NIAID Workshop on Flavivirus Immunity

Abstract

On September 16, 2009, the National Institute of Allergy and Infectious Diseases (NIAID), part of the U.S. National Institutes of Health, convened a workshop to discuss current knowledge of T- and B-cell immune epitopes for members of the Flavivirus genus (family Flaviviridae), and how this information could be used to increase our basic understanding of host-pathogen interactions and/or advance the development of new or improved vaccines and diagnostics for these pathogens. B-cell and T-cell responses to flaviviruses are critical components of protective immunity against these pathogens. However, they have also been linked to disease pathogenesis. A detailed understanding of the biological significance of immune epitope information may provide clues regarding the mechanisms governing the induction of protective versus pathogenic adaptive immune responses.

Introduction

T he Flavivirus genus is composed of enveloped viruses with a single-stranded positive-sense RNA genome. Flavivirus members discussed at the workshop were, in order of presentation: dengue virus (DENV); yellow fever (YFV) and tick-borne encephalitis viruses (TBEV); West Nile virus (WNV); and Japanese (JEV), St. Louis (SLEV), and Murray Valley (MVEV) encephalitis viruses. Members of the Flavivirus genus are responsible for a significant amount of morbidity and mortality worldwide. For example, nearly 3 billion people are at risk of contracting DENV. The World Health Organization (WHO) estimates that 50–100 million DENV infections occur annually (WHO fact sheet: http://www.who.int/mediacentre/factsheets/fs117/en/index.html). YFV cases number approximately 20,000 per year, mostly in tropical regions of Africa and parts of South America; mortality ranges from 15–50% (WHO fact sheet: http://www.who.int/mediacentre/factsheets/fs100/en/index.html). WNV is the leading cause of human arboviral infection in the United States, with over 28,000 cases and 1200 deaths reported since 1999 (9). JEV is the leading cause of viral encephalitis in Asia, with approximately 50,000 reported cases per year and up to a 30% mortality rate, mostly in children (WHO fact sheet: http://www.who.int/vaccine_research/diseases/vector/en/index1.html#disease%20burden). SLEV infection is relatively rare, with 4651 cases reported throughout the United States from 1964 to 2005, and exhibits a mortality range of 5–30%, with higher rates seen in the elderly (CDC fact sheet: http://www.cdc.gov/ncidod/dvbid/sle/Sle_FactSheet.html). Vaccines exist for YFV, JEV, and TBEV, and are in development for DENV and WNV. However, limited information exists about the mechanisms of protection of the licensed vaccines, and of adaptive immune responses to natural infections in humans.

The NIAID supports the Immune Epitope Database and Analysis Resource (IEDB, www.immuneepitope.org) (23), to provide the research community with a web-based warehouse of immune epitope information for infectious and immune-mediated diseases, and access to immune epitope prediction and analysis tools. The entire IEDB currently contains nearly 65,000 unique immune epitopes (over 13,000 mouse and over 20,000 human, with the rest derived from non-human primate and other species for which epitope information is available), curated from approximately 8500 research articles, or directly submitted by researchers to the IEDB website. The IEDB defines immune epitopes as chemical structures recognized by specific immune receptors, including T-cell receptors (TCR), major histocompatibility (MHC) molecules, and antibodies. Further criteria are applied to identify antibody and T-cell epitopes for curation into the IEDB. Epitopes for linear B-cell and CD8 T-cell/class I MHC epitopes are defined as being no more than 11 residues long, with a minimum of 7 residues for the CD8 T-cell epitopes. Linear CD4 T-cell/class II MHC epitopes are defined as being 15 residues or fewer in length. Epitope-containing regions are defined as 12 residues or greater for B-cell and CD8 T-cell/class I MHC epitopes, and 16 residues or greater for CD4 T-cell/class II MHC epitopes (see the IEDB curation manual: http://tools.immuneepitope.org/wiki/index.php/Curation_Manual2.0#Minimum_Criteria_for_Curation). The IEDB includes both the epitope sequences and accompanying experimental data from the published literature, and direct submissions from investigators. These epitopes have been characterized to varying degrees in the literature. To be included in the IEDB, minimal characterization criteria are currently required, which are: B-cell epitopes–antibody binding data; T-cell epitopes–MHC binding data and/or evidence of T-cell stimulation (in vivo or in vitro). Additional criteria are curated into the IEDB, if available, including: evidence of in-vivo immune protection, structural or co-crystallization data; adoptive transfer studies; and mechanisms of immune-mediated pathogenesis.

