The Scripps Research Institute:

University of California at San Diego:

University of Texas Medical Branch:

  • Thomas W. Geisbert, Ph.D., Associate Director and Professor 

Harvard University and the Broad Institute

Autoimmune Technologies, LLC

  • Luis M. Branco, Project Director

Corgenix Medical Corporation, Inc.

  • F. Jon Geske, Ph.D., Project Director

Kenema Government Hospital

  • Sheik Humarr Khan, M.D., Director, Lassa Ward
  • Augustine Goba, Director, Lassa Laboratory
  • James Bangura, Team Leader, LF Outreach Team

Tulane-International Center for Research on Tropical Infections, N’Zérékoré, Guinea

  • Mamadou Coulibaly, Ph.D., Director

The Irrua Specialist Teaching Hospital, Irrua, Nigeria

  • Christian T. Happi, M.D., Director, Infectious Disease Laboratory


Rationale for the experimental system chosen for epitope identification.
The Viral Hemorrhagic Fevers and Lassa Fever.  Because of their high case fatality rates, ability to
spread easily by human-human contact, and potential for aerosol release, most hemorrhagic fever
viruses are classified as BSL-4 and are included on the NIAID Category A Select Agents list of potential
bioterrorist threats. Members of the virus families Arenaviridae (Lassa, Junin, and Machupo), Filoviridae
(Ebola and Marburg), and Bunyaviridae (Rift Valley fever and Crimean Congo hemorrhagic fever) are the
agents of greatest concern (3). The low number of viral particles required to initiate infection (estimated at
10-100) and their relative ease of culture in vitro make hemorrhagic fever viruses “ideal” candidates for
weaponization. This threat has increased considerably in the last few years, as these viruses can be
readily isolated from a variety of biological sources, including infected humans, other mammals, and
insects, depending upon the particular virus, legally and illegally. No country or group is presently
recognized as actively developing hemorrhagic fever viruses as weapons, but concerns persist over
clandestine activity, such as the reported attempt by the Japanese cult group Aum Shinrikyo to acquire
EBOV, and the whereabouts of virus stocks produced during the Soviet era (4).

Lassa fever (LF) is a severe and often fatal

hemorrhagic illness caused by Lassa virus (LASV), a
member of the Arenaviridae (5). The disease is
found across West Africa (Fig. 1). Humans contract
LASV primarily through contact with contaminated
excreta of rodents of the genus Mastomys, which is
the natural reservoir (6, 7). LASV is maintained in
the rodents through vertical (mother to offspring)
transmission (5). There is actually little known
regarding the transmission of virus from the rodent
reservoir to the human host, although there is
compelling evidence that arenaviruses (LASV (8)
and Junin virus (9)) are stable and infectious by the
aerosol route in nonhuman primates. The aerosol
route has also been occasionally suspected in
human transmission, but infection via contact to
abrasions in the skin, mucous membranes and
ingestion are also viable routes of spread.
Secondary transmission of LASV between humans
occurs through direct contact with infected blood or
bodily secretions (5). Nosocomial transmission and
outbreaks have been described in healthcare
facilities in endemic areas (10-13). The antiviral drug
ribavirin is effective for LF, but is most effective only
when administered within the first 6 days after
disease onset (14-18). 


 Why choose LF for human immunological study? Unlike many VHFs, LF is not a rare disease that emerges only as sporadic cases or in outbreak form. Although surveillance is inadequate to determine the true incidence, up to 300,000 infections and 5,000 deaths from LF are estimated to occur yearly across West Africa, where the disease is endemic (5, 7, 19). There is also potential for outbreaks of novel arenaviruses as documented recently in South Africa (20). After dengue hemorrhagic fever, LF has the highest of any incidence of any VHF in the world. The highest incidence of LF is in the “Mano River Union (MRU)” countries of Sierra Leone, Liberia, and Guinea (13, 21-26). However, the numbers of LF cases appear to be increasing dramatically in parts of Nigeria (27) (Happi, personale observations).


