OUP user menu

Altered T helper 1 reaction but not increase of virus load in patients with dengue hemorrhagic fever

Rong-Fu Chen, Jien-Wei Liu, Wen-Ting Yeh, Lin Wang, Jen-Chieh Chang, Hong-Ren Yu, Jiin-Tsuey Cheng, Kuender D. Yang
DOI: http://dx.doi.org/10.1016/j.femsim.2004.11.012 43-50 First published online: 1 April 2005


To investigate whether dengue-2 patients with and without dengue hemorrhagic fever had different virus load, immune mediators, or T helper (Th) reaction, we simultaneously measured virus load, immune mediators and the Th1/Th2 transcription factors T-bet/GATA-3 mRNA expression in a large outbreak of dengue-2 infections in Southern Taiwan. Results showed that virus load was not significantly different between patients with and without dengue hemorrhagic fever. Patients with dengue fever had higher IFN-γ levels, but patients with dengue hemorrhagic fever had significantly higher IL-10 levels. Further studies showed that patients with dengue hemorrhagic fever had a significantly lower T-bet than those with dengue fever, but GATA-3 mRNA expression in peripheral blood leukocytes was not significant difference between both groups. In conclusion, altered Th1 reaction as reflected by lower T-bet mRNA expression associated with higher IL-10 levels might be involved in the pathogenesis of dengue hemorrhagic fever.

  • Virus load
  • Dengue hemorrhagic fever
  • T-bet
  • GATA-3
  • Real-time RT-PCR

1 Introduction

Dengue (DEN) virus transmitted by mosquitoes usually caused a self-limited febrile illness, called dengue fever (DF). A life threatening complication of DF called dengue hemorrhagic fever (DHF) has been prevalent in Southeast Asia since 1950s. DHF is now becoming a global disease, reported from many countries in the Pacific Rim and the Americas [1]. The characteristic features of DHF include altered immune reactions, vascular leakage, hemoconcentration and thrombocytopenia [2]. DHF has been classified into four grades on the basis of clinical presentation. The mildest is grade I and the most severe is grade IV [3]. The first indication of an immunological mechanism for DHF was the observation in a Bangkok outbreak of DHF in 1960s [4]. In that outbreak, over 85% of children with DHF had high DEN heterotypic cross-reactive antibody titers, suggesting an antibody-dependent enhancement (ADE) of DEN infection in the pathogenesis [5]. Several prospective studies have concluded that DHF is more common in secondary DEN infections than in primary DEN infections [4]. Despite extensive studies, the pathogenesis of DHF cannot be fully attributed to the ADE. In another study with mononuclear cells from children prior to secondary DEN infections, an augmented cytokine response was correlated to the severity upon secondary DEN infections [6,7]. This suggests that immune enhancement of DEN infections may be linked to cytokine production by activated dengue virus-specific T cells. Activation of dengue virus-specific T cells and dengue virus-infected monocytes may result in increased capillary permeability in patients with DHF [8,9].

T helper (Th) cells have two major subsets, Th1 and Th2 cells, based on cytokine production profiles. Th1 cells secrete IFN-γ and TNF-β responsible for cell-mediated reactions, delayed-type hypersensitivity and tissue injury [10]. Th2 cells secrete IL-4, IL-10, and IL-13 responsible for protective or offensive antibody production by B cells [11]. Earlier studies have shown that cell cytotoxicity and IFN-γ might be implicated in complication of DEN infections [9,12]. In contrast, Chaturvedi et al. [13] had reported a shift from Th1-type cytokine response to Th2-type cytokine response in patients with DHF. A similar cytokine response was also observed in DEN-infected human peripheral blood leukocyte (PBL) cultures [14]. In an in vitro model, we have previously found that ADE of DEN-2 infections was associated with the suppression of Th1 reaction [15]. Additional studies simultaneously testing virus load and immune mediators in the leukocyte reaction showed that ADE of virus replication was correlated to PGE2 but not to IFN-γ or IL-4 levels [16].

