OUP user menu

A model of the real-time correlation of viral titers with immune reactions in antibody-dependent enhancement of dengue-2 infections

Rong-Fu Chen, Wen-Ting Yeh, Ming-Yu Yang, Kuender D Yang
DOI: http://dx.doi.org/10.1111/j.1574-695X.2001.tb01542.x 1-7 First published online: 1 February 2001


We simultaneously assessed dengue-2 virus (DEN-2) titers by real-time quantitative reverse transcription-polymerase chain reaction (RT-PCR) and immune reactions including interleukin-4 (IL-4), interferon-γ (IFN-γ) and prostaglandin E2 (PGE2) production by human mononuclear cells (MNLs) in a model of antibody-dependent enhancement (ADE). We found that DEN-1 immune sera at 1:100 and 1:250, but not those at 1:10 or control sera, enhanced DEN-2 infections in human MNLs as assessed by the fluorogenic RT-PCR technique. The enhanced profiles of DEN-2 infections determined by the RT-PCR in 6 h were reproducible by the standard plaque-forming unit (PFU) measurement established after 7 days. The ADE-enhanced DEN-2 titers determined by the RT-PCR were 5.5–33-fold higher than those detected by PFU assay, suggesting that total virions during infections were much higher than the viable ones detected by PFU assay. MNLs in response to DEN-2 infections had higher IFN-γ and PGE2 production. However, the enhancement of DEN-2 infections by DEN-1 immune sera in MNLs was not associated with further enhancement of IFN-γ production. In contrast, the presence of subneutralizing DEN-1 immune sera that enhanced DEN-2 infections also enhanced PGE2 but not IL-4 production. The results of this study suggest that ADE of DEN-2 infections associated with induction of immunosuppressive mediators such as PGE2 and IL-4 can be simultaneously assessed in a real-time fashion.

  • Dengue virus
  • Reverse transcription-polymerase chain reaction
  • Cytokine
  • Antibody-dependent enhancement

1 Introduction

Dengue viruses are members of the flavivirus family and have four distinct serotypes (DEN-1, 2, 3 and 4) with numerous isolated strains [1]. Dengue fever (DF) was a benign viral illness before World War I, after which it became a worldwide infectious disease due to troops and transportation carrying the infected mosquitoes and larva throughout the world [24]. The prevalence of dengue hemorrhagic fever (DHF) combined with shock is increasing in Southeast Asia, China and South America, with regional variations in the prevalence of four types of dengue infections [3,57]. Thus, with the arrival of the era of the global village, DHF has become a worldwide problem [45,8].

The extent of involvement of DEN virulence or immune enhancement in fatal DHF remains unclear. Molecular epidemiological studies have suggested that certain genotypes might contribute to the occurrence of DHF in humans [9,10]. However, others have suggested that antibody-dependent enhancement (ADE) may mediate DHF severity [4,11]. Moreover, epidemiological studies have also shown that DHF can occur in infants, with transplacental DEN-1 immune serum followed by DEN-2 infection [12]. It has been proposed that antibody neutralization or enhancement of dengue infections may be determined by the interaction between antibodies and viral epitopes at different antibody:virus ratios [13]. Thus, the role of ADE has become a great concern in areas with a high prevalence of multiple types of DEN infection. Another concern is related to the recent introduction of DEN vaccination [14]. These vaccinations may raise different antibody titers in different age populations resulting in the possibility of ADE-mediated fulminant DHF.

