In contrast, neomycin treatment did not affect EBV⁻ KSHV⁻ (Akata, Ramos, Loukes, and BJAB) and EBV⁺ (Akata-EBV, LCL, and Raji) cell survival even at higher concentrations (Fig. 2F). Paromomycin treatment did not affect survival of KSHV⁺ and KSHV⁻ B-cell lines (Fig. 2D, E, and F). These results supported our hypothesis that only KSHV⁺ PEL cells expressing elevated angiogenin levels were affected by neomycin treatment and not KSHV⁻ B-cell lines that expressed negligible levels of angiogenin. Similar results were also obtained with the trypan blue assay (see Fig. S1 in the supplemental material). It is important to note that even 2 mM neomycin had no effect on uninfected HMVEC-d cell survival (37). These results suggested that nuclear translocation of angiogenin plays an important role(s) in the survival of BCBL-1 cells latently infected with KSHV but not the survival of EBV⁻ and EBV⁺ cells.

Neomycin treatment reduces latent gene expression and induces lytic gene expression in BCBL-1 cells.

We hypothesized that reduction in BCBL-1 survival by neomycin could be due to a combination of the following reasons: (i) induction of apoptosis by blockage of angiogenin function; (ii) reduction in latent gene expression that is essential for cell survival, such as NF-κB activation by vFLIP (28), modulation of p53, and activation of c-Myc by LANA-1 (11) (24); and/or (iii) KSHV lytic cycle activation. Since antiapoptosis in BCBL-1 cells is in part mediated by KSHV latent gene expression, we next tested the effect of blocking angiogenin's nuclear translocation on viral gene expression. For this, BCBL-1 cells treated with various concentrations of neomycin were tested for ORF73 and ORF50 gene expression at 3 and 6 days posttreatment by quantitative real-time RT-PCR. At 3 days posttreatment, we observed a dose-dependent inhibition in ORF73 expression, with reductions of about 35, 40, 52, 56, 57, and 61% at neomycin concentrations of 25, 50, 100, 200, 500, and 1,000 μM, respectively (Fig. 3A). Neomycin treatment for 6 days had similar results (see Fig. S1B in the supplemental material). Surprisingly, at 3 days posttreatment, there was also a dose-dependent increase in ORF50 gene expression, with increases in ORF50 expression of about 2-, 1.3-, 4.3-, 51-, 99-, and 104-fold at neomycin concentrations of 25, 50, 100, 200, 500, and 1,000 μM, respectively (Fig. 3B). Neomycin treatment for 6 days showed similar results, with ORF50 expression peaking at a 55-fold increase with 200 μM neomycin (see Fig. S1C in the supplemental material).

FIG. 3.

Effect of neomycin treatment on latent ORF73, early lytic ORF50, ORF59, and late lytic gpK8.1A expression and virus production. cDNA prepared from BCBL-1 cells treated with the indicated concentrations of neomycin after 3 days was used for measurement of ORF73 (A) and ORF50 (B) gene expression by quantitative real-time PCR and normalized to expression of tubulin. (C) BCBL-1 cells treated with the indicated concentrations of neomycin for 3 days were harvested after 6 days posttreatment, washed, fixed with acetone, and incubated overnight with antibodies against LANA-1, ORF59, and K8.1A in 2% BSA. Cells were washed with PBS, incubated with secondary antibody conjugated with Alexa Fluor 488 (green) for 1 h at room temperature, washed, mounted, and visualized under a fluorescence microscope. Magnification, ×20. (D) Quantitative representation of immunofluorescence analysis. The numbers of LANA-1-, ORF59-, and gpK8.1A-positive cells were counted and represented as a percentage by considering the number of DAPI-positive live cells as 100%. (E) BCBL-1 cells were treated with increasing concentrations of neomycin. At 6 days posttreatment, supernatants were harvested, overlaid on HMVEC-d cells, and incubated for 2 h, DNA was extracted, and ORF73 DNA copy numbers were determined by quantitative real-time DNA PCR. (F) HMVEC-d cells pretreated with the indicated concentrations of neomycin for 1 h at 37°C were infected with KSHV (20 DNA copies/cell) for 2 h, the DNA was extracted, and ORF73 copy numbers were measured by quantitative real-time DNA PCR. **, P < 0.001; ***, P < 0.0001.

To confirm these results, BCBL-1 cells treated with different concentrations of neomycin for 6 days were fixed, permeabilized, and examined by IFA for latent LANA-1, early lytic ORF59, and late lytic envelope gpK8.1A proteins. A dose-dependent reduction in LANA-1 staining was observed in neomycin-treated cells (Fig. 3C). In contrast, the number of cells expressing ORF59 and gpK8.1A increased upon neomycin treatment in a dose-dependent manner (Fig. 3C). To ascertain that there was a significant inhibition in LANA-1 expression and increase in lytic gene expression upon neomycin treatment, the numbers of LANA-1⁺, ORF59⁺, and K8.1A⁺ cells were counted and normalized to the number of DAPI-positive live cells. In untreated BCBL-1 cells, more than 90% of cells expressed LANA-1, which was reduced to 60, 52, 42, and 35% with neomycin concentrations of 25, 100, 200, and 1,000 μM, respectively. KSHV ORF59 expression was seen in 6, 15, 25, 42, and 57% of untreated cells and cells treated with neomycin concentrations of 25, 100, 200, and 1,000 μM, respectively. KSHV gpK8.1A expression was seen in 4% of untreated cells, which increased to 7, 15, 25, and 38% with 25, 100, 200, and 1,000 μM neomycin, respectively (Fig. 3D). These results confirmed our quantitative RT-PCR results and suggested a potential role for angiogenin in the maintenance of KSHV latency and/or inhibition of the lytic cycle.

Neomycin treatment induces infectious KSHV virion production from BCBL-1 cells.

To determine whether increased lytic cycle gene expression also results in production of virion particles, uninfected HMVEC-d cells were infected with supernatants from BCBL-1 cells treated with various concentrations of neomycin for 6 days. After 2 h of infection, DNA from infected HMVEC-d cells was extracted, and the amount of internalized viral DNA was estimated by quantitative real-time DNA PCR for the ORF73 gene. Compared to untreated BCBL-1 cell culture supernatant, an increased quantity of internalized KSHV DNA was observed from neomycin-treated BCBL-1 cell supernatant (Fig. 3E). About 3.9-fold-more internalized virus was observed in the supernatants of BCBL-1 cells treated with 100 and 200 μM neomycin and about 2-fold more was observed in the supernatants of cells treated with 500 and 1,000 μM neomycin (Fig. 3E). Interestingly, in supernatant of cells treated with 100 μM neomycin, 1.2-fold more KSHV DNA was detected than in supernatant from TPA-treated BCBL-1 cells, and there was 5-fold more KSHV internalization than in untreated cell culture supernatant. Similar results were observed when supernatants from BCBL-1 cells treated with neomycin for 3 days were used for infection of HMVEC-d cells (data not shown). These results demonstrated that neomycin induced the lytic cycle of KSHV, resulting in the production of virus particles.