Recently, a model for molecular Th1/Th2 polarization that involves a reciprocal regulation of the Th1-specific transcription factor T-bet (T-box expressed in T cells) and the Th2-specific transcription factor GATA-3 (GATA-binding protein 3) mRNA expression has been deciphered [17]. In order to clarify whether Th1/Th2 cytokine induction or virus load was involved in the DHF outbreak in Southern Taiwan, we simultaneously measured the virus load, the Th1/Th2 cytokine profiles in blood and the Th1/Th2 transcription factor (T-bet/GATA-3) mRNA expression. We analyzed the relationships among the immune reactions, virus load and disease severity in a large DEN-2 outbreak in Southern Taiwan.

2 Materials and methods

2.1 Case-control study design

We utilized a complicated and uncomplicated case-control design in this study [18,19]. Upon admission, patients with suspicious DEN infections were recruited for this study. This study was approved by the Institution Review Board of this hospital. Once we enrolled 1 to 2 cases of DHF case, we simultaneously included 2–4 DEN infected patients without DHF and one normal control for this study during the 2002–2003 outbreak of DEN-2 in Kaohsiung, Taiwan. DEN-2 infections were confirmed by the DEN-2 virus detection in blood by real-time quantitative RT-PCR [16]. Patients with DHF were defined by the WHO criteria showing DF complicated with reduced platelets (<100,000/mm3), petechial or hemorrhagic manifestations, and plasma leakage showing hemoconcentration ≥20%, pleural effusion, ascites or hypoalbuminemia [20]. According to WHO criteria, we classified patients with DHF grade I and grade II as mild DHF (n= 24), and those with DHF grade III and grade IV as severe DHF (n= 9). The normal controls were normal adult volunteers who had no detectable dengue virus in blood by RT-PCR and absence of dengue antibody determined by dengue IgG Capture enzyme-linked immunosorbent assay (ELISA) kits (described below).

2.2 Collection and separation of blood samples

Heparinized-blood samples (5 ml) from febrile patients who were hospitalized with suspicious DEN infections within 2–7 days of illness were collected for studies. A part of the blood (0.5 ml) was subjected to extraction of virus RNA for real-time quantitative RT-PCR analysis of virus load [16]. The rest of the blood was separated into plasma and blood cells by centrifugation at 3000 rpm (150g) for 10 min. After collecting the plasma for aliquots and storage at −70 °C, the leukocytes were separated from red blood cells by 4.5% dextran sedimentation as previously described [16].

2.3 Assessment of dengue primary and secondary infections

Serological methods to detect DEN antibodies have been the most commonly diagnostic procedures for differentiation of primary and secondary DEN infections [21]. We used ELISA kits to detect IgG antibodies for the differentiation between primary and secondary DEN-2 infections in this study. Since our blood samples were collected between 2 and 7 days of the illness, we could define a definite primary DEN-2 infection by RT-PCR detection of DEN-2 virus in the blood, and defined a secondary DEN-2 infection by detectable DEN IgG and RT-PCR detection of DEN-2 virus in the blood. The dengue IgG antibodies were assessed by the IgG Capture ELISA Kits (Panbio, Queensland, Australia). The cut-off value for a positive DEN IgG detection was determined by a calibrator Antibody Index (AI) of each sample greater than 22, based on the ratio of the calibrator optical density to a standard value as the manufacturer's recommendation.

