Flaviviruses can mediate fusion from without in Aedes albopictus mosquito cell cultures.

Flavivirus-induced polykaryocytes were detected in monolayers of Aedes albopictus (clone C6/36) mosquito cells as early as 20 min after adsorbing virus to these cells. A high multiplicity of infection with dengue (DEN)-1, 2, 3, 4, Japanese encephalitis, and yellow fever viruses was required to demonstrate fusion from without (FFWO) with these flaviviruses. Optimal conditions for FFWO included exposure of adsorbed virus to pH 6.0 and an incubation temperature of 39 degrees C. DEN-2 monoclonal antibodies to the envelope E glycoprotein inhibited cell fusion, whereas monoclonal antibodies to the prM and NS1 proteins did not inhibit cell fusion. These results indicate that flaviviruses cause FFWO soon after adsorption to C6/36 mosquito cells and the process is most likely mediated by the virion envelope E glycoprotein.

Expression of the envelope antigen of dengue virus in vaccine strains of Salmonella.

The envelope gene of dengue 4 virus (DEN) was cloned in a plasmid under the control of Escherichia coli expression signals. A clone that expressed 93% of the gene was found to be detrimental to the bacterial host. Another clone which carried only 76% of the E gene was found to be quite stable in vitro as well as in vivo. The killed recombinant bacteria induced antibodies in mice which recognized native DEN virus. Attenuated Salmonella typhimurium (SAL) strains carrying the DEN-E plasmid were tested for their efficacy as orally administered live vaccines. Protective immunization was assessed in a mouse model by immunizing three-week old BALB/c mice followed by challenge with DEN virus. It was found that these young mice were highly susceptible to the carrier SAL strains (M206 and aroA SL3261). Moreover, the SAL-infected mice were more susceptible to DEN virus challenge than control mice, suggesting that the SAL infection caused immunosuppression in these young mice.

Dengue 3 virus infection of Aedes albopictus and Aedes aegypti: comparison of parent and progeny candidate vaccine viruses.

DEN-3 parent (strain CH53489) and progeny candidate vaccine (clone 24/28) viruses were compared in their abilities to interact with Aedes aegypti and Ae. albopictus. The parent and progeny virus were equivalent in their ability to infect, to replicate in, and to be transmitted by both species of mosquitoes. The candidate vaccine DEN-3 clone was temperature sensitive and had small plaque morphology. These phenotypic markers remained stable during mosquito passage. Thus, temperature sensitivity and small plaque morphology are not reliable biological markers for attenuation.

Infection of Aedes albopictus and Aedes aegypti mosquitoes with dengue parent and progeny candidate vaccine viruses: a possible marker of human attenuation.

Dengue (DEN-1) and DEN-4 parent (P) and progeny candidate vaccine (CV) viruses were compared in their abilities to infect and to replicate in Aedes aegypti and Aedes albopictus mosquitoes. The DEN CV clones were temperature sensitive (ts) and had small plaque morphology. The DEN-1 and DEN-4 CV viruses differed in their ability to infect, to replicate in, and to be transmitted by mosquitoes. The DEN-1 CV virus was not attenuated for the vector mosquitoes; oral infection rates with the CV virus were as high as or higher than the P virus, and the CV virus replicated efficiently in mosquitoes after oral infection. The DEN-4 CV virus was attenuated; it was less efficient than its P virus in infection and replication in mosquitoes. Thus, the ts phenotype and small plaque morphology are not reliable biological markers for prediction of vector attenuation. Similar results were reported by others for attenuation in man and monkeys. These studies with DEN-1 and DEN-4 viruses, and previously reported studies with DEN-2 virus and with DEN-3 virus suggest that vector and vertebrate host attenuation are genetically linked. Thus, vector attenuation may be a biological marker for human attenuation.

Monoclonal antibodies against dengue 2 virus E-glycoprotein protect mice against lethal dengue infection.

A panel of 11 murine monoclonal antibodies directed against dengue type 2 was evaluated for antigen specificity by dot immunobinding assay and Western blot analysis and for in vitro and in vivo biological activities. Nine of the 11 monoclonal antibodies reacted with viral E-glycoprotein based on the Western blot analysis; one reacted with a 36 Kd protein present in dengue-infected C6/36 mosquito cells. The nine E-glycoprotein-reactive monoclonal antibodies also neