Virology 421 (2011) 245–252

Contents lists available at SciVerse ScienceDirect

Virology

j ourna l homepage: www.e lsev ie r .com/ locate /yv i ro

Infection and activation of human peripheral blood monocytes by dengue virusesthrough the mechanism of antibody-dependent enhancement☆ ,☆☆ ,★ ,★★

Peifang Sun a,⁎, Karolis Bauza a, Subhamoy Pal a, Zhaodong Liang a, Shuenn-jue Wu b, Charmagne Beckett b,Timothy Burgess b, Kevin Porter b

a The Henry M. Jackson Foundation for the Advancement of Military Medicine, Inc. Rockville, MD 20850, USAb Viral and Rickettsial Diseases Department, Naval Medical research Center, 503 Robert Grant Ave, Silver Spring, MD 20910, USA

☆ The views expressed in this article are those of thereflect the official policy or position of the Department ofense, nor the U.S. Government.☆☆ This work was supported by work unit number: 6★ The study protocol was approved by the Naval M

tional Review Board in compliance with all applicablethe protection of human subjects.★★ I am a military service member (or employee of thwas prepared as part of my official duties. Title 17 U.S.Cprotection under this title is not available for any workment.’ Title 17 U.S.C. §101 defines a U.S. Government woitary service member or employee of the U.S. Governofficial duties. This study protocol was approved by theReview Board in compliance with all applicable Federaltection of human subjects.

⁎ Corresponding author.E-mail address: Peifang.sun@med.navy.mil (P. Sun

0042-6822/$ – see front matter © 2011 Published by Eldoi:10.1016/j.virol.2011.08.026

a b s t r a c t

a r t i c l e i n f o

Article history:Received 30 June 2011Returned to author for revision 29 August 2011Accepted 31 August 2011Available online 26 October 2011

Keywords:DengueHuman monocytesAntibody-dependent enhancementCytokineCo-stimulatory markers

Human monocytes are susceptible to dengue virus (DV) infection through an FcR-dependent pathwayknown as antibody-dependent enhancement (ADE). In this study, infection enhancement was observedwhen purified monocytes were infected with DV serotypes in the presence of serially diluted immuneserum antibodies. Analyzing binding of the DV-antibody immune complexes to monocytes by quantifyingthe amount of viruses attached to monocytes, we found that binding did not correlate with the input amountof antibodies; rather, it peaked at suboptimal antibody concentrations, correlating with the observed infec-tion enhancement. These results suggested that immune complexes are involved in hindering DV from bind-ing to FcR-bearing cells; when such a protective feature is weakened, enhancement of viral attachment andADE are observed. Further, increased cytokine production (TNF-alpha and IFN-alpha), and costimulatorymarker expression (CD86 and CD40), were found to be associated with infection enhancement, suggestinga pathological role of ADE-affected monocytes in dengue hemorrhagic diseases.

author and do not necessarilyf the Navy, Department of De-

000.RAD1.S.A0312.edical Research Center Institu-Federal regulations governing

e U.S. Government). This work. §105 provides that ‘Copyrightof the United States Govern-

rk as a work prepared by a mil-ment as part of that person'sNaval Medical Research Centerregulations governing the pro-

).

sevier Inc.

© 2011 Published by Elsevier Inc.

Introduction

DV comprises 4 closely related but antigenically distinct serotypes(1, 2, 3, 4) of arthropod-borne single stranded RNA viruses in thegenus flaviviruses. All 4 serotypes cause a range of disease in humans, in-cluding classic dengue fever (DF, acute febrile illness with headache,body aches, and rash), dengue hemorrhagic fever (DHF), and the life-threatening dengue shock syndrome (DSS) (Gubler, 1989; Halstead,1988). Dengue is the second most important mosquito-borne diseaseafter Malaria. It is endemic throughout large parts of tropical and

subtropical regions of America and Asia. More than 2.5 billion peoplein more than 100 endemic countries are at risk of infection. DHF/DSSrepresents 2.5 to 20% of total hospitalized dengue cases, and is responsi-ble for 30,000 deaths every year.

Typically, acute DF begins 5–7 days after a bite from an infectedmos-quito. Viremia levels usually peak during the first 3 days of the febrileperiod, after which they decline rapidly. In most patients, as the viremiadeclines, fever ends and a period of recovery follows; in a small percent-age of patients, however, severe plasma leakage leads to clinical mani-festations of thrombocytopenia, hypotension, and bleeding, which areprimary signs of DHF and DSS (Libraty et al., 2002; Srikiatkhachorn,2009; Srikiatkhachorn et al., 2007; Vaughn et al., 2000).

One of themajor risk factors for DHF/DSS is the patient's previous ex-posure to DV. Primary DV infections usually result in non-complicated DFand the development of both humoral and cellular immunity; both arelong-term and protect the host from re-infection from the same serotype.Although this anti-dengue immune response is cross-reactive in nature, itdoes not confer long-term cross-protection to other serotypes. Instead, anassociation of DHF with secondary infections was first reported by Hal-stead et al. in a group of Thai children in the 1960s (Russell et al.,1967), and then observed in several other outbreaks in Cuba (Guzmanet al., 2002), and Southeast Asia (Kliks et al., 1988). A phenomenonknown as antibody-dependent enhancement (ADE) where immunesera can enhance infection when diluted to suboptimal concentrations(Goncalvez et al., 2007; Halstead et al., 2005; Moi et al., 2010) was dem-onstrated in in vitro experiments. With ADE, cross-reactive non-

246 P. Sun et al. / Virology 421 (2011) 245–252

neutralizing antibodies acquired from a previous (primary) infection canmediate an enhancement of secondary heterologous infection through aFcγR-dependent mechanism (Mathew and Rothman, 2008).

Histochemistry of autopsy samples from fatal dengue cases and invitro assays using peripheral blood cells revealed that macrophages/monocytes are primary targets for DV infection (Jessie et al., 2004;Kangwanpong et al., 1995). Macrophages/monocytes are potent phago-cytic cells responsible for removing infectious pathogens. However,they are often attacked by invasive pathogens. These cells express allthree classes of FcγR, (FcγRI, II and III) which serve as receptors for den-gue virus entry in the forms of virus-antibody immune complexes. Ele-vated levels of TNF-α, a cytokine produced by activated monocytes,have been detected in DHF/DSS patients (Hober et al., 1993; Nguyenet al., 2004; Vitarana et al., 1991). In this study, functional aspects ofADE-affected monocytes, cell activation and cytokine production, werecharacterized in order to allow a better understanding of the role ofADE and monocytes in the pathology of DHF.

In an in vitro model, infection enhancement was investigated usingpurified primary monocytes from random healthy blood donors andserum IgG antibodies purified froma human subjectwhowas previous-ly exposed to DV3 (D3IgG). Infection enhancement of DV3 andDV1wasobserved at different antibody concentrations. A positive correlationwas observed between the magnitude of ADE levels and cell activationand cytokine secretion.

Results

ELISA and neutralization activity of the purified D3IgG

The serotype cross-reactivity of the purified D3IgG was evaluated byELISA andneutralization assay. ELISA result showed that theD3IgG bondto both DV3 and DV1 with significantly higher O.D. values (Fig. 1a)

0

20

40

60

80

100

1:3 1:9 1:27 1:81 1:243 1:723

EL

ISA

O.D

.%

neu

tral

izat

ion

a

b

0

0.5

1

1.5

2

2.5

3

1:3 1:9 1:27 1:81 1:243 1:7291:24301:7190 0

D3IgG DV3 D3IgG DV1ControlIgG DV3 ControlIgG DV1

IgG dilution

Fig. 1. ELISA and neutralization measurement for D3IgG and ControlIgG. Purified IgGantibodies from a dengue serotype-3 immune subject (D3IgG, close symbol) and a den-gue non-immune subject (ControlIgG, open symbol) were tested for binding (a) andneutralization activities (b) to DV1 (circle symbol) and DV3 (triangle symbol) respec-tively. The neutralization assay was performed using a L-SIGN Raji cell line. The NT50for the D3IgG on DV3 is 44.

compared to the control IgG, suggesting that D3IgG is capable to bindto both DV3 and DV1. By contrast, neutralization activity of the D3IgGagainst DV3 and DV1 was clearly different (Fig. 1b). No neutralizationactivitywas seenwithDV1. Using non-linear regression, a neutralizationtiter-50 (NT50) of 44 was obtained, meaning that at D3IgG dilution of44, the antibodies are capable to reduce 50% of DV3 infection. The con-trol IgG did not show any binding in ELISA, nor infection neutralization.

D3IgG-mediated enhancement of infection in human monocytes

Infection enhancement experiments were set up using serially dilut-ed D3IgG and the control IgG. Infection of monocytes by DV alonewith-out the presence of the D3IgG (corresponding to antibody concentrationof 0) was set as the infection baseline from which infection enhance-ment was determined. As shown in Figs. 2b and c, infection baselinesfor both DV1 and DV3 were low: 1.03% for DV1 from an average of 10donors, and 0.13% for DV3 from an average of 3 donors, indicating thatvery few cells were directly infected by DV. Infection enhancement toheterotypic serotype DV1 was readily detected with undiluted D3IgGand peaked at antibody dilution of 1:9 (14.41%), showing a N14 fold en-hancement. On the other hand, infection enhancement with the homo-typic serotype DV3 (Fig. 2b) was not seen in the presence of higherantibody titers (from 1 to 1:27); infection enhancement appeared atlower antibody titers (1:81–1:243). The highest infection rate was0.86%, a 6.6 fold enhancement above the baseline. Interestingly, forthis particular case, the peak ADE (1:81) for DV3 appeared when neu-tralization antibody titers were below 44, the titer of NT50.

The L-SIGNdetectionmethod thatmeasures viruses in the culture su-pernatants showed a similar pattern of infection enhancement (Figs. 2dand e). Infectious virus titers in culture supernatants increased from 24 hto 48 h post-infection. For DV1, infectious virus titers in culture superna-tants declined from 48 h to 72 h. Due to the shortage of monocytes, wewere unable tomeasure supernatant virus titers for DV3 at 72 h post-in-fection. Results suggested that antibody enhancement of DV infection onFcγR-bearing cells is dependent on antibody titers and virus serotypes.

Immune complex binding to monocytes

The initial step that leads to an infection is the formation of virus–antibody immune complexes and binding the immune complexes toFcγR on target cells. By quantifying the amount of viruses attached tomonocytes, we were able to indirectly evaluate the amount of DV-anti-body immune complexes bond to monocytes. The quantitative mea-surement of viruses was done by measuring the amount of viral RNAcopy numbers in a qPCR method. For binding assay, cells exposed tovirus alone (antibody concentration of 0) were set as assay baselines.As shown in Fig. 3, the peak of DV3 binding was at IgG dilution of1:27, which was slightly miss-matched with the peak of infection en-hancement of 1:81 (Figs. 2b and d); the peak of binding of DV1 (1:9)waswell-matchedwith infection enhancement (Figs. 2c and e). Regard-less, the virus binding curves (Fig. 3) showed an overall similar trendwith the infection enhancement curves (Fig. 2): the highest binding ac-tivity corresponded to the suboptimal concentrations of D3IgG. For thecontrol IgG, baseline binding was detected, but no significant enhance-ment of bindingwas observed at any concentration. In all of the bindingexperiments performed, untreated monocyte or monocytes treatedonlywith immune antibodies (Mock control) showed 0 RNA copy num-bers in qPCR (data not shown).

