Session Chairs: Denise Whitby & Christian Münz
Oral Talk #50
Malaria as a Risk Factor for Early Age of KSHV Infection in a
Cohort of Children in Western Kenya
Rosemary Rochford1, Katherine Sabourin1, Ibrahim Daud3, Peter O. Sumba3, Nazzarrena Labo2, Wendell Miley2, Denise Whitby2
1University of Colorado, 2Frederick National Laboratory for Cancer Research, 3Kenya Medical Research Institute
Kaposi sarcoma herpes virus (KSHV) is endemic to Sub-Saharan Africa. However, the geographic distribution is not uniform within the region suggesting a difference in factors that promote transmission. Earlier studies suggest that malaria may be important for the transmission of KSHV. To follow-up on this observation, we evaluated samples from a cohort of Kenyan infants, followed through age three, to explore whether children with Plasmodium infection were more likely to become KSHV seropositive. Plasma from 167 Kenyan children was tested for KSHV and P. falciparum antibodies. Children were considered seropositive for P. falciparum if antibodies to MSP-1 proteins from any of three Plasmodium strains were detected using a bead-based Luminex multiplex assay. Seropositivity for KSHV was determined by presence of antibodies to ORF73, ORF61, ORF38, ORF65, ORF59, K8.1, K8.1b, or K5 as detected by Luminex (Plos Pathogen, 2014 10(3)e1004046). To determine if Plasmodium infection in children was associated with earlier KSHV seroconversion, we utilized a cox proportional hazards regression model including Plasmodium seropositivity as a time varying covariate and KSHV sero-status at 12, 18, 24, 30 and 36 months as the outcome. When measured by K8.1 and ORF73 only, 16 (10%) children were KSHV seropositive by age 3 versus 52(31%) when using the broader array of KSHV proteins. The majority of children had at least one Plasmodium infection by 3 years of age (n=157, 94%). When KSHV seropositivity was measured using antibodies to eight KSHV proteins, there was a significant increase in the hazard of being KSHV seropositive among children that had a Plasmodium infection prior to KSHV seroconversion (HR=2.3, 95%CI:1.1-4.9). Our research suggests that malaria may have a role in enhancing KSHV transmission and ongoing studies are exploring this possibility.
Cohort of Children in Western Kenya
Rosemary Rochford1, Katherine Sabourin1, Ibrahim Daud3, Peter O. Sumba3, Nazzarrena Labo2, Wendell Miley2, Denise Whitby2
1University of Colorado, 2Frederick National Laboratory for Cancer Research, 3Kenya Medical Research Institute
Kaposi sarcoma herpes virus (KSHV) is endemic to Sub-Saharan Africa. However, the geographic distribution is not uniform within the region suggesting a difference in factors that promote transmission. Earlier studies suggest that malaria may be important for the transmission of KSHV. To follow-up on this observation, we evaluated samples from a cohort of Kenyan infants, followed through age three, to explore whether children with Plasmodium infection were more likely to become KSHV seropositive. Plasma from 167 Kenyan children was tested for KSHV and P. falciparum antibodies. Children were considered seropositive for P. falciparum if antibodies to MSP-1 proteins from any of three Plasmodium strains were detected using a bead-based Luminex multiplex assay. Seropositivity for KSHV was determined by presence of antibodies to ORF73, ORF61, ORF38, ORF65, ORF59, K8.1, K8.1b, or K5 as detected by Luminex (Plos Pathogen, 2014 10(3)e1004046). To determine if Plasmodium infection in children was associated with earlier KSHV seroconversion, we utilized a cox proportional hazards regression model including Plasmodium seropositivity as a time varying covariate and KSHV sero-status at 12, 18, 24, 30 and 36 months as the outcome. When measured by K8.1 and ORF73 only, 16 (10%) children were KSHV seropositive by age 3 versus 52(31%) when using the broader array of KSHV proteins. The majority of children had at least one Plasmodium infection by 3 years of age (n=157, 94%). When KSHV seropositivity was measured using antibodies to eight KSHV proteins, there was a significant increase in the hazard of being KSHV seropositive among children that had a Plasmodium infection prior to KSHV seroconversion (HR=2.3, 95%CI:1.1-4.9). Our research suggests that malaria may have a role in enhancing KSHV transmission and ongoing studies are exploring this possibility.