Approximately 6% of febrile patients tested at Irrua Specialist Teaching Hospital (ISTH) in the northern part of Edo, Nigeria had PCR-confirmed LF, which extrapolates to hundreds of patients with Lassa fever per year, considering the number of patients with febrile illnesses seen at ISTH. LF also occurs frequently in UN peacekeepers, aid workers, missionaries, and travelers in West Africa and is the most frequently imported person-to-person communicable VHF (28). Recent importation of cases of LF into Germany, the Netherlands, the United Kingdom, and the United States by travelers on commercial airlines illustrates the potential for the spread of this highly dangerous and contagious pathogen (28-31). In 2004, for example, a man suffering from fever, chills, severe sore throat, and diarrhea flew from Liberia to Newark, NJ. He was subsequently found to have LF, exposing 188 persons to the virus before dying (32). Ultimately, the high incidence of LF, along with the unique infrastructure and resources of our recently established public health and research network in West Africa offer one of the only opportunities in the world to conduct detailed and integrated studies of B cell epitopes and other potential correlates of protective immunity in a VHF and Category A Select Agent. Such studies, as proposed here, are imperative to guide the rational design of therapies, clinical management strategies and vaccines. 
Evidence that both humoral and cellular immunity to LASV is required to prevent LF. LASV is enveloped with a genome that consists of two single-stranded ambisense RNA segments, large (L, 7.2 Kb) and small (S, 3.4Kb). L encodes the viral polymerase and zinc binding protein (Z, matrix protein). S encodes the nucleoprotein (NP) and glycoprotein precursor (GPC - processed to GP1 and GP2). Antibodies against LASV structural proteins were induced in a study in which inactivated LASV failed to protect from lethal challenge (33). These results indicate that humoral responses to LASV structural proteins alone are insufficient to protect against LF. In contrast to results with inactivated LASV, recombinant vaccinia viruses (rVV) expressing LASV glycoproteins used as live vaccines, which likely induce both humoral and cellular immune responses, were effective at preventing signs of LF in macaques (34). Expression of both GP1 and GP2 by the rVV was necessary to confer protection (34-37). rVV expressing only NP was protective in guinea pigs, but not macaques (35-37). Previously, members of our team (TWG and coworkers) demonstrated that a recombinant vesicular stomatitis virus (VSV) engineered to express LASV GP was able to confer protection against lethal LASV infection in macaques  (1). The VSV-LASV GP vaccine induced strong humoral and cellular immune responses in four vaccinated and challenged monkeys. Despite a transient LASV viremia in vaccinated animals 7 days after challenge, the vaccinated animals showed no evidence of clinical disease. In contrast, two control animals developed severe LF symptoms including rashes, facial edema, and elevated liver enzymes, and ultimately succumbed to the LASV infection. Both neutralizing antibodies and virus-specific T cells were detected following challenge. The basis for protection with this vaccine was likely due to a priming of both synergistic humoral and cell- mediated immune responses. This conclusion is consistent with prior investigations demonstrating that passive transfer of neutralizing antibodies early after infection can be an effective treatment for LF and other arenaviral hemorrhagic fevers (38-43). 