Immune epitope knowledge can be used to measure immune responses to natural infections or vaccines. For example, the production of MHC class I and class II tetramers, for monitoring CD8 and CD4 T-cell responses, requires detailed information about the composition of the immune epitopes (i.e., peptides) recognized by the T cells (1). Immune epitope information can also contribute to understanding mechanisms of immune protection and immune-mediated pathogenesis triggered by infections, or may facilitate the development of immune-based diagnostics or subunit vaccines. The IEDB investigators have conducted detailed meta-analyses of immune epitope information for a number of pathogens (4,7,29,32). These analyses highlight current knowledge of antibody and T-cell epitopes to the particular pathogens, describe association of specific epitopes with protective immunity or disease state, and identify opportunities for future studies to fill knowledge gaps.

Induction of B-cell and T-cell responses to flaviviruses is a critical component of protective immunity against these pathogens. It is generally thought that people who have recovered from flavivirus infections have lifelong immunity to that particular virus; this is a debatable issue and in need of further study. Adaptive immune responses have also been linked to disease pathogenesis for several members of the Flavivirus genus (2,15,17,21). NIAID program staff requested the IEDB contractors to conduct a meta-analysis of flavivirus immune epitopes and convened a panel of flavivirus experts (listed in the Acknowledgments) to discuss the biological implications and research opportunities to improve our knowledge of flavivirus immunity (30). The main areas of discussion at the workshop were the use of immune epitope information for characterizing protective immunity, and understanding mechanisms of immune-mediated disease pathogenesis and advancing vaccine development, including generation of cross-protective vaccines. This article summarizes the major issues for advancing knowledge of adaptive immune responses to the Flavivirus members discussed at the workshop, which focused on DENV, YFV, TBEV, WNV, JEV, SLEV, and MVEV.

Dengue Virus Group

The DENV group is composed of four antigenically distinct, but closely related, serotypes (DENV 1, 2, 3, and 4). The virus is transmitted to humans by Aedes aegypti and Aedes albopictus mosquitoes (28). Many infected individuals exhibit no clinical signs (asymptomatic), but tens of millions of cases of the debilitating but self-limiting dengue fever (DF), and up to 500,000 cases of potentially fatal dengue hemorrhagic fever (DHF)/dengue shock syndrome (DSS), occur annually (11). In 2007, over 890,000 cases of DENV infections were reported in the Americas alone, including approximately 26,000 cases of DHF (31). The majority of DHF/DSS cases occur in children and adults experiencing a secondary infection with a different serotype, with a minority of DHF/DSS cases occurring in infants born to dengue-immune mothers (8). However, the mechanisms underlying the development of DHF and DSS are not fully understood. While some antibody and T-cell epitopes have been defined for dengue viruses, and selected HLA alleles correlate with disease severity (27), additional epitopes need to be defined to link immune reactivity to various disease states or to identify correlates of immune protection. The development of DHF may be influenced by the functional characteristics of B- and T-cells after immune activation. For example, individual T-cells may recognize identical or related epitopes through their unique TCR, but produce different cytokines or exhibit diverse phenotypes based on TCR avidity (25). Similarly, B-cells may produce antibodies with distinct functional properties based on the viral antigenic epitope responsible for B-cell activation (24). Therefore the repertoire of antibodies directed against specific epitopes and/or their affinity may influence their protective or pathogenic potential upon subsequent infection with a distinct DENV serotype. The IEDB currently contains approximately 400 B-cell (approximately130 mouse and 280 human), and 200 T-cell epitopes (approximately 50 mouse and 160 human) for DENV. The bulk of antibody epitopes have been defined in the viral envelope (E) protein, and the majority of the T-cell epitope data are from studies of DENV 2 strains. This information suggests that more immune epitope studies are needed for other viral proteins, and for DENV serotypes 1, 3, and 4. The DENV serotypes 1, 2, 3, and 4 are really four different viruses, and exhibit a significant amount of nucleic acid sequence diversity. In addition to reagents for immune analysis, reagents are also needed to study DENV pathophysiology. Finally, additional animal and human studies would improve our understanding of host immune responses to DENVs, their contributions to disease pathogenesis, and methods to apply this knowledge to vaccine development. Existing animal models can be used to investigate these questions while new animal models are developed that more closely correlate with human responses to DENV infection. Prospective human cohort studies and the development and use of sample-sparing assays would also improve knowledge of human adaptive immune responses to DENV infections, furthering the development of safe, cross-serotype protective vaccines. The risk of a dengue vaccine inducing protection against some but not all four serotypes could increase the risk for more severe infection by the unprotected serotypes. This major public health concern and challenge requires the generation of protective immunity to all four serotypes, and would be assisted by a better understanding of the human B- and T-cell epitopes responding to DENV infections, and/or DENV vaccines.

