Cultured mononuclear peripheral blood leukocytes (PBL) from nonimmune human beings and monkeys are nonpermissive to dengue 2 virus (D2V) infection at multiplicities of infection of 0.001-0.1, but become permissive when non-neutralizing dengue antibody is added to medium. D2V infection occurred in PBL prepared from anti-coagulated but not from defibrinated plasma. Infection enhancement was produced by multiple lots of heterotypic anti-dengue raised in several mammalian species. Homotypic anti-dengue neutralized D2V at high concentrations but enhanced at low concentrations; enhancement end point in one serum was 1:320,000. The infection-enhancing factor was a noncytophilic antibody of the IgG class. D2V infection occurred in the absence of heat-labile complement components but did not occur when complexes were prepared with anti- dengue F(ab)(2). Treatment of PBL with several proteases increased permissiveness to D2V infection by immune complexes but not by virus alone. Two rhesus monkey serums collected 14 days after D2V infection contained an IgG antibody with high-titered enhancing activity but with no hemagglutination-inhibition or neutralizing activity. Virus-antibody complexes are irreversibly attached to PBL within 15 min and completely internalized in 60 min. There was considerable variation in cellular infection in different experiments, however, maximum virus yields usually exceeded 1,000 plaque-forming units per 1 x 10(6) PBL occurring between 2 and 4 days in culture. In vitro antibody-dependent infection of PBL provides a possible model for study of pathogenetic mechanisms in infants with dengue shock syndrome who passively acquire maternal anti-dengue IgG.

Dengue viruses occur as four antigenically related but distinct serotypes transmitted to humans by Aedes aegypti mosquitoes. These viruses generally cause a benign syndrome, dengue fever, in the American and African tropics, and a severe syndrome, dengue hemorrhagic fever/dengue shock syndrome (DHF/DSS), in Southeast Asian children. This severe syndrome, which recently has also been identified in children infected with the virus in Puerto Rico, is characterized by increased vascular permeability and abnormal hemostasis. It occurs in infants less than 1 year of age born to dengue-immune mothers and in children 1 year and older who are immune to one serotype of dengue virus and are experiencing infection with a second serotype. Dengue viruses replicate in cells of mononuclear phagocyte lineage, and subneutralizing concentrations of dengue antibody enhance dengue virus infection in these cells. This antibody-dependent enhancement of infection regulates dengue disease in human beings, although disease severity may also be controlled genetically, possibly by permitting and restricting the growth of virus in monocytes. Monoclonal antibodies show heterogeneous distribution of antigenic epitopes on dengue viruses. These epitopes serve to regulate disease: when antibodies to shared antigens partially neutralize heterotypic virus, infection and disease are dampened; enhancing antibodies alone result in heightened disease response. Further knowledge of the structure of dengue genomes should permit rapid advances in understanding the pathogenetic mechanisms of dengue.

In the previous paper we demonstrated for peripheral blood leukocytes (PBL) 1 from nonimmune primate donors the dependence of dengue 2 virus (D2V) permissiveness upon the presence of non-neutralizing anti-dengue IgG in culture medium (1). Infection occurs in the absence of complement and is dependent upon a protease-resistant leukocyte Fc receptor. When the infectious dose is at or above a multiplicity of infection (MOI) of 0.001 and mononuclear PBL separated from anti-coagulated blood are standardized to 1 × 10e/ml, antibodydependent D2V infection is highly reproducible. The significance of antibodyenhanced infection as a possible mechanism in the pathogenesis of severe dengue infections in infants born in dengue endemic areas was discussed. In this paper we report studies on the identification and the enumeration of human and simian blood and simian tissue leukocytes supporting antibody-enhanced dengue infection, in vitro. Materials and MethodsPreparation of virus, antibody, separation of leukocytes, and methods of leukocyte propagation are as previously described (1). Unless otherwise specified, PBL were infected with D2V at a MOI between 0.1 and 0.01 in a final concentration of 1:200 of H-187 anti-dengue 4 (aD4) and were grown in RPMI-1640 medium supplemented with 10% heat-inactivated fetal bovine serum (FBS), glutamine, and antibiotics (complete RPMI).Leukocyte Counts and Identification. Leukocyte suspensions diluted in trypan blue and in leukocyte diluent were counted in a hemacytometer. Viable cells were those which excluded 0.2% trypan blue after 5 rain incubation at 37°C. Methods for counting monocytes and sheep erythrocyte (SRBC)-rosetted (T) and membrane immunoglobulin-bearing (B) lymphocytes are described separately. 2 Briefly, 1 x 105 PBL in 0.1 ml medium were incubated with an equal volume of 0.5% neuraminidase-treated SRBC at 37°C for 5 min, then centrifuged at 100 g for 5 rain, and incubated

