Session 8A Latency 10:30 AM – 12:00 PM
Session Chairs: Jennifer Totonchy & Maria Masucci
EBV Persistence without its EBNA3A and 3C Oncogenes in vivo Anita Murer
Oral Talk #62
EBV Persistence without its EBNA3A and 3C Oncogenes in
vivo
Anita Murer1, Donal McHugh1, Nicole Caduff1, Jens Kalchschmidt2, Mario Barros3, Andrea Zbinden4, Riccarda Capaul4, Gerald Niedobitek3, Martin Allday5, Obinna Chijioke1,6, Christian Münz1 1Viral Immunobiology, Institute of Experimental Immunology, University of Zürich, 2Genomics and Immunity, NIAMS, National Institutes of Health, 3Institute of Pathology, Unfallkrankenhaus Berlin, 4Institute of Medical Virology, University of Zürich, 5Molecular Virology, Department of Medicine, Imperial College London, 6Institute of Pathology and Molecular Pathology, University Hospital Zürich
The oncogenic Epstein Barr virus (EBV) infects the majority of the human population and usually persists within its host for life without symptoms. The EBV oncoproteins nuclear antigen 3A (EBNA3A) and 3C (EBNA3C) are required for B cell transformation in vitro and are expressed in EBV associated immunoblastic lymphomas in vivo. In order to address the necessity of EBNA3A and EBNA3C for persistent EBV infection in vivo, we infected NOD-scid γc null mice with reconstituted human immune system components (huNSG mice) with recombinant EBV mutants devoid of EBNA3A or EBNA3C expression. These EBV mutants established latent infection in secondary lymphoid organs of infected huNSG mice for at least 3 months, but did not cause tumor formation. Low level viral persistence in the absence of EBNA3A or EBNA3C seemed to be supported primarily by proliferation with the expression of early latent EBV gene products transitioning into absent viral protein expression without elevated lytic replication. In vitro, EBNA3A and EBNA3C deficient EBV infected B cells could be rescued from apoptosis through CD40 stimulation, mimicking T cell help in secondary lymphoid tissues. Thus, even in the absence of the oncogenes EBNA3A and 3C, EBV can access a latent gene expression pattern that is reminiscent of EBV persistence in healthy virus carriers without prior expression of its whole growth transforming program.
vivo
Anita Murer1, Donal McHugh1, Nicole Caduff1, Jens Kalchschmidt2, Mario Barros3, Andrea Zbinden4, Riccarda Capaul4, Gerald Niedobitek3, Martin Allday5, Obinna Chijioke1,6, Christian Münz1 1Viral Immunobiology, Institute of Experimental Immunology, University of Zürich, 2Genomics and Immunity, NIAMS, National Institutes of Health, 3Institute of Pathology, Unfallkrankenhaus Berlin, 4Institute of Medical Virology, University of Zürich, 5Molecular Virology, Department of Medicine, Imperial College London, 6Institute of Pathology and Molecular Pathology, University Hospital Zürich
The oncogenic Epstein Barr virus (EBV) infects the majority of the human population and usually persists within its host for life without symptoms. The EBV oncoproteins nuclear antigen 3A (EBNA3A) and 3C (EBNA3C) are required for B cell transformation in vitro and are expressed in EBV associated immunoblastic lymphomas in vivo. In order to address the necessity of EBNA3A and EBNA3C for persistent EBV infection in vivo, we infected NOD-scid γc null mice with reconstituted human immune system components (huNSG mice) with recombinant EBV mutants devoid of EBNA3A or EBNA3C expression. These EBV mutants established latent infection in secondary lymphoid organs of infected huNSG mice for at least 3 months, but did not cause tumor formation. Low level viral persistence in the absence of EBNA3A or EBNA3C seemed to be supported primarily by proliferation with the expression of early latent EBV gene products transitioning into absent viral protein expression without elevated lytic replication. In vitro, EBNA3A and EBNA3C deficient EBV infected B cells could be rescued from apoptosis through CD40 stimulation, mimicking T cell help in secondary lymphoid tissues. Thus, even in the absence of the oncogenes EBNA3A and 3C, EBV can access a latent gene expression pattern that is reminiscent of EBV persistence in healthy virus carriers without prior expression of its whole growth transforming program.