Yellow Fever and Tick-Borne Encephalitis Viruses

Yellow fever virus (YFV) is spread to humans through the bite of infected Aedes or Haemogogus mosquitoes. Clinical manifestations range from a self-limiting acute febrile phase that lasts for 3–4 days, accompanied by severe muscle aches, nausea, and vomiting, to a more toxic form of the disease that affects multiple organ systems and is often fatal. Individuals living or traveling to Africa and jungle regions of Latin America are at highest risk of exposure to YFV. The WHO estimates that there are approximately 20,000 cases annually of yellow fever worldwide, with a 15–50% mortality rate. A licensed vaccine for yellow fever has been available since the 1930s. The current vaccine is composed of the live attenuated 17D YFV strain, and a single dose appears to confer protection in all inoculated individuals for up to 10 years. Serious, and at times fatal, adverse events have occurred in a small percentage of YFV vaccine recipients, and have been increasingly reported in recent years (3,10,12,19,22, and http://wwwnc.cdc.gov/travel/content/yellowbook/home-2008.aspx)

The IEDB currently contains four B-cell (1 mouse and 3 human) and 165 T-cell (158 mouse and 7 human) epitopes for YFV. Further examination of this information shows that the YFV epitope information is focused on the 17D virus and mouse immune responses, with very limited information on human immunity. Although the published literature, from which the IEDB obtained their data, contains some valuable information on T-cell responses to the 17D YFV strain, the mechanisms of 17D vaccine-mediated B- and T-cell immunity are poorly defined, and identification of critical protective B- and T-cell epitopes is lacking. In addition, very little information exists about immune responses and immune epitopes for wild-type YFV, and how these compare with vaccine-induced immunity. This lack of information is probably due largely to difficulties in obtaining blood samples from infected individuals, since many patients do not seek medical care because of their remote location. In addition, there has been very little research activity on the human immune responses to YFV due in part to the availability of a very effective live vaccine. Further research in this area is warranted, especially since serious adverse events associated with the YFV 17D vaccine do occur. Immunological studies of YFV 17D vaccinees could serve as a model system to characterize an effective protective immune response to a vaccine. Additional understudied areas of research mentioned at the workshop included the effects of pre-existing immunity to YFV infection or vaccination on generation or maintenance of immunity to other flaviviruses and vice versa, the mechanism of antibody-mediated enhancement of disease associated with passive antibody transfer, identification of T-cell epitopes presented by HLA A, B, and C alleles and development of HLA-C reagents and the need for MHC typing, and identification/development of appropriate animal models and the best viral/vaccine strain to use to study immune protection. Primate models may be important, but there is also a need to develop and use viral strains that will cause disease in animal models to support virus challenge studies.