Cytokine production

As shown in Fig. 4, monocytes infectedwith either DV3 or DV1 alonewithout the treatment of D3IgG (IgG dilution of 0) did not significantlyshow cytokine production; higher amounts of cytokine production cor-responded to infection enhancement. While TNF-α, IFN-α and IL-10production was all detected from DV1 ADE-affected monocytes during

0

1b c

d e

% o

f in

fect

ed m

onoc

ytes

%

of i

nfec

ted

L-S

IGN

cel

ls

0

10

20

30

40

1 0

SS

C

a DV+D3IgG DV alone Cell alone

0

6

12

18

0

20

40

60

80

1 0

24 h

48 h

72 h

0.75

0.5

0.25

1:2431:811:271:91:3

IgG dilution

1:2431:811:271:91:3

IgG dilution

DV1DV3

2H2

Fig. 2. Infection enhancement in monocytes. a) Dot plot images showing infected monocytes. Infected cells were directly stained with a fluorescent monoclonal anti-dengue antibody2H2. b–e) Infection enhancement of monocytes with DV3 and DV1 in the presence of serially diluted D3IgG (close symbols) and ControlIgG (open symbols). For the x-axis, the firstmark, 1, represents undiluted IgG,whereas 0 represents no IgG control. Infectedmonocyteswas directly detected by immunofluoresecent stainingwith anti-denguemonoclonal antibody2H2 and percentages of infectedmonocytes detected by 2H2 48 h post infection were shown (b, c). Background staining (monocytes treatedwithMock antigen at each IgG dilution)wassubtracted. The supernatants collected frommonocyte cultures at 24, 48 and 72 h post-infection from the same sets of experiments shown in b and c were tested on L-SIGN Raji cells. Thepercentages of infected L-SIGN Raji cells were shown (d, e). Background staining (themock control)was subtracted. Data for c and ewere a summary of 10 experiments usingmonocytesfrom 10 different subjects. Data for b and d were a summary of 3 experiments. Error bars representing SEM.

DV3

0

3000

6000

9000

1 00

1000

2000

3000

4000

5000

1 0

ControlIgG

DV

gen

ome

RN

A c

opy

#

1:2431:811:271:91:3

IgG dilution

DV1

IgG dilution

1:2431:811:271:91:3

D3IgG

Fig. 3. Assessing the amount of immune complex binding to monocytes using DV RNA copy numbers by the qPCRmethod. Serotype of DVwas indicated. Data is from one experiment torepresent more than 3 experiments. Cells treated with mock antigen and serially diluted IgG showed 0 RNA copies (data not shown).

247P. Sun et al. / Virology 421 (2011) 245–252

0

20

40

60

0

30

60

90

120

150

1 0

0

30

60

90

120

1 0

TN

F-α

pg/

ml

IFN

-α p

g/m

l

IL-1

0 pg

/ml

c

e

b

a

d

0

30

60

90

120

0

30

60

90

120

1:2431:811:271:91:3

IgG dilution

DV1

DV3

IgG dilution

1:2431:811:271:91:3

Fig. 4. Cytokine production by infected monocytes. Monocytes were infected with DV1 or DV3 in the presence of serially diluted D3IgG (close symbols) and ControlIgG (open sym-bols). Supernatants from monocyte cultures were collected for cytokine detection at 3 time points, 24 24 ( ) , 48 48 ( ) and 72 72 ( ) h post infection. X-axis shows serialdilutions of IgG where 1 represents undiluted IgG and 0 represents no IgG control. Data from cells treated with mock antigen and serially diluted IgG was subtracted. Data wassummarized from 3 experiments with error bars representing SEM.

248 P. Sun et al. / Virology 421 (2011) 245–252

the whole assay period (24–72 h post-infection), the highest cyto-kine production was seen at 48 h post-infection. TNF-α and IFN-αproduction was also seen in DV3 ADE-affected monocytes at 48 hpost-infection. No IL-10 was detected in DV3 infected monocytes.No cytokine production was seen with the control IgG.

Costimulatory markers on ADE-affected monocytes

As shown in Figs. 5a and b, CD86 and CD40 were upregulated onmonocytes treated with DV1 and D3IgG after 2 days of infection.The kinetics of CD86 and CD40 expression corresponded to the kinet-ics of infection enhancement (shown in Fig. 2). Cells treated with an-tibodies alone, and cells treated with DV1 and the control IgG did notshow upregulation of these markers.

Discussion

Antibodies elicited by dengue infectionmay play dual roles: blockinginfection through a mechanism of antibody neutralization or enhance

viral infection via a phenomenon of ADE. Characterizing the functionalnature of dengue immune sera has been an ongoing effort for researchand vaccine development. The phenomenon of ADE has been investigat-ed in vitrowith a variety of cell types, THP-1 humanmonocytic leukemiacell line (Chareonsirisuthigul et al., 2007; Diamond et al., 2000; Fust,1997), K562 human erythroleukemia cell line (Guy et al., 2004; Littauaet al., 1990), and primary human dendritic cells (Boonnak et al., 2008).In this study, we chose to use primary bloodmonocytes in an in vitro in-fectionmodel to study ADE.Monocytes are precursors ofmacrophages intissues. Monocytes and monocyte-derived macrophages are scavengingand antigen-presenting cells, and they secrete cytokines, chemokines,growth-regulatingmolecules; they are actively involved inmany inflam-matory diseases, such as autoimmune diseases, bacterial and viral infec-tious diseases, etc. The abundance of monocytes in human peripheralblood (about 30% of human PBMCs) suggested the potential importanceof these cells for dengue infection and disease pathogenesis.

Monocytes are prime targets for DV replication in vivo (Blackley et al.,2007; Jessie et al., 2004; Kangwanpong et al., 1995). Besides FcγR,mono-cytes express mannose receptor, a member of C type lectin, shown to

0

2000

4000

6000

8000

10000

1 00

2000

4000

6000

8000

1 1:3 1:27 0

D3IgG+DV1

D3IgG+Mock

ControlIgG+DV1

ControlIgG+Mock

Geo

met

ric M

ean

Flu

ores

cent

Inte

nsity

a) CD86 b) CD40

1:2431:811:271:91:3

IgG dilution

1:2431:811:9

IgG dilution

Fig. 5.Monocyte activation markers. Monocytes were infected with DV1 in the presence of D3IgG (close symbols) and ControlIgG (open symbols). Cells were collected after 48 h ofinfection for cell surface marker CD40 and CD86 staining. The geometric mean fluorescent intensity of indicated marker was shown. Data represents one of 3 experiments.