G-quadruplexes Regulate mRNA Translation of LANA to Control Antigen Presentation during Latency Prerna Dabral
ral Talk #63
G-Quadruplexes Regulate mRNA Translation of LANA to
Control Antigen Presentation during Latency
Prerna Dabral1, Subhash C. Verma1
1Department of Microbiology and Immunology, University of Nevada, Reno School of Medicine
Kaposi’s sarcoma associated herpesvirus, like other herpesviruses establishes life-long latent infection in the B- and endothelial cells. During latency, only a handful viral genes are expressed, which play important roles in maintaining latency and escaping the infected cells from the host immune surveillance system. Latency Associated Nuclear Antigen (LANA) is one of the predominantly expressed proteins, which is critical for viral genome replication and persistence of the episomes during cell division. LANA self-regulates its synthesis and maintains the protein levels sufficient for persistence below a threshold triggering host’s immune recognition. However, the mechanism of how LANA regulates its synthesis is not fully understood. We previously showed that G-quadruplex in nucleic acid sequences affects biological process including DNA replication fork progression. Sequence analysis of LANA mRNA showed the presence of GC-rich regions that are required for formation of G-quadruplex structures. Various biophysical and biochemical assays including CD spectrometry and gel mobility shift assay of oligonucleotides with GC-rich region of LANA mRNA identified G-quadruplex. Furthermore, Crosslinking and Immunoprecipitation (CLIP) assay with G-quadruplex binding antibody confirmed the presence of G-quadruplex structures in LANA mRNA. Our experiments with LANA sequence (forming G-quadruplex) and full-length LANA treated with G-quadruplex stabilizing compounds showed a reduction in LANA translation and reduced activation of CD8+T cells owing to a lesser antigen presentation on the Antigen Presenting cells. Mass spectrometry of proteins interacting with G-quadruplex forming sequence of LANA mRNA revealed the binding of heterologous ribonucleoproteins (HNRNPs), which we propose to play critical role in destabilizing those G-quadruplexes for controlled expression of LANA. These data confirmed the importance of G-quadruplex in regulated LANA expression and evasion from host immune surveillance system.
Control Antigen Presentation during Latency
Prerna Dabral1, Subhash C. Verma1
1Department of Microbiology and Immunology, University of Nevada, Reno School of Medicine
Kaposi’s sarcoma associated herpesvirus, like other herpesviruses establishes life-long latent infection in the B- and endothelial cells. During latency, only a handful viral genes are expressed, which play important roles in maintaining latency and escaping the infected cells from the host immune surveillance system. Latency Associated Nuclear Antigen (LANA) is one of the predominantly expressed proteins, which is critical for viral genome replication and persistence of the episomes during cell division. LANA self-regulates its synthesis and maintains the protein levels sufficient for persistence below a threshold triggering host’s immune recognition. However, the mechanism of how LANA regulates its synthesis is not fully understood. We previously showed that G-quadruplex in nucleic acid sequences affects biological process including DNA replication fork progression. Sequence analysis of LANA mRNA showed the presence of GC-rich regions that are required for formation of G-quadruplex structures. Various biophysical and biochemical assays including CD spectrometry and gel mobility shift assay of oligonucleotides with GC-rich region of LANA mRNA identified G-quadruplex. Furthermore, Crosslinking and Immunoprecipitation (CLIP) assay with G-quadruplex binding antibody confirmed the presence of G-quadruplex structures in LANA mRNA. Our experiments with LANA sequence (forming G-quadruplex) and full-length LANA treated with G-quadruplex stabilizing compounds showed a reduction in LANA translation and reduced activation of CD8+T cells owing to a lesser antigen presentation on the Antigen Presenting cells. Mass spectrometry of proteins interacting with G-quadruplex forming sequence of LANA mRNA revealed the binding of heterologous ribonucleoproteins (HNRNPs), which we propose to play critical role in destabilizing those G-quadruplexes for controlled expression of LANA. These data confirmed the importance of G-quadruplex in regulated LANA expression and evasion from host immune surveillance system.