Tick-borne encephalitis virus, or TBEV, is spread by the bite of infected hard-bodied (ixodid) ticks. Ixodes ricinus and Dermacentor marginatus ticks are the most common TBEV carriers. Three substrains of TBEV have been identified: European or Western tick-borne encephalitis virus, Siberian tick-borne encephalitis virus, and Far-Eastern tick-borne encephalitis virus (also known as Russian Spring Summer encephalitis virus). These viruses induce meningitis, encephalitis, or meningoencephalitis in infected individuals (20). Several thousand cases are reported annually and disease severity has been linked to the virus substrain. For example, individuals infected with the Far-Eastern subtype generally exhibit more severe disease and a higher fatality rate, whereas the Siberian subtype may induce a more chronic encephalitis and neuropsychiatric sequelae (14). The licensed vaccine is not available in the United States, but is available in some endemic countries.

The IEDB currently contains 22 mouse TBEV B-cell epitopes curated from the scientific literature. No mouse T-cell or human B- and T-cell epitopes were found by IEDB staff after a search of the published literature. The limited immunological information on TBEV may be due to the fact that these viruses are classified as HHS select agents and require biosafety level 3–4 facilities for study. Further immunological studies are warranted because the potential exists for development of a cross-serogroup vaccine due to high amino acid sequence homology among the TBEV groups. Therefore additional studies are needed to identify cross-protective epitopes and to develop and test novel cross-serogroup vaccines that are safe for children and adults.

Japanese Encephalitis Virus Group (West Nile Virus and Japanese, St. Louis, and Murray Valley Encephalitis Viruses)

The Japanese encephalitis virus complex members discussed at the workshop include WNV and SLEV, MVEV, and JEV (13). These viruses are transmitted to humans primarily through bites from infected Culex and Aedes mosquitoes. WNV is the leading cause of human arboviral infections in the United States, with over 28,000 cases and 1200 deaths reported since 1999 (9). Although most infections are asymptomatic, a mild febrile syndrome and more severe neuroinvasive disease may occur. For example, 3630 cases of WNV neuroinvasive disease and 124 deaths were reported in 2007, mainly in elderly patients. SLEV infections are mainly seen in the United States, but cases have been reported in Mexico and Canada. Clinical disease is relatively rare, with 4651 cases reported throughout the United States from 1964 to 2005, with a mortality range of 5–30%, although higher rates have been observed among the elderly. MVEV is endemic in northern Australia and Papua New Guinea. The majority of MVEV infections result in subclinical disease. However, a small number of MVEV-infected individuals develop encephalitis, with a 25% mortality rate. JEV is the most clinically relevant virus in this group, and is the leading cause of viral encephalitis in Asia. Licensed JEV vaccines exist in Asia, Australia, the European Union, and the United States. However, approximately 50,000 new cases are reported each year, mainly in children in Asia, with a mortality rate of approximately 25%.

A search for JEV serocomplex immune epitopes curated within the IEDB demonstrates that WNV is the most extensively studied virus, with 181 T-cell (158 mouse and 23 human) and 29 B-cell (23 mouse and 6 human) epitopes identified. Although JEV induces higher morbidity and mortality worldwide, only 19 mouse T-cell (all mouse) and 22 B-cell epitopes (14 mouse and 8 human) have been described. The numbers are even lower for SLEV and MVEV, with no T-cell and 9 mouse B-cell epitopes reported for SLEV, and only 12 mouse T-cell and 17 mouse B-cell epitopes described for MVEV.