Since we observed a relatively reduced level of internalized virus in the supernatants of BCBL-1 cells treated with 500 and 1,000 μM neomycin, we next determined whether neomycin carried over with the BCBL-1 cell culture supernatant could have inhibited KSHV entry. HMVEC-d cells pretreated with 200, 500, or 1,000 μM neomycin or with 10 μM phosphatidylinositol 3-kinase (PI3K) inhibitor LY294002 were infected with stock KSHV (purified from TPA-induced BCBL-1 cells) for 2 h, and DNA internalization was measured. There was no significant inhibition in KSHV entry with all three concentrations of neomycin. In contrast, as previously demonstrated by us (40), LY294002 inhibited KSHV entry by about 80% (Fig. 3F). These results demonstrated that neomycin treatment had no effect on KSHV entry into the target endothelial cells.

To confirm the above finding, KSHV was purified from BCBL-1 cells treated with 200 μM neomycin or 20 ng/ml TPA for 3 days, and viral copy numbers were quantitated by real-time DNA PCR for the ORF73 gene. Compared to TPA treatment, we observed about a 2-fold increase in virus production upon neomycin treatment (Fig. 4A). Virus purified at 6 days post-neomycin or post-TPA treatment showed similar results (data not shown). To further confirm the infectivity of purified virus, the kinetics of viral DNA internalization in HMVEC-d cells was measured by real-time DNA PCR. There was a time-dependent increase in KSHV internalization (Fig. 4B) with virus purified upon neomycin treatment (3 days), which indicated that KSHV released from neomycin-treated cells was intact and infectious and 200 μM neomycin treatment was enough to induce a significant increase in virus production.

FIG. 4.

Quantitative real-time PCR demonstrating KSHV infection in HMVEC-d cells. (A) BCBL-1 cells were treated with 200 μM neomycin or 20 ng/ml TPA for 3 days, virus was purified from the supernatant, DNA was extracted, and ORF73 DNA copy numbers were determined. (B) HMVEC-d cells were infected with KSHV (10 DNA copies/cell) purified from neomycin-treated BCBL-1 cells for the indicated time points, DNA was extracted, and ORF73 DNA copy numbers were determined. (C) HMVEC-d cells were infected for 2, 8, and 24 h with KSHV (10 DNA copies/cell) purified from neomycin-treated BCBL-1 cells. Total RNA was extracted, converted to cDNA, and used for measurement of ORF73 and ORF50 gene expression by quantitative real-time PCR and normalized to that of tubulin. ***, P < 0.0001.

During de novo infection of HMVEC-d cells, KSHV latent ORF73 gene expression increased over time with transient concurrent expression of immediate-early expression of the lytic switch ORF50 gene, which subsequently declined (21). To determine whether KSHV purified from neomycin-treated BCBL-1 cells (3 days) follows similar kinetics, viral gene expression was monitored upon de novo infection in HMVEC-d cells. Latent ORF73 gene expression increased with time (Fig. 4C), whereas ORF50 expression was very high at 2 h p.i., reduced after 8 h, and reached a very low level at 24 h p.i. (Fig. 4C). These results confirmed our findings that blocking of nuclear translocation of angiogenin in BCBL-1 cells by neomycin leads to the inhibition of latent gene expression and induction of lytic gene expression, resulting in infectious virion production.

Neomycin treatment of latently infected TIVE-LTC cells results in blockage of cell survival, inhibition of latent gene expression, and induction of the lytic cycle.

Next, we determined whether neomycin treatment of latently infected endothelial cells also results in reduction of latent gene expression, lytic cycle activation, and a block in cell survival. An et al. (1) infected telomerase-immortalized human umbilical vein endothelial (TIVE) cells and isolated TIVE-LTC cells. Similar to BCBL-1 cells, TIVE-LTC cells express LANA-1 and other latent KSHV genes and thus serve as a good model system to study latency in endothelial cells (1). A dose-dependent reduction in TIVE-LTC cell survival was observed with neomycin treatment (Fig. 5A), with about 50% reduction in cell survival at 3 days posttreatment at 1,000 μM neomycin, while 100 μM paromomycin did not have any impact on TIVE-LTC cell survival (Fig. 5A). A trypan blue assay with 200 μM neomycin at 1, 2, and 3 days posttreatment confirmed this finding (Fig. 5B). Neomycin treatment had no significant effect on control TIVE cells (Fig. 5B).

FIG. 5.

Effect of neomycin treatment on TIVE-LTC cell survival and viral gene expression. (A) TIVE-LTC cells were treated with increasing concentrations of neomycin for 3 days and assayed for cell survival by MTT assay. Paromomycin (100 μM) treatment was used as a control. (B) The trypan blue assay was done with TIVE and TIVE-LTC cells at 0, 1, 2, and 3 days posttreatment with 200 μM neomycin, and the results are represented as the percentage of cell survival by considering the number of cells in untreated cells on each day as 100%. (C and D) cDNA prepared from TIVE-LTC cells left untreated (UT) or treated with 200 μM neomycin after 24, 36, and 48 h was used for measurement of ORF73 (C) and ORF50 (D) gene expression by quantitative real-time PCR and normalized to expression of tubulin. (E) TIVE-LTC cells treated with the indicated concentrations of neomycin for 3 days were stained with antibodies against ORF73, ORF59, and gpK8.1A. Cells were washed with PBS, incubated with secondary antibody conjugated with Alexa Fluor 488 (green) for 1 h at room temperature, washed, mounted, and visualized under a fluorescence microscope. Magnification, ×20. (F) TIVE-LTC cells were left untreated or treated with 200 μM neomycin for 3 days, virus was purified from the supernatant, DNA was extracted, and ORF73 DNA copy numbers were determined. **, P < 0.001; ***, P < 0.0001.

Analysis of viral gene expression in TIVE-LTC cells revealed an inhibition in ORF73 gene expression, with about 43, 39, and 40% inhibition at 24, 36, and 48 h post-neomycin treatment, respectively (Fig. 5C). In contrast, upon neomycin treatment ORF50 expression increased similar to that in BCBL-1 cells, with about 1.3-, 1.6-, and 2.5-fold increases at 24, 36, and 48 h, respectively: note the peak increase of 2.5-fold at 48 h post-neomycin treatment compared to the level in untreated cells (Fig. 5D). These results further substantiated our finding that, as in BCBL-1 cells, blocking nuclear translocation of angiogenin by neomycin also inhibited latent gene expression in latently infected TIVE-LTC cells.

To confirm these results, TIVE-LTC cells treated with different concentrations of neomycin for 3 days were fixed, permeabilized, and examined by IFA for LANA-1, ORF59, and gpK8.1A protein expression. A dose-dependent reduction in LANA-1 staining was observed in neomycin-treated cells (Fig. 5E). In contrast, the number of cells expressing ORF59 and gpK8.1A increased upon neomycin treatment in a dose-dependent manner (Fig. 5E). These results not only confirmed our quantitative RT-PCR data but also clearly demonstrated that nuclear translocation of angiogenin plays critical roles in the maintenance of KSHV latency and/or inhibition of the lytic cycle.

To determine if the increase in lytic gene expression resulted in increased virus production, supernatant collected from TIVE-LTC cells left either untreated or treated with neomycin was harvested and concentrated, DNA was extracted, and 500 ng of DNA was used to measure ORF73 DNA copy numbers. There was a 4-fold increase in virus production into the supernatant compared to the level in untreated cells (Fig. 5F). This indicated that similar to BCBL-1 cells, neomycin treatment in TIVE-LTC cells also increased lytic KSHV gene expression, which resulted in an increase in virus production into the supernatant. However, the viral copy numbers secreted into the supernatant from TIVE-LTC cells (Fig. 5F) were significantly lower than those in BCBL-1 cells (Fig. 4A) when similar cell numbers were used for neomycin treatment. This was similar to the difference observed in levels of ORF50 expression between these two cell types. This could be due to a combination of factors, such as the low percentage of TIVE-LTC cells (10 to 20%) that carry latent viral DNA, compared to 100% of BCBL-1 cells; lower KSHV genome copy numbers in TIVE-LTC cells, compared to >80 copies in BCBL-1 cells; and/or the inherent nature of KSHV latency in TIVE-LTC cells. Nevertheless, lytic cycle induction and increased virus production upon neomycin treatment in both these cell types confirmed the importance of nuclear angiogenin in maintaining KSHV latent gene expression.

Neomycin blocks BCBL-1 cell PLC-γ phosphorylation.

Regulation of the switch from the KSHV latent to lytic cycle is a very complex phenomenon, and several studies, including the use of TPA, have shown the roles played by PKC, ERK, and NF-κB in the lytic cycle switch. Neomycin is an aminoglycoside used to treat bacterial infection, which acts by blocking 70S ribosome biogenesis in bacteria. To determine whether neomycin-mediated lytic cycle activation could also be due to signal pathway activation as with TPA, BCBL-1 cells were treated with various concentrations of neomycin and Western blotted for phosphorylated forms of PLC-γ, AKT, and ERK.

Hu (17) reported in 1998 that neomycin treatment blocked nuclear translocation of angiogenin by blocking PLC-γ activation in endothelial cells. PLC plays a significant role in transmembrane signaling. In response to extracellular stimuli like hormones, neurotransmitters, and growth factors, PLC hydrolyzes phosphatidylinositol 4-5 bisphosphate (PIP2) to generate two second messengers: inositol 1,4,5-triphosphate (IP3) and diacylglycerol (DAG). At least four families of PLCs have been identified: PLC-β, PLC-γ, PLC-δ, and PLC-ɛ. PLC-γ is activated by both receptor and nonreceptor tyrosine kinases (27). We analyzed the effect of neomycin treatment on PLC-γ phosphorylation. In BCBL-1 cells, PLC-γ phosphorylation decreased with increasing neomycin concentrations, with about 60% inhibition at 100 and 200 μM neomycin compared to that in untreated cells (Fig. 6A). PLC-γ inhibition by neomycin, reduction in latent gene expression, and lytic cycle activation suggested that PLC-γ may play roles in maintenance of KSHV latency. In contrast, TPA treatment resulted in a 2.8-fold increase in PLC-γ phosphorylation (Fig. 6A), which suggested that both neomycin and TPA could be inducing the lytic cycle via different pathways.

FIG. 6.

Western blot analysis of phospho-PLC-γ, phospho-ERK, and phospho-AKT in neomycin-treated BCBL-1 cells and effect of PLC-γ inhibitor U73122 on BCBL-1 cell survival and viral gene expression. BCBL-1 cells were treated with the indicated concentrations of neomycin. At 3 days posttreatment, cells were harvested with radioimmunoprecipitation assay (RIPA) lysis buffer and Western blotted with phospho- and total PLC-γ (pPLCγ and tPLCγ, respectively) (A), phospho- and total AKT (pAKT and tAKT, respectively) (B), and phospho- and total ERK (pERK and tERK, respectively) (C) antibodies. BCBL-1 cells induced with TPA for 3 days were used as a positive control. (D) Nuclear and cytoplasmic extracts were prepared from BCBL-1 cells either untreated (lane 1) or treated with 5 μM U73122 (lane 2) or 20 ng/ml TPA (lane 3) and Western blotted with antiangiogenin antibody (top two panels). The nuclear extract blot was striped and probed with anti-lamin B and antitubulin antibodies (bottom two panels). (E) The MTT assay was done with BCBL-1 cells treated with increasing concentrations of the PLC-γ inhibitor U73122. OD, optical density. (F and G) cDNA was prepared from BCBL-1 cells treated with the indicated concentrations of U73122 and subjected to quantitative real-time RT-PCR for ORF73 (F) and ORF50 (G) gene expression and normalized to tubulin. Neomycin (200 μM) and TPA (20 ng/ml) treatment was used for comparison. (H) Supernatants harvested from BCBL-1 cells treated with the indicated concentrations of U73122 were overlaid on HMVEC-d cells, DNA was extracted after 2 h, and ORF73 DNA copy numbers were determined by quantitative real-time DNA PCR. **, P < 0.001; ***, P < 0.0001.

Neomycin blocks BCBL-1 cell AKT phosphorylation.

AKT, also referred to as PKB, is a protein kinase activated by insulin and various growth factors and plays a critical role in controlling the balance between cell survival and apoptosis (14). AKT is activated by a phospholipid activation loop involving PIP2 and PIP3 via PLC-γ and phosphoinositide-dependent kinase (PDK). Wang and Damania (45) showed that KSHV confers a survival advantage to endothelial cells via activation of the phosphatidylinositol 3-kinase (PI3K)/AKT pathway, and Peng et al. (36) demonstrated that inhibition of the PI3K/AKT pathway enhances gamma-2 herpesvirus lytic replication and facilitates reactivation from latency. Since AKT is one of the downstream survival signaling targets of PLC-γ, we tested the effect of neomycin on AKT activation. Phospho-AKT was diminished by about 40 and 60% with 100 and 200 μM neomycin, respectively, while TPA treatment did not show any significant reduction (Fig. 6B).

Neomycin induces ERK phosphorylation in BCBL-1 cells.

ERK1/2 is a mitogen-activated protein kinase (MAPK) family of signal molecules that is known to regulate vital cellular events, including transcription, cell division, apoptosis, and angiogenesis. Cohen et al. (7) described the essential role played by ERK signaling in TPA-induced lytic cycle reactivation. Our studies have shown that during de novo infection, KSHV induces ERK1/2 to modulate the initiation of viral gene expression and host cell genes (40). Hence, we determined the effect of neomycin on ERK1/2 phosphorylation in BCBL-1 cells. Neomycin treatment resulted in significant activation of ERK of about 2.7- and 3.4-fold with 100 and 200 μM concentrations, respectively (Fig. 6C). As shown previously by Cohen et al. (7) and Ford et al. (10), TPA treatment led to a 7.8-fold increase in phospho-ERK levels. These results demonstrated that irrespective of the upstream pathway mediating ERK phosphorylation, this activation is essential for lytic cycle induction by both TPA and neomycin.

These results suggested that induction of the KSHV lytic cycle by neomycin treatment is probably mediated by blocking PLC-γ, as well as AKT, and activation of ERK pathways.

PLC-γ inhibitor U73122 reduces BCBL-1 cell survival, blocks latent gene expression, and induces the lytic cycle and infectious virus production.

Since these results suggested a role for a block in PLC-γ phosphorylation by neomycin in inhibiting KSHV latency and inducing the lytic cycle, we next analyzed the effect of conventional PLC-γ inhibitor U73122 on BCBL-1 cell survival and viral gene expression. Nuclear extracts from BCBL-1 cells treated with U73122 showed a significant reduction in angiogenin levels compared to those in untreated cells (Fig. 6D, lanes 1 and 2). Interestingly, TPA treatment did not inhibit the nuclear translocation of angiogenin (Fig. 6D), thus suggesting that neomycin and TPA might be activating the lytic cycle via two different pathways. Cytoplasmic extracts from U73122-treated cells showed higher angiogenin levels than those from untreated and TPA-treated BCBL-1 cells. Blocking of angiogenin nuclear translocation by a conventional PLC-γ inhibitor clearly demonstrated that nuclear translocation of angiogenin is mediated by PLC-γ phosphorylation.

An MTT cell survival assay was done with increasing concentrations of U73122. At 3 days posttreatment, we observed a dose-dependent reduction in BCBL-1 cell survival with about 14, 17, 27, 39, and 42% inhibition at U73122 concentrations of 1, 2, 5, 10, and 20 μM, respectively (Fig. 6E). ORF73 expression was inhibited by about 82, 83, and 89% with U73122 concentrations of 1, 2, and 10 μM, respectively, and by 53% with 200 μM neomycin treatment (Fig. 6F). TPA treatments led to a 46% increase in ORF73 gene expression (Fig. 6F). In contrast, compared to untreated BCBL-1 cells, ORF50 gene expression increased about 80-, 76-, and 490-fold with U73122 concentrations of 1, 2 and 10 μM, respectively (Fig. 6G). Neomycin (200 μM) treatment induced the lytic cycle to similar levels nearly 70-fold, and TPA treatment upregulated the lytic cycle 10-fold. Interestingly, lytic cycle activation was 7-fold higher with neomycin treatment and 49-fold higher with U73122 than with TPA treatment. These results were interpreted as suggestive of a novel role for PLC-γ in mediating KSHV latent gene expression and inhibiting lytic gene expression.

To determine whether lytic cycle activation by PLC-γ inhibitor U73122 also results in productive virus replication, supernatants harvested from BCBL-1 cells treated with increasing concentrations of U73122, 200 μM neomycin, and TPA were used to infect HMVEC-d cells. Compared to untreated BCBL-1 cell culture supernatant, there were about 2-, 3.2-, and 4.3-fold increases in ORF73 DNA internalization with U73122 concentrations of 1, 2, and 10 μM, respectively, and 3.6- and 2.4-fold increases with 200 μM neomycin and 20 ng/ml TPA, respectively (Fig. 6H). Compared to supernatants from TPA-treated cells, increased internalized KSHV DNA was detected with supernatants from both neomycin- and U73122-treated cells. These results clearly substantiated our finding that neomycin and U73122 treatment could block latent gene expression and induce the lytic cycle by inhibiting PLC-γ phosphorylation. This is the first demonstration of PLC-γ and AKT downregulation in the inhibition of KSHV latent gene expression and induction of lytic genes in BCBL-1 cells.

ERK inhibitor U0126 blocks lytic cycle activation induced by neomycin, U73122, and TPA.

We observed an increase in ERK phosphorylation during both neomycin treatment and TPA treatment in BCBL-1 cells (Fig. 6C). To confirm the importance of ERK activation in mediating lytic cycle induction during neomycin treatment, BCBL-1 cells were treated with either neomycin alone or in combination with U0126. As shown in Fig. 7A, there was a significant increase in ORF50 expression upon neomycin treatment, which was inhibited by about 50% when U0126 was added along with neomycin (Fig. 7A). Similar results were also observed with TPA and U0126 treatment (Fig. 7B and C). PLC-γ inhibitor U73122 treatment showed very high ORF50 gene expression levels, and with U0126 treatment expression was reduced to nearly basal levels (Fig. 7C). These results further supported our observation that ERK activation is critical for lytic cycle induction by neomycin and PLC-γ inhibitor treatment in BCBL-1 cells.

FIG. 7.

Effect of ERK inhibitor U0126 on ORF50 expression in neomycin-treated BCBL-1 cells. BCBL-1 cells were left untreated or were treated with 200 μM neomycin with and without 10 μM U0126 (A), 20 ng/ml TPA with and without 10 μM U0126 (B), and 5 μM U73122 with and without 10 μM U0126 (C). Three days posttreatment, the cells were harvested, cDNA was prepared and used for measurement of ORF50 gene expression by quantitative real-time PCR, and the results were normalized to those for tubulin. **, P < 0.001; ***, P < 0.0001.

Silencing of angiogenin in BCBL-1, TIVE-LTC, and HeLa cells inhibits latent gene expression and induces lytic cycle activation.

It is reasonable to argue that reduced cell survival, inhibition of latency, and lytic cycle activation could have been merely due to neomycin's effect on signal pathway activation and not due to prevention of nuclear translocation of angiogenin. To test this possibility, cell survival, signal pathway activation, and viral gene expression were monitored after silencing of angiogenin. 293T cells were transfected with lentivirus vectors expressing si-angiogenins 1, 2, 3, 4, and 5 (si-ang1 to -5), and real-time PCR was done to quantitate the knockdown in angiogenin levels. There was an 80% reduction in angiogenin expression with all five antiangiogenin siRNAs tested (see Fig. S2A in the supplemental material), and ang-ELISA substantiated this finding (see Fig. S2B in the supplemental material). To confirm these results, 293T cells were transfected with pcDNA-GFP or ang-GFP, along with si-ang1 and -2, for 24 h and observed by IFA for green fluorescent protein (GFP) expression. si-ang did not have any effect on pcDNA-GFP expression, while there was a significant reduction in ang-GFP expression with si-ang1 and -2 (see Fig. S2C in the supplemental material).

From transfected HEK 293T cells, lentivirus-expressed si-ang1, si-ang2, si-lamin, and si-RL were purified. BCBL-1 cells were transduced with lentivirus-expressed si-ang, si-RL, or si-lamin. An MTT assay was done to determine the effect of silencing of angiogenin on BCBL-1 cell survival. BCBL-1 cells treated with si-ang1 and si-ang2 showed 50% less survival than si-RL- or si-lamin-transduced BCBL-1 cells (Fig. 8A). These results correlated well with the neomycin-treated BCBL-1 cell survival assay (Fig. 2C) and confirmed the importance of angiogenin in mediating BCBL-1 cell survival.

FIG. 8.

Viral gene expression in si-angiogenin-treated cells. (A) BCBL-1 cells were transduced with lentivirus expressing si-RL, si-lamin, si-ang1, and si-ang2 and assayed for cell survival by the MTT assay. (B, C, and D) BCBL-1 cells were transduced with si-RL, si-lamin, si-ang1, and si-ang2 and Western blotted for phospho- and total PLC-γ (pPLCγ and tPLCγ, respectively) (B), phospho- and total AKT (pAKT and tAKT, respectively) (C), and phospho- and total ERK (pERK and tERK, respectively) (D). cDNA prepared from BCBL-1 cells (E and F) and TIVE-LTC cells (G and H) transduced with lentivirus harboring si-ang, si-RL, or si-lamin was used for quantitative real-time PCR to measure ORF73 (E and G) and ORF50 (F and H) gene expression and normalized to expression of tubulin. (I and J) HeLa cells were transfected with si-ang, si-RL, or si-lamin and infected with KSHV (10 DNA copies/cell) 24 h posttransfection. Forty-eight hours postinfection, total RNA was collected, cDNA was prepared, and ORF73 (I) and ORF50 (J) copy numbers were estimated by quantitative real-time PCR. **, P < 0.001; ***, P < 0.0001.

Identical to our observation with neomycin treatment, si-ang transduction resulted in about a 50 to 60% reduction in PLC-γ phosphorylation compared to the level in si-RL-transduced cells (Fig. 8B). Phospho-AKT levels were also inhibited about 60 to 70% with si-angiogenin (Fig. 8C), while ERK phosphorylation increased 3.5-fold and 3-fold with si-ang1 and si-ang2, respectively (Fig. 8D). The nearly identical phosphorylation profiles of PLC-γ, AKT, and ERK with neomycin as well as si-angiogenin further supported our finding that angiogenin could be involved in latency establishment via phosphorylation of the PLC-γ pathway and neomycin could be inhibiting latent gene expression by blocking PLC-γ phosphorylation.

To determine the effect of silencing angiogenin on viral gene expression, RNA was extracted and quantitated by real-time PCR for ORF73 and ORF50 gene expression. Both si-ang1- and si-ang2-treated cells displayed a reduction in latent ORF73 gene expression. Compared to si-RL-transduced BCBL-1 cells, we observed 88 and 87% inhibition in ORF73 expression with si-ang1 and si-ang2, respectively (Fig. 8E). In contrast, ORF50 expression increased 12- and 10-fold with si-ang1 and si-ang2, respectively (Fig. 8F). These results were in concordance with our neomycin and U73122 studies, suggesting angiogenin expression and nuclear translocation are critical for the maintenance of latent gene expression.

We observed a reduction in latent gene expression in TIVE-LTC cells with neomycin, which was similar to the level in BCBL-1 cells. To further validate our findings, we analyzed the effect of si-ang on viral gene expression in TIVE-LTC cells. Similar to the results observed with BCBL-1 cells, compared to si-RL-transduced TIVE-LTC cells, ORF73 expression was inhibited by 48 and 40% with si-ang1 and si-ang2, respectively (Fig. 8G). ORF50 gene expression was elevated 2.7- and 2.0-fold with si-ang1 and si-ang2, respectively (Fig. 8H). These results confirmed that in both BCBL-1 and TIVE-LTC cells that are latently infected with KSHV, angiogenin had a critical role to play in the maintenance of latency or in the suppression of lytic cycle reactivation.

As both BCBL-1 cells and TIVE-LTC cells were latently infected, we next analyzed the effect of silencing of angiogenin during de novo KSHV infection. We could not use HMVEC-d cells as survival of these cells was severely impaired with si-ang treatment (data not shown). HeLa cells have been very well characterized for angiogenin expression, and angiogenin is known to translocate into the nucleus, bind to ribosomal DNA (rDNA), and promote ribosome biogenesis (20). Hence, we transfected HeLa cells with lentivirus-expressed si-ang, si-RL, or si-lamin for 24 h, infected them with KSHV (10 DNA copies/cell) for 48 h, and analyzed the expression of viral genes by quantitative real-time PCR. As shown in Fig. 8I, ORF73 expression was inhibited by 65 and 87% with si-ang1 and si-ang2, respectively, while ORF50 expression was upregulated 2.7- and 2.1-fold with si-ang1 and si-ang2, respectively (Fig. 8J).

