Leukemia virus long terminal repeat activates NFkappaB pathway by a TLR3-dependent mechanism

Abstract

The long terminal repeat (LTR) region of leukemia viruses plays a critical role in tissue tropism and pathogenic potential of the viruses. We have previously reported that U3-LTR from Moloney murine and feline leukemia viruses (Mo-MuLV and FeLV) upregulates specific cellular genes in trans in an integration-independent way. The U3-LTR region necessary for this action does not encode a protein but instead makes a specific RNA transcript. Because several cellular genes transactivated by the U3-LTR can also be activated by NFkappaB, and because the antiapoptotic and growth promoting activities of NFkappaB have been implicated in leukemogenesis, we investigated whether FeLV U3-LTR can activate NFkappaB signaling. Here, we demonstrate that FeLV U3-LTR indeed upregulates the NFkappaB signaling pathway via activation of Ras-Raf-IkappaB kinase (IKK) and degradation of IkappaB. LTR-mediated transcriptional activation of genes did not require new protein synthesis suggesting an active role of the LTR transcript in the process. Using Toll-like receptor (TLR) deficient HEK293 cells and PKR(-/-) mouse embryo fibroblasts, we further demonstrate that although dsRNA-activated protein kinase R (PKR) is not necessary, TLR3 is required for the activation of NFkappaB by the LTR. Our study thus demonstrates involvement of a TLR3-dependent but PKR-independent dsRNA-mediated signaling pathway for NFkappaB activation and thus provides a new mechanistic explanation of LTR-mediated cellular gene transactivation.

Figure 1. Activation of NFκB dependent gene expression by the LTR

A. Analysis of transactivational activity of the 61E-LTR towards CAT-reporters with promoters from different genes. Balb-3T3 cells were cotransfected in 100mm plates with 7.5 μg of 61E-LTR or the vector plasmid pTZ19 and 7.5 μg of various CAT-reporter constructs by DEAE-dextran method and were harvested 48 hrs post-transfection for reporter assays. This experiment was performed three times and a representative chromatogram is presented. For quantitation purpose, individual spots on the chromatogram were collected by scraping and radioactivity was measured by liquid scintillation counting. Fold-induction for any particular CAT construct was expressed as the ratio of acetylated ¹⁴C-chloramphenicol generated for that construct by the LTR to that generated by the vector. B. Requirement of NFκB binding-site in the reporter for activation by the LTR. Two different CAT-reporter plasmids, with or without NFκB binding sites in the promoter (NFBCO-CAT or OBCO-CAT, respectively) were cotransfected separately with 61E-LTR or pTZ19 plasmid in Balb-3T3 cells and CAT assays were performed. To determine dose dependence, 2.5, 5.0 or 7.5 μg of 61E-LTR were transfected. One representative chromatogram and fold-induction values are presented. C. Analysis of transactivational activity of 61E-LTR towards two different NFκB-dependent luciferase reporter constructs. Six-well plates of Balb-3T3 cells were cotransfected with 61E-LTR or pTZ19 (300 ng) and 3XKB-luc (100 ng) or PBIIX-Luc reporter (100 or 250 ng) by lipofectamine plus method and harvested 48 hrs post-transfection for luciferase assays. Total amount of plasmid transfected in each well in both transfection methods was maintained equal using pTZ19 vector plasmid.

Figure 2. Translocation of p65 and expression of endogenous NFκB dependent genes by the LTR

A. Analysis of nuclear p65 following expression of LTR. Nuclear extracts were prepared from Balb-3T3 cells transfected with 61E-LTR or pTZ19 by lipofectamine method at indicated time post-transfection. The time denotes period after the initial 3 hr incubation with DNA and lipofectamine. Twenty micrograms of protein for each sample were separated in 10% SDS-PAGE and analyzed for p65 or PCNA by western immunoblotting. Band intensities were determined directly on the film using LabWorks Image Analysis Program (UVP Inc, Upland, CA). Fold inductions were calculated after the band intensities were normalized for equal PCNA loading. B. Activation of endogenous cyclin D1 and Bcl-2 protein. Twenty microgram of whole cell lysates from Balb-3T3 cells transfected with either 61E-LTR or pTZ19 vector were analyzed for cyclin D1 or Bcl-2 level by western immunoblotting. Lysates were prepared 40 hrs after transfection. Same blots were further analyzed for β-actin protein following stripping of the antibodies used in first immunoblotting.