Further studies are recommended to determine the requirements for generation of protective immunity to JEV serocomplex members, including the role of innate immunity and immunopathology by adaptive T- and B-cell responses, since immune responses may also play a role in disease pathogenesis. WNV is the most studied virus in this group, and recent evidence suggests that regulatory T cells play a role in controlling WNV symptoms in mice and humans (18). Protective CD4 and CD8 T-cell functions have been described in mice and humans (5,6,26). Further analyses are required to decipher the mechanisms regulating protective or pathogenic T-cell responses in WNV infections. In terms of B-cell responses to WNV and other Flavivirus members, it would be valuable to know whether antibodies directed against different regions of a specific viral protein (e.g., WNV E protein) also have different functions, and how such observations relate to human infection and disease. It would also be valuable to know whether the T-cell epitope hierarchy seen in WNV is true for other flaviviruses, how data obtained from analysis of T-cell epitope responses from peripheral blood relate to neuroinvasive disease, and to define antibody and T-cell epitopes in the natural host for Flavivirus infection (e.g., avian).

Among the JEV serocomplex members, with the exception of WNV, there is very limited knowledge of T-cell and B-cell epitopes, and it is not clear whether WNV is good model for the other viruses in this serocomplex. Antibodies to JEV proteins NS3 and NS5 should be defined, and improved efforts are needed to identify human B-cell and T-cell epitopes for JEV, MVEV, and SLEV. Another important question regarding JEV immunity is elucidation of the mechanisms controlling the lack of immune-mediated pathology after subsequent infection with another JEV serotype, as is seen in DENV infections. As with TBEV, there is a concern that the listing of JEV as a select agent hinders research efforts. Studies demonstrating that antibodies to JEV can be generated in pigs may help support the removal of this virus from the HHS select agent list.

Conclusion

The primary goals of this NIAID-supported workshop were to discuss our current knowledge of B- and T-cell epitopes derived from flaviviruses and how this knowledge could be applied to improving our understanding of the mechanisms of protective immunity or immune-mediated pathogenesis, as well as the development of new or improved vaccines for members of the Flavivirus genus. Suggestions for additional research efforts include: development of prospective human cohort studies of natural infections and vaccinated populations to better define human T- and B-cell epitopes and their roles in immunopathogenesis of disease and natural and vaccine-induced protection; the use of other more tractable flaviviruses like WNV to develop technologies to study host immunity/disease pathogenesis followed by translation of these technologies to studies of other flaviviruses; correlation of results from DENV antibody epitope studies with antibody recognition of other flaviviruses; translation of in-vitro findings to studies of protection or enhancement of infection and disease in animal models and human populations; increased use of non-human primates in studies of flavivirus immunity and disease pathogenesis; and the development of collaborations with immunologists, flavivirus experts, and clinical investigators to increase access to and studies of samples from infected individuals. Curation of B-cell epitopes within the IEDB is currently limited to those epitopes for which molecular structures or amino acid sequences have been defined. There is a significant body of literature describing antibody responses to various flaviviruses, including protective antibodies, for which the molecular structures have not been defined. Analysis of this information could very likely provide novel insights into the mechanisms of protective immunity and immune-mediated pathogenesis to flaviviruses. It was suggested that such information should be considered for inclusion in the IEDB or another public resource. In addition to the IEDB, the NIAID provides numerous research resources to assist basic and applied research on flavivirus immunity and vaccine development (Table 1). Use of these reagents and enhanced interest in the Flaviviridae family by immunologists will advance our understanding of the mechanisms required for immune protection, and those involved in immune-mediated pathogenesis.

Table 1.

