Kaposi's sarcoma-associated herpesvirus latency-associated nuclear antigen interacts with multifunctional angiogenin to utilize its antiapoptotic functions

Abstract

Kaposi's sarcoma-associated herpesvirus (KSHV) is etiologically associated with the angioproliferative Kaposi's sarcoma (KS). KSHV infection and the expression of latency-associated nuclear antigen (LANA-1) upregulates the angiogenic multifunctional 123-amino-acid, 14-kDa protein angiogenin (ANG), which is detected in KS lesions and in KSHV-associated primary effusion lymphoma (PEL) cells. ANG knockdown or the inhibition of ANG's nuclear translocation resulted in decreased LANA-1 gene expression and reduced KSHV-infected endothelial and PEL cell survival (Sadagopan et al., J. Virol. 83:3342-3364, 2009). Further studies here demonstrate that LANA-1 and ANG colocalize and coimmunoprecipitate in de novo infected endothelial cells and in latently infected PEL (BCBL-1 and BC-3) cells. LANA-1 and ANG interaction occurred in the absence of the KSHV genome and other viral proteins. In gel filtration chromatography analyses of BC-3 cell lysates, ANG coeluted with LANA-1, p53, and Mdm2 in high-molecular-weight fractions, and LANA-1, p53, and Mdm2 also coimmunoprecipitated with ANG. LANA-1, ANG, and p53 colocalized in KSHV-infected cells, and colocalization between ANG and p53 was also observed in LANA-1-negative cells. The deletion constructs of ANG suggested that the C-terminal region of amino acids 104 to 123 is involved in LANA-1 and p53 interactions. Silencing ANG or inhibiting its nuclear translocation resulted in decreased nuclear LANA-1 and ANG levels, decreased interactions between ANG-LANA-1, ANG-p53, and LANA-1-p53, the induction of p53, p21, and Bax proteins, the increased cytoplasmic localization of p53, the downregulation of Bcl-2, the increased cleavage of caspase-3, and the apoptosis of cells. No such effects were observed in KSHV-negative BJAB cells. The phosphorylation of p53 at serine 15, which is essential for p53 stabilization and for p53's apoptotic and cell cycle regulation functions, was increased in BCBL-1 cells transduced with short hairpin RNA targeting ANG. Together, these studies suggest that the antiapoptosis observed in KSHV-infected cells and the suppression of p53 functions are mediated in part by ANG, and KSHV has probably evolved to utilize angiogenin's multiple functions for the maintenance of its latency and cell survival. Thus, targeting ANG to induce the apoptosis of cells latently infected with KSHV is an attractive therapeutic strategy against KSHV infection and associated malignancies.

Fig 1

KSHV LANA-1 interaction with angiogenin in various KSHV-infected cells. (A and B) HMVEC-d cells grown to ∼60 to 70% confluence were serum starved for 8 h and either infected with 20 KSHV DNA copies/cell for 48 h or left uninfected. An equal quantity of protein from each lysate was immunoprecipitated with equal concentrations of either rabbit anti-ANG or anti-LANA-1 antibodies, separated with a 7.5% SDS-PAGE gel, and subjected to Western blotting with rat anti-LANA-1 monoclonal antibodies (A) or separated with a 12.5% gel and subjected to Western blotting with goat anti-ANG antibodies (B). (C and D) Nuclear extracts were prepared from BCBL-1 cells and immunoprecipitated with rabbit-IgG, anti-ANG, or anti-LANA-1 antibodies, separated with a 7.5% gel, and subjected to Western blotting for LANA-1 (C) or separated with a 12.5% gel and subjected to Western blotting for angiogenin (D). Five percent input and lamin loading controls are shown. (E) BCBL-1 cell lysates were immunoprecipitated with rabbit-IgG or anti-LANA-1 antibodies, separated with a 10% gel, and subjected to Western blotting for PML. Five percent input control for LANA-1, PML, and β-actin loading controls are shown.

Fig 2

LANA-1 colocalization with ANG in the nucleus of KSHV-infected cells. (A) Serum-starved ∼60 to 70% confluent HMVEC-d cells were either infected with 20 KSHV DNA copies/cell for 48 h or left uninfected. After fixation and permeabilization, HMVEC-d, BCBL-1, and BC-3 cells were stained for LANA-1 (red, rabbit anti-LANA-1 antibodies) and angiogenin (green, goat anti-ANG antibodies). DAPI (blue) was used as a nuclear stain and merged with LANA-1 and ANG images. The white inserts are enlarged on the right. (B) HMVEC-d cells infected with KSHV as described for panel A were stained for LANA-1, PML bodies, and DAPI, and the images were merged. White inserts are shown as enlarged images. (C) HMVEC-d cells infected with KSHV were stained for LANA-1, ANG, DAPI, and the nucleolar marker fibrillarin, and the images were merged.

