|Home | About | Journals | Submit | Contact Us | Français|
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.
Breeding pairs of NOD.Cg-Prkdcscid IL2rgtm1Wjl/SzJ (abbreviated as NOD-scid IL2rγ null) mice were purchased from the Jackson Laboratory (Bar Harbor, ME) and housed in a specific-pathogen-free facility under sterile conditions (microisolator in biological safety cabinet; sterile food, water, and bedding). All animal procedures were reviewed and approved by our Institutional Animal Care and Use Committee. Newborn and adult manipulations (transplantation, virus inoculation, clinical sign measurements, etc.) were performed under sterile conditions in a biological safety cabinet while mice were under inhalation anesthesia (1 liter/min O2 plus 2% isoflurane for light manipulations; 2 liters/min O2 plus 3% isoflurane for deep anesthesia). To reduce variation in experimental measurements in mice (due to stress), all procedures were done by the same person at the same time of day.
Human CB from anonymous donors was obtained from the South Texas Blood and Tissue Center (San Antonio, TX). CB mononuclear cells were separated by Ficoll-Hypaque density gradient, and CD34+ hematopoietic stem cells were isolated using a CD34+ progenitor cell selection system (Dynal Biotec) according to the manufacturer's instructions. The purity of positively selected CD34+ cells ranged from 85 to 90% and was confirmed by flow cytometry analysis (see below). Newborn mice were sublethally irradiated with 100 cGy from a cesium source located at the UT Health Sciences Center (San Antonio, TX). To avoid contamination, mice were transported from the mouse room to the irradiation facilities in a Rad Disk rodent microisolation irradiator cage (Braintree Scientific), designed to fit into the Gammacell 40 irradiator loading chambers. Twenty-four hours later, mice were given transplants by intrahepatic inoculation with 3 × 105 purified CB CD34+ cells.
The purity of CB-derived CD34+ cells was analyzed using flow cytometry by staining the cells with phycoerythrin-conjugated anti-human CD34 (clone 563) antibody (BD Biosciences), which is different from that used on beads for positive selection. Engraftment levels were evaluated in peripheral blood 6 weeks after transplantation. Blood (50 μl) was collected by the retro-orbital route in phosphate-buffered saline (PBS) containing heparin and stained with direct labeled anti-human antibodies: CD45-allophycocyanin, CD3-Pac Blue, CD8-PerCP.Cy5.5 (BD Biosciences), CD16-Alexa Fluor 700 (Invitrogen), CD20-fluorescein isothiocyanate, and CD14-phycoerythrin (Beckman Coulter). Red blood cells were lysed with 1× lysing buffer (BD Biosciences) and washed twice in PBS, and the remaining cells were fixed with 1.6% methanol-free formaldehyde. Control isotype antibodies were used for background staining. Samples were acquired using a CyAn ADP analyzer, and data were analyzed using Summit software (Beckman Coulter).
Eight viral strains representing the four genotypes (SE Asian, American, West African, and Indian) (16) of dengue virus serotype 2 were used in this study (Table (Table1).1). Viral stocks were prepared as follows: monolayers of the C6/36 (A. albopictus) cell line were grown to 90% confluence in 75-cm2 flasks, infected with dengue virus at a multiplicity of infection of about 10 genome equivalents (0.01 PFU) per cell in 3 ml of maintenance medium (minimal essential medium with 2% fetal bovine serum, 1× nonessential amino acids, 100 U/ml of penicillin, and 100 μg/ml of streptomycin), and incubated 1 h at 28°C with 5% CO2. Fresh maintenance medium was added to reach 15 ml, and flasks were incubated as described above for 7 to 9 days, until infection occurred in more than 90% of the cells, as evidenced by indirect fluorescent antibody (IFA) assay. For daily IFA monitoring, 10 to 20 μl suspensions of C6/36 cells were placed on multiwell slides, air dried, and fixed in ice-cold acetone. Slides were then incubated for 30 min at 37°C with a polyclonal anti-dengue virus serotype 2 (DEN-2) mouse hyperimmune ascitic fluid (CDC, Fort Collins, Co) diluted 1:200 in phosphate-buffered saline (PBS). Slides were rinsed twice in PBS and wells loaded with fluorescein isothiocyanate-labeled goat anti-mouse immunoglobulin (Sigma) diluted 1:100 in PBS. Slides were incubated for 30 min at 37°C and washed twice in PBS. Slides were examined at 200 to 400x magnification with an inverted fluorescence microscope. Cell supernatants were harvested and individual aliquots were prepared by adding 30% of gelatin (Sigma) and stored at −70°C until use. One aliquot of each virus stock was thawed and 50 μl was used for RNA extraction using Trizol reagent (Invitrogen) according to the manufacturer's instructions; eluted RNA in DEPC-water was used to estimate viral RNA concentration by real time quantitative reverse transcription (RT)-PCR (see below). Concentrations (in genome equivalents per ml) are shown in Table Table11.