249P. Sun et al. / Virology 421 (2011) 245–252

bind all 4 serotypes of DV (Kliks et al., 1989). Another C-type lectin, DC-SIGN (Dendritic Cell-Specific Intercellular adhesionmolecule-3-GrabbingNon-integrin), abundantly expressed on human immature myeloid line-age dendritic cells, is also characterized as a receptor with high bindingaffinity to all 4 serotypes of DV (Tassaneetrithep et al., 2003). For cellsthat express both FcγR and DC-SIGN, such as human dendritic cells, DC-SIGN seem to interfere with FcγR-mediated virus uptake (Boonnaket al., 2008; Marovich et al., 2001). Only when DC-SIGN expression ondendritic cells was reduced, ADE phenomenon appeared. Unlike humandendritic cells,we found that infectingmonocytes through cell surface re-ceptor such as mannose receptors was not efficient, but through FcγRwas muchmore efficient. Using DV1 as an example, among 10 randomlyselected donors, without the presence of antibodies, an average of 1.03%monocytes was infected. In the presence of D3IgG, there was an averageof 14.41% (N14 fold increase) infection. The results suggested that mono-cytes may be poor target cells in primary infections, but they may play aprominent role in secondary infections (Durbin et al., 2008; Kou et al.,2008).

Infection of human monocytes requires the presence of antibodies;therefore, FcγR expressed on the monocyte surface is the key receptormolecule for viral binding and entry in the form of antibody–DV im-mune complex. Since FcγR has high binding affinity to immune com-plexes than to free antibody molecules, we asked if there existed arelationship between antibody concentration and immune complexbinding under a fixed amount of viruses, and how these factors wouldaffect infection enhancement. Immune complex binding wasmeasuredindirectly by quantitating the amount of viruses bound to monocytes.We found that binding of the immune complexes does not necessarilycorrespond to input amount of antibodies, rather, the peak of immunecomplex binding was at suboptimal concentrations of antibodies. Com-paring the infection curves (Fig. 2) with the binding curves (Fig. 3), wefound that they shared a similar trend. However, while binding (Fig. 3b)and infection of DV1 (Figs. 2c and d) seemed to peak at the same anti-body dilution, the binding and infection curves of DV3 appeared to beslightly mismatched: the binding peaked at 1:27 (Fig. 3a) whereas theinfection peaked at 1:81 (Figs. 2b and d). This may reflect the complexnature of serum antibodies. A dengue immune serum is a mixture ofneutralization and enhancement antibodies. Based on several well-documented theories regarding molecular mechanisms of antibodyneutralization (Burton et al., 2000, 2001; Diamond et al., 2008; Klasseand Sattentau, 2002; Oliphant et al., 2006; Pierson et al., 2007), antibodyneutralization can occur at several levels: 1) blocking the virus fromattaching to the receptors; 2) deactivating the virus by targeting to crit-ical functional molecular spots which are essential for the replication

cycle of the virus, for example, binding to the fusion peptide (a se-quence on domain II of dengue E protein), thus hindering the virus–host cell membrane fusion process; and 3) hindering other replicationsteps inside the host cells (Burton et al., 2001). The imperfectmismatchbetween the binding and infection curves of DV3 may suggest that in-fection enhancement is determined not only by the amount of viralbinding, but also by a delicate balance of other serum neutralizationand enhancement activities. Since the immune serum used for thisstudy is from a DV3 immune donor. This serum contains neutralizationantibodies to DV3, but not to DV1 (Fig. 1). Therefore, it is understand-able that binding of DV3 did not immediately result in infection.

The FcγRs are distributed throughout the immune system, particu-larly on effector cells of the innate immune system, such asmacrophage,monocytes, and dendritic cells (Ravetch and Bolland, 2001). The bindingof the immune complexes to FcγRs provides powerful triggering signals,leading to an activation of these effector cells. FcγR play important rolesin the pathophysiology of autoimmune diseases and hypersensitivity(Ravetch and Bolland, 2001). We showed that production of IL-10,TNF-α and IFN-α frommonocyte cultures correlated with the enhance-ment of infection. These cytokines were nearly undetectable when in-fection of monocytes was done without the antibodies. TNF-α is apluripotent vasoactive immunomodulator. DHF diseases have been sug-gested to be a consequence of immunopathology caused by overproduc-tion of inflammatory mediators (Hober et al., 1993; Nguyen et al., 2004;Rothman, 2004). Increased serum and plasma TNF-α in patients withDV infection were reported (Hober et al., 1993; Vitarana et al., 1991).Relatively higher levels of TNF-α have been detected in ethnic Thai pa-tients with DHF, compared with patients with DF. Single nucleotidepolymorphism of TNF gene allele ismore frequently detected in DHFpa-tients of secondary infections (Fernandez-Mestre et al., 2004). Likewise,IFN-α was detected in both DHF and DF patients' sera of Thai children:higher percentages of DHF children had detectable serum IFN-α com-pared to DF children. Taken together, these observations emphasizedthe role of monocytes in secondary dengue infections. Although mono-cytes can be activated during primary infections, theymight be poor cy-tokine secreting cells.

The fundamental mechanism underlying severe dengue clinicalmanifestations (bleeding, thrombocytopenia and shock) is vesicularleakage. It is possible that a complex interaction between virus andhost immune response plays roles to impact the vesicular endothelialbarrier integrity, leading to plasma leakage (Mathew and Rothman,2008; Srikiatkhachorn, 2009). Endothelial cell permeability in the con-text of dengue infection has been studied in vitro using human umbilicalvein endothelial cells (HUVEC) (Dewi et al., 2004). TNF-α at 1 and

250 P. Sun et al. / Virology 421 (2011) 245–252

0.1 μg/ml increased HUVAC permeability, but TNF-α at lower dose didnot. However, another study showed that high dose of TNF-α alonedid not affect HUVAC cell permeability; TNF-α in combination with cy-tokines produced by T cells played a role on HUVAC cell permeability(Seynhaeve et al., 2006). Our future effort will be focused on under-standing how cytokines and cellular immune responses impact endo-thelial cell permeability and vesicular leakage.