KSHV Requires EBV for Infection of Peripheral B Cells Aurélia Faure
Oral Talk #51
KSHV Requires EBV for Infection of Peripheral B Cells
Aurelia Faure1, Bill Sugden1
1McArdle Laboratory for Cancer Research, University of Wisconsin-Madison
Primary Effusion Lymphoma (PEL) is the only human cancer known to be associated with two tumor viruses. PELs are causally associated with Kaposi’s sarcoma-associated herpesvirus (KSHV), and 90% of them are also co-infected with Epstein-Barr virus (EBV). Despite its association with B cell malignancy, it has notoriously been difficult to infect peripheral B cells with KSHV in vitro. We have monitored infection of peripheral CD19+ B cells with a strain of KSHV that encodes GFP (KSHV BAC16) by flow cytometry. We show that KSHV cannot establish infection in peripheral B cells in the absence of B cell stimulation or EBV co-infection. B cell stimulation only supports inefficient and non- reproducible KSHV infection. EBV infection, however, supports co-infection of up to 3.5 % of primary B cells by KSHV. We found that the order and timing of infection with the two viruses significantly alters the efficiency of infection with KSHV. KSHV infection is most efficient when cells are exposed to it prior to EBV infection whereas this efficiency decreases rapidly following infection with EBV. We have isolated and characterized pure populations of co-infected peripheral B cells. KSHV is lost from them at a rate of ~3.5% per generation indicating that KSHV is not providing these cells sufficient advantages to be maintained in them. Our data reveal that EBV plays an essential but complex role allowing infection of peripheral B cells with KSHV and more generally, that EBV is likely to be essential for establishment of PEL. Our data also indicate that additional genetic and/or epigenetic changes must occur in these co-infected cells for KSHV to be maintained and for them to evolve into PELs.
Aurelia Faure1, Bill Sugden1
1McArdle Laboratory for Cancer Research, University of Wisconsin-Madison
Primary Effusion Lymphoma (PEL) is the only human cancer known to be associated with two tumor viruses. PELs are causally associated with Kaposi’s sarcoma-associated herpesvirus (KSHV), and 90% of them are also co-infected with Epstein-Barr virus (EBV). Despite its association with B cell malignancy, it has notoriously been difficult to infect peripheral B cells with KSHV in vitro. We have monitored infection of peripheral CD19+ B cells with a strain of KSHV that encodes GFP (KSHV BAC16) by flow cytometry. We show that KSHV cannot establish infection in peripheral B cells in the absence of B cell stimulation or EBV co-infection. B cell stimulation only supports inefficient and non- reproducible KSHV infection. EBV infection, however, supports co-infection of up to 3.5 % of primary B cells by KSHV. We found that the order and timing of infection with the two viruses significantly alters the efficiency of infection with KSHV. KSHV infection is most efficient when cells are exposed to it prior to EBV infection whereas this efficiency decreases rapidly following infection with EBV. We have isolated and characterized pure populations of co-infected peripheral B cells. KSHV is lost from them at a rate of ~3.5% per generation indicating that KSHV is not providing these cells sufficient advantages to be maintained in them. Our data reveal that EBV plays an essential but complex role allowing infection of peripheral B cells with KSHV and more generally, that EBV is likely to be essential for establishment of PEL. Our data also indicate that additional genetic and/or epigenetic changes must occur in these co-infected cells for KSHV to be maintained and for them to evolve into PELs.