LMP2A Regulates BART miRNAs Transcription by Translational Activation of ETS Family Genes Hyoji Kim
Oral Talk #64
LMP2A Regulates BART miRNAs Transcription by
Translational Activation of ETS Family Genes
Hyoji Kim1, Hisashi Iizasa1, Yuichi Kanehiro1, Hironori Yoshiyama1
1Shimane University, Faculty of Medicine
BamHI A rightward transcript microRNAs (BART miRNAs) are expressed more abundantly in EBV-infected epithelial tumors than in EBV-infected B lymphomas. It is considered that BART miRNAs are important for maintenance of viral latency in EBV-infected epithelial cells. However, the mechanism of their transcriptional regulation is not fully understood. To clarify detailed regulation of BART miRNAs transcription in epithelial cells, BART promoter region was cloned to the luciferase vector.
The promoter activity was higher in EBV-infected epithelial cells than uninfected cells. Mutagenesis assay revealed 4 important regions for BART promoter activity, which were ETS family binding sites. The qPCR analysis showed 4 out of 28 ETS family genes were strongly expressed in EBV-infected epithelial cells. The siRNA inhibition clarified 3 out of 4 ETS family genes enhanced BART transcription. The most functional gene (ETS A) was deleted using CRISPR-Cas9 system. ETS A deletion resulted dramatic reduction in BART miRNAs transcription and induction of EBV lytic infection. Since EBV-infected cells showed higher ETS A expression than uninfected cells, ETS A expression was examined in epithelial cells transfected with different EBV latent gene plasmids. Cells expressing LMP2A showed upregulation of ETS A expression. Finally, epithelial cells infected with EBV mutant lacking LMP2A showed reduced ETS A expression as well as BART miRNAs transcription.
Suggesting, LMP2A enhances BART miRNAs transcription through upregulation of ETS family genes in epithelial cells showing latent EBV infection.
Translational Activation of ETS Family Genes
Hyoji Kim1, Hisashi Iizasa1, Yuichi Kanehiro1, Hironori Yoshiyama1
1Shimane University, Faculty of Medicine
BamHI A rightward transcript microRNAs (BART miRNAs) are expressed more abundantly in EBV-infected epithelial tumors than in EBV-infected B lymphomas. It is considered that BART miRNAs are important for maintenance of viral latency in EBV-infected epithelial cells. However, the mechanism of their transcriptional regulation is not fully understood. To clarify detailed regulation of BART miRNAs transcription in epithelial cells, BART promoter region was cloned to the luciferase vector.
The promoter activity was higher in EBV-infected epithelial cells than uninfected cells. Mutagenesis assay revealed 4 important regions for BART promoter activity, which were ETS family binding sites. The qPCR analysis showed 4 out of 28 ETS family genes were strongly expressed in EBV-infected epithelial cells. The siRNA inhibition clarified 3 out of 4 ETS family genes enhanced BART transcription. The most functional gene (ETS A) was deleted using CRISPR-Cas9 system. ETS A deletion resulted dramatic reduction in BART miRNAs transcription and induction of EBV lytic infection. Since EBV-infected cells showed higher ETS A expression than uninfected cells, ETS A expression was examined in epithelial cells transfected with different EBV latent gene plasmids. Cells expressing LMP2A showed upregulation of ETS A expression. Finally, epithelial cells infected with EBV mutant lacking LMP2A showed reduced ETS A expression as well as BART miRNAs transcription.
Suggesting, LMP2A enhances BART miRNAs transcription through upregulation of ETS family genes in epithelial cells showing latent EBV infection.
EBV Latency IIb is Phenotypically Distinct from Latency III Independent of LMP1 Expression
Joshua E. Messinger
Oral Talk #65
EBV Latency IIb is Phenotypically Distinct from Latency III
Independent of LMP1 Expression
Joshua E Messinger1, Joanne Dai1, Lyla J Stanland1, Alexander M Price1, Micah A Luftig1
1Duke University
EBV infection of primary B cells in vitro generates immortalized lymphoblastoid cell lines (LCLs) expressing the full repertoire of latency proteins (latency III). However, we recently demonstrated that early after infection, LMP1 and LMP2A are expressed at reduced levels until ~2-3 weeks post-infection. This period of EBNA-positive, LMP-low gene expression is termed latency IIb.