Immunology-Related Resources for Flavivirus Research

Resource/URL
Immune Epitope Database and Analysis Resource
Comprehensive database of antibody and T-cell epitope information curated from the literature. Analysis resource includes a suite of epitope data analysis and visualization tools.
http://www.immuneepitope.org/home.do
Bioinformatics Resource Center (BRC)
Relational databases containing a variety of data types, such as genome sequencing, comparative genomics, genome polymorphisms, gene expression, proteomics, host/pathogen interactions, and pathways.
http://www3.niaid.nih.gov/research/resources/brc
Biodefense and Emerging Infectious Diseases Research Resource Repository (BEI Resources)
Access to high-quality reagents, such as monoclonal antibodies, peptides, pathogen stocks, and cell lines, for infectious disease research. Reagent production costs covered by the BEI Resources.
http://www.beiresources.org/Catalog/tabid/248/Default.aspx
World Reference Center for Emerging Viruses and Arboviruses (WRCEVA)
Access to numerous arboviruses, including over 1700 flaviviruses and virus-related reagents, such as antibodies, ascites fluids, sera, antigens, viral RNA and cDNA, and oligonucleotides. Viruses and reagent production costs are covered by the WRCEVA.
http://www3.niaid.nih.gov/LabsAndResources/resources/dmid/wrceva/
NIH Tetramer Core Facility
Sponsored by NIAID and NCI for production and distribution of MHC class I, class II and non-classical MHC tetramers for analysis of T-cell responses. MHC alleles are available for humans, non-human primates, and mice. Tetramer production costs are covered by the Tetramer Facility.
http://tetramer.yerkes.emory.edu/
Vaccine Trial and Evaluation Units
Provide assistance in testing vaccines and treatments for infectious diseases (at phase I and phase II clinical trial stages).
http://www3.niaid.nih.gov/LabsAndResources/resources/vteu/
Taconic Emerging Mouse Models Program
Immunologically-related, gene-targeted mouse strains (in partnership with Taconic Farms, Inc.). The costs of this program are partially underwritten by NIAID funds to provide ready access to emerging mouse models.
http://www.taconic.com/wmspage.cfm?parm1 = 1661#NIAID
NIH Nonhuman Primate Reagent Resource
Sponsored by NIAID and NCRR to facilitate access to existing and new immunological and other reagents used in nonhuman primate models. Information is also available on commercial antibody reactivity and other reagents with different primate species and documented SOPs for working with reagents, and for selected nonhuman-primate-specific assays and techniques.
http://nhpreagents.bidmc.harvard.edu/NHP/default.aspx

Footnotes

Author Disclosure Statement

No competing financial interests exist.

Acknowledgments

We thank the following investigators for their participation in the Flavivirus Immune Epitope Workshop on September 16, 2009: Alan Barrett, Jonathan Bramson, Michael Diamond, Daved Fremont, Sharone Greene, John Roehrig, Alessandro Sette, Tom Solomon, and Kerrie Vaughan.

References

Altman

, Moss

, Goulder

et al. Phenotypic analysis of antigen-specific T lymphocytes. Science, 1998; 274:94–96.

Barrett

, Teuwen

. Yellow fever vaccine—how does it work and why do rare cases of serious adverse events take place? Curr Opin Immunol, 2009; 21:308–313.

Belsher

, Gay

, Brinton

et al. Fatal multiorgan failure due to yellow fever vaccine-associated viscerotropic disease. Vaccine, 2007; 25:8480–8485.

Blythe

, Zhang

, Vaughan

et al. An analysis of the epitope knowledge related to Mycobacteria. Immunome Res, 2007; 3:10.5.

Brien

, Uhrlaub

, Hirsch

, Wiley

, Nikolich-Zugich

. Key role of T cell defects in age-related vulnerability to West Nile virus. J Exp Med, 2009; 206:2735–2745.

Brien

, Uhrlaub

, Nikolich-Zugich

. West Nile virus-specific CD4 T cells exhibit direct antiviral cytokine secretion and cytotoxicity and are sufficient for antiviral protection. J Immunol, 2008; 181:8568–8575.

Bui

, Peters

, Assarsson

, Mbawuike

, Sette

. Ab and T cell epitopes of influenza A virus, knowledge and opportunities. Proc Natl Acad Sci USA, 2007; 104:246–251.

Capeding

, Brion

, Caponpon

, Gibbons

, Jarman

, Yoon

, Libraty

. The incidence, characteristics, and presentation of dengue virus infections during infancy. Am J Trop Med Hyg, 2010; 82:330–336.

Diamond

, Mehlhop

, Oliphant

, Samuel

. The host immunologic response to West Nile encephalitis virus. Front Biosci, 2009; 14:3024–3034.

10.

Doblas

, Domingo

, Bae

et al. Yellow fever vaccine-associated viscerotropic disease and death in Spain. J Clin Virol, 2006; 36:156–158.