Male and female NOD-scid IL2rγ null adult mice (6 to 8 weeks) with engraftment levels of 16 to 80% were used to make groups of five or six individuals for infection experiments with DEN-2. Each experimental group, to be infected by one virus strain, consisted of mice which had received transplants of CD34+ cells from different CB donors, to avoid intradonor variability in virus replication. Viral stocks were diluted with sterile PBS, to a concentration of 9 log10 genome equivalents (about 6 log PFU) in a final volume of 100 μl, and inoculated subcutaneously on the backs of mice under anesthesia. For control groups, aliquots of C6/36 cell culture medium were prepared with 30% gelatin as described above and diluted in PBS in the same way as virus for mouse inoculation. Mouse temperature and erythema were monitored daily; temperature was measured using a RET-3 rectal probe coupled to a BAT-12 Microprobe thermometer (Physitemp Instruments), and erythema was assessed with a DSMII ColorMeter (Cortex Technology, Hadsund, Denmark) by measuring every day on the same shaved area of skin (armpit); the erythema index (9) was obtained by measuring the intensity of reflected light in the spectral range of green and red light (520 to 580 nm, and 660 to 690 nm, and expressed in O.D.). To determine viremia, blood was collected (25 μl) on even days starting at day 2 postinfection (p.i.) by the retro-orbital route using calibrated capillary micropipettes (Drummond Scientific), alternating right and left eyes; serum was separated by centrifugation at 10,000 × g for 5 min, and 10 μl of serum was used for RNA extraction as described above. For thrombocytopenia determination, 20 μl of blood was collected as described above and diluted 1:100 in 1% buffered ammonium oxalate for cell lysis using a BD Unopette system test reservoir (BD Biosciences); the diluted sample was then used to charge a Neubauer hemacytometer, and platelets were counted under ×400 magnification using bright-light microscopy.
Viral RNA copies in stocks or sera from infected humanized mice were estimated using a previously reported protocol (24) with some modifications. RNA template (5 μl of the viral stock or 10 μl of the serum samples) was amplified in duplicate or triplicate (viral stocks and sera, respectively) using an RNA Ultrasense one-step quantitative RT-PCR system (Invitrogen) in a final reaction volume of 25 μl containing 1.25 μl of enzyme mix, 5 μl of 5× Ultrasense reaction mix, 0.5 μl of 5′-carboxy-X-rhodamine reference dye, 5 μM of d2C16A primer (5′-GCTGAAACGCGAGAGAAACC-3′), 5 μM of d2C46B primer (5′-CAGTTTTAITGGTCCTCGTCCCT-3′), and 5 μM of VICd2C38B probe (FAM-5′-AGCATTCCAAGTGAGAATCTCTTTGTCAGCTGT-3′-6-carboxytetramethylrhodamine). Amplification was performed on a 7500 real-time PCR system (Applied Biosystems) as follows: one cycle at 48°C for 30 min, one cycle at 95°C for 15 min, and 40 cycles at 95°C for 15 s and 60°C for 1 min. To estimate RNA copy numbers, a standard curve was generated using in vitro-transcribed RNA standards. RNA standards were produced as follows. A 94-bp fragment of the DEN-2 K0049 SE Asian strain was amplified by RT-PCR and cloned into the pCR2.1 plasmid using a TOPO cloning kit (Invitrogen). The recombinant plasmid was linearized with HindIII, and RNA transcripts were generated with a T7 Megascript kit (Ambion) following the manufacturer's instructions. Concentration of transcribed RNA was determined with a Ribogreen RNA quantitation kit (Molecular Probes), and 10-fold serial dilutions were prepared and used to construct a standard curve. The sensitivity of the assay is 240 RNA copies per ml or 1.2 copies per reaction tube.
An indirect enzyme-linked immunosorbent assay (ELISA) was used to determine the production of specific human antibodies (total immunoglobulins) against DEN-2 in the hu-NOD-scid IL2rγ null mice. ELISA plates were coated with DEN-2 antigen (Microbix Biosystems, Inc.) at a concentration of 0.5 μg/ml in carbonate-bicarbonate buffer, pH 9.6, and incubated overnight at 4°C. Plate wells were blocked using 300 μl of 1% bovine serum albumin blocking solution (KPL) for 2 h at 37°C or overnight at 4°C. Serum samples from infected mice obtained as described above were serially diluted starting at 1:20. Diluted serum (100 μl) was added to the plates, which were then incubated 1 h at room temperature and washed three times with 1× washing solution (KPL). Total human immunoglobulin against dengue virus was detected by adding 100 μl per well of a peroxidase-labeled antibody (mouse serum adsorbed) against human immunoglobulin A (IgA)+IgG+IgM (heavy plus light chain) diluted 1:500 (KPL). Plates were incubated for 1 h at room temperature and then washed as described above. Binding antibody was detected by adding 100 μl per well of peroxidase substrate (SureBlue TMB; KPL) to the plates and incubated for 10 min at room temperature. The reaction was stopped by adding 100 μl per well of 1% HCL solution (TMB stop solution; KPL), and results were read at a wavelength of 450 nm. The cutoff for a positive result was three times the average of the optical density of negative controls.
Viremia, erythema index, temperature, and platelet values were pooled for the mice inoculated with each of the two viruses representing each genotype. For multiple comparisons of the four different genotypes, a two-way analysis of variance with Bonferroni's posttest was used, to account for variation within and between genotypes. For comparison of two genotypes for each time point, a one-way analysis of variance with Bonferroni's posttest was used. For comparison of individual experimental groups, a Mann-Whitney test was used for each pair. For platelet level comparisons, a t test was used. All statistical analyses were performed using GraphPad Prism, version 5, for Mac OSX.
To study the development of human-like dengue fever in this mouse strain, we used a panel of eight strains, representing the four genotypes within serotype 2 (Table (Table1).1). Experimental groups consisted of five or six mice given transplants of CD34+ cells from different donors, and reconstitution levels (CD45+ cells in peripheral blood at 6 weeks posttransplantation) for the SE Asian genotype groups (n = 24) ranged from 20 to 80% (median, 55), for the American genotype groups (n = 18) from 17 to 67% (median, 45), for the Indian genotype groups (n = 11) from 19 to 76% (median, 42), and for the African genotype groups (n = 10) from 16 to 68% (median, 64). A detailed list of engraftment levels for each experimental group are shown in Table Table1.1. For instance, the experimental groups infected with SE Asian viruses, K0049 and 429557, consisted of four groups, one with six individuals, three with five individuals each, and one with three individuals, for a total of 24 individuals (n = 24). After virus inoculation (or culture medium for controls), mice were monitored daily to measure temperature and erythema, and bleeds occurred every 2 days to determine viremia in the sera. All mice receiving a virus inoculation recuperated from the disease and are alive, to be used for secondary infection studies.
Viremia in all infected hu-NOD-scid IL2rγ null mice was detectable beginning on day 2 p.i., and the peaks were different for each genotype. Mice infected with the SE Asian genotype showed a two-peak viremia, with the first on days 8 to 10 and the second and highest at day 16 p.i. (Fig. (Fig.1A).1A). These peaks were statistically significant compared with viremias in mice infected with viruses of the American genotype (Fig. (Fig.1B).1B). These results are in agreement with our previous observations that SE Asian viruses replicate in human cells and disseminate in mosquitoes at higher rates than viruses of the American genotype (2, 6). Mice infected with virus strains of the Indian and West African genotypes showed the shortest viremia periods, with viremia falling to undetectable levels by day 16 and 14 p.i., respectively (Fig. 1C and D), and the Indian and American viruses showed the lowest levels of viremia (around 103 RNA copies/ml at peaks). Overall, viruses from the SE Asian genotype of DEN-2 replicated at much higher levels and longer (up to 105 RNA copies/ml, and beyond day 18 p.i.). The viremia curves for individual viruses of a genotype and the mean derived from them, which are the data used for comparisons in Fig. Fig.11 (see Fig. S1 in the supplemental material), demonstrated that virus curves are not significantly different (Mann-Whitney pairwise comparisons) within a genotype. As described above, to avoid donor variability in replication, we used mice which had received transplants of cells from different donors in the same experimental group; however, there were no significant differences in virus levels for comparison of infection of mice receiving different CB cells or according to donor (see Fig. S2 in the supplemental material).
Infected and control mice were measured daily for erythema as described above; however, since our observations showed no major changes from one day to another, and to simplify graphics and to make these analogous to viremia data, we showed data for even days only. To avoid any shave-related skin redness, each mouse was carefully shaved 24 h before the first measurement, and when necessary, additional shavings were performed right after the last erythema determination to allow skin to recover for the next day's measurement. Erythema was consistently higher in infected mice than in control groups during days 2 to 10 p.i. (Fig. (Fig.2A).2A). And similar to the viremia curves, the comparison of erythemas in mice infected with SE Asian and American viruses showed statistically significant differences at days 12 to 16 (Fig. (Fig.2B).2B). The lowest erythema index was observed in the groups infected with the West African viruses; compared to the other three genotypes, differences are statistically significant at days 2 to 10 p.i., and they were barely above negative control levels (Fig. (Fig.2C).2C). Comparison of the mice infected with the Indian genotype to those infected with all other genotypes showed no significant differences (Fig. (Fig.2D),2D), and these mice were thus intermediate in their responses. As reported previously using the humanized NOD/SCID strain (3), increases in the erythema index represent one of the most consistent clinical signs of infection in hu-NOD-scid IL2rγ null mice.
Similar to the erythema determination, temperature was measured in the control and infected mice on a daily basis, but data from only even days are plotted, to simplify the graphs. Temperature was the most variable clinical sign, both individually and by group, showing no statistical differences between genotypes (Fig. (Fig.3A).3A). However there was a consistent increase in temperature in the infected mice, with some time points (days 2, 10, 12, and 14 p.i.) being statistically different for all genotypes compared to control mice (Fig. (Fig.3B).3B). Mice infected with the SE Asian and Indian genotypes showed the highest temperature increases, and in the case of the SE Asian and American genotypes, the temperature curves showed two peaks similar to those observed with viremia (Fig. (Fig.1B1B).
To first determine when thrombocytopenia reached its peak in the hu-NOD-scid IL2rγ null mouse strain, platelet levels were measured every 2 days from day 0 to 16 p.i., in a group of mice infected with SE Asian strain K0049. This strain was chosen because it replicates to the highest titers in human monocytes and dendritic cells and it was used in our previous studies in NOD/SCID mice (3). We observed a significant reduction in platelets, reaching the lowest point at day 10 p.i. (Fig. (Fig.4A).4A). Based on these results, and due to the limited amounts of blood that we are permitted to take from live mice, in all subsequent experiments platelets were measured only at day 10 p.i. We observed that mice infected with the SE Asian genotype viruses showed the lowest platelet counts, followed by the groups infected with the West African and American genotypes; these differences were statistically significant compared to the control group (Fig. (Fig.4B).4B). However, mice infected with the Indian genotype viruses showed no significant difference from the control group, and results were statistically different from those for mice infected with the SE Asian genotype viruses (Fig. (Fig.4B4B).
The production of specific antibodies against dengue virus was measured in sera from infected mice using an indirect screening ELISA. Total immunoglobulins were measured at 4 weeks p.i., and positive mice were bled again 4 weeks later for a second screening. Anti-dengue virus antibodies were detected only in one group (n = 8) of mice infected with the K0049 strain of the SE Asian genotype, the most virulent virus. Results of this ELISA at week 4 p.i. showed two positive mice; when the same group was tested again at week 8 p.i., three mice were positive (data not shown). The failure to produce IgM, IgG, or IgA by most of the infected hu-NOD-scid IL2rγ null mice could be due to many factors, including variations due to stem cell donor (i.e., implicit differences in virus replication levels) (6), but most likely this was due to the lack of human immune factors (interleukins) required to support B-lymphocyte survival. However, these results demonstrated that some adequate conditions (possibly provided by the stem cell inocula themselves) for immunostimulation can develop in these mice to produce antibodies in response to a dengue virus challenge.
Although the number of test subjects used in these studies was limited by the procedures necessary for individual engraftment, results for eight viruses were consistent with genotype groupings. Therefore, we are confident in our conclusions regarding in vivo differences between dengue virus genotypes, although we used only two viruses to represent each of these groups. Any variation in virus replication due to CB donor, transplantation level, or virus passage differences was accounted for by analysis of variance, and our data graphs include the standard error of the mean to demonstrate these ranges. However, we have limited our interpretations only to the statistically significant trends in differences between genotypes, to eliminate the effect of any measurable differences in reagents.
In terms of the validity of using this mouse model to mirror dengue virus pathogenesis in humans, these mice are still the only system described to date that mimics dengue fever (28). Other systems, using immunocompromised and immunodeficient mice, have produced some clinical signs of dengue (viremia, fever, vascular leakage, or hemorrhaging), but only by using specific strains of virus that have high passage levels (and therefore genetic differences introduced by selection), routes of inoculation that are not natural (i.e., intravenous or intraperitoneal), or virus doses that are extremely high (8 to 9 log10 PFU). Kuruvilla et al. (14) recently described the RAG2−/− γc−/− strain of mice that was engrafted with human fetal stem cells, and these mice responded with fever and viremia after inoculation with pooled dengue viruses of different serotypes; there was no rash or thrombocytopenia. However, one-half of the inocula (of one or three to four pooled DEN-2 viruses) were given via peritoneum (the other half were subcutaneous), and these were the mice that produced detectable (two- to fourfold over background) antibodies to dengue (3 of 16 mice had IgM at 4 weeks and 6 of 16 had IgG at 8 weeks p.i.). Our system offers the advantage of avoiding the use of embryonic stem cells, and we have confirmed that viruses of low passage number (two to four cell culture passes from one patient sample) do produce consistent clinical signs. Although the dose of virus used in the humanized NOD-scid IL2rγ null mice described here was higher than what was used in our previous model (6 log10 versus 4.7 log10), these are still doses that could theoretically be delivered by a single mosquito bite. We are currently performing studies where single infected mosquitoes will be tested for the limits of dengue virus transmission to these humanized mice.
The results for two West African strains shown here are the first clear in vivo phenotypic differences between this genotype and others belonging to serotype 2. Other authors have suggested that these sylvatic viruses might emerge to become the agents of new outbreaks, and because they are so distinct genetically, they might overcome protective immunity induced by vaccination (22). The evidence shown here suggests that this is not the case, because the viremias are much shorter and the skin pathology implicitly measured here as rash is much lower. That is, these viruses seem to be at an evolutionary disadvantage, even more so than the American genotype viruses that have been displaced by the SE Asian viruses in many countries. We have shown that differences in replicative ability in human skin target cells and whole mosquitoes result in much lower viremias and vectorial capacity, respectively, and thus in drastically reduced virus transmission (1, 6). Other investigators in West Africa have shown that A. aegypti mosquitoes collected there are poor vectors for the sylvatic viruses; these viruses have very low rates of mosquito dissemination and therefore reduced potential to be transmitted (8). Whether vaccine preparations would protect against this genotype remains inconclusive, but there should be less concern because these viruses have not caused outbreaks in West Africa, and countries in that area have actually imported their epidemic viruses, while their sylvatic dengue foci are being eliminated by human environmental disruption (17, 27).
