Humanized Mice Show Clinical Signs of Dengue Fever according to Infecting Virus Genotype

We demonstrated that the infection of humanized NOD-scid IL2rγ ^null mice with different strains (representing the four genotypes) of dengue virus serotype 2 (DEN-2) can induce the development of human-like disease, including fever, viremia, erythema, and thrombocytopenia. Newborn mice were irradiated and received transplants by intrahepatic inoculation of human cord blood-derived hematopoietic progenitor cells (CD34⁺). After 6 weeks, mouse peripheral blood was tested by flow cytometry to determine levels of human lymphocytes (CD45⁺ cells); rates of reconstitution ranged from 16 to 80% (median, 52%). Infection (with approximately 10⁶ PFU, the equivalent of a mosquito bite) of these humanized mice with eight low-passage-number strains produced a high viremia extending to days 12 to 18 postinfection. We observed a significant decrease in platelets at day 10 in most of the mice and an increase in body temperature (fever) and erythema (rash) in comparison with humanized mice inoculated with cell culture medium only. Comparison of Southeast (SE) Asian and other genotype viruses (American, Indian, and West African) in this model showed significant differences in magnitude and duration of viremia and rash, with the SE Asian viruses always being highest. Indian genotype viruses produced lower viremias and less thrombocytopenia than the others, and West African (sylvatic) viruses produced the shortest periods of viremia and the lowest rash measurements. These results correlate with virulence and transmission differences described previously for primary human target cells and whole mosquitoes and may correlate with epidemiologic observations around the world. These characteristics make this mouse model ideal for the study of dengue pathogenesis and the evaluation of vaccine attenuation and antivirals.Dengue viruses, which cause the disease dengue fever (DF) and its more severe form, dengue hemorrhagic fever (DHF), in humans, have been spreading to more areas of the world along with their mosquito (Aedes aegypti and Aedes albopictus) vectors. Now over 100 countries are affected, including some areas of the United States (Texas and Hawaii) (5, 26). Due to the fact that only humans show clinical signs and symptoms of disease, it has been difficult to directly test the mechanisms of pathogenesis of these viruses (4). Through decades of research, including clinical, epidemiologic, and laboratory studies, the factors involved in producing disease, whether it be DF or DHF, have remained unproved. However, there are many indications that both the virus and the host contribute to the occurrence and severity of disease: there are genetic differences in the virus and host immune response that can be measured in vitro, and these factors seem to lead to immunopathology in addition to the damage done by virus replication. Because there are four antigenically distinct dengue viruses (serotypes 1 to 4), humans can theoretically have dengue virus infections leading to clinical disease up to four times, and the immunity to the first virus enhances the probability of developing severe dengue after a subsequent infection. Thus, the development of vaccines has been hampered by the unknown effects of inoculating with a tetravalent preparation that might cause immunopathology or severe disease, and there are no appropriate animal models in which to test vaccine attenuation and efficacy for human applications.In 2005 we reported the development of humanized NOD/SCID (nonobese diabetic/severe combined immunodeficient) mice that produced signs of DF upon infection with one strain of dengue virus (3). The mice were humanized by giving them transplants of purified hematopoietic stem cells from human umbilical cord blood (CB) samples taken from normal births. After subcutaneous infection with a low dose of a Southeast (SE) Asian virus, the viremia, rash, and thrombocytopenia were significantly higher, longer lasting, and more like human disease than in any other animal model described at the time. We concluded that this model could be used to test antiviral treatments, since these mice did not produce measurable human antibodies. Since then, many other immunodeficient mouse strains have been produced that can have enhanced human engraftment levels, and they develop functional human immune system cells, including some level of adaptive immunity (20). It has been reported that some of these mouse strains develop immunoglobulins specific for human immunodeficiency virus and dengue virus, albeit at low levels (14, 25).Here we present results of dengue virus pathogenesis studies in a new mouse strain, NOD-scid IL2rγ ^null, that has a much higher degree of human lymphocyte development (median of 52%, versus 14% previously). The comparison of viruses from different genetic subgroups of dengue serotype 2 has led us to conclude that this model is reflective of actual human dengue pathogenesis, and this development might bring us to a new era in testing the factors that contribute to dengue disease.