Session 5B: Epidemiology

Session Chairs: Charles Wood & Alan Chiang  (Play Circle)
Newly Derived endemic Burkitt Lymphoma Cell Lines from Kenyan Patients and the Makings of an Avatar Mouse Model Ann M. Moormann

Oral Talk #43
Newly Derived Endemic Burkitt Lymphoma Cell Lines from
Kenyan Patients and the Makings of an Avatar Mouse Model
Ann M. Moormann1, Priya Saikumar Lakshmi1, Cliff Oduor2, Rachel Gerstein1, Catherine Forconi1, Juliana A . Otieno3, Micah Luftig4, Christian Münz5, John Michael Ong’echa2, Jeffrey A. Bailey1, Michael Brehm1
1University of Massachusetts Medical School, 2Kenya Medical Research Institute, 3Jaramogi Oginga Odinga Teaching and Referral Hospital, 4Duke University, 5University of Zurich
Advances in pre-clinical research on endemic Burkitt lymphoma (eBL) have been limited by the lack of translational model systems that reproduce human immunity to EBV, capture clinically relevant viral variation, and target patient-derived eBL tumors. We have optimized methods to establish patient-derived eBL cell lines, developed a transcriptomics pipeline, and created a novel eBL Avatar mouse model by implanting patient-derived xenografts (PDX) into NOD-scid IL2rγnull (NSG) mice engrafted with human hematopoietic stem cells (HSC). HSC-engrafted NSG mice develop functional human immune system components and enable the study of interactions between human immune cells and eBL tumors. In parallel, we have characterized the tumor microenvironment (TME) associated with eBL patient outcomes by immunohistochemistry (IHC) staining of ex vivo tissue blocks and by multiparameter flow cytometric analysis of tumor infiltrating lymphocytes (TILs). Clinically relevant immune parameters, including CD8+ T cells and tumor-associated macrophages, have been replicated in preliminary studies using our eBL Avatar mouse model. The HSC-engrafted NSG mice supported growth of eBL and are being used to test the efficacy of immune checkpoint inhibitors to improve anti-tumor responses. In addition, this translational pipeline is able to interrogate human and viral gene expression profiles using single cell RNA sequencing. Combined, these next-generation methods aim to increase our understanding of eBL pathology and provide empirical guidance for new therapies to improve survival for children diagnosed with eBL. 
Oral Talk #44
Potential Correlates of Mother to Child Transmission in the
Humoral Response to KSHV in Zambia
Lisa K Poppe1, Landon N Olp1, Veenu Minhas1,2, Clement Gondwe3, Chipepo Kankasa3, John T West1, Charles Wood1
1Nebraska Center for Virology and School of Biological Sciences, University of Nebraska-Lincoln, 2Department of Epidemiology, College of Public Health, University of Nebraska Medical Center, 3Department of Pediatrics and Child Health, University Teaching Hospital
Background: While Kaposi’s sarcoma-associated herpesvirus has been recognized as the causative agent for Kaposi’s sarcoma for over 20 years, little is known about the factors that influence transmission in early childhood. Our lab has previously shown a high prevalence of early childhood infection, with 40% of Zambian children seroconverting by the age of four. Mother to child transmission is believed to be the most common route of transmission. Whether qualitative and quantitative aspects of the maternal humoral immune response against KSHV affect transmission has not been determined.
Objective: The goals of this study were to longitudinally characterize the humoral response in a cohort of HIV-infected Zambian mothers without KS, whose children were previously studied for primary KSHV infection, to identify potential factors that so-associate with KSHV transmission to the child.
Results: 89 of 127 (70.1%) mothers were determined to be KSHV seropositive at one or more time points during two years of follow-up. 81 of these mothers had at least one child who became KSHV seropositive during follow-up, while the children of the remaining 46 women never seroconverted. Mothers of KSHV seropositive children had a lower seroprevalence (54/81, 66.7%) and lower antibody titers compared to mothers of children who did not seroconvert (35/46, 76.1%). Preliminary data also suggests a higher prevalence of neutralizing antibodies among mothers of children who did not seroconvert compared to mothers of those who did.
Conclusions: KSHV seroprevalence is about 70% among HIV-infected women in Zambia, and specific antibody titers are generally low. However, KSHV seroprevalence and antibody titers are lower among mothers whose children seroconvert to KSHV compared to mothers of children who remain KSHV seronegative. 