Previous work has demonstrated that LCLs display a wide distribution of LMP1 expression. We found that LMP1low expressing cells within an LCL population overlap with LMP1 expression in latency IIb. Immuno-histochemical staining of EBV-associated lymphomas indicates heterogeneity in EBV latent gene expression with many cells expressing EBNA2, but lacking LMP1. Due to the similar expression levels of LMP1 in LMP1low LCLs and early infected latency IIb cells, it is difficult to discern these two molecular phenotypes in vivo. A better understanding of what defines these latency stages is needed to understand EBV-associated tumor composition in vivo and how this influences disease progression and response to therapy.
To address these questions, we performed total mRNA sequencing of sorted latency IIb cells and donor-matched LCLs stratified by LMP1 expression using ICAM-1 as a proxy for LMP1. Hierarchical clustering revealed that latency IIb is phenotypically distinct from LCLs independent of donor and LMP1 expression. Gene Set Enrichment Analysis identified DNA replication as a hallmark of latency IIb. We also identified and validated host biomarkers of latency IIb distinct from LMP1low latency III including CCR6 and FcRL4. Conversely, CCR7 and PLOD2 were confirmed to be host markers of latency III. These data demonstrate that latency IIb is a unique stage from latency III independent of LMP1 and provides host markers of these stages. This research will allow for a better understanding of the heterogeneity of EBV+ tumors and its role in pathogenesis, immunity, and therapeutic responses.
Independent of LMP1 Expression
Joshua E Messinger1, Joanne Dai1, Lyla J Stanland1, Alexander M Price1, Micah A Luftig1
1Duke University
EBV infection of primary B cells in vitro generates immortalized lymphoblastoid cell lines (LCLs) expressing the full repertoire of latency proteins (latency III). However, we recently demonstrated that early after infection, LMP1 and LMP2A are expressed at reduced levels until ~2-3 weeks post-infection. This period of EBNA-positive, LMP-low gene expression is termed latency IIb.
Previous work has demonstrated that LCLs display a wide distribution of LMP1 expression. We found that LMP1low expressing cells within an LCL population overlap with LMP1 expression in latency IIb. Immuno-histochemical staining of EBV-associated lymphomas indicates heterogeneity in EBV latent gene expression with many cells expressing EBNA2, but lacking LMP1. Due to the similar expression levels of LMP1 in LMP1low LCLs and early infected latency IIb cells, it is difficult to discern these two molecular phenotypes in vivo. A better understanding of what defines these latency stages is needed to understand EBV-associated tumor composition in vivo and how this influences disease progression and response to therapy.
To address these questions, we performed total mRNA sequencing of sorted latency IIb cells and donor-matched LCLs stratified by LMP1 expression using ICAM-1 as a proxy for LMP1. Hierarchical clustering revealed that latency IIb is phenotypically distinct from LCLs independent of donor and LMP1 expression. Gene Set Enrichment Analysis identified DNA replication as a hallmark of latency IIb. We also identified and validated host biomarkers of latency IIb distinct from LMP1low latency III including CCR6 and FcRL4. Conversely, CCR7 and PLOD2 were confirmed to be host markers of latency III. These data demonstrate that latency IIb is a unique stage from latency III independent of LMP1 and provides host markers of these stages. This research will allow for a better understanding of the heterogeneity of EBV+ tumors and its role in pathogenesis, immunity, and therapeutic responses.
Uracil DNA Glycosylase of Gammaherpesvirus and Mammals Differentially Support Mutational Outcomes Kevin McBride
Oral Talk #66
Uracil DNA Glycosylase of Gammaherpesvirus and Mammals
Differentially Support Mutational Outcomes
Yunxiang Mu1, Qiwen Dong2, Monika Zelazowska1, Joshua Plummer1, Laurie Krug2, Kevin Mcbride1
1University of Texas MD Anderson Cancer Center, 2Stony Brook University
All herpesviruses encode a viral homolog of the mammalian Uracil DNA-glycosylase (UNG) gene. Host UNG removes mutagenic uracils from genomic DNA and normally triggers high- fidelity base excision repair (BER). An exception is in B cells, where host UNG triggers mutagenic repair of U:G lesions created by activation-induced cytidine deaminase (AID) during class-switch recombination (CSR) and somatic hypermutation (SHM). The mechanism that shifts high-fidelity UNG/BER to error-prone repair is not understood. Gammaherpesvirus infect many B cell subsets including germinal center cells. Using murine gammaherpesvirus 68 infections in mice we find that germinal center cells are the major viral targets during acute infection. Analysis of infected cells demonstrates overlapping expression of both vUNG and AID. Neither the viral influence on CSR and SHM nor the effect of the mutagenic germinal center environment on the virus has not been well-characterized. Here, we investigated the consequences of the vUNG encoded by ORF46 of murine gammaherpesvirus (MHV68) on UNG-dependent processes of the host. Using an in vitro class switch recombination assay we find that vUNG from gammaherpesviruses can process genomic uracils during CSR. We analyzed the effect of vUNG on AID induced mutations in a fibroblast cell line with an indicator gene. We find that compared to mUNG, ORF46 dramatically suppressed AID induced mutations abating both transition and transversion mutations. Biochemical analysis revealed differences in the UNGs including abasic site binding and efficiency in catalyzing single stranded versus double stranded uracil containing substrates. Analysis of domain swaps between host and vUNG revealed domains that differentially supported catalytic activity, abasic site binding and support for mutational repair. These results demonstrate a divergence of UNG structure and function in supporting high-fidelity repair and demonstrate that biochemical properties of UNG influence repair outcome of AID lesions.