No LF vaccine is currently available for use in humans, and the currently available drug, ribavirin, is only effective if administered early in infection. While this proposal focuses on identification and characterization of B cell epitopes, it is essential for development of an effective vaccine or immune- based therapeutics (for example: neutralizing antibodies) to investigate the responses to LASV infection of both arms of the human immune system.  The importance of identifying and characterizing both humoral and cellular responses to viruses in humans was revealed with unfortunate clarity in the failure of the recent Merck HIV vaccine candidate, a recombinant nonreplicating adenovirus vector encoding Gag, Pol, and Nef proteins from HIV-1 clade B. The trial was halted in September 2007 because the vaccine was unable to protect volunteers from HIV-1 infection or lower viral load in infected individuals (44, 45). Until the results of the recent human clinical trial using this vaccine were known, HIV vaccine design strategies had focused on inducing strong and diverse T cell responses. The Merck vaccine effectively elicited broad cellular CD4+ and CD8+ T cell responses to multiple HIV-1 epitopes as measured with traditional ELISPOT assays, but failed to provide any detectable protection against HIV-1 infection. It is apparent that there must be concerted efforts to identify the most protective and functional responses, both humoral and cellular, against critical regions of viral pathogens, such as HIV-1 and LASV. Ineffective immune responses would not be expected to elicit LASV control, and may compromise important responses that effectively eliminate virally-infected cells or even enhance infection. 
Acquired humoral immunity in LF. The humoral arm of the adaptive immune system involves the production of antibodies by cells of the B cell lineage, which bind and in some cases neutralize or enhance infectivity of pathogens, including viruses. In previous studies, one of us (DGB) demonstrated that in patients confirmed to have LF, LASV antigens indicative of viremia typically appear in the blood within the first week of illness (46). LASV-specific IgM appears in the second week of illness, and in most cases correlates with a rapid drop in the amount of viral antigens and nucleic acid. Whether viral clearance results from the IgM response or as appears more likely from a combination of innate and adaptive humoral and cellular immune mechanisms is unknown. The presence of LASV-specific IgG in the absence of clinical symptoms and either IgM or viremia suggests previous exposure to LASV. Our recent development of improved LF immunodiagnostics based on recombinant antigens (see below) will facilitate the dissection and further characterization of the humoral immune responses to LASV, and identify responses to B cell epitopes that provide the strongest correlates of protection from LF. 
Humoral immune responses to each of the LASV structural proteins have been detected (47). However, there have been few efforts to perform fine structure epitope mapping of the antigenic sites recognized by LASV-specific antibodies. Murine monoclonal antibodies (MAbs) have been produced against several arenaviruses (48-50). Competitive binding analysis of such MAbs identified at least four antigenic sites on NP, two on GP1 and six on GP2 of LASV (Josiah strain). There was wide divergence in the reactivities of the MAbs to LASV isolates from across the geographic range of the virus. Considerable genetic diversity exists among LASV isolates. LASV isolates in the western part of the range in Sierra Leone, Liberia, and Guinea are fairly conserved and comprise a single genetic lineage, while LASV isolates in the eastern part of the range in Nigeria are more diverse, comprising three lineages (51, 52).  MAbs to Lassa virus GP1 and NP uniformly distinguished viruses from the West African countries of Sierra Leone, Liberia and Guinea from those of Nigeria. On the other hand, certain LASV GP2-directed MAbs reacted broadly with all LASV isolates as well as South American arenaviruses demonstrating that an epitope or epitopes on GP2 may be highly conserved in arenaviruses. In another study among the few that have attempted to define B cell epitopes on LASV proteins, synthetic peptides were used to define a minimal epitope recognized by certain broadly cross-reactive MAbs. The epitope (GPC aa 374-378, KFWYL – a GP2 epitope) is highly conserved amongst LASV and other arenaviruses (53). Polyclonal sera from humans and from animals experimentally infected with Junin, lymphocytic choriomeningitis, and Lassa viruses bound specifically to the peptide. 

One of the hallmarks of LASV infection is the apparent absence of detectable neutralizing antibodies during acute infection. Low-titer neutralizing antibodies, if formed, typically appear several weeks to months after the resolution of infection (42, 54). It is important to note however, that present therapy of Argentine hemorrhagic fever, caused by Junin virus, a New World arenavirus, involves transfusion of immune plasma in defined doses of neutralizing antibodies during the prodromal phase of illness (39). Treatment of LASV-infected individuals using pooled immunoglobulin containing high titers of anti-LASV antibodies was effective in some cases, but not others. The efficacy of passive immunization for treatment of LF appears dependent on the titer of the neutralizing antibody infused (41, 42, 54). Geographic origin is also a factor as geographically matched plasma is more likely to contain adequate neutralizing titers against homologous LASV strains. Despite their potential importance, neutralizing antibody epitopes of LASV recognized by humans remain essentially unexplored. A fundamental understanding of the mechanisms of antibody binding and antibody-mediated neutralization of LASV may have significant implications for the generation of antibody-based therapeutics or epitope-targeted vaccines.


Tulane University, New Orleans, LA 504-865-4000