In addition, we examined monocyte activation and found that co-stimulatory markers which are important for T cell stimulation wereup-regulated on ADE-affected monocytes. A study aimed to define pri-mary cell types permissible to DV infection within PBMCs of endemicchildren naturally infectedwith dengue found that activatedmonocytes(CD14+CD11c+CD32+CD86+CD4−CD8−) were the primary celltype in which DV antigen was detected. A higher percentage of PBMCsfrom DHF patients co-expressed DV antigen, CD86, CD32, and CD11cthan did those from DF patients, indicating that increased activation ofmonocytes and greater numbers of DV-infected cells were associatedwith more severe dengue illnesses (Durbin et al., 2008). Our in vitrodata showing ADE-affected monocytes have elevated CD86 and CD40expression not only supported this study, but also provided explana-tion: increased infection through ADE caused CD86 upregulation andcell activation. CD86 and CD40 are costimulatory molecules expressedmainly on antigen-presenting cells such as monocytes, macrophages,dendritic cells, and B cells. Inflammatory cytokines, TNF-α and IFN-α,can act synergistically with antigens to promote the expression ofthese costimulatory markers (Steinbrink et al., 2000). Therefore, it ispossible that higher levels of CD40 and CD86 expression in ADE-affect-ed monocytes maybe a consequence of increased exposure to both ofthe antigens and inflammatory cytokines. Further, CD40 and CD86 arecostimulatory markers important for T cell activation. Higher levels ofsoluble markers of cellular immunity including TNF-α, IL-2 receptorsand CD8 were found in sera of patients with more severe disease;higher levels of activated DV-specific CD8+ T cells were found in pa-tients of more severe disease; increased production of both Th1 andTh2 cytokines were found during acute disease (Libraty et al., 2002;Rothman, 2003). Thesefindings suggest that T cells play a role in diseasepathogenesis. Increased expression of cytokine and costimulatorymarkers on ADE-affected monocytes may increase T cell activation,thus contribute to disease severity.

In conclusion, monocytes can be more efficiently infected throughFcγR-dependent mechanism than through cell-receptor such as man-nose receptors. In response to ADE, monocytes produced high levelsof IFN-α and TNF-α and expressed higher co-stimulatory markers. Wethink that they play a more prominent pathological role in secondaryinfections than in primary infections. On the other hand, pre-existingserum antibodies can block viruses, protecting FcγR-bearing cellsfrom DV infection. Future investigations into how to augment the anti-bodies-neutralization of FcγR-bearing cells, by serum complements orbyusing syntheticmolecules that can change the course of phagocytosisand its down-stream cellular signaling cascades, are the keys for mini-mizing immune enhancement.

Materials and methods

Virus preparation

DV1 strain West Pac 74 and DV3 strain CH53489 were grown andpropagated in mycoplasma-free Vero cell culture. The viral titers weredetermined by limiting-dilution plaque assay on Vero cells. If not specif-ically stated, both viruses were used at titers of 3×106 PFU/ml. Theinfecting MOI at this concentration was close to 1.

Medium

Sterile Complete Lymphocyte Medium was prepared as following:RMPI 1640 w/o L-Glutamine (Mediatech) supplemented with 1% of

Penicillin 10,000 IU/ml and Streptomycin 10,000 μg/ml (Invitrogen),10% Heat-inactivated Fetal Bovine Serum (Quality Biologics), 1%L-glutamine 200 mM in 0.85% NaCl, and 1% Non-essential AminoAcid (Mediatech).

Monocyte preparation

Human peripheral blood from unidentified healthy donors wasobtained under an approved human use protocol NMRC.2007-0009. Pe-ripheral blood mononuclear cells (PBMCs) were isolated using Ficoll-Hypaque (GE Healthcare) density gradient centrifugation. CD14+monocytes were positively selected fromwhole PBMCs using magneticmicrobeads for CD14 isolation (Miltenyi Biotec). After isolation, CD14+cells were placed overnight in 96-well plate (2×105 cells/well) at 37 °Cin a 5% CO2 incubator before being used for ADE experiments.

Purification of serum IgG

The IgGwas purified from a DV3-immune serum and a dengue non-immune serum (obtained through the approved human use protocolNMRC.2007.2009) using IgG purification columns and Protein Gcoupled agarose. Briefly, 15 ml columonocytess (Pierce) were packedwith 5 ml of Immobilized protein G plus (Thermo Fisher Scientific).The human serum was diluted 1:2 with Binding/Wash buffer (Pierce)and allowed to bind to the Protein G during 2 h incubation period atroom temperature. Unbound substances were washed away with15 ml of wash buffer. The bound IgG was eluted with 15 ml of gentleelution buffer. Purified IgG was first dialyzed in Tris-buffered saline for24 h, to be followed by another 24 h dialysis in Phosphate-buffered sa-line using 15–30 ml slide-a-lyzer dialysis cassettes (Pierce). The dia-lyzed IgG was then concentrated by utilizing Millipore concentrationcolumonocytess. The concentration was recorded using spectropho-tometer reading at 280 nm.

ELISA and antibody neutralization assay

The binding and the neutralization activity of the D3IgG to DV3 andDV1weremeasured as previously described (Kochel et al., 2000;Martinet al., 2006). Briefly, 96-well ELISA plate was coated with DV1 andDV3 respectively and serially diluted D3IgG was added. The plate wasdeveloped following common procedures of ELISA and the bindingwas scaled by optical density (O.D.) values. For neutralization assay,30 μl/well of serially diluted D3IgG was cocultured with 30 μl/well ofDV at 37 °C. After 30 min incubation, 60 μl/well of liver/lymph node-spe-cific intercellular adhesion molecule-3-grabbing integrin (L-SIGN)transfected Raji cells were added to the plate which was then placedinto the 37 °C 5% incubator for 20 h. Infection was determined byintracellular immuno-fluorescent staining using a dengue group specificmonoclonal antibody, 2H2. The neutralization activity was shown asreduction of infection at each D3IgG concentration on the bases of aninfection control which was set as infection without the interference ofany serum.