Fig 3

KSHV LANA-1 and angiogenin interaction in overexpressed 293T cells. (A) Nuclear (nuc) and cytoplasmic (cyto) extracts were prepared from 293T cells transfected with full-length GFP tagged angiogenin (FL-ANG GFP), separated with a 10% gel, and subjected to Western blotting for GFP. The purity of the extracts was checked by Western blotting for β-actin and lamin. (B) U2OS cells were transfected with FL-ANG GFP plasmid and stained for GFP and DAPI after fixing and permeabilization. The white box is shown as an enlarged image. (C) U2OS cells transfected with ANG-GFP and LANA-1 plasmids were stained for ANG (green), LANA-1 (red), or DAPI (blue), and the images were merged. The white box is shown as an enlarged image. (D and E) 293T cells were transfected with pcDNA-GFP empty vector, FL-ANG GFP, or pcINeo full-length LANA-1 plasmids and subjected to IP with rabbit anti-LANA-1 IgG antibodies, separated with a 10% gel, and subjected to Western blotting with mouse anti-GFP monoclonal antibodies (D) or separated with a 12.5% gel and subjected to Western blotting with goat anti-ANG antibodies (E). (F) 293T cells transfected as described above were subjected to IP with rabbit anti-ANG antibodies, separated with a 7.5% gel, and subjected to Western blotting with rabbit anti-LANA-1 antibodies. Fiver percent input and β-actin loading controls are shown.

Fig 4

Gel chromatography for LANA-1, ANG, and p53 in KSHV-infected BC-3 cells. (A) Cell lysates from BC-3 cells were separated by Superdex 200 HR column gel filtration chromatography. Ten gel filtration fractions, including the whole-cell lysate (WCL), were immunoblotted for LANA-1 (7.5% gel), ANG (12.5% gel), and p53 (10%). (B and C) Nuclear extracts from BC-3 cells were subjected to IP with rabbit IgG and rabbit anti-angiogenin IgG antibodies, separated with a 7.5% gel, and subjected to Western blotting for LANA-1 (B) and for p53 (C). (D) Whole-cell lysates from 293T cells were subjected to IP as described above and subjected to Western blotting for p53. Input controls for p53 and ANG are shown. (E and F) Whole-cell lysates from SaOS-2 (p53 null) and U2OS (wild-type p53) cells were subjected to IP with ANG, separated with a 10% gel, and subjected to Western blotting for p53 (E) and Mdm2 (F). Input controls for p53, Mdm2, and β-actin loading controls are shown. (G) SaOS-2 and U2OS cells were transfected with pcDNA empty vector or full-length LANA-1 plasmids, subjected to IP with ANG, separated with a 7.5% gel, and subjected to Western blotting for LANA-1. Input control for p53 as well as LANA-1 and β-actin loading controls are shown.

Fig 5

Immunofluorescence analyses of LANA-1, ANG, and p53 interactions and mapping of ANG domain interacting with LANA-1 and p53. (A) Permeabilized TIVE-LTC cells were stained for ANG (green), LANA-1 (red), and p53 (blue). The red-bordered box image from a LANA-1-negative cell is merged, and the enlarged image of p53 and ANG is shown on the left. Red arrows point to p53 and ANG colocalization. (B) The white-bordered box image from a LANA-1-positive cell was merged, and enlarged images of double and triple colocalization of LANA-1, ANG, and p53 are shown. White arrowheads indicate the triple colocalization of LANA-1, ANG, and p53 (white spots), while the long white arrows point to LANA-1 and ANG colocalization (yellow spots). A Venn diagram indicating the effects of different color combinations is provided to interpret double and triple colocalization results. 1, LANA-1 and ANG; 2, LANA-1 and p53; 3, LANA-1, ANG and p53; 4, Ang and p53. (C) Schematic diagram showing full-length (FL) angiogenin indicating the amino acid (aa) residues with reported important functions, nuclear localization signal area, and N- and C-terminal deletion constructs. (D) 293T cells were transfected with FL ANG-GFP or four C-terminal GFP-tagged ANG deletion constructs, and their expression was checked by Western blotting for GFP. (E) 293T cells were transfected with full-length LANA-1 along with the empty vector, full-length ANG, or four ANG deletion constructs. The lysates were subjected to IP with rabbit anti-LANA-1 IgG antibodies, separated with a 10% gel, and subjected to Western blotting for GFP. LANA-1 expression levels were checked by Western blotting for LANA-1 in the samples. (F) 293T cells were transfected with the empty vector, FL ANG-GFP, or four of its deletion constructs, and the lysates were subjected to IP with rabbit anti-p53 IgG antibodies, separated with a 10% gel, and subjected to Western blotting for GFP. Input control for p53 and β-actin loading controls are shown.

Fig 6

Effects of neomycin on LANA-1 and ANG in BCBL-1 cells. (A) BCBL-1 cells were left untreated or were treated with 200 or 500 μM neomycin for 3 days and stained for ANG (green), LANA (red), and DAPI for nuclei (blue), and the images were merged. White boxes are shown as enlarged images. (B) The percentage of LANA-1-positive cells was determined by manually counting the number of LANA-1 dot-positive cells in an average of five different fields with 20 cells per view. Values shown represent the means ± standard deviations from three independent experiments. (C) Levels of LANA-1 protein per cell were measured by examining the number of LANA-1 dots/BCBL-1 cell in an average of five different fields by manually counting the number of punctate nuclear LANA-1 dots in cells from three independent experiments. (D) DNA was extracted from BCBL-1 cells treated with neomycin as described above, and KSHV DNA copy numbers were quantitated by estimating the ORF73 copies by real-time DNA PCR.