Figure 3. Full length FeLV activates NFκB in an LTR-dependent manner

A. Analysis of NFκB activation by full-length and U3-LTR constructs from FeLV and Mo-MuLV. Balb-3T3 cells were cotransfected with 7.5 μg of NFBCO-CAT reporter and 7.5 μg of 61E-LTR (FeLV U3-LTR) or GMNX (Mo-MuLV U3-LTR) or 61E (full-length FeLV) or Mov9 (full-length Mo-MuLV) or pTZ19 vector plasmid by DEAE-dextran method. Cells were harvested 48 hrs post-transfection for CAT assays as described in Figure 1. Similar results were obtained in three independent experiments and the fold-induction values for the presented chromatogram are indicated. B. Analysis of NFκB activation by the full-length FeLV containing mutation in the U3-LTR. Balb-3T3 cells were cotransfected with 100 ng of 3XKB-luc reporter and 300 ng of either wild type or mutant FeLV U3-LTR constructs (61E-LTR or EDD2, respectively), or wild type or mutant full-length FeLV constructs (61E or 61E-Mut, respectively). Cells were harvested 48 hrs post-transfection for luciferase assays as in Figure 1.

Figure 4. Activation of NFκB by the LTR require proteasomal activity, phosphorylation and degradation of IκB

A. Effect of proteasomal inhibitor lactacystin on LTR-mediated activation of NFκB. Balb-3T3 cells were cotransfected with 100 ng 3XKB-luc reporter and 300 ng 61E-LTR or p65/relA expression plasmid by lipofectamine plus method. Lactacystin (5μM) or DMSO solvent control was added on to the cells along with DNA-lipofectamine mixture. Cells were harvested 24 hr post-transfection for luciferase assay. B. Induction of IκB-α degradation by the LTR. Balb-3T3 cells transfected with 61E-LTR or pTZ19 similary as in Figure 2A but instead cytoplasmic extracts were prepared at indicated time period and analyzed for IkB-α by western immunoblotting. Same blots were later analyzed for β-actin. C. Effect of expression of super repressor dn-IκB. Balb-3T3 cells were cotransfected with 100 ng 3XKB-luc reporter and 300 ng 61E-LTR and various amounts of dn-IκB plasmid (50 ng, 100 ng or 200 ng) by lipofectamine plus method as indicated in Figure 1. Cells were harvested 24 hrs post-transfection for luciferase assay. D. Effect of inhibition of IκB-kinase (IKK) activity. Balb-3T3 cells were cotransfected with 3XKB-luc reporter and 61E-LTR along with expression plasmids for dominant negative form of IKK subunits (dn-IKK1 and dn-IKK2, 300 ng) as appropriate and processed for luciferase assay.

Figure 5. New protein synthesis is not necessary for LTR-mediated gene transactivation