ADE experiment setup

DV was co-incubated with serially diluted D3IgG at a ratio of 20 μlD3IgG and 80 μl of virus for 30 min at 37 °C in a 5% CO2 incubator allow-ing the formation of immune complexes. Mock controls were set up thesameway, but Vero cell supernatant containing no viruses were used toreplace DV. Following the incubation period, 100 μl of the DV (orMock)and antibody mixtures was transferred onto monocytes and incubatedfor 3 h at 37 °C in a 5% CO2 incubator. Cells were washed twice withplain RPMI 1640 and once with complete growth media (CM). At theend of 3 washes, 200 μl of CM was added to the plate and the cultureswere placed in 37 °C 5% CO2 incubator for 3 days. Monocyte cultures

251P. Sun et al. / Virology 421 (2011) 245–252

and supernatants were collected at times for assessment of infectivityand cytokines.

Detection of DV-infection in monocytes and in culture supernatants

Viral infectionwasmeasured directly inmonocytes and inmonocyteculture supernatants. Briefly,monocytes were harvested, fixed and per-meabilized using a Perm/Fix kit (BD). After staining the cells intracellu-larlywith fluorescent 2H2 for 30 min, percentage of infectedmonocyteswas determined by immuno-flow cytometry. Infectious progeny virionsthat were produced by infected monocytes into the culture superna-tants were quantified using a Raji cell line expressing L-SIGNmolecules.In a 96-U-bottomed plate, 50 μl of monocyte culture supernatants wascultured with 50 μl L-SIGN Raji cells containing 5×105 cell counts.After 24 h of incubation for DV1 supernatants and 48 h for DV3 super-natants, percentage of infected L-SIGN cells was determined by intracel-lular staining of 2H2 antibodies and immuno-flow cytometry.

Binding assay

Binding of DVwas performed as previously described (Takhampunyaet al., 2009). Briefly, D3IgG was co-incubated with DV for 30 min at 4 °C.Before the addition of D3IgG–DV mixtures, monocytes were chilled onice for 30 min. Cells were incubated with D3IgG–DV mixture for 1.5 hat 0 °C (binding assay) in serum-free RPMI 1640 medium containingminimumessential amino acids (1%), L-glutamine (1%), sodiumpyruvate(1%), and Penicillin/Streptomycin (1%). Binding of DV to monocytes wasquantitated by qPCR and results were presented as DV RNA copy #.

Quantitative PCR for measurement of DV positive sense RNA copies

The positive sense RNA copy numberswere determined by Quantita-tive PCR using reaction conditions modified from a PCR assay describedpreviously (Callahan et al., 2001). Briefly, the following primer sequenceswere used:

Forward primer: AAG GAC TAG AGG TTA KAG GAG ACC CReverse primer: GGC GYT CTG TGC CTG GAW TGA TGFlouregenic probe: FAM– AAC AGC ATA TTG ACG CTG GGA GAGACC –TAMRA

This probe recognizes Dengue serotypes 1 and 3 only. The probeswere labeledwith FAM as the flourophore and TAMRA as the quencher.The reactionwas carried out using a TaqMan EZ-RT PCR kit with the fol-lowing final reagent concentrations: 1× TaqMan Buffer, 0.1 Units/μl ofRecombinant Tth DNA Polymerase, 0.1 mM of dNTP's, 0.15 μM probe,0.75 μMof forward primer, 0.75 μMof reverse primer, 115 mMof Potas-sium Chloride (KCl) and 3 mM of Manganese Acetate. The assay wasrun on an ABI 7000 (Applied Biosystems) using the following cyclingconditions: One 30 minute reverse transcription step at 60 °C followedby 45 cycles of a 95 °C denaturation step lasting 15 s and a 60 °C exten-sion step lasting 30 s.We used a housekeeping geneβ-actin as a loadingcontrol in every experiment (Pfaffl, 2001).

Cytokine production in culture supernatants

IFN-αwas measured by ELISA, according to the manufacturer's rec-ommendations (PBL Biomedical Laboratories). Samples were read on aSpectraMax 340 plate reader (Molecular Devices). All other cytokineswere quantified by BD inflammatory cytometric bead array (CBA) kiton FACScan or FACSArray using CBA software (BD Pharmingen).

Expression of co-stimulatory markers

Monocyte co-stimulatory marker expression was determined usingmonoclonal antibodies, CD86 (BD), CD40 (Beckman Coulter) and HLA-

A, B, C (BD). Staining of these markers was performed after 48 h postinfection.

Acknowledgment

This work was supported by Military Infectious Disease ResearchProgram (MIDRP) funding #: S0165_08_NM.

We thank Dr. Dauner Allison for her generous consultation on theqPCR work.

References

Blackley, S., Kou, Z., Chen, H., Quinn, M., Rose, R.C., Schlesinger, J.J., Coppage, M., Jin, X.,2007. Primary human splenic macrophages, but not T or B cells, are the principaltarget cells for dengue virus infection in vitro. J. Virol. 81 (24), 13325–13334.

Boonnak, K., Slike, B.M., Burgess, T.H., Mason, R.M., Wu, S.J., Sun, P., Porter, K., Rudiman,I.F., Yuwono, D., Puthavathana, P., Marovich, M.A., 2008. Role of dendritic cells inantibody-dependent enhancement of dengue virus infection. J. Virol. 82 (8),3939–3951.

Burton, D.R., Saphire, E.O., Parren, P.W., 2001. A model for neutralization of virusesbased on antibody coating of the virion surface. Curr. Top. Microbiol. Immunol.260, 109–143.

Burton, D.R., Williamson, R.A., Parren, P.W., 2000. Antibody and virus: binding and neu-tralization. Virology 270 (1), 1–3.

Callahan, J.D., Wu, S.J., Dion-Schultz, A., Mangold, B.E., Peruski, L.F., Watts, D.M., Porter,K.R., Murphy, G.R., Suharyono, W., King, C.C., Hayes, C.G., Temenak, J.J., 2001. Devel-opment and evaluation of serotype- and group-specific fluorogenic reverse tran-scriptase PCR (TaqMan) assays for dengue virus. J. Clin. Microbiol. 39 (11),4119–4124.

Chareonsirisuthigul, T., Kalayanarooj, S., Ubol, S., 2007. Dengue virus (DENV) antibody-dependent enhancement of infection upregulates the production of anti-inflammatorycytokines, but suppresses anti-DENV free radical and pro-inflammatory cytokine pro-duction, in THP-1 cells. J. Gen. Virol. 88 (Pt 2), 365–375.