Projects Goals
Characterize longitudinal humoral immune response in KSHV

No socio-demographic or maternal characteristics were found to affect KSHV infection rates, however, the maternal immune response were not monitored
Questions to ask: differentiate children who are KSHV + vs KSHV-

  • Mother CD4 counts: not to much difference
  • Lower KSHV Ab titers in ART-mothers
    • KSHVAb titers are fairly stable and fall into three categories (Low, intermediate, high)
    • neutralizing antibodies (nAb) positive correlation between binding Ab titers and 

Summary and conclusions
~ 70% KSHV seroprevalence in HIV+ KS- mothers
suggest that mother with higher antibody

Oral Talk #45
Genomic Analyses Support the Association of Pathogenic
Variants of Epstein-Barr Virus with Nasopharyngeal Carcinoma
Alan K.S. Chiang1,4, Tsz-Fung Chan1,4, Kwai-Fung Hui1,4, Jane J. Shen1, Pak C. Sham2, Dora L. Kwong3,4, Maria L. Lung3,4, Wanling Yang1
1Department of Pediatrics and Adolescent Medicine, Li Ka Shing Faculty of Medicine, The University of Hong Kong, 2Centre for Genomic Sciences, The University of Hong Kong, 3Department of Clinical Oncology, Li Ka Shing Faculty of Medicine, The University of Hong Kong, 4Center for Nasopharyngeal Carcinoma Research, The University of Hong Kong
Undifferentiated nasopharyngeal carcinoma (NPC) is highly prevalent in Southern Chinese populations. Epstein-Barr virus (EBV) is universally harbored in the tumour cells of NPC suggesting a pathogenic role of the virus. This project aimed to delineate pathogenic variants of EBV associated with NPC. We sequenced EBV genomes harbored in saliva of 142 population carriers and in primary tumor biopsies of 62 NPC patients of Hong Kong and compared them together with published EBV sequences derived from 3 NPC cell lines and 11 NPC biopsies. Cluster analysis identified 5 subgroups (1A-C & 2A-B) among which 1A-C mainly consisted of type 1 EBV and 2A-B consisted of type 2 and intertypic EBV. Subgroups 1A-C & 2A-B accounted for 17%, 25%, 44%, 3% and 11% of the EBV genomes harbored in saliva of the population carriers, respectively. In contrast, the majority of the EBV genomes harbored in tumor biopsies belonged to subgroups 1A and 1B, accounting for 33% and 63%, respectively, with the remaining 4% belonging to subgroup 1C. EBV genomes from the NPC cell lines, C666-1, M81 and NPC43, all belonged to subgroup 1B. Principal component analysis stratified the five subgroups along the first two principal components. Neighbour-joining tree showed that subgroups 1A and 1B were more closely related. Specific single nucleotide variations (SNVs) were found to be highly associated with subgroups 1A and 1B, distinguishing them from 1C. In conclusion, whilst 5 subgroups of EBV were discovered in the saliva of population carriers, subgroups 1A and 1B, which are characterized by specific SNVs, predominated in the tumor biopsies, supporting the association of pathogenic variants of EBV with NPC. 

He declared that : although I am the on the Scientific Advisory Board, I am the Session Chair… I have nothing to do in selecting me as the speaker! ^o^

EBV enters at the precancerous lesions

Geographyic Distributiuon of NPC are limited to severak areas (Suthern Chine, Alaska, Greenland/Eskomos/Iñupiaq/Iñupiat)

EBV variants (and diseases)
delta LMP1 is found in 

HKNPC1, 2, 3, 9 (Kwok et al, J Virol 2014)


non-sense of BILF1?


To carry out a case-control study, AoE of NPC 
  • 67 NPC biopsies 
  • 380nNPC saliva -> 71 saliva -> 24 saliva with 
  • __?? population controls (142 were used in the present study
The genomes of EBV isolated in HK Chinese are genetically distinct from those derived from other geographic regions
  • type I  A-C and 2A-B in controls
  • type 1A and 1B in 62 NPC samples
  • PC1 separates EBV genomes of NPC and control (submitted for review)
  • Potential NPC-associated SNPs (submitted for review)

Oral Talk #46
Genome Sequencing Analysis Identifies Epstein-Barr Virus
Subtypes Associated with the Risk of Nasopharyngeal
Carcinoma
Miao Xu1
1Sun Yat-sen University Cancer Center   (後由 Dr. Mu Sheng Zeng代講)
Nasopharyngeal carcinoma (NPC) is one of the most common cancers in Southern China and Southeast Asia. The traditionally accepted risk factors are at most weakly and perhaps not causally associated with NPC risk in the present era. Epstein-Barr virus (EBV) is closely related to NPC worldwide. However, the genetic basis of its association with NPC remains unclear. Here we performed whole-genome analysis of EBV genomes sequenced from 269 clinical samples, combined with publicly available EBV sequences, to reveal population-related clustering of EBV genomes and the existence of NPC-dominant strains in an endemic population. Based on an EBV genome-wide association study, we identified the strong association of EBV variants with NPC. We developed a 5-site classifier and validated it in a population-based case-control study to discriminate between high-risk and low-risk EBV subtypes for NPC. The most common subtype conferred an odds ratio of 13.11 for NPC, making it the strongest risk factor identified to date. For functional distinction between NPC high-risk and low-risk EBV subtypes, this study highlights the importance of the genes and regions that regulate viral life cycle, DNA replication and transcription, packaging, and reproduction. Our genome-wide EBV sequencing analysis establishes the first disease-associated EBV classification and provides new insights into the genetic basis for EBV-related NPC carcinogenesis and potential risk prediction. 