These results demonstrated that silencing of angiogenin could block latency and induce lytic gene expression in latently infected BCBL-1 cells or TIVE-LTC cells and during de novo KSHV infection, and more importantly, all of these results further substantiated our finding that angiogenin plays a critical role in the maintenance of KSHV latency.

Neomycin treatment inhibits latent ORF73 gene expression and induces lytic ORF50 gene expression in HMVEC-d cells.

To determine the effect of neomycin on viral gene expression during primary infection of HMVEC-d cells, serum-starved HMVEC-d cells (∼60 to 70% confluence) were pretreated with 200 μM neomycin for 1 h and infected with KSHV (10 DNA copies/cell). Neomycin treatment had no effect on latent ORF73 gene expression until 48 h p.i., while a 40% inhibition was observed at 72 h p.i. (Fig. 9A). Interestingly, similar to BCBL-1 and TIVE-LTC cells, ORF50 gene expression increased by 1.4-, 2.5-, and 5.9-fold at 24, 48, and 72 h p.i., respectively, by neomycin treatment in HMVEC-d cells (Fig. 9B). A dose-response study with 200, 500, and 1,000 μM neomycin treatment yielded similar results (data not shown). These results clearly demonstrate that nuclear angiogenin could be required for the establishment of latency during de novo KSHV infection of HMVEC-d cells (at later time points), and neomycin treatment could act either by inhibiting ORF73 expression or by inducing signal pathway activation, resulting in the upregulation of lytic cycle activation.

FIG. 9.

Effect of neomycin treatment on latent and lytic KSHV gene expression during de novo KSHV infection of endothelial cells and effect of angiogenin treatment on PLC-γ, AKT, and ERK activation and ORF73 gene expression in HMVEC-d cells. (A and B) HMVEC-d cells were left untreated or were pretreated with 200 μM neomycin for 1 h at 37°C and infected with KSHV (10 DNA copies/cell) for 24, 48, and 72 h. cDNA was prepared from these cells and assayed for ORF73 (A) and ORF50 (B) expression by quantitative real-time RT-PCR and normalized to tubulin. (C to E) HMVEC-d cells were treated with 1 μg/ml angiogenin, and at the indicated time posttreatment, cells were harvested and Western blotted for phospho- and total PLC-γ (pPLCγ and tPLCγ, respectively) (C), phospho- and total AKT (pAKT and tAKT, respectively) (D), and phospho- and total ERK1/2 (pERK and tERK, respectively) (E). (F) Serum-starved HMVEC-d cells were pretreated with 1 μg/ml angiogenin for 1 h at 37°C and infected with KSHV (10 DNA copies/cell). At the indicated time points, total RNA was extracted and then converted to cDNA, and ORF73 copy numbers were measured by quantitative real-time PCR and normalized to those for tubulin. (G) Serum-starved HMVEC-d cells were pretreated with 200 μM neomycin or 2 μM U73122 for 1 h at 37°C before being treated with 1 μg/ml angiogenin and infected with KSHV (20 DNA copies/cell). At 24 h postinfection, total RNA was extracted and converted to cDNA, and ORF73 copy numbers were measured by quantitative real-time PCR and normalized to those for tubulin. *, P < 0.01; **, P < 0.001; ***, P < 0.0001.

Angiogenin treatment induces PLC-γ, AKT, and ERK1/2 activation and induces latent gene expression.

We hypothesized that if neomycin treatment and si-angiogenin could block latency by inhibiting PLC-γ and AKT phosphorylation, then angiogenin should be able to activate PLC-γ and AKT in these cells to aid in the establishment of latency. BCBL-1 and TIVE-LTC cells express LANA-1, which augmented angiogenin expression (37); hence, BCBL-1 and TIVE-LTC are not the ideal cells with which to study signal pathway activation by angiogenin as they have increased endogenous angiogenin expression. Hu (17) in 1998 reported that angiogenin might be controlling its own nuclear translocation via the activation of PLC-γ in endothelial cells. Our previous studies showed that HMVEC-d cells expressed minimal levels of angiogenin and that KSHV infection upregulated both expression and secretion of angiogenin in these cells (37). Hence, to test the effect of angiogenin addition on PLC-γ activation, HMVEC-d cells were treated with 1 μg of angiogenin and the lysates were Western blotted for phospho-PLC-γ. We observed 1.4-, 1.5-, and 2.0-fold increases in PLC-γ phosphorylation at 5, 10, and 15 min post-angiogenin treatment, respectively (Fig. 9C). There was no change in total PLC-γ levels at all time points.

Both our previous studies (37) and our current observations indicate the importance of angiogenin in cell survival. Moreover, we also observed an inhibition of AKT phosphorylation during both neomycin treatment and during si-angiogenin transduction. To determine whether angiogenin treatment could induce AKT phosphorylation, angiogenin-treated HMVEC-d cells were tested by Western blotting for phospho-AKT. Angiogenin induced 2.2-, 2.4-, and 2.8-fold increases in AKT phosphorylation at 5, 10, and 15 min post-angiogenin treatment, respectively (Fig. 9D), while there was no change in total AKT levels.

ERK phosphorylation is critical for both latent and lytic gene expression during de novo KSHV infection of HMVEC-d cells (40). Studies by Xie et al. (47) also showed that the MEK/ERK, JNK, and p38MAPK pathways were constitutively activated in BCBL-1 cells, while Liu et al. (25) demonstrated the activation by angiogenin of ERK1/2 in human umbilical vein endothelial cells (HUVECs). Similarly, we observed about 1.2-, 3.1-, and 3.4-fold increases in ERK phosphorylation at 5, 10, and 15 min post-angiogenin treatment in HMVEC-d cells, while the total ERK levels remain unchanged (Fig. 9E).

To determine whether PLC-γ, AKT, and ERK activation mediated by angiogenin could result in latent gene expression, HMVEC-d cells were pretreated with 1 μg/ml angiogenin, and ORF73 gene expression was measured by quantitative real-time PCR. As shown in Fig. 9F, ORF73 expression increased about 1.3-, 1.4-, and 1.8-fold at 2, 8, and 24 h post-KSHV infection with angiogenin treatment compared to the level in untreated cells. These results further confirmed our finding that angiogenin has roles to play in the maintenance of latency, which could be mediated via the activation of PLC-γ and AKT. To ascertain that the reduction in latent gene expression is primarily due to inhibition of nuclear translocation of angiogenin by neomycin and to prove the effect is mediated by PLC-γ inhibition, serum-starved HMVEC-d cells were pretreated with 200 μM neomycin or 2 μM U73122 for 1 h before being treated with 1 μg/ml angiogenin for 1 h and then infected with KSHV (20 DNA copies/cell). Twenty-four hours postinfection, cDNA was prepared and used to measure ORF73 RNA copy numbers. There was a 2-fold increase in ORF73 expression with angiogenin addition, while expression was inhibited by 60 and 75% by neomycin and U73122 treatment (Fig. 9G). These results, together with those shown in Fig. 6A and D, substantiated our finding that nuclear translocation of angiogenin is critical for latent gene expression, and both neomycin and U73122 blocked ORF73 expression by preventing the nuclear translocation of angiogenin by inhibiting the PLC-γ pathway.

DISCUSSION

Our previous studies with endothelial HMVEC-d cells focused predominantly on the functional significance of angiogenin during KSHV infection (37). When we sought to understand the importance of angiogenin in latently infected B cells, we noticed accidentally that neomycin treatment results in inhibition of latent gene expression and induction of the lytic cycle. The comprehensive exploration of this observation presented here demonstrates the surprising novel findings that angiogenin, involved in angiogenesis, plays important roles in maintenance of KSHV latency and identified a so-far-unexplored PLC-γ pathway in KSHV latency (Fig. 10). Detection of more substantial quantities of infectious KSHV in the supernatants of neomycin-treated BCBL-1 cells than the currently used TPA treatment also identified a potential new methodology to obtain a higher yield of infectious KSHV by using significantly less expensive neomycin without using TPA, a known carcinogen.

FIG. 10.

Schematic representation depicting the mode of action of angiogenin in maintaining latency. The results presented in the present studies demonstrate the following: KSHV-upregulated angiogenin is expressed both during de novo KSHV infection and in latently infected cells. Angiogenin induces PLC-γ phosphorylation, which in turn is required for nuclear translocation of angiogenin. PLC-γ phosphorylation also results in AKT phosphorylation, which is required for cell survival, and ERK phosphorylation, which is necessary for latent gene expression and for the establishment of latency. Neomycin treatment blocked nuclear translocation of angiogenin and inhibited PLC-γ and AKT phosphorylation, leading to cell death, reduction in latent gene expression, and an increase in lytic gene expression. U73122, a conventional PLC-γ inhibitor, reduced cell survival, blocked latent gene expression, and induced lytic gene expression. Angiogenin silencing also reduced cell survival, blocked latent gene expression, and increased lytic gene expression. Collectively, these studies show that KSHV-induced angiogenin plays roles in the maintenance of latency via the PLC-γ pathway.

The hallmark of KSHV infection, as in other herpesviruses, is the establishment of its latent infection, periodic reactivation, and reinfection. Beside relying on its latent gene products, KSHV must also have evolved strategies to regulate host gene expression to create an intracellular environment that is conducive for successful establishment and maintenance of a latent infection that includes methods to overcome apoptosis, autophagy, transcriptional restriction, and lytic cycle control, as well as host intrinsic, innate, and adaptive immune responses (3, 12). Our earlier and recent studies provide evidence for the above notion in that KSHV-induced COX-2, and its product, prostaglandin E₂ (PGE₂), play roles in latent ORF73/vCyclin/vFLIP promoter activity (42). Latent gene expression was significantly reduced by COX-2 inhibitors, and this inhibition was relieved by exogenous supplementation with PGE₂ (42). Several lines of evidence presented here demonstrate that KSHV has evolved to utilize another host factor, the growth factor angiogenin, for its advantage. Detection of higher levels of angiogenin in KSHV-infected PEL cells, negligible levels in KSHV⁻ and EBV⁺ B-lymphoma cells, and reduction of KSHV⁺ B-cell survival but not that of KSHV⁻ or EBV⁺ B cells by preventing nuclear transport of angiogenin, together with the ability of LANA-1 to induce angiogenin expression and secretion (37), clearly suggest a specific association of angiogenin with KSHV biology.

Inhibition of PLC-γ by neomycin, resulting in reduced latent gene expression and activation of the lytic cycle, clearly indicates a role for a PLC-γ pathway in the maintenance of KSHV latency. Activation of PLC-γ and upregulation of ORF73 expression by angiogenin substantiated these findings and further validate the importance of signal molecules in controlling both latent and lytic gene expression. It could be argued that the reduction in LANA-1 expression could have been due to neomycin cytotoxicity at higher concentrations, as it was normalized to tubulin in our real-time PCR analysis. To rule out this possibility, we counted the number of LANA-1-positive cells and normalized it to the number of live cells as seen by DAPI staining and observed identical results. Additionally, si-ang treatment showed similar results, substantiating the importance of angiogenin in KSHV latency establishment. Moreover, the cell survival assay studies with TIVE, TIVE-LTC, KSHV⁺, KSHV⁻, EBV⁺, and EBV⁻ cells clearly suggest that neomycin treatment specifically targets only cells latently infected with KSHV and does not affect normal uninfected cells or cells infected with EBV. Our angiogenin expression studies also showed (Fig. 1A and B) that only KSHV-infected cells express and secrete significant amounts of angiogenin, thus emphasizing a role for angiogenin in KSHV biology and the establishment of latency.

Our studies indicate a highly interlinked loop between LANA-1 induction of angiogenin gene expression, PLC-γ/AKT phosphorylation, angiogenin nuclear translocation, LANA-1 expression, latency, and cell survival. The existence of such a loop is supported by several lines of evidence, such as inhibition of PLC-γ and AKT activation by neomycin, activation of both of these molecules by angiogenin, KSHV gene expression and cell survival profiles seen with si-angiogenin transduction studies identical to those observed with neomycin and PLC-γ inhibitor U73122, and identical results observed both in latently infected cells and during de novo KSHV infection. These studies suggest that KSHV could be inducing angiogenin for PLC-γ activation that is required for latency, as well as for the transport of angiogenin into the nucleus and subsequent functions, including maintenance of latency mediated by angiogenin.