Dewi, B.E., Takasaki, T., Kurane, I., 2004. In vitro assessment of human endothelial cellpermeability: effects of inflammatory cytokines and dengue virus infection. J. Virol.Methods 121 (2), 171–180.

Diamond, M.S., Edgil, D., Roberts, T.G., Lu, B., Harris, E., 2000. Infection of human cellsby dengue virus is modulated by different cell types and viral strains. J. Virol. 74(17), 7814–7823.

Diamond, M.S., Pierson, T.C., Fremont, D.H., 2008. The structural immunology of anti-body protection against West Nile virus. Immunol. Rev. 225, 212–225.

Durbin, A.P., Vargas,M.J.,Wanionek, K., Hammond, S.N., Gordon, A., Rocha, C., Balmaseda, A.,Harris, E., 2008. Phenotyping of peripheral blood mononuclear cells during acute den-gue illness demonstrates infection and increased activation of monocytes in severecases compared to classic dengue fever. Virology 376 (2), 429–435.

Fernandez-Mestre, M.T., Gendzekhadze, K., Rivas-Vetencourt, P., Layrisse, Z., 2004.TNF-alpha-308A allele, a possible severity risk factor of hemorrhagic manifestationin dengue fever patients. Tissue Antigens 64 (4), 469–472.

Fust, G., 1997. Enhancing antibodies in HIV infection. Parasitology 115, S127–S140Suppl..

Goncalvez, A.P., Engle, R.E., St Claire, M., Purcell, R.H., Lai, C.J., 2007. Monoclonal anti-body-mediated enhancement of dengue virus infection in vitro and in vivo andstrategies for prevention. Proc. Natl. Acad. Sci. U.S.A. 104 (22), 9422–9427.

Gubler, D.J., 1989. Surveillance for dengue and dengue hemorrhagic fever. Bull. Pan Am.Health Organ. 23 (4), 397–404.

Guy, B., Chanthavanich, P., Gimenez, S., Sirivichayakul, C., Sabchareon, A., Begue, S.,Yoksan, S., Luxemburger, C., Lang, J., 2004. Evaluation by flow cytometry of anti-body-dependent enhancement (ADE) of dengue infection by sera from Thai chil-dren immunized with a live-attenuated tetravalent dengue vaccine. Vaccine 22(27–28), 3563–3574.

Guzman, M.G., Kouri, G., Valdes, L., Bravo, J., Vazquez, S., Halstead, S.B., 2002. Enhancedseverity of secondary dengue-2 infections: death rates in 1981 and 1997 Cubanoutbreaks. Rev. Panam. Salud Publica 11 (4), 223–227.

Halstead, S.B., 1988. Pathogenesis of dengue: challenges to molecular biology. Science239 (4839), 476–481.

Halstead, S.B., Heinz, F.X., Barrett, A.D., Roehrig, J.T., 2005. Dengue virus: molecularbasis of cell entry and pathogenesis, 25–27 June 2003, Vienna, Austria. Vaccine23 (7), 849–856.

Hober, D., Poli, L., Roblin, B., Gestas, P., Chungue, E., Granic, G., Imbert, P., Pecarere, J.L.,Vergez-Pascal, R., Wattre, P., et al., 1993. Serum levels of tumor necrosis factor-alpha (TNF-alpha), interleukin-6 (IL-6), and interleukin-1 beta (IL-1 beta) in den-gue-infected patients. Am. J. Trop. Med. Hyg. 48 (3), 324–331.

Jessie, K., Fong, M.Y., Devi, S., Lam, S.K., Wong, K.T., 2004. Localization of dengue virus innaturally infected human tissues, by immunohistochemistry and in situ hybridiza-tion. J. Infect. Dis. 189 (8), 1411–1418.

Kangwanpong, D., Bhamarapravati, N., Lucia, H.L., 1995. Diagnosing dengue virus infec-tion in archived autopsy tissues by means of the in situ PCR method: a case report.Clin. Diagn. Virol. 3 (2), 165–172.

Klasse, P.J., Sattentau, Q.J., 2002. Occupancy andmechanism in antibody-mediated neu-tralization of animal viruses. J. Gen. Virol. 83 (Pt 9), 2091–2108.

Kliks, S.C., Nimmanitya, S., Nisalak, A., Burke, D.S., 1988. Evidence that maternal dengueantibodies are important in the development of dengue hemorrhagic fever in in-fants. Am. J. Trop. Med. Hyg. 38 (2), 411–419.

252 P. Sun et al. / Virology 421 (2011) 245–252

Kliks, S.C., Nisalak, A., Brandt, W.E., Wahl, L., Burke, D.S., 1989. Antibody-dependent en-hancement of dengue virus growth in human monocytes as a risk factor for denguehemorrhagic fever. Am. J. Trop. Med. Hyg. 40 (4), 444–451.

Kochel, T.J., Raviprakash, K., Hayes, C.G., Watts, D.M., Russell, K.L., Gozalo, A.S., Phillips,I.A., Ewing, D.F., Murphy, G.S., Porter, K.R., 2000. A dengue virus serotype-1 DNAvaccine induces virus neutralizing antibodies and provides protection from viralchallenge in Aotus monkeys. Vaccine 18 (27), 3166–3173.

Kou, Z., Quinn, M., Chen, H., Rodrigo, W.W., Rose, R.C., Schlesinger, J.J., Jin, X., 2008. Mono-cytes, but not T or B cells, are the principal target cells for dengue virus (DV) infectionamong human peripheral blood mononuclear cells. J. Med. Virol. 80 (1), 134–146.

Libraty, D.H., Endy, T.P., Houng, H.S., Green, S., Kalayanarooj, S., Suntayakorn, S.,Chansiriwongs, W., Vaughn, D.W., Nisalak, A., Ennis, F.A., Rothman, A.L., 2002.Differing influences of virus burden and immune activation on disease severityin secondary dengue-3 virus infections. J. Infect. Dis. 185 (9), 1213–1221.

Littaua, R., Kurane, I., Ennis, F.A., 1990.Human IgG Fc receptor IImediates antibody-depen-dent enhancement of dengue virus infection. J. Immunol. 144 (8), 3183–3186.