are there NPC-associated EBV subtypes?
269 EBV isolates were sequenced
EBV whole-genome targeted seq
EBV-positive cell line


The high-risk EBV subtype is the strongest NPC rish factor identified to date
Function of the three coding variants
  • high-rosk variants suppress EBV lytic replication, therby promote viral latency
  • the amount of EBV DNA in saliva from NPC-endemic population
  • EBV genome replication
The worldwide distribution of NPC associated variants
EBV vaccune for cancer prevenation and personalized Gary Nabel)
ENV indmiunes host inflammatory respinse and DNA damage (HLA mileculaes recognize EBV and control infection
quantifying the impact of NPC risk factors 
interaction between host SNPs and EBV subtypes

HLA-B/C genotype

           EBV low risk  |   EBV high risk (~ 20 / 10 ^ 5)
low risk A/A             4         vs           12
Intermediate A/C
high risk C/C
Oral Talk #47
The Zp-V3 Variant of the Epstein-Barr Virus BZLF1 Promoter
Enhances Lytic Reactivation in Response to B-Cell Receptor
Stimulation and is Over-Represented in Burkitt Lymphomas
Jillian A. Bristol1, Reza Djavadian1, Carrie B. Coleman2, Makoto Ohashi1, Mitchell Hayes1, James C. Romero-Masters1, Elizabeth A. Barlow1, Paul J. Farrell3, Rosemary Rochford2, Eric C. Johannsen1, Shannon C. Kenney1
1University of Wisconsin-Madison, 2University of Colorado School of Medicine, 3Imperial College Faculty of Medicine
Latent Epstein-Barr virus (EBV) infection contributes to both B-cell and epithelial-cell malignancies. Whether lytic EBV infection also contributes to EBV-induced tumors is unclear, although there is some evidence that excessive lytic EBV replication during malaria infection promotes Burkitt lymphomas (BLs). A particular variant (Zp-V3) of the viral immediate-early BZLF1 promoter (Zp) that controls lytic EBV reactivation is over- represented, relative to its frequency in non-malignant tissue, in EBV-positive nasopharyngeal carcinomas and AIDS-related lymphomas (Tong et al., JNCI 2003 and Martini et al., Journal of Infection 2007). However, no functional differences between the prototype Zp (Zp-P) and the cancer-associated variant (Zp-V3) have been identified. Here we show that a single nucleotide difference between the Zp-V3 and Zp-P promoters creates a binding site for the cellular transcription factor, NFATc1, in the Zp-V3 (but not Zp- P) variant, and greatly enhances Zp activity and lytic viral reactivation in response to NFATc1-inducing stimuli such as B-cell receptor activation and ionomycin. Furthermore, we demonstrate that restoring this NFATc1-motif to the Zp-P variant in the context of the intact EBV B95.8 strain (Zp-P containing) genome greatly enhances lytic viral reactivation in response to the NFATc1-activating agent, ionomycin. We also show that the Zp-V3 variant is over-represented in EBV-positive BLs and gastric cancers, and in EBV- transformed B-cell lines derived from EBV-infected breast milk of Kenyan mothers that had malaria during pregnancy. These results demonstrate that the Zp-V3 enhances EBV lytic reactivation to physiologically-relevant stimuli, and suggest that increased lytic infection may contribute to the increased prevalence of this variant in certain EBV-associated malignancies. 

Does lytic EBV infection contribute to malignancy?
Lytic EBV reactivation cascade: two small studies found that Zp-V3 is highly prevalent in regions where NPC and BL are common, but rare in other parts of the world
Are there functional differences between Zp-P (prototype) and Zp-V3?