AKT is downregulated by neomycin and upregulated by angiogenin. The role of AKT in the maintenance of latency has been shown in gamma-2 herpesvirus MHV-68, where inhibition of the PI3K/AKT pathway enhanced lytic replication and facilitated reactivation from latency (36). AKT, a prosurvival molecule, is also controlled by a phospholipid pathway and is known to be critical for cell proliferation and antiapoptosis (14). Our AKT phosphorylation results are in concordance with the MTT cell survival assay. Neomycin-mediated reduction in cell survival could be multifactorial, as nuclear translocation of angiogenin is known to be critical both for survival of endothelial cells and for cell proliferation mediated by angiogenin and other growth factors, such as VEGF and FGF, whose functions are also controlled by angiogenin. KSHV might be producing angiogenin for activation of the AKT cell survival pathway, which when blocked during neomycin or PLC-γ inhibitor treatment, results in cell death. This is the first report showing the effect of blocking nuclear translocation of angiogenin on cell survival mediated by AKT.

ERK1/2 phosphorylation was activated by both neomycin treatment and angiogenin addition. ERK1/2 activation is known to be critical for the downstream phosphorylation of a number of transcription factors, such as Elk-1, c-Fos, c-Jun, ATF, CREB, ETS-2, NF-κB, serum response factor (SRF), and components of the AP1 complex (9). It is interesting that the ORF73 promoter/enhancer region possesses several transcription factor binding motifs which include sites for c-Myc, c-Jun, IRF1/2, AP-1, SP-1, Nf-1, Oct-6, AP-2, and Oct-1 (39). Similarly, Oct-1, CREB, AP-1, c-Jun, cellular CCAAT/enhancer binding protein, NF-κB, and other transcription factor binding motifs are also present in the ORF50 promoter region. This clearly indicates the importance of ERK1/2 activation and substantiates our previous finding (40) that ERK1/2 activation plays key roles in both latent and lytic gene expression. Further studies will be required to identify the upstream signal molecule leading to ERK1/2 activation that is responsible for lytic cycle activation upon neomycin treatment.

The viral and host factors dictating KSHV latency maintenance, lytic cycle control, and the switch to the lytic cycle are the most sought after questions in the KSHV field and remain to be fully understood (13). Since KSHV ORF50 (RTA) is the master initiator and regulator of the lytic cycle, latency maintenance is intimately linked to controlling ORF50 gene expression, as well as antiapoptosis, cell cycle regulation, and cell survival. Although several studies indicate roles for signal pathways in lytic cycle activation, it is clear that no single pathway could decide the fate of KSHV latency versus lytic switch (13). LANA-1 has been shown to directly downregulate RTA transcription by targeting its promoter (22). The purpose of LANA-1-mediated upregulation of angiogenin expression upon establishment of latency (37) could be to aid in the maintenance of latency via PLC-γ activation and for regulation of nuclear translocation of angiogenin, which is also mediated by PLC-γ activation. Since angiogenin is known to bind DNA to induce 45S RNA synthesis, angiogenin's contribution to KSHV latent gene expression could be due to a combination of the following possibilities: (i) angiogenin may be modulating KSHV latent/lytic gene expression via its direct interactions with latency/lytic/host promoters, through interactions with host factors, or both; (ii) induction of signal pathways modulating viral/host gene expression; and/or (iii) induction of transcription factors modulating viral/host genes.

Whether angiogenin binds directly or indirectly to latency and ORF50 promoters and whether it augments or suppresses their transcription, respectively, is not known at present. Reduction of angiogenin levels in the nucleus might have reduced angiogenin-dependent latent gene expression and/or direct suppression of the ORF50 promoter, resulting in lytic gene expression. The effect on ORF50 gene expression could also be indirectly through the reduction in LANA-1 expression, thus relieving LANA-1-mediated negative regulation of the ORF50 promoter. Reduction in cell survival by neomycin and the PLC-γ inhibitor U73122 could be a combination of the above mentioned factors regulating latent gene expression, which are themselves powerful mediators of antiapoptosis and cell survival.

The design of studies to elucidate the role played by angiogenin and PLC-γ in the regulation of ORF73 and ORF50 promoters is hampered by the fact that all primary cells express low levels of angiogenin, these are poorly transfectable, and unfortunately all transformed or cancerous cell lines that are easily transfectable express high levels of endogenous angiogenin. Studies are in progress to determine the impact of angiogenin on the ORF73 promoter, using si-ang-transduced cells, which are beyond the scope of the present work.

Besides the interlinked loop between angiogenin induction by LANA-1, angiogenin's ability to induce PLC-γ/AKT phosphorylation, angiogenin nuclear translocation, and LANA-1 expression, there is evidence for the possibility of additional factors in these interlinked loops. For example, vFLIP, coded for by another latent gene generated from a common promoter and message coding for LANA-1, is a well-known inducer of NF-κB, which is also involved in COX-2 induction (29). We have shown that NF-κB inhibition in KSHV-infected endothelial cells inhibits angiogenin (38). Our recent studies with COX-2 inhibitor and siRNA also show that COX-2 plays roles in angiogenin induction, latent gene expression, and cell survival (41) and that vFLIP is involved in COX-2 induction (unpublished data). Further extensive experiments are essential to fully comprehend these links and the roles of the various factors in KSHV genome epigenetics, latency, and cell survival. Our study also exposes an interesting regulatory loop between KSHV induction of angiogenin expression and angiogenin's role in the establishment and maintenance of KSHV latency. Hence the observed effect of inhibition of angiogenin by si-ang, inhibition of nuclear transport of angiogenin, and inhibition of PLC-γ on KSHV latent gene expression and cell survival in the context of KSHV-infected cells is probably not just due to angiogenin's role as the direct activator of these processes but is probably due to the combinatorial effect on reduction in KSHV latent gene expression and its downstream consequences, including reduction in angiogenin expression and cell survival. In addition, PLC-γ by inducing AKT may also be mediating cell survival, and consequently, inhibition of PLC-γ by neomycin or by U73122 may affect cell survival due to AKT reduction (Fig. 10). The observed regulatory loop of angiogenin in KSHV biology opens up a new avenue that could be potentially exploited for effective control of KSHV and PEL.

Overall, our studies demonstrating induction of angiogenin by LANA-1, angiogenin detection in KSHV latently infected cells that is tightly regulated by PLC-γ activation, and their role in latent gene expression indicate the plasticity of KSHV to utilize an angiogenic factor for its survival advantage and have opened up a new paradigm in KSHV latency and cell survival. Further studies are in progress to define the mechanism behind the roles of angiogenin and PLC-γ in KSHV latency and cell survival. Since these studies also demonstrate that the potential use of neomycin, or a PLC-γ inhibitor, could lead to the elimination of KSHV latent infection, studies are also in progress to test the efficacies of neomycin and the PLC-γ inhibitor in preventing BCBL-1 growth in an SCID/NOD mice model.