Marovich, M., Grouard-Vogel, G., Louder, M., Eller, M., Sun, W., Wu, S.J., Putvatana, R.,Murphy, G., Tassaneetrithep, B., Burgess, T., Birx, D., Hayes, C., Schlesinger-Frankel,S., Mascola, J., 2001. Human dendritic cells as targets of dengue virus infection. J.Investig. Dermatol. Symp. Proc. 6 (3), 219–224.

Martin, N.C., Pardo, J., Simmons, M., Tjaden, J.A., Widjaja, S., Marovich, M.A., Sun, W.,Porter, K.R., Burgess, T.H., 2006. An immunocytometric assay based on dengue in-fection via DC-SIGN permits rapid measurement of anti-dengue neutralizing anti-bodies. J. Virol. Methods 134 (1–2), 74–85.

Mathew, A., Rothman, A.L., 2008. Understanding the contribution of cellular immunityto dengue disease pathogenesis. Immunol. Rev. 225, 300–313.

Moi, M.L., Lim, C.K., Takasaki, T., Kurane, I., 2010. Involvement of the Fc gamma receptorIIA cytoplasmic domain in antibody-dependent enhancement of dengue virus in-fection. J. Gen. Virol. 91 (Pt 1), 103–111.

Nguyen, T.H., Lei, H.Y., Nguyen, T.L., Lin, Y.S., Huang, K.J., Le, B.L., Lin, C.F., Yeh, T.M., Do, Q.H.,Vu, T.Q., Chen, L.C., Huang, J.H., Lam, T.M., Liu, C.C., Halstead, S.B., 2004. Dengue hem-orrhagic fever in infants: a study of clinical and cytokine profiles. J. Infect. Dis. 189 (2),221–232.

Oliphant, T., Nybakken, G.E., Engle, M., Xu, Q., Nelson, C.A., Sukupolvi-Petty, S., Marri, A.,Lachmi, B.E., Olshevsky, U., Fremont, D.H., Pierson, T.C., Diamond, M.S., 2006. Anti-body recognition and neutralization determinants on domains I and II of West NileVirus envelope protein. J. Virol. 80 (24), 12149–12159.

Pfaffl, M.W., 2001. A new mathematical model for relative quantification in real-timeRT-PCR. Nucleic Acids Res. 29 (9), e45.

Pierson, T.C., Xu, Q., Nelson, S., Oliphant, T., Nybakken, G.E., Fremont, D.H., Diamond,M.S., 2007. The stoichiometry of antibody-mediated neutralization and enhance-ment of West Nile virus infection. Cell Host Microbe 1 (2), 135–145.

Ravetch, J.V., Bolland, S., 2001. IgG Fc receptors. Annu. Rev. Immunol. 19, 275–290.Rothman, A.L., 2003. Immunology and immunopathogenesis of dengue disease. Adv.

Virus Res. 60, 397–419.Rothman, A.L., 2004. Dengue: defining protective versus pathologic immunity. J. Clin.

Invest. 113 (7), 946–951.Russell, P.K., Udomsakdi, S., Halstead, S.B., 1967. Antibody response in dengue and den-

gue hemorrhagic fever. Jpn. J. Med. Sci. Biol. 20, 103–108 Suppl..Seynhaeve, A.L., Vermeulen, C.E., Eggermont, A.M., ten Hagen, T.L., 2006. Cytokines and

vascular permeability: an in vitro study on human endothelial cells in relation totumor necrosis factor-alpha-primed peripheral blood mononuclear cells. Cell Bio-chem. Biophys. 44 (1), 157–169.

Srikiatkhachorn, A., 2009. Plasma leakage in dengue haemorrhagic fever. Thromb. Hae-most. 102 (6), 1042–1049.

Srikiatkhachorn, A., Krautrachue, A., Ratanaprakarn, W., Wongtapradit, L., Nithipanya,N., Kalayanarooj, S., Nisalak, A., Thomas, S.J., Gibbons, R.V., Mammen Jr., M.P.,Libraty, D.H., Ennis, F.A., Rothman, A.L., Green, S., 2007. Natural history of plasmaleakage in dengue hemorrhagic fever: a serial ultrasonographic study. Pediatr. In-fect. Dis. J. 26 (4), 283–290 discussion 291–2.

Steinbrink, K., Paragnik, L., Jonuleit, H., Tuting, T., Knop, J., Enk, A.H., 2000. Induction ofdendritic cell maturation and modulation of dendritic cell-induced immune re-sponses by prostaglandins. Arch. Dermatol. Res. 292 (9), 437–445.

Takhampunya, R., Palmer, D.R., McClain, S., Barvir, D.A., Lynch, J., Jarman, R.G., Thomas,S., Gibbons, R.V., Burgess, T.H., Sun, P., Kamau, E., Putnak, R., Zhang, C., 2009. Phe-notypic analysis of dengue virus isolates associated with dengue fever and denguehemorrhagic fever for cellular attachment, replication and interferon signalingability. Virus Res. 145 (1), 31–38.

Tassaneetrithep, B., Burgess, T.H., Granelli-Piperno, A., Trumpfheller, C., Finke, J., Sun,W., Eller, M.A., Pattanapanyasat, K., Sarasombath, S., Birx, D.L., Steinman, R.M.,Schlesinger, S., Marovich, M.A., 2003. DC-SIGN (CD209) mediates dengue virus in-fection of human dendritic cells. J. Exp. Med. 197 (7), 823–829.

Vaughn, D.W., Green, S., Kalayanarooj, S., Innis, B.L., Nimmannitya, S., Suntayakorn, S.,Endy, T.P., Raengsakulrach, B., Rothman, A.L., Ennis, F.A., Nisalak, A., 2000. Dengueviremia titer, antibody response pattern, and virus serotype correlate with diseaseseverity. J. Infect. Dis. 181 (1), 2–9.

Vitarana, T., de Silva, H.,Withana, N., Gunasekera, C., 1991. Elevated tumour necrosis factorin dengue fever and dengue haemorrhagic fever. Ceylon Med. J. 36 (2), 63–65.