Oral Talk #52
Tumor Epstein-Barr Virus Status is Prognostic in Primary
Effusion Lymphoma
Kathryn Lurain1, Mark N. Polizzotto1, Priscila H. Goncalves1, Ramya Ramaswami1, Armando C. Filie2, Seth M. Steinberg3, Vickie Marshall4, Wendell Miley4, Richard Little1, Denise Whitby4, Elaine S. Jaffe2, Stefania Pittaluga2,Robert Yarchoan1, Thomas S. Uldrick1
1HIV & AIDS Malignancy Branch, Center for Cancer Research, National Cancer Institute, 2Laboratory of Pathology, Center for Cancer Research, National Cancer Institute, 3Biostatistics and Data Management Section, Center for Cancer Research, National Cancer Institute, 4Viral Oncology Section, AIDS and Cancer Virus Program, Leidos-Biomedical, Frederick National Laboratory for Cancer Research
Background: Kaposi sarcoma herpesvirus (KSHV) is the etiologic agent of primary effusion lymphoma (PEL), an aggressive HIV-associated non-Hodgkin B-cell lymphoma, and a form of multicentric Castleman disease (KSHV-MCD). The role of Epstein-Barr virus (EBV) coinfection, which is observed in 70-90% of PEL cells, is poorly understood. We evaluated clinical correlates of peripheral blood mononuclear cell (PBMC)-associated EBV viral load (VL) and tumor EBV status in PEL patients.
Methods: We identified 19 HIV-infected PEL patients who received antiretroviral therapy and chemotherapy, 19 KSHV-MCD patients, and 28 HIV-associated lymphoma (HIV-L) patients. Tumor EBV status in PEL was evaluated by in situ hybridization against EBV- encoded small RNA (EBER). PBMC EBV VL was evaluated by PCR for pol, normalized to ERV3 (units, copies/106 PBMC). EBV VL in PEL, KSHV-MCD, and HIV-L were compared using two-tailed rank sum tests. Cancer-specific survival analyses were performed with Kaplan-Meier methods, two-sided log-rank test and exploratory Cox modeling using backward selection were performed to identify prognostic viral, immunologic or clinical factors.
Results: Baseline EBV VL in PEL (median=1,580 copies/mL) was comparable to that in KSHV-MCD (median=488 copies/mL; p=0.09) and HIV-L (median 467 copies/mL; p=0.25). There was a plateau in overall survival (OS) in PEL patients after 2 years, with 3-year cancer-specific survival of 47%. Tumor EBER positivity was associated with significantly decreased mortality (HR=0.265; 95% CI: 0.076-0.931; p=0.038). At 3 years of follow up, the median OS for 14 EBER+ PEL was not reached (64% 3-year survival rate) and for 5 EBER- patients was 10.4 months (p=0.029). However, baseline EBV VL was not prognostic (p=0.70).
Conclusions: EBV viral load is comparable in PEL, KSHV-MCD and HIV-L. Tumor EBV status, but not peripheral blood EBV VL, is prognostic in PEL. EBV-positive PEL has a significantly improved prognosis compared to EBV-negative PEL, with a plateau in long- term survival at 2 years.
Effusion Lymphoma
Kathryn Lurain1, Mark N. Polizzotto1, Priscila H. Goncalves1, Ramya Ramaswami1, Armando C. Filie2, Seth M. Steinberg3, Vickie Marshall4, Wendell Miley4, Richard Little1, Denise Whitby4, Elaine S. Jaffe2, Stefania Pittaluga2,Robert Yarchoan1, Thomas S. Uldrick1
1HIV & AIDS Malignancy Branch, Center for Cancer Research, National Cancer Institute, 2Laboratory of Pathology, Center for Cancer Research, National Cancer Institute, 3Biostatistics and Data Management Section, Center for Cancer Research, National Cancer Institute, 4Viral Oncology Section, AIDS and Cancer Virus Program, Leidos-Biomedical, Frederick National Laboratory for Cancer Research
Background: Kaposi sarcoma herpesvirus (KSHV) is the etiologic agent of primary effusion lymphoma (PEL), an aggressive HIV-associated non-Hodgkin B-cell lymphoma, and a form of multicentric Castleman disease (KSHV-MCD). The role of Epstein-Barr virus (EBV) coinfection, which is observed in 70-90% of PEL cells, is poorly understood. We evaluated clinical correlates of peripheral blood mononuclear cell (PBMC)-associated EBV viral load (VL) and tumor EBV status in PEL patients.