2.4 Real-time quantitative RT-PCR analysis of virus load

We subjected viral RNA extracted from whole blood of the patients to fluorogenic quantitative RT-PCR detection of total virions as previously described [16]. In brief, 0.5 ml of blood was individually combined with 0.5 ml of TRIZOL® solution (Invitrogen, CA, USA) to separate RNA from DNA and protein fractions. After thorough vortexing, 0.1 ml of chloroform was added to samples (Scharlau, sa, Barcelona, European Union) for phase separation. After centrifugation, the upper aqueous phase was transferred to a fresh DEPC-treated eppendorf and the same volume of isopropanol (Merck KGaA, Darmstadt, Germany) was added for RNA precipitation at −20 °C for 1 h. The RNA was harvested by centrifugation at 12,000g for 10 min at 4 °C, followed by 75% ethanol (Merck KGaA, Darmstadt, Germany) precipitation. Fluorescent RT-PCR was carried out in an ABI 7700 quantitative PCR machine (Applied Biosystems, Foster City, CA) for 40 cycles using TaqMan technology [16,22]. The forward primer, the reverse primer and the TaqMan probe sequence for detecting DEN-2 were shown in Table 1[16].

View this table:
Table 1

RT-PCR primers and TaqMan probes used in this study

DEN-2Forward primer: 5′-GGC TTA GCG CTC ACA TCC A-3′
Reverse primer: 5′-GCT GGC CAC CCT CTC TTC TT-3′
T-betForward primer: 5′-AAC ACA GGA GCG CAC TGG AT-3′
Reverse primer: 5′-TCT GGC TCT CCG TCG TTC A-3′
GATA-3Forward primer: 5′-ACC GGC TTC GGA TGC AA-3′
Reverse primer: 5′-TGC TCT CCT GGC TGC AGA C-3′
β-ActinForward primer: 5′-GGC CAA CCG CGA GAA GAT-3′
Reverse primer: 5′-CGT CAC CGG AGT CCA TCA C-3′

2.5 Measurement of blood IFN-α, IFN-γ, IL-13 and IL-10 levels

The blood cytokines including IFN-α, IFN-γ, IL-10 and IL-13 were measured to reflect Th1/Th2 reactions. The IFN-α, IFN-γ, IL-13 and IL-10 levels were measured by ELISA kits purchased from Bender MedSystems Inc. (Vienna, Austria). The results were calculated from interpolation in a standard curve made from a series of well-known concentrations of commercial standards [15,16].

2.6 Real-time quantitative RT-PCR analysis of T-bet and GATA-3 mRNA expression

We subjected total RNA extracted from PBLs to quantitative analysis of leukocyte mRNA expression. As described above, the blood leukocyte pellet was mixed with 0.5 ml of TRIZOL® solution (Invitrogen, CA, USA). Finally, the RNA was subjected to the real-time RT-PCR detection with TaqMan probes using the ABI PRISM 7700 instrument (Applied Biosystems, Foster City, CA) as previously described [22]. TaqMan probes and primers for the quantitative detection of target mRNAs were designed by using Primer Express computer software (Applied Biosystem, Foster City, CA) as shown in Table 1. There were three reproducible experiments performed, the data presented were analyzed from a representative triplicate experiment. RT-PCR products were also visualized on ethidium bromide-stained 1.5% agarose (Pierce Co., Rockford, IL, USA) gel with a 100-bp ladder (Pharmacia Biotech, Piscataway, NJ, USA) as a reference. The increase of the T-bet and GATA-3 mRNA expression was therefore calculated assuming 100% efficient PCR where each Ct was normalized to β-actin mRNA expression as shown by the equation at Embedded Image. The Ct1 (target) and Ct2 (target) represent the Ct values for the target gene expression from patients and normal control samples, respectively. Ct1 (actin) and Ct2 (actin) represent the Ct values for the β-actin gene expression in patients and normal control samples, respectively.

2.7 Data presentation and statistics

Data from subjects with confirmation of DEN-2 infections were classified into patients with DF and DHF. Results from the measurement of immune mediators were presented as means ± SE. Virus load was presented as copies per milliliter calculated by standard curve of known copies. Differences in sex and proportion of secondary DEN-2 infections between patients with DF and DHF are analyzed by χ2 test. Analysis of variance (ANOVA) followed by Student's t-test is used to analyze immune mediator differences between patients with and without DHF. Correlations of virus load to clinical severity and cytokines are determined by Pearson correlation.