Differentially Support Mutational Outcomes
Yunxiang Mu1, Qiwen Dong2, Monika Zelazowska1, Joshua Plummer1, Laurie Krug2, Kevin Mcbride1
1University of Texas MD Anderson Cancer Center, 2Stony Brook University
All herpesviruses encode a viral homolog of the mammalian Uracil DNA-glycosylase (UNG) gene. Host UNG removes mutagenic uracils from genomic DNA and normally triggers high- fidelity base excision repair (BER). An exception is in B cells, where host UNG triggers mutagenic repair of U:G lesions created by activation-induced cytidine deaminase (AID) during class-switch recombination (CSR) and somatic hypermutation (SHM). The mechanism that shifts high-fidelity UNG/BER to error-prone repair is not understood. Gammaherpesvirus infect many B cell subsets including germinal center cells. Using murine gammaherpesvirus 68 infections in mice we find that germinal center cells are the major viral targets during acute infection. Analysis of infected cells demonstrates overlapping expression of both vUNG and AID. Neither the viral influence on CSR and SHM nor the effect of the mutagenic germinal center environment on the virus has not been well-characterized. Here, we investigated the consequences of the vUNG encoded by ORF46 of murine gammaherpesvirus (MHV68) on UNG-dependent processes of the host. Using an in vitro class switch recombination assay we find that vUNG from gammaherpesviruses can process genomic uracils during CSR. We analyzed the effect of vUNG on AID induced mutations in a fibroblast cell line with an indicator gene. We find that compared to mUNG, ORF46 dramatically suppressed AID induced mutations abating both transition and transversion mutations. Biochemical analysis revealed differences in the UNGs including abasic site binding and efficiency in catalyzing single stranded versus double stranded uracil containing substrates. Analysis of domain swaps between host and vUNG revealed domains that differentially supported catalytic activity, abasic site binding and support for mutational repair. These results demonstrate a divergence of UNG structure and function in supporting high-fidelity repair and demonstrate that biochemical properties of UNG influence repair outcome of AID lesions.