Zp-P and Zp-V3 have only 7 nucleotides differences
  • Luc reporting assay
  • Both had similar responses to KLF4
  • Residue 141 in the Zp-V3 us critical for 
  • Converting A to G at -141 in Zp-P rescuers BCR responsiveness
  • The Zp-V3 variant contains an NFAT binding motif on slapping -141 residue
  • NFAT is a cellular transcription factor that is activated by BCR stimulation, but is constitutively active in many lymphoma
  • NFAT, cFos, and cJun bind cooperatively to DNA *Chen et al., Nature 392 (1998)
NFAT binds to Xp0-V3, but not Zp-P in EMSA
Gel Super shift
NFATc1 binding required for Ap1 binding to the ZIIISA motif
BCR mimic LMP2A activates the Zp-V3 more strongly than Zp-P, via the NFAT binding suite
In BJAB B cells, mutation of -141 diminishes LMP2A expression, this is also requires NFATc1 and AP1 binding

Converting the -141 Zp nucleotide in the intact B95-8 increases lytic replication
NFATc1-inducing ionomycin increases lytic protein expression in stably infected Burkett cells with Zp-V3

Endogenous NFATc1 binds more efficiently to the Zp-V3 variant in 

Is Zp-V3 over-represented in EBV+ BL?

Zp-V3 is over-represented in type 1 EBV, EBV+gastric carcinoma 

Zp-V3 is over-represented in LCLs derived from EBV+ breast milk from women that had malaria during pregnancy, suggesting that Zp-V3 containing in strains may be more prone to lytic reactivation in breast milk
Oral Talk #48
Epstein-Barr Virus (EBV) Serological Profile and Hodgkin
Lymphoma (HL) Risk
Zhiwei Liu1, Ruth Jarrett2, Henrik Hjalgrim3, Carla Proietti4, Ellen T. Chang5,6, Karin E. Smedby7, Kelly J. Yu1, Ruth Pfeiffer1, Allan Hildesheim1, Denise Doolan4, Anna E. Coghill11Division of Cancer Epidemiology and Genetics, National Cancer Institute, 2MRC, University of Glasgow Centre for Virus Research, , 3Statens Serum Institut, 4Australian Institute of Tropical Health & Medicine, James Cook University, 5Stanford Cancer Institute, 6Center for Health Sciences, Exponent, Inc., 7Department of Medicine Solna, Clinical Epidemiology Unit, Karolinska Institutet
Background. EBV infection is linked to the development of a subset of HL cases, but the association between comprehensive patterns of the humoral response to EBV and HL risk has not been fully characterized, and antibody markers often have not been evaluated separately for EBV-positive and EBV-negative HL cases.
Methods. We probed sera from 139 EBV-positive HL cases, 70 EBV-negative HL cases, and 141 apparently healthy controls frequency matched to EBV-positive HL cases on sex and age, selected from 3 studies conducted in the UK, Denmark, and Sweden. Sera were tested using a peptide microarray to measure IgG and IgA antibody responses against 199 EBV protein sequences representing 86 EBV proteins. We used unconditional logistic regression to estimate the associations between anti-EBV antibody response and HL status, adjusted for sex, age, and study area.
Findings. We focused on the 159 IgG and 193 IgA probes with coefficients of variation (CVs) less than 20%. Levels of 73 IgG and 36 IgA anti-EBV antibodies differed between EBV-positive HL cases and healthy controls (Pnegative HL cases and healthy controls (P≥0.0003).
Conclusion. We identified 12 anti-EBV antibody elevations in individuals with EBV-positive HL tumors, but not those with EBV-negative HL tumors. These data extend beyond the limited number of anti-EBV antibodies evaluated to date and suggest that differences in EBV antibody patterns among HL cases are specific to EBV-positive tumors rather than to a systematically dysregulated immune response present in all HL cases.

VCA, EA and EBNA1
Using and EBV antibody microarray, to valuate the association between 
Serum samples collected before cancer treatment were selected from 

  • Two case-control studies and three case series in Scotland in 1990s
  • Scandinavian
  • Protein microarray platform (developed by Austrian Doolean)

Boxplot of BGLF5 and EBV positive more strongest marker IgG and IgA (cool check this out)
Distinguish EBV status of HL using serological markers

Two methods were used to select antibodies that can distinguish between EBV + HL and EBV-HL

Random Forrest metrics and Stepwise logistic regression 

Top 5 BVFR2, BFRF1, BGLF5 (seq1, seq2, seq3)

Preliminary conclusions

  • We identified additional antibodies associated with EBV -positive HK, including those targeting viral lytic proteins such as BGLF5
  • Difference in EBV antibody patterns amount HL cases are specific to EBV positive tumor rather than to a systematically dysregulated immune response present 
  • in all HL cases are derived by IgG rather than IgA

—poster 122 and 125 (and pdf 5152, 5271)



Leave a Comment

Scroll to Top