3 Results

3.1 Demographic data of patients with DF and DHF

During a large DEN-2 outbreak in Southern Taiwan between 2002 and 2003, we recruited 128 suspected DEN-2 patients who were admitted to this hospital for this study in a complicated vs. uncomplicated case control design. Out of the 128 patients, 99 were proven to be DEN-2 infected with detectable DEN-2 by real-time quantitative RT-PCR in blood. The age of the patients studied ranged from 20 to 81. Of the patients, 66 had DF, and 33 had DHF. Twenty-four out of the 33 DHF patients had mild DHF (grade I/II) and the other 9 DHF patients had severe DHF (grade III/IV). The main characteristics of the study population were summarized in Table 2. The gender distribution was not different between patients with DF and DHF. Secondary DEN-2 infections were more frequently found in patients with DHF than those with DF (57% vs. 25%, p= 0.010; Table 2).

View this table:
Table 2

Characteristics of patients with DF and DHF

DiseaseGenderAge (range)Infection history
MaleFemalePrimary (%)Secondary (%)
DF25/6641/6646.8 (20–81)7525
DHF13/3320/3357.8 (30–76)4357
Total38/9961/9950.5 (20–81)6634
  • Abbreviations: DF, dengue fever; DHF, dengue hemorrhagic fever.

  • Indicates a significant difference between both groups as analyzed by χ2 test; p= 0.010.

3.2 Viral load in patients with DF and different severity of DHF

A real-time fluorogenic RT-PCR assay was used to quantify the DEN-2 virus load in blood. Results showed that there was no significant difference of virus load among patients with DF, mild DHF I/II and severe DHF III/IV (173.8 ± 37.4, 358.0 ± 154.8, and 269.0 ± 112.0 copies/ml, respectively; Fig. 1(a)). To investigate the possibility of different patterns of viremia during the transition from fever to defervescence, we examined sequential virus load in PBLs of patients with DF and DHF. It was found that the earlier blood samples taken from the DEN-2 infected patients the higher virus titers detected. Patients with DHF revealed a higher virus load in the early febrile stage than those with DF, but the difference did not reach a significant difference (Fig. 1(b)).

Figure 1

A real-time fluorogenic RT-PCR detection of the DEN-2 genome copies in blood. Blood samples collected upon admission (2–7 days after onset of symptoms) were subjected to RNA extraction for real time RT-PCR analysis of virus load. (a) The virus titers in blood determined by the TaqMan fluorogenic RT-PCR were 173.8 ± 37.4; 358.0 ± 154.8; and 269.0 ± 112.0 copies/ml, respectively, in patients with DF (n= 66), DHF I/II (n= 24), and DHF III/IV (n= 9). There was no significant difference among three groups. (b) Kinetic changes of the virus load detected by real-time fluorogenic RT-PCR in patients with DF (open bar) and DHF (solid bar).

3.3 Cytokine profiles in the blood from patients with DF and DHF

In response to virus infections, human beings usually release interferons (IFNs), such as IFN-α and IFN-γ to promote cellular immunity (Th1 reaction) [10], and secret IL-10 and IL-13 to promote antibody production (Th2 reaction) [11]. We assessed IFN-α, IFN-γ, IL-10 and IL-13 in blood to reflect the Th1/Th2 profiles. It was found that patients with DF elicited higher IFN-γ production than those with DHF (130.2 ± 16.9 vs. 84.2 ± 12.6 pg/ml, p= 0.01; Table 3). In contrast, patients with DHF had higher IL-10 levels than those with DF (117.0 ± 52.8 vs. 11.7 ± 3.5 pg/ml, p= 0.03). On the other hand, there were no difference in the IFN-α and IL-13 levels between patients with DF and DHF.

View this table:
Table 3

Blood Th1/Th2 cytokines in patients with DF and DHF

Cytokines (pg/ml)DFDHFp
IFN-α94.1 ± 28.0126.8 ± 49.30.28
IFN-γ130.2 ± 16.984.2 ± 12.60.01
IL-1011.7 ± 3.5117.0 ± 52.80.03
IL-1311.2 ± 4.76.0 ± 5.40.23
  • Abbreviations: DF, dengue fever; DHF, dengue hemorrhagic fever. IFN, interferon; IL, interleukin.