Identification of the relationship between the virulence of DEN and immune enhancement has been hindered by the lack of appropriate animal models for studies and by limitation in the ability to simultaneously assess viral load and immune reactions. A variety of rodent models have been used in the study of infection and immunity induced by dengue viruses [1517]. Unfortunately, most animals are not susceptible to DEN unless an animal-adapted DEN is used [15]. The fact that severe combined immunodeficiency mice are not susceptible to DEN infections unless they receive reconstituted human peripheral mononuclear leukocytes (MNLs) or leukemia cells [16,17], indicates the importance of the kinetic relationship between human leukocyte reactions and DEN replication. Studies with dengue patients by Kurane et al. [18] showed that plasma interferon-γ (IFN-γ) levels were higher both in patients with DF or DHF, whereas the IFN-? levels were significantly higher in patients with DHF than those with DF 1 day before defervescence. We also showed that MNLs from dengue patients in response to DEN-2 released a higher IFN-γ than those from non-immune controls[19]. Recently, many studies have shown that cytokine responses in virus infections influence host recovery and viral persistence [20]. These studies, however, did not have an opportunity to simultaneously compare the cytokine response and virus replication. In this study, we investigated ADE in human MNLs with DEN-2 infections in the presence or absence of heterotypic DEN-1 immune sera by simultaneous monitoring of virus replication and immune reactions. DEN-1 immune sera were selected due to its ready availability following a recent epidemic DEN-1 outbreak in Taiwan [21]. DEN-2 viruses were selected in this study because the most prevalent heterotypic DHF and dengue shock syndrome (DSS) has occurred in classical DEN-1 prevalence followed by DEN-2 outbreak [12,22]. Employing the in vitro ADE of DEN-2 infections by DEN-1 subneutralizing sera, we have previously shown that MNLs from DEN-1 immune patients in response to DEN-2 could release different cytokine and lipid mediator profiles in comparison to those from non-immune controls [23]. This study further explored other possible mechanisms involved in the relationship between virus replication and immune reactions such as Th1/Th2 cytokines (IFN-γ and interleukin-4 (IL-4)) and prostaglandin E2 (PGE2) production in a real-time fashion.

2 Materials and methods

2.1 Preparation of DEN-1 immune and non-immune sera

DEN-1 immune sera were collected from DEN-1-infected donors who were confirmed to have DF in an outbreak of DEN-1 infection in 1991 [21]. After informed consent was obtained, 10 DEN-1 immune sera were harvested from a 10-ml peripheral blood sample collected from each patient. These sera were tested for specific anti-DEN antibody titers using a dengue blot detection kit purchased from Gene Labs Diagnostics, Singapore. Non-immune control sera were obtained from normal volunteers and confirmed to be free of DEN-specific antibody titer by dengue blot detection. These sera were heat-inactivated at 56°C for 30 min to deplete the complement activity before use in neutralization and enhancement experiments.

2.2 Preparation of DEN-2 viruses

DEN-2 (New Guinea C strain) obtained from the Institute of Preventive Medicine, National Defense Medical Center was used in this study. Viruses were propagated in Aedes albopictus C6/36 cells as previously described [23]. Virus titers were determined by a standard plaque-forming unit (PFU) assay on BHK-21 cells. The virus titers were adjusted to 1.2×107 pfu ml−1 in RPMI 1640 (Gibco BRL, Gaithersburg, MD, USA) with 10% fetal calf serum (FCS) in a large-scale preparation. The same lot of viruses were aliquoted and stored at −70°C before use.

2.3 Viral PFU assay

Viral PFU assays were performed in BHK-21 cells [24,25]. BHK-21 cells (2×105/ml) suspended in modified Eagle's medium were seeded into each 24-well culture plate (Nalge Ncnu, Rochester, NY, USA) and incubated at 37°C in a humidified CO2 incubator. The cells were cultured overnight before viral adsorption. A series of viral dilutions made in the culture medium at 0.1 ml were adsorbed onto the monolayer of BHK-21 cells for 1 h at 28°C. Each virus-infected culture well was overlaid with 0.8 ml of 1.5% carboxylmethyl cellulose containing 2% FCS for an additional 6 days. The plaque-forming wells were fixed with 10% formalin for 60 min followed by staining with 0.5% crystal violet in normal saline solution containing 50% alcohol and 5% formalin. Results were calculated by counting plaques on the four replicate wells, and the final results were corrected by individual dilution factors.