Methods: We identified 19 HIV-infected PEL patients who received antiretroviral therapy and chemotherapy, 19 KSHV-MCD patients, and 28 HIV-associated lymphoma (HIV-L) patients. Tumor EBV status in PEL was evaluated by in situ hybridization against EBV- encoded small RNA (EBER). PBMC EBV VL was evaluated by PCR for pol, normalized to ERV3 (units, copies/106 PBMC). EBV VL in PEL, KSHV-MCD, and HIV-L were compared using two-tailed rank sum tests. Cancer-specific survival analyses were performed with Kaplan-Meier methods, two-sided log-rank test and exploratory Cox modeling using backward selection were performed to identify prognostic viral, immunologic or clinical factors.
Results: Baseline EBV VL in PEL (median=1,580 copies/mL) was comparable to that in KSHV-MCD (median=488 copies/mL; p=0.09) and HIV-L (median 467 copies/mL; p=0.25). There was a plateau in overall survival (OS) in PEL patients after 2 years, with 3-year cancer-specific survival of 47%. Tumor EBER positivity was associated with significantly decreased mortality (HR=0.265; 95% CI: 0.076-0.931; p=0.038). At 3 years of follow up, the median OS for 14 EBER+ PEL was not reached (64% 3-year survival rate) and for 5 EBER- patients was 10.4 months (p=0.029). However, baseline EBV VL was not prognostic (p=0.70).
Conclusions: EBV viral load is comparable in PEL, KSHV-MCD and HIV-L. Tumor EBV status, but not peripheral blood EBV VL, is prognostic in PEL. EBV-positive PEL has a significantly improved prognosis compared to EBV-negative PEL, with a plateau in long- term survival at 2 years.
PEL
- rare, aggressive HIV-associated B-cell non-HL median life < 1 yr
- ~ 80% of PEL cells ar co-infected with EBV
- Inflammatory markers in PEL
- IL6 and IL10 are prognostic in PEL
Biomarkers:
- EBV status: EBER
- PBMC-associated KSHV and EBV viral load PCR for K6 and Pol
- In HIB-L patients, KSHV serostatus – K8.1 +/-or LANA
- Statistical
19 PEL patients (14 EBV 5 EBV-)
- 1 patient KSHV-MCD
- 10 PELs
- 8
Baseline of EBV and KSHV viral loads are not prognostic
Tumor EBV status is prognostic in PEL
Multi…. EBV associated PEL has better survival
Notes: PBMC-associated KSHV VL is useful in evaluating PEL patients
Oral Talk #53
Epstein-Barr Virus (EBV) Enhances Latent Plasmid
Maintenance of Kaposi Sarcoma-Associated Herpesvirus
(KSHV)
Rachele Bigi1, Hyowon An1, Nancy Raab-Traub1, Dirk Dittmer1
1University of North Carolina at Chapel Hill
Primary effusion lymphoma (PEL) is a B-cell lymphoma, which is always associated with Kaposi’s Sarcoma-associated herpesvirus (KSHV). In many cases PEL are also co- infected with Epstein-Barr virus (EBV), in others they are not. Thus, the requirement for EBV co-infection is not clear. We find that adding EBV genomes to KSHV+ single positive PEL lead to increased KSHV genome maintenance and KSHV Latency-Associated Nuclear Antigen (LANA) expression on a single cell basis. In naturally co-infected PEL, EBV was necessary for KSHV plasmid maintenance. When KSHV+/EBV+ PEL were transfected with EBV- CRISPR/Cas9 plasmids targeting different regions in EBV to deplete the viral genome, we observed a dramatic decrease in cell viability, KSHV genome copy number and LANA protein expression. Ectopic expression of EBNA-1 ameliorated this effect, even though LANA and EBNA-1 did not colocalize in PEL. This implies that EBV plays an important role in the pathogenesis of co-infected PEL by increasing KSHV viral load and LANA expression. It also suggests that small molecules that target EBV EBNA-1 may have efficacy against dually-infected PEL, as these cells have become addicted to EBV.
PEL harbors latent KSHV, many also latent EBV
JSC1, BC1 (EBV + KSHV +)
BC3, CRO-AP6 (KSHV+)
BC3.EBV : EBV is maintained without selection in naturally co-infected PEL, but lost from ex Vito infected PEL ex Vito
KSHV copy number correlated with EBV copy number
KSHV copy number correlated with the number of “LANA dots”
EBNA1 and LANA do not co-localized
KSHV copy number is higher in EBV-positive cells (by digital PCR for single cell, ex Vito EBNV infected
There is a set-point of KSHV genome copy number, which is modulated by EBV
EBV-specific CRISPR/cas-0 destroy EBV genomes (Wang and Quake PNAS 2014)
PCEP4
CRISPR/Cas-9 mediated EBV deletion in PEL: EBV is necessary to maintain KSHV in duality-infected PEL cells
Removing EBV from EBV-pos PEL causes cell death
EBNA1 in-trans rescues genome loss and survival
Genome-wide CGH (Blood 2011, —> Dirk Dittmer’s group)
Maintenance of Kaposi Sarcoma-Associated Herpesvirus
(KSHV)
Rachele Bigi1, Hyowon An1, Nancy Raab-Traub1, Dirk Dittmer1
1University of North Carolina at Chapel Hill
Primary effusion lymphoma (PEL) is a B-cell lymphoma, which is always associated with Kaposi’s Sarcoma-associated herpesvirus (KSHV). In many cases PEL are also co- infected with Epstein-Barr virus (EBV), in others they are not. Thus, the requirement for EBV co-infection is not clear. We find that adding EBV genomes to KSHV+ single positive PEL lead to increased KSHV genome maintenance and KSHV Latency-Associated Nuclear Antigen (LANA) expression on a single cell basis. In naturally co-infected PEL, EBV was necessary for KSHV plasmid maintenance. When KSHV+/EBV+ PEL were transfected with EBV- CRISPR/Cas9 plasmids targeting different regions in EBV to deplete the viral genome, we observed a dramatic decrease in cell viability, KSHV genome copy number and LANA protein expression. Ectopic expression of EBNA-1 ameliorated this effect, even though LANA and EBNA-1 did not colocalize in PEL. This implies that EBV plays an important role in the pathogenesis of co-infected PEL by increasing KSHV viral load and LANA expression. It also suggests that small molecules that target EBV EBNA-1 may have efficacy against dually-infected PEL, as these cells have become addicted to EBV.
PEL harbors latent KSHV, many also latent EBV
JSC1, BC1 (EBV + KSHV +)
BC3, CRO-AP6 (KSHV+)
BC3.EBV : EBV is maintained without selection in naturally co-infected PEL, but lost from ex Vito infected PEL ex Vito
KSHV copy number correlated with EBV copy number
KSHV copy number correlated with the number of “LANA dots”
EBNA1 and LANA do not co-localized
KSHV copy number is higher in EBV-positive cells (by digital PCR for single cell, ex Vito EBNV infected
There is a set-point of KSHV genome copy number, which is modulated by EBV
EBV-specific CRISPR/cas-0 destroy EBV genomes (Wang and Quake PNAS 2014)
PCEP4
CRISPR/Cas-9 mediated EBV deletion in PEL: EBV is necessary to maintain KSHV in duality-infected PEL cells
Removing EBV from EBV-pos PEL causes cell death
EBNA1 in-trans rescues genome loss and survival
Genome-wide CGH (Blood 2011, —> Dirk Dittmer’s group)