Abstract. A prospective study on dengue (DEN) viruses was initiated in October 1995 in Gondokusuman kecamatan, Yogyakarta, Indonesia. This report presents data from the first year of the study. The studied cohort included all children 4-9 years of age living in the kecamatan. Blood samples for serology were collected from 1,837 children in October 1995 and again in October 1996. Blood samples for virus isolation and serology were collected from cohort children who were seen in municipal health clinics with febrile syndromes or admitted to hospitals with a provisional diagnosis of dengue hemorrhagic fever. Dengue serotype antibody prevalence and 1995-1996 infection rates were calculated using a single dilution (1:60) 70% plaque reduction endpoint neutralization test. Prevalence of dengue antibody at the beginning of the study was DEN 1 ϭ 12%, DEN 2 ϭ 16%, DEN 3 ϭ 2%, DEN 4 ϭ 4%, and two or more dengue infections ϭ 22%. Total dengue antibody prevalence increased from 38% in 4-year-old children to 69% in 9-year-old children. During the observation period, primary dengue infection rates were DEN 1 ϭ 4.8%, DEN 2 ϭ 7.7%, DEN 3 ϭ 4.2%, and DEN 4 ϭ 3.4%, while two or more dengue infections occurred in 6.7% of the study population. The secondary dengue infection rate was 19.0%. From febrile cases, all four dengue viruses were isolated with DEN 3 predominating. Seven children were hospitalized, including one fatal case with a hospital diagnosis of dengue shock syndrome. Based upon presence of antibody in the initial cohort bleeding and the serologic response both weeks and several months following illness, all had secondary dengue infections. Neutralizing antibody patterns in the initial cohort bleeding and in late convalescent serum samples permitted recognition of dengue infection sequence in five patients: DEN 2-DEN 1 (3), DEN 2-DEN 4 (1), DEN 1-DEN 3 (1), and none in the sequence DEN 1-DEN 2. In the total cohort 6.5% of the observed secondary infections were of the sequence DEN 2-DEN 1, while 4.9% were DEN 1-DEN 2, a highly pathogenic sequence in previous studies. Reduced pathogenic expression of secondary DEN 2 with enhanced pathogenic expression of secondary DEN 1 infections was an unexpected finding. Further studies will be required to understand the respective contributions to pathogenicity of antibody from initial dengue infections versus the biological attributes of the second infecting dengue viruses.The dengue (DEN) viruses (serotypes 1, 2, 3, and 4) are transmitted in nearly all tropical countries with a total population at risk in excess of 2.5 billion people.

Although its underlying mechanisms are poorly understood, data comparing each of the four dengue virus serotypes suggest that in vitro antibody-dependent infection enhancement is a reproducible and measurable phenomenon related to other serological measures of antibody-virus binding. Information characterizing infection enhancement may provide clues to disease pathogenesis for dengue and other viruses that exhibit antibody-enhanced infection. We propose criteria for the detection and quantification of in vitro antibodydependent enhancement of flavivirus infection based on observations using all four dengue virus serotypes, macrophage-like cell lines and human peripheral blood monocytes, and various immune sera and monoclonal antibodies. It is proposed that antibody-dependent infection enhancement is defined by the following findings: (i) significantly increased virus production is measured in quantitative assays at different points on the growth curve; (ii) assays of the virus output of cells infected with mixtures of constant amounts of virus and serial dilutions of the pre-existing antibody source produce characteristic 'enhancement profiles' of rising and falling virus output over at least a 10-3-fold dilution range; (iii) for each enhancing antibody source the dilution producing maximal infection enhancement is related to other serological measures of binding to the envelope, or another virus component; (iv) infection enhancement is detected with different antibody sources and virus strains (when available) tested over a range of m.o.i.; (v) other causes of enhanced virus production are ruled out.