11.

Endy

, Yoon

, Mammen

. Prospective cohort studies of dengue viral transmission and severity of disease. Curr Top Microbiol Immunol, 2010; 338:1–13.

12.

Engel

, Vasconcelos

, McArthur

, Barrett

. Characterization of a viscerotropic yellow fever vaccine variant from a patient in Brazil. Vaccine, 2006; 24:2803–2809.

13.

Gould

, Solomon

. Pathogenic flaviviruses. Lancet, 2008; 371:500–509.

14.

Gritsun

, Lashkevich

, Gould

. Tick-borne encephalitis. Antiviral Res, 2003; 57:129–146.

15.

Gubler

, Kuno

, Markoff

. Flaviviruses. Fields Virology, 5th 1 Knipe

, Howley

. Lippincott Williams & Wilkins: Philadelphia, 2007; 1153–1253.

16.

Hayes

, Komar

, Nasci

, Montgomery

, O'Leary

, Campbell

. Epidemiology and transmission dynamics of West Nile virus disease. Emerg Infect Dis, 2005; 11:1167–1173.

17.

King

, Getts

, Rana

, Shrestha

, Kesson

. Immunopathology of flavivirus infections. Immunol Cell Biol, 2007; 85:33–42.

18.

Lanteri

, O'Brien

, Purtha

et al. Tregs control the development of symptomatic West Nile virus infection in humans and mice. J Clin Invest, 2009; 119:3266–3277.

19.

Lindsey

, Schroeder

, Miller

et al. Adverse event reports following yellow fever vaccination. Vaccine, 2008; 26:6077–6082.

20.

Mansfield

, Johnson

, Phipps

, Stephenson

, Fooks

, Solomon

. Tick-borne encephalitis virus—a review of an emerging zoonosis. J Gen Virol, 2009; 90:1781–1794.

21.

Matthew

, Rothman

. Understanding the contribution of cellular immunity to dengue disease pathogenesis. Immunol Rev, 2008; 225:300–313.

22.

McMahon

, Eidex

, Marfin

et al. Neurologic disease associated with 17D-204 yellow fever vaccination: a report of 15 cases. Vaccine, 2007; 25:1727–1734.

23.

Peters

, Sidney

, Bourne

et al. The immune epitope database and analysis resource: from vision to blueprint. PLoS Biol, 2005; 3:e91.

24.

Pierson

, Diamond

. Molecular mechanisms of antibody-mediated neutralisation of flavivirus infection. Expert Rev Mol Med, 2008; 10:e12.

25.

Rothman

. T lymphocyte responses to heterologous secondary dengue virus infections. Ann NY Acad Sci, 2009; 1171,Suppl 1:E36–E41.

26.

Sitati

, Diamond

. CD4+ T-cell responses are required for clearance of West Nile virus from the central nervous system. J Virol, 2006; 80:12060–12069.

27.

Stephens

. HLA and other gene associations with dengue disease severity. Curr Top Microbiol Immunol, 2010; 338:99–114.

28.

Vasilakis

, Weaver

. The history and evolution of human dengue emergence. Adv Virus Res, 2008; 72:1–76.

29.

Vaughan

, Blythe

, Greenbaum

et al. Meta-analysis of immune epitope data for all plasmodia: overview and applications for malarial immunobiology and vaccine-related issues. Parasite Immunol, 2009; 32:78–97.

30.

Vaughn

, Greenbaum

, Blythe

, Peters

, Sette

. Meta-analysis of All Immunie Epitope Data in the Flavivirus Genus: Inventory of Current Immune Epitope Data Status in the Context of Virus Immunity and Immunopathology. Viral Immunol, 2010; 23:259–284.

31.

World Health Organization. Fact sheet no. 117. Revised May 2008. http://www.who.int/mediacentre/factsheets/fs117/en/. 2008 11 21.

32.

Zarebski

et al. Analysis of epitope information related to Bacillus anthracis and Clostridium botulinum. Expert Rev Vaccines, 2008; 7:55–74.