2.4 Preparation of MNLs

Human MNLs were prepared from the peripheral blood of healthy volunteers who were confirmed to be free of past DEN infections by serology using the dengue blot detection kit. The blood was mixed with 4.5% dextran at a ratio of 1:5 to separate leukocytes from red blood cells. Leukocytes were separated into MNLs and polymorphonuclear cells by a Ficoll-Paque density centrifugation as previously described [2627]. MNLs were suspended at 2×106 cells ml−1 for studies.

2.5 ADE of DEN-2 infections in MNL cells

Human MNLs (2×106 cells ml−1) were incubated with DEN-2 at a multiplicity of infection (MOI)=0.1. DEN-2 was pre-opsonized with and without different titers of DEN-1 immune sera at 37°C for 60 min and adsorbed onto MNLs for another 60 min. The infected MNLs were washed to remove extracellular viruses and cultured in RPMI 1640. After 3 days, the culture supernatants were harvested for measurement of viral PFU formation in BHK-21 cells as described above.

2.6 Isolation of viral RNA

One volume of the DEN-2 from the original well-known DEN-2 preparation or the supernatants from DEN-2-infected MNLs were collected and mixed with 3 volumes of Tri-reagent solution (Sigma, St. Louis, MO, USA). After thorough vortexing, samples were added at one-fifth the volume of chloroform (Scharlau, sa, Barcelona, Spain) for phase separation. The mixtures were shaken vigorously and allowed to stand for 10 min at room temperature before centrifugation at 12 000×g for 15 min at 4°C. After centrifugation, the upper aqueous phase was transferred to a fresh diethylpyrocarbonate (DEPC)-treated tube and the same volume of isopropanol (Merck KGaA, Darmstadt, Germany) was added for RNA precipitation at −20°C for 10 min. The RNA was harvested by centrifugation at 12 000×g for 10 min at 4°C, followed by precipitation with 75% ethanol (Merck KGaA, Darmstadt, Germany). Finally, the RNA was resuspended in 20 µl of 0.1% DEPC-treated H2O.

2.7 Measurement of viral RNA titers by real-time quantitative reverse transcription-polymerase chain reaction (RT-PCR)

The TaqMan fluorogenic RT-PCR technique (Perkin Elmer, Foster City, CA, USA) allows performance of RT and PCR in one tube. The real-time detection of the RT-PCR products is established by targeting specific release of a fluorescent reporter molecule during the PCR reaction [2830]. During the course of PCR, processing in the primer extension by the 5′ exonuclease activity of the rTth DNA polymerase degrades the internally hybridizing probe [29,30]. The degradation of the probe leads to an increase in the level of fluorescence in the reaction mixture. A fluorescent threshold was manually set across all samples in the experiment starting with the exponential phase of the fluorescent signal increase [2729]. As the threshold is set at 10 times the standard deviation of the mean baseline fluorescence emission calculated from the first three to 15 PCR cycles [29,31], the fractional number of RT-PCR cycles demonstrating fluorescence exceeding the threshold values was designated the cycle threshold (Ct). DEN-2 primers were designed from the available published sequences with the aid of a computer program for sequence analysis (Primer Express, PE Applied Biosystem). Sense and antisense primers as well as the nested TaqMan probe were derived from the envelope (E) region sequence of DEN-2 from the gene bank (sense primer: 5′-tgaca atgcg ttgca tagga at-3′; antisense primer: 5-cgtcg tcaca cagct tccat-3′; and nested probe sequence: 5′-tttca ggagg aagct gggtt gacat agtct t-3′). The primers and nested fluorogenic probe could detect DEN-2 but not other subtypes of DEN or Japanese encephalitis viruses.