Successful Short Term Infection of Humanized Mice with EBV Lacking EBNA3 Genes Contrasts Their Essential Role in B Cell Transformation In Vitro and Their Role in Favouring Latency in Memory B Cells Christine T. Styles
Oral Talk #67
Successful Short Term Infection of Humanized Mice with EBV Lacking EBNA3 Genes Contrasts Their Essential Role in B Cell Transformation In Vitro and Their Role in Favouring Latency in
Memory B Cells
Christine T. Styles1, Quentin Bazot1, Gillian Parker1, Robert E. White1, Kostas Paschos1, Paul J. Farrell1, Martin J. Allday1
1Section of Virology, Imperial College Department of Medicine
Following infection of mature human B cells, EBV induces proliferation and differentiation, and the EBNA3 proteins are central to this process. The relationship between in vitro B cell transformation and in vivo persistent infection can be investigated in humanized mice, whose immune system has been reconstituted from human haematopoietic stem cells. In vitro infection of primary B cells with recombinant EBVs where the EBNA3 locus is deleted, or where EBNA3A and EBNA3C are conditionally inactivated leads to cell death/arrest around 15-20 days post infection. We have previously shown the EBNA3 proteins support B cell transformation and establishment of LCLs through repression of tumour suppressors including p16INK4a and BIM. However, we have established an additional function for the oncogenic EBNA3 proteins, whereby EBNA3A and EBNA3C epigenetically suppress the plasma cell differentiation factors p18INK4c and BLIMP-1. We speculate that this favours memory B cell differentiation and subsequent viral persistence. Western blotting and qPCR indicates EBNA3A and EBNA3C repress transcription of p18INK4a and BLIMP-1. Chromatin immunopreciptation shows that EBNA3A and EBNA3C bind within regulatory elements of these differentiation factors (suggesting direct regulation) and that EBNA3- mediated regulation involves PRC2 and the repressive histone mark H3K27me3. Concurrently, 20 days post infection EBNA3A/EBNA3C-null cells develop a plasma cell-like phenotype, characterised by elevated CD138 and CD38 levels alongside increased immunoglobulin production. Although EBNA3A is important and EBNA3C essential for in vitro B cell transformation and likely favour memory B cell differentiation and latency in vivo, in humanised mice we have identified splenic tumour development and persistence of virus-infected cells 6 weeks post infection with recombinant EBVs, both with single deletions of EBNA3A or EBNA3C, and with a deletion of the entire EBNA3 locus. Interestingly, splenic lymphocytes derived from these infections did not lead to LCL expansion ex vivo, in contrast to cells where all EBNA3s were expressed.
Memory B Cells
Christine T. Styles1, Quentin Bazot1, Gillian Parker1, Robert E. White1, Kostas Paschos1, Paul J. Farrell1, Martin J. Allday1
1Section of Virology, Imperial College Department of Medicine
Following infection of mature human B cells, EBV induces proliferation and differentiation, and the EBNA3 proteins are central to this process. The relationship between in vitro B cell transformation and in vivo persistent infection can be investigated in humanized mice, whose immune system has been reconstituted from human haematopoietic stem cells. In vitro infection of primary B cells with recombinant EBVs where the EBNA3 locus is deleted, or where EBNA3A and EBNA3C are conditionally inactivated leads to cell death/arrest around 15-20 days post infection. We have previously shown the EBNA3 proteins support B cell transformation and establishment of LCLs through repression of tumour suppressors including p16INK4a and BIM. However, we have established an additional function for the oncogenic EBNA3 proteins, whereby EBNA3A and EBNA3C epigenetically suppress the plasma cell differentiation factors p18INK4c and BLIMP-1. We speculate that this favours memory B cell differentiation and subsequent viral persistence. Western blotting and qPCR indicates EBNA3A and EBNA3C repress transcription of p18INK4a and BLIMP-1. Chromatin immunopreciptation shows that EBNA3A and EBNA3C bind within regulatory elements of these differentiation factors (suggesting direct regulation) and that EBNA3- mediated regulation involves PRC2 and the repressive histone mark H3K27me3. Concurrently, 20 days post infection EBNA3A/EBNA3C-null cells develop a plasma cell-like phenotype, characterised by elevated CD138 and CD38 levels alongside increased immunoglobulin production. Although EBNA3A is important and EBNA3C essential for in vitro B cell transformation and likely favour memory B cell differentiation and latency in vivo, in humanised mice we have identified splenic tumour development and persistence of virus-infected cells 6 weeks post infection with recombinant EBVs, both with single deletions of EBNA3A or EBNA3C, and with a deletion of the entire EBNA3 locus. Interestingly, splenic lymphocytes derived from these infections did not lead to LCL expansion ex vivo, in contrast to cells where all EBNA3s were expressed.