    Data presented are means ± SE, and p between both groups as tested by the Student's t-test.

3.4 Correlation of the severity of DHF to virus load and host immune status

Elevated circulating levels of soluble TNF receptors and soluble IL-2 receptors have been shown to correlate with the disease severity of DHF [23,24]. While we classified the patients' disease severity into simple DF, mild DHF with grade I/II and severe DHF with III/IV classes, the disease severity was significantly correlated to age (p= 0.001), secondary infections (p= 0.017) but not to sex or virus load (Table 4). We also found a significant correlation between the disease severity and IL-10 (p < 0.001), but not IFN-α, IFN-γ or IL-13 levels (Table 4).

View this table:
Table 4

Correlation of disease severity with age and immune status

Host parametersDisease severity
r valuep value
Secondary infection0.2480.017
Virus load0.0550.612
  • Correlation of the disease severity to different demographic characteristics and cytokines in patients with and without DHF was determined by Pearson correlation.

  • Disease severity was classified into simple DF (n= 66), mild DHF I/II (n= 24), and severe DHF III/IV (n= 9).

3.5 The Th1/Th2 transcription factors T-bet/GATA-3 mRNA expression in patients with DF and DHF

In order to explore why patients with DF tended to have higher IFN-γ levels and patients with DHF tended to have higher IL-10 levels, studies were next performed to investigate the molecular basis of the Th1/Th2 polarization by studying T-bet/GATA-3 mRNA expression. Results showed that the Th1 specific transcription factor, T-bet mRNA expressing levels were significantly higher in PBLs from patients with DF compared to those with DHF (increase fold: 204.1 ± 31.9 vs. 125.4 ± 20.3; p= 0.040; Fig. 2(a)). In contrast, the Th2 specific transcription factor, GATA-3 mRNA expressing levels were not different between these two groups (increase fold in DF vs. DHF: 6.1 ± 0.7 vs. 5.9 ± 1.2, respectively; Fig. 2(b)). Further studies on the kinetic changes of T-bet mRNA expression in patients with DF and DHF showed that patients with DF tended to have higher T-bet mRNA expression than those with DHF in early febrile stage (Fig. 2(c)). In the late stage (5–7 days after onset of symptoms), the T-bet mRNA expressing levels were not different between patients with and without DHF.

Figure 2

Real-time quantitative RT-PCR analysis of T-bet/GATA-3 mRNA expression in PBLs from patients with and without DHF. The increase of T-bet mRNA expression was calculated as normalization by β-actin mRNA expression and comparison to normal control. (a) A summary of the T-bet mRNA expression in PBLs from patients with DF and DHF. The increase fold of T-bet mRNA expression in patients with DHF was significantly lower than those with DF (increase fold: 125.4 ± 20.3 vs. 204.1 ± 31.9; p= 0.040). (b) A summary of the GATA-3 mRNA expression in PBLs from patients with DF and DHF. The GATA-3 mRNA expression in patients with DF and DHF was not significantly different as compared to the normal control in each independent experiment (increase fold: 6.1 ± 0.7 vs. 5.9 ± 1.2, respectively). (c) Kinetic changes of T-bet mRNA expression in patients with DF (open bar) and DHF (solid bar).

4 Discussion

Studies to explore the relationship between virus load and the severity of DHF have raised controversial results [2328]. Some studies have shown that DHF in certain DEN-2 and DEN-3 infections was related to an increase in virus load [2628]. A previous study had, however, shown that virus load was higher in patients with DF than those with DHF [25]. Vaughn et al. [23] showed that virus load was not significantly different between DF and DHF patients. We also found that the virus load was not different between patients with and without DHF (Fig. 1 (a), Table 4). Earlier studies in children have shown that Th1 cytokines, such as IL-2 as well as IFN-γ were associated with complication of DEN infections [9,12]. Chaturvedi et al. [13,14] showed that Th2 cytokines might be more related to development of DHF. Taken together, these results suggest that pathogenesis of the DHF in different DEN outbreaks relates to virus load or skewed Th1/Th2 reaction depending on host age, virus serotypes and probably sequential infections of different virus serotypes.