The RT-PCR reagents were added following the protocol of TaqMan EZ RT-PCR Kit (Perkin Elmer, Foster City, CA, USA). All reactions were performed in an ABI-7700 sequence detector (Perkin Elmer, Foster City, CA, USA). The RT-PCR cycling parameters were set as follows: the RT reaction at 50°C, 2 min; 60°C, 30 min; and 95°C, 5 min; followed by 40 cycles of PCR reactions at 94°C, 20 s, and 60°C, 1 min. The results were detected in a real-time fashion and recorded on a plot showing fluorescence versus time. Reaction products (104 bp) were also visualized on ethidium bromide-stained 2.5% agarose (Pierce Co., Rockford, IL, USA) gel with a 100-bp ladder (Pharmacia Biotech, Piscataway, NJ, USA) as a reference.

2.8 Measurement of IFN-γ, IL-4 and PGE2 levels

Many immune mediators have been proposed to mediate vascular leakage and shock syndrome. In this study, the relations between the Th1/Th2 profiles demonstrated by IFN-γ/IL-4 levels and vascular leakage mediator demonstrated by PGE2 levels were examined. IFN-γ and IL-4 levels were measured by an enzyme-linked immunosorbent assay (ELISA) kit purchased from R&D System (Minneapolis, MN, USA). Similarly, PGE2 levels were also measured by a competitive ELISA kit purchased from R&D System. The results were calculated from interpolation in a standard curve made from a series of well-known concentrations of IFN-γ, IL-4 and PGE2.

3 Results

3 Assessment of ADE of DEN-2 replication by standard viral PFU and real-time quantitative RT-PCR assays

To determine whether different DEN-1 immune serum titers could have different effects on DEN-2 replication in MNLs, we investigated the effect of heterotypic antibody enhancement of DEN-2 infections by the viral PFU assay and by the TaqMan RT-PCR assay. We used specific DEN-2 primers and a nested fluorescent probe to detect DEN-2 RNA in 6 h. This technique could detect DEN-2 titers closely correlated to the titers determined by viral PFU assay. The appearance of fluorescence was reflected in a real-time fashion as indicated by the PCR-reacting cycles called Ct values, as shown in Fig. 1a. A series of 10-fold dilutions of DEN-2 titers was reflected by the appearance of fluorescence after an average of 3.48 PCR cycles in the fluorogenic RT-PCR. This appearance of fluorescence approximately 3.48 PCR later indicates a 23.48=11.2-fold decrease of viral RNA titers in each 10-fold diluting factor. Employing a classical viral PFU assay established after 7 days, we found that DEN-1 immune serum at 1:10, similar to control serum, did not neutralize or enhance DEN-2 replication in MNLs, whereas the ADE of DEN-2 infections occurred in the titers at 1:100 and 1:250 dilutions (Fig. 2a). The DEN-1 immune sera at 1:250 induced the best enhancement of DEN-2 infections (Fig. 2b).

Figure 1

Correlation of quantitative RT-PCR analysis of DEN-2 titers with PFU titers. (a) Sensitivity of the RT-PCR analysis of DEN-2 titers. RNA harvested from a series of well-known DEN-2 titers (PFU) was subjected to the real-time quantitative RT-PCR. The lower limit of detection could be as low as detecting limit downward to 1.4 pfu reaction−1, but not as low as 0.7 pfu. (b) Correlation of original PFU titers with Ct values of the result from RT-PCR. A series of RT-PCR analysis of DEN-2 titers were closely correlated to the PFU titers (r=0.997, p<0.001; Pearson coefficient) performed with the same virus dilutions. There were three reproducible experiments performed, the data presented were calculated from a representative triplicate study.

Figure 2

ADE of DEN-2 infection in MNLs by heterotypic DEN-1 immune serum. Two assay systems including the standard PFU assay (a) and real-time quantitative RT-PCR (b) were simultaneously analyzed. MNLs at 2×106 cells ml−1 were infected with DEN-2 at an MOI=0.1 after the DEN-2 were pre-opsonized with no human serum (negative control; blank bar), non-immune control serum (dotted bar) or DEN-1 immune serum (striped bar) at ratios of 1:10, 1:100, 1:250 or 1:1000. Using the RNA harvested from MNLs infected by DEN-2 at an MOI=0.1 in the presence of control or DEN-1 immune serum, the virus titers were determined by interpolating their Ct values onto a standard curve. The viral titer in the DEN-1 immune sera at 1:250 was higher than that in control sera (2.5±1.2×104 versus 2.3±1.2×103 pfu ml−1; P<0.05, t-test; *P<0.05). Results are presented as a mean±S.E.M. of 10 experiments.