Enhancement of in vivo Latency by Substitution of the KSHV LANA DNA Binding Domain with that of MHV68 in Chimeric Virus Kenneth M. Kaye
Oral Talk #68
Enhancement of in vivo Latency by Substitution of the KSHV
Lana DNA Binding Domain with that of MHV68 in Chimeric
Virus
Marta Pires de Miranda1, Ana Quendera1, Aline C. Habison2, Chantal Beauchemin2, Min Tan2, Sofia A. Cerqueira, Bruno Correia3, Rajesh Ponnusamy3, Edward J. Usherood4, Colin E. McVoy3, J. Pedro Simas, Kenneth M. Kaye2
1Instituto de Microbiologia e Instituto de Medicina Molecular, Faculdade de Medicina, Universidade de Lisboa, 2Brigham and Women’s Hospital, Harvard Medical School, 3Instituto de Tecnologia Quimica e Biologica, Universidade Nova de Lisboa, 4Department of Microbiology and Immunology, Geisel School of Medicine at Dartmouth
Kaposi’s sarcoma-associated herpesvirus (KSHV) LANA (kLANA) and murine gammaherpesvirus 68 (MHV68) LANA (mLANA) mediate episome persistence for their respective viruses. In order to persist during latent infection in proliferating cells, viral episomal genomes replicate and then partition to progeny cell nuclei following mitosis. C- terminal LANA is the DNA binding domain (DBD) and binds its recognition element in terminal repeat (TR) DNA. This binding is essential for LANA mediated DNA replication and episome persistence. N-terminal kLANA binds histones H2A/H2B on the nucleosome surface to attach to mitotic chromosomes. LANA simultaneously binds TR DNA and mitotic chromosomes to serve as a molecular tether to segregate viral genomes to progeny cell nuclei following mitosis. kLANA (1162 residues) and mLANA (314 residues) have substantial size and sequence differences. For instance, mLANA lacks the large internal repeat region present in kLANA. The KSHV TR (0.8kb) and MHV68 TR (1.2kb) elements also differ in size, and although both are GC rich, they lack sequence homology. Despite these differences, we recently found that kLANA can substitute for mLANA in recombinant MHV68 and that this chimeric virus is capable of establishing latent infection in mice. The mLANA and kLANA DBDs share sequence and structural homology. We now find that substitution of the kLANA DBD with the mLANA DBD enhances the ability of kLANA chimeric virus to establish latency in vivo.
Lana DNA Binding Domain with that of MHV68 in Chimeric
Virus
Marta Pires de Miranda1, Ana Quendera1, Aline C. Habison2, Chantal Beauchemin2, Min Tan2, Sofia A. Cerqueira, Bruno Correia3, Rajesh Ponnusamy3, Edward J. Usherood4, Colin E. McVoy3, J. Pedro Simas, Kenneth M. Kaye2
1Instituto de Microbiologia e Instituto de Medicina Molecular, Faculdade de Medicina, Universidade de Lisboa, 2Brigham and Women’s Hospital, Harvard Medical School, 3Instituto de Tecnologia Quimica e Biologica, Universidade Nova de Lisboa, 4Department of Microbiology and Immunology, Geisel School of Medicine at Dartmouth
Kaposi’s sarcoma-associated herpesvirus (KSHV) LANA (kLANA) and murine gammaherpesvirus 68 (MHV68) LANA (mLANA) mediate episome persistence for their respective viruses. In order to persist during latent infection in proliferating cells, viral episomal genomes replicate and then partition to progeny cell nuclei following mitosis. C- terminal LANA is the DNA binding domain (DBD) and binds its recognition element in terminal repeat (TR) DNA. This binding is essential for LANA mediated DNA replication and episome persistence. N-terminal kLANA binds histones H2A/H2B on the nucleosome surface to attach to mitotic chromosomes. LANA simultaneously binds TR DNA and mitotic chromosomes to serve as a molecular tether to segregate viral genomes to progeny cell nuclei following mitosis. kLANA (1162 residues) and mLANA (314 residues) have substantial size and sequence differences. For instance, mLANA lacks the large internal repeat region present in kLANA. The KSHV TR (0.8kb) and MHV68 TR (1.2kb) elements also differ in size, and although both are GC rich, they lack sequence homology. Despite these differences, we recently found that kLANA can substitute for mLANA in recombinant MHV68 and that this chimeric virus is capable of establishing latent infection in mice. The mLANA and kLANA DBDs share sequence and structural homology. We now find that substitution of the kLANA DBD with the mLANA DBD enhances the ability of kLANA chimeric virus to establish latency in vivo.