Several cytokines in blood have been implicated in certain complications of viral infections [29,30]. Antibody-dependent enhancement of DEN infections has long been implicated in facilitating in vivo DEN complication and in vitro replication in Fc-receptor bearing cells [31]. The Fc receptor-mediated reactions can be associated with altered production of pro-inflammatory cytokines. Pro-inflammatory cytokines, such as TNF-α and IL-8 were elevated in patients with DHF/DSS [32], whereas certain studies did not find any difference of the blood TNF-α concentrations between patients with DF and DHF [33]. Recently, cytokines related to dominant Th2 reaction have been related to the pathogenesis of DHF [34,35]. In support of this hypothesis, we previously demonstrated that human mononuclear cells in response to DEN-2 virus released higher IL-4 but not IFN-γ production in the presence of DEN-1 immune serum [16]. In this study, we have further demonstrated that an early commitment of lower Th1 transcription factor, T-bet mRNA, expression compatible with lower IFN-γ and higher IL-10 was found in patients with DHF. Mustafa et al. [35] had reported that high levels of IL-13 might contribute to the pathogenesis of DHF. We, however, found no difference of IL-13 levels between patients with DF and DHF in our studies. We did find that IL-10 plays an important role in the pathogenesis of DHF. The blood IL-10 levels were higher in patients with DHF than those with DF (Table 3), and were correlated to the disease severity (p < 0.001, Table 4). This study is the first to demonstrate that a lower Th1 transcription factors, T-bet mRNA, associated with higher IL-10 levels in blood, may be involved in immunopathogenesis of DHF. T-bet, a member of T-box family of transcription factors, is a master regulator of Th1 lineage commitment [36]. An earlier augmentation of T-bet mRNA expression may be considered while developing therapeutics for the prevention of DHF.

Another important finding identified in this study is that the first large DHF outbreak of DEN-2 in Taiwan occurred in an adult patient population. A previous DEN-1 outbreak in Kaohsiung [37] did not have any DHF case reported and another small series of DEN-3 outbreak reported sporadic DHF cases in Tainan, Taiwan [38]. All the DHF patients in this outbreak were adult patients. This is quite different from previous reports showing higher incidence and severity of DHF in young infants [3941]. Differences in host immune response between children and adults have been observed in other viral infections, such as lymphocytic choriomeningitis virus [42]. It, however, may be speculated that both infants and aged people tend to have an innate Th2 preponderance [43,44] that favors the development of DHF as shown in our and other studies [7,13,14,45]. It is not excluded that adults who had been exposed to previous DEN-1 infections tended to have Th2 reaction in response to subsequent DEN-2 infections.

Another question unresolved in this study is the relationship between DHF and secondary DEN infections. Several prospective studies have shown that DHF is more commonly found in secondary DEN infections than in primary DEN infections [4,46]. Although a significantly higher frequency of secondary DEN-2 infections was found in patients with DHF, patients with DF were not exclusively found in patients with secondary DEN-2 infections. The fact that a larger portion of secondary DEN-2 infections and a smaller portion of primary DEN-2 infections can cause DHF suggesting that certain interactions among virus virulence, host genetics and immune status determine the complication of DHF. Further studies are needed to differentiate whether hosts with an innate Th2 tendency or hosts in secondary DEN-2 infections prone to Th2 polarization are susceptible to DHF.


This study was in part supported by a grant NSC92-2314-B-182A-110 from National Science Council, 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].
  39. [39].
  40. [40].
  41. [41].
  42. [42].
  43. [43].
  44. [44].
  45. [45].
  46. [46].
View Abstract