Fig 7

Effects of 3 days of neomycin treatment on LANA-1, ANG, and p53 (A to C) nuclear extracts were prepared from BCBL-1 cells that were either left untreated or treated with 200 μM neomycin for 3 days. Treated and untreated nuclear lysates were subjected to IP with rabbit anti-ANG IgG antibodies, separated with a 7.5% gel, and subjected to Western blotting for LANA-1 (A); subjected to IP with rabbit anti-LANA-1 IgG antibodies, separated in a 10% gel, and subjected to Western blotting for p53 (B); or subjected to IP with rabbit anti-ANG IgG antibodies and Western blotting for p53 (C). Input controls for ANG, LANA-1, and p53 are shown for all IP reactions. (D and E) BCBL-1 cells were treated with neomycin as described above, and the expression of ORF73 and ORF50 genes was determined by real-time RT-PCR. Values shown represent the means ± standard deviations from three independent experiments. *, P < 0.05; ***, P < 0.005. BJAB (F and G) and BCBL-1 (H and I) cells were either left untreated or were treated with 200 or 500 μM neomycin for 3 days, and the expression of p53 and p21 genes was measured by real-time RT-PCR.

Fig 8

Early time point effects of neomycin treatment on p53, p21, and LANA-1 protein levels. BJAB cells (A), BCBL-1 cells (B), and BC-3 cells (C) were treated with 200 μM neomycin for 4, 6, 8, or 24 h or were left untreated, and cell lysates were subjected to Western blotting for p53 (10% gel), p21 (12.5% gel), LANA-1 (7.5% gel), and β-actin proteins.

Fig 9

Distribution of p53 upon neomycin treatment. BCBL-1 cells were either left untreated or were treated with 200 μM neomycin for 4, 8, and 24 h, fixed with acetone, and stained for p53 (red) and for nuclei (DAPI; blue). White boxes are shown as enlarged images.

Fig 10

Apoptotic effects of neomycin and antiapoptotic effects of angiogenin. (A) Serum-starved (8 h) HMVEC-d cells in 8-well chamber slides were either left untreated, treated with 1 μg/ml ANG, infected with 20 KSHV DNA/cell, or pretreated with 200 μM neomycin (N) for 1 h and then infected with the virus or treated with 1 μg/ml ANG. At 24 h, the slides were stained with anti-cleaved caspase-3 antibody and visualized under the fluorescence microscope. (B) HMVEC-d cells were treated similarly, and the samples were separated with a 10% gel and subjected to Western blotting for cleaved caspase-3.

Fig 11

Reduction of cell survival upon angiogenin knockdown. (A) 293T cells were transfected with control sh-GFP and sh-Renilla (RL) plasmids or two sh-ANG (sh-ANG1 and sh-ANG2) plasmids, and ANG knockdown was checked by separating the lysates with a 12.5% gel and performing Western blotting for ANG. (B) ANG knockdown was also checked by real-time RT-PCR in sh-GFP- and sh-ANG1-transduced BCBL-1 cells. (C to E) BCBL-1 cells transduced with either sh-GFP or sh-ANG1 were subjected to Western blotting for p53, p-p53, p21, Bcl-2, Bax, caspase 3, and β-actin. Fold changes are indicated. (F) BCBL cells transduced with either sh-GFP or sh-ANG1 were stained with YO-PRO dye and analyzed by FACS for the percentage of live cells. Values shown represent the means ± standard deviations from three independent experiments. ***, P < 0.005.

Fig 12

Schematic model showing the events occurring in B cells latently infected with KSHV in the context of angiogenin and the consequences of silencing angiogenin or inhibiting PLCγ activation by neomycin. In BCBL-1 cells, LANA-1 expression induces ANG, which activates the PLCγ, ERK, and AKT pathways. PLCγ activation is required for the nuclear translocation of ANG. The present study shows that in the nucleus, LANA-1 interacts with ANG and p53 and forms a complex. There are also other complexes consisting of ANG-p53, LANA-1-p53, and LANA-1-ANG. These interactions suppress p53 functions, leading to enhanced cell survival and latency maintenance. When ANG is silenced or its nuclear transport via PLCγ activation is inhibited by neomycin treatment, AKT signaling is decreased while the ERK pathway is unaffected. ANG's nuclear translocation is inhibited, which results in decreased LANA-1 expression; decreased interactions between LANA-1-ANG-p53, LANA-1-p53, ANG-p53, and LANA-1-ANG, leading to the activation of p53; and increased detection of p53 free of ANG and LANA-1 in the cytoplasm, resulting in the induction of apoptosis, KSHV lytic cycle, and cell death.