Results of ELISA experiments showed that some mice have detectable antibodies against dengue virus, but only when infected with the most virulent SE Asian strain (K0049). Infection with other genotypes failed to induce detectable levels of immunoglobulins in a screening assay. Since each experimental group contains mice that had received transplants of progenitor cells from different donors, implicit host genetic differences could influence the production of antibodies after a challenge; however, further analyses are needed to address this question. For example, we could use CD34+ cells expanded in vitro (10) in order to have enough cells to carry out transplantation with a larger group of mice and directly compare different genotype strains. In addition, even the reported success in producing antibodies in other strains of humanized mice (13, 14, 25) has been qualified by no or low titers of detectable antibodies. This could be due to the fact that B-cell homeostasis is abnormal in transplant recipient mice, and human B cells could be lost within 2 weeks, suggesting that the murine environment does not provide the necessary cytokine environment for their survival (10). Even when we obtain higher engraftment levels, the numbers of human immune cells are very low in the total blood of these mice (see Fig. S3 in the supplemental material). Recent studies have demonstrated that treatment of these mice with human recombinant tumor necrosis factor alpha and human B-lymphocyte stimulator (BLyS/BAFF) can enhance engraftment of B and T cells and promote B-cell survival with a concomitant increase in antibodies against T-dependent and T-independent antigens (10, 19). Because of the key role the antibody response plays in the immunopathology of dengue, we are now establishing protocols to use these cytokines in our humanized NOD-scid IL2rγ null mouse model, to improve the anti-dengue virus antibody titers. This would allow us to measure the role of the humoral response in sequential infections and to develop a mouse that could be used to test vaccine efficacy in a human immune system milieu.
Prospective clinical studies in humans have shown that there is a correlation between the dengue disease course (and sometimes severity) and viremia (23); the results of our study are in agreement with that observation. Infected mice showed a correlation between viremia and fever peaks; in general, infected mice showed a rapid increase of temperature by day 2 p.i., with a slight decrease by day 8 to 10 p.i. and a second increase by day 12 p.i. The mouse group infected with the SE Asian genotype viruses showed the maximum fever peak (about 2°C higher than controls) and, along with the mice infected with American genotype viruses, showed a two-peak viremia and fever curves which were statistically different from controls. In contrast, mice infected with viruses of the Indian and West African genotypes showed only one peak of fever and viremia. In addition, the latter two genotypes produced lower levels of viremia and/or thrombocytopenia in these mice. These observations could provide evidence for lower virus replication or virulence in humans, which could also reflect some epidemiologic differences. There is evidence from serotype 2 and 3 viruses that specific genotypes have increased transmission, can displace other genotypes, and can be associated with epidemics of more severe disease (SE Asian genotype for DEN-2, genotype IIIb for DEN-3) (15, 18). The Indian, American, and West African genotypes of DEN-2 are currently being displaced by the SE Asian genotype in many regions affected by dengue, as evidenced by numerous, continually updated phylogenies (7, 21, 22, 29). The mice analyzed here therefore show clinical signs that might reflect this epidemiology, where more virus replication (higher and longer, when infected with SE Asian viruses) implies a greater opportunity to be transmitted. Gubler et al. reached similar conclusions in the 1970s, when studying dengue outbreaks in the Pacific Islands (11, 12). We believe we now have the laboratory systems in which to test these hypotheses.
We thank Dennis A. Bente for establishing the breeding colony of mice.
Funding was provided by the Robert J. Kleberg Jr. and Helen C. Kleberg Foundation, Instituto Nacional de Salud Publica de Mexico, and NIH grant AI50123.
Published ahead of print on 17 June 2009.
†Supplemental material for this article may be found at http://jvi.asm.org/.