3.2 Th1/Th2 profiles in the enhancement of DEN-2 infections by DEN-1 immune sera

In addition to studying the ADE of DEN-2 replication in MNLs by real-time quantitative RT-PCR, we also simultaneously explored whether the ADE of DEN-2 replication in MNLs was associated with a change of cytokine profiles related to Th1/Th2 reaction. The Th1/Th2 profiles as determined by IFN-γ/IL-4 production revealed that DEN-2 alone elicited higher IFN-γ production by MNLs (P<0.05; n=10), but IFN-γ levels were not significantly (P>0.05; n=10) enhanced in the ADE of DEN-2 infections (Fig. 3a). The IL-4 levels were decreased after DEN-2 infections followed by elevation in the ADE of DEN-2 infections by subneutralizing DEN-1 immune sera, but did not reach a significant difference (P>0.05; n=10) (Fig. 3b). Results from this study suggest that the ADE of DEN-2 infections by DEN-1 immune serum did not significantly alter the Th1/Th2 reactions.

Figure 3

IFN-γ and IL-4 levels in the supernatants of DEN-2-infected MNLs in the presence of control (ctrl) or DEN-1 immune sera at 0, 1:100 and 1:250 dilutions. (a) IFN-γ levels were not significantly elevated in the ADE of DEN-2 infection although DEN-2 alone elicited higher IFN-γ production (*P<0.05; n=10). (b) IL-4 levels were decreased after DEN-2 infection, but elevated in the ADE of DEN-2 infections by subneutralizing DEN-1 immune sera although there was no significant difference in statistics.

3.3 Association of PGE2 production with the enhancement of DEN-2 infections by DEN-1 immune sera

In addition to the measurement of cytokine production in the ADE of DEN-2 infection, we also studied production of the vasodilatation mediator, PGE2. We found that in response to DEN-2 infections, MNLs could elicit PGE2 production, and that the ADE of DEN-2 infections by subneutralizing DEN-1 immune sera was associated with further elevation of PGE2 levels (P<0.05; n=10) (Fig. 4). These results suggest that PGE2 may be involved in the pathogenesis of ADE.

Figure 4

PGE2 levels in the supernatants of DEN-2-infected MNLs in the presence of control or DEN-1 immune sera. DEN-2 infections in MNLs could elicit PGE2 production, and the ADE of DEN-2-infected MNLs induced by subneutralizing DEN-1 immune sera at 1:100 was associated with further elevation of PGE2 levels (*P<0.05; n=10).