A. Effect of cycloheximide on the activation of NFκB-dependent luciferase reporters by the LTR. Cotransfection experiments were carried out with 3XKB-luc reporter and 61E-LTR or vector pTZ19 plasmid in Balb-3T3 cells by lipofectamine plus method. Cycloheximide (5 μM or 1μM CHX) was added to cells in appropriate wells 30 min after the DNA-lipofectamine complex was added onto the cells. Cells were harvested for luciferase assay at 18 hr (for 5 μM CHX treatment) or 24 hr (for 1 μM CHX treatment) post-transfection. B. Effect of cycloheximide on transcription of luciferase reporter gene by the LTR. Total cellular RNA (2 μg) from a second set of Balb-3T3 cells cotransfected and cycloheximide treated essentially similarly as above, were subjected to RT-PCR analysis for luciferase mRNA. RT-PCR figure shows data for 24 hr treatment group only. Essentially similar results were obtained with 18 hr treatment group. In order to demonstrate absence of DNA contamination in RNA preparations, one set of PCR amplification for each sample was carried out without any RT and loaded onto the lanes marked as ‘-’. All RNA samples were also tested for β-actin mRNA by RT-PCR performed in similar manner as for luciferase mRNA. Lane C represents PCR control with no RNA sample. Products were separated on 2% agarose gel. The 100 bp DNA ladder was included in the left lane. The amplified product for luciferase mRNA (250 bp) and β-actin mRNA (353 bp) are indicated by an arrow. C. Effect of cycloheximide on the inhibitory effect of dn-IκB. Balb-3T3 cells were cotransfected with 3XKB-luc reporter and 61E-LTR or vector pTZ19 plasmid as above. In addition, some wells were also cotransfected with 200 ng of dn-IκB as indicated. One set of LTR and LTR+dn-IκB transfected cells were treated with 1 μM cycloheximide. Cells from all transfected wells were harvested at 24 hr post transfection for luciferase assay. D. Effect of cycloheximide and dn-IκB on transcriptional activation of luciferase reporter gene by the LTR. Total cellular RNA from a similar set of transfected cells as in section C were subjected to RT-PCR analysis for luciferase and β-actin mRNA essentially as described in section B.

Figure 6. Role of PKR and TLR3 in the activation of NFκB by the LTR

A. Analysis of PKR expression in the cell lines used in this study. Whole cell lysates (15 μg total protein) from actively growing Balb-3T3, PKR^+/+ or PKR^−/− MEFs were separated on 10% SDS-PAGE, and western immunoblotting analysis was performed using polyclonal anti-PKR serum. Lysates were also analyzed in separate SDS-PAGE for β-Actin using specific monoclonal antibody. B. Analysis of NFκB-dependent reporter gene activation in PKR^−/− cells. PKR^−/− and PKR^+/+ cells were cotransfected by lipofectamine plus method with 100 ng 3XKB-luc and 300 ng of either wild type or mutant FeLV U3-LTR constructs (61E-LTR or EDD2, respectively), wild type or mutant full-length FeLV constructs (61E or 61E-Mut, respectively) or pTZ19 vector plasmid. Cells were harvested for luciferase assay 48 hrs after transfection. As a control, same set of transfection was also carried out on Balb-3T3 cells as in Figure 3B. C. Role of TLR3 in the activation of NFκB pathway by the LTR. TLR deficient regular HEK293 cells and HEK293 cells stably expressing mouse TLR3 (from Invivogen) were cotransfected by lipofectamine plus method with 100 ng of 3XKB-Luc reporter and 300 ng of pTZ19 vector, 61E-LTR or p65 expression plasmid as indicated. Synthetic dsRNA polyI:C (100 μg/ml) or loxoribine (1μM) were added to the media 8 hrs before harvesting the cells for luciferase assay at 24 hrs post-transfection.

Figure 7. Analysis of upstream signaling pathways leading to the activation of NFκB by the LTR

A. Inhibition of NFκB activation does not affect collagenase IV and MCP-1 activation by the LTR. Balb-3T3 cells were cotransfected with −517coll-CAT, NFBCO-CAT or MCP-1-CAT together with pTZ19 vector or 61E-LTR and dn-IκB as described in Figure 1A. B. Analysis of role of Ras-Raf-MAPK pathway in NFκB activation. Balb-3T3 cells were cotransfected with 3XKB-luc and 61E-LTR or pTZ19 vector along with inhibitors as indicated. The dominant negative Ras (RasN17) and dnRaf-1 (Raf-BXB-K375W) constructs were used at 300 ng per well. The MEK1/2 inhibitor PD98059 (50 μM) was added to the cells in appropriate wells along with DNA-lipofectamine mixture.