Identification of ARKL1 as a Jun-interacting Factor that Negatively Regulates EBV Reactivation Umama Zia Siddiqi
Oral Talk #69
Identification of ARKL1 as a Jun-interacting Factor that
Negatively Regulates EBV Reactivation
Umama Zia Siddiqi1, Anup Vaidya1, Xinliu Li1, Lori Frappier1 1Department of Molecular Genetics, University of Toronto
EBV reactivation from the latent to the lytic cycle starts with activation of the Zp promoter controlling expression of the EBV BZLF1 transcription factor. Zp includes a binding site for c-Jun which was previously shown to be an important activator of Zp. Through proteomic approaches, we identified an uncharacterized cellular protein, Arkadia-like 1 (ARKL1), as a protein that binds the cellular CK2 kinase regulatory subunit (β) through the same binding pocket as EBNA1. Investigation of the effect of ARKL1 on EBV infection showed that silencing ARKL1 in EBV-positive gastric carcinoma (AGS-EBV), nasopharyngeal carcinoma (NPC43) or B cells (containing M81 EBV) resulted in BZLF1 expression and EBV reactivation, while overexpression of ARKL1 decreased EBV reactivation. To gain insight into the mechanism of action of ARKL1, we purified FLAG-tagged ARKL1 from human cells and identified interacting proteins by tandem mass spectrometry. This confirmed the interaction of ARKL1 with CK2 and also revealed an interaction with Jun, suggesting that ARKL1 might suppress BZLF1 expression by inhibiting Jun function at Zp. Consistent with this hypothesis, overexpression of ARKL1 decreased Zp activity in a Zp reporter assay, while ARKL1 silencing increased Zp activity, and ARKL1 overexpression decreased Jun activity in a Jun reporter assay. In addition, chromatin-immunoprecipitation assays showed that ARKL1 localized to Zp within the EBV genome. The ARKL1 sequence that binds Jun was mapped and shown to be the same sequence that binds CK2β, suggesting that ARKL1 might bind Jun indirectly through CK2β. In keeping with this model, silencing of CK2β abrogated the ability of ARKL1 to immunoprecipitate endogenous Jun and mimicked the effect of ARKL1 silencing on causing EBV reactivation. Together these findings suggest that ARKL1 binds Jun indirectly through CK2β, and that this interaction inhibits Jun activity at Zp, turning off BZLF1 expression and subsequent reactivation.
Negatively Regulates EBV Reactivation
Umama Zia Siddiqi1, Anup Vaidya1, Xinliu Li1, Lori Frappier1 1Department of Molecular Genetics, University of Toronto
EBV reactivation from the latent to the lytic cycle starts with activation of the Zp promoter controlling expression of the EBV BZLF1 transcription factor. Zp includes a binding site for c-Jun which was previously shown to be an important activator of Zp. Through proteomic approaches, we identified an uncharacterized cellular protein, Arkadia-like 1 (ARKL1), as a protein that binds the cellular CK2 kinase regulatory subunit (β) through the same binding pocket as EBNA1. Investigation of the effect of ARKL1 on EBV infection showed that silencing ARKL1 in EBV-positive gastric carcinoma (AGS-EBV), nasopharyngeal carcinoma (NPC43) or B cells (containing M81 EBV) resulted in BZLF1 expression and EBV reactivation, while overexpression of ARKL1 decreased EBV reactivation. To gain insight into the mechanism of action of ARKL1, we purified FLAG-tagged ARKL1 from human cells and identified interacting proteins by tandem mass spectrometry. This confirmed the interaction of ARKL1 with CK2 and also revealed an interaction with Jun, suggesting that ARKL1 might suppress BZLF1 expression by inhibiting Jun function at Zp. Consistent with this hypothesis, overexpression of ARKL1 decreased Zp activity in a Zp reporter assay, while ARKL1 silencing increased Zp activity, and ARKL1 overexpression decreased Jun activity in a Jun reporter assay. In addition, chromatin-immunoprecipitation assays showed that ARKL1 localized to Zp within the EBV genome. The ARKL1 sequence that binds Jun was mapped and shown to be the same sequence that binds CK2β, suggesting that ARKL1 might bind Jun indirectly through CK2β. In keeping with this model, silencing of CK2β abrogated the ability of ARKL1 to immunoprecipitate endogenous Jun and mimicked the effect of ARKL1 silencing on causing EBV reactivation. Together these findings suggest that ARKL1 binds Jun indirectly through CK2β, and that this interaction inhibits Jun activity at Zp, turning off BZLF1 expression and subsequent reactivation.