4 Discussion

Many mechanisms have been proposed to explain DHF and DSS [4,32,33]. ADE and virulence of DEN are two major possible mechanisms. The associations between the results of molecular studies and clinical analyses suggest that certain virus mutations and genotypes might contribute to serious DHF and DSS [9,10]. However, molecular studies have also isolated viruses belonging to two distinct genotype groups from both DF and DHF cases [10]. This clearly indicates that virulence of DEN is not the only factor contributing to serious dengue infections. ADE has been shown to mediate serious dengue infections with DHF or DSS in both in vitro and ex vivo models [11,13]. Employing a variety of dengue viruses, monoclonal and polyclonal antibodies, Morens [13] demonstrated that ADE of DEN infections occurred both in macrophage-like cells and human monocytes. The fact that ADE exists in different strains of viruses as well as different types of antibodies indicates that antigen specificity or antibody epitope may not be as important or unique for inducing ADE, as the presence of viral receptors of target cells close to FcR. In support of the involvement of both viral and host factors in a variety of cell tropism of DEN infections, Chen et al. [34] recently showed that a chimeric protein with viral and immunoglobulin Fc epitopes could effectively block DEN infections. In fact, we also found that the ratios of total viral RNA titers detected by quantitative RT-PCR to viable viral PFU titers were affected by different subneutralizing immune serum titers. These results indicate that both standard PFU assay and real-time quantitative RT-PCR can detect ADE of DEN-2 infections induced by certain subneutralizing titers of DEN-1 immune sera. However, the viable viral PFU titers were 25- to 33-fold lower than the total viral RNA titers detected by quantitative RT-PCR in the reactions with no enhancement or limited enhancement. In the highest enhancement which was induced by 1:250 subneutralizing DEN-1 immune serum, the ratio of total RNA titers detected by RT-PCR to viable PFU titer decreased to 5.5-fold. The total viral RNA titers determined by the TaqMan quantitative RT-PCR were variably higher than those determined by standard viable PFU titers indicating that kinetic viral replication and killing of viruses occurred during the DEN infections in human MNLs in the presence of different subneutralizing immune serum titers. This finding indicates that viral replication and the killing of DEN viruses are kinetically influenced by immune parameters such as antibody titers and immune mediators.

Several lines of evidence raise disturbing questions about the possible role of ADE in the natural history of viral infections with flaviviruses and HIV [4,3537]. The ADE of DEN infections has been previously demonstrated [4,38]. Similarly, in vitro studies have linked ADE to a variety of HIV strains in different cell types using polyclonal and monoclonal antibodies directed against various gp120 and gp41 antigens [36,37]. If ADE is truly involved in the pathogenesis of dengue and HIV diseases, the introduction of vaccines that induce humoral immunity to certain viral epitopes may also become worrisome potential causes for worsening HIV and DEN infections upon natural exposure after vaccination. The use of a real-time quantitative RT-PCR analysis for assessing antibody-mediated viral neutralization or enhancement in MNLs in conjunction with simultaneous assessment of immune reactions may provide a suitable method for determining how vaccine-elicited antibodies directed against HIV or DEN might modify immunity and infectivity.

The importance of IFN-γ in the control of a number of virus infections has been demonstrated in mice dysfuctional for the activity of IFN-γ[20]. We have previously shown that MNLs, obtained from DEN-1 immune donors, in response to DEN-2 released higher levels of IFN-γ than those from non-immune donors [19]. In the present study, we also found that IFN-γ production increased after DEN-2 infections. Th1-type immune responses appear to play an important role in the clearance of many viruses and their subversion may represent an important means of establishing chronic infection [20]. In the presence of subneutralizing heterotypic antibodies that promoted DEN-2 replication, we found that Th1/Th2 profiles as demonstrated by IFN-γ/IL-4 production were not significantly affected in the ADE by DEN-1 immune sera although IL-4 levels appeared to be increased in the ADE reactions. Moreover, we found that subneutralizing ADE of DEN-2 infection was associated with the elevation of PGE2, suggesting that an abnormal surge of PGE2 levels may play an important role in the pathogenesis of ADE. It is not known why the subneutralizing antibody titer at 1:100 had a higher induction of PGE2 production than the titer at 1:250, while enhancing the best ADE of DEN-2 replication. It is not excluded that production of cytokines and PGE2 show a kinetic change depending on time course, viral load and antibody titers studied. Further studies are needed to kinetically correlate the PGE2 levels with viral load and clinical outcomes in patients with and without DHF/DSS.


This work was supported by grants DOH88-TD-1019 and DOH87-TD-1028 from the Department of Health, Executive Yuan, and by the Section of Techniques and Reagents of PE Biosystems, Taiwan.


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
  26. [26].
  27. [27].
  28. [28].
  29. [29].
  30. [30].
  31. [31].
  32. [32].
  33. [33].
  34. [34].
  35. [35].
  36. [36].
  37. [37].
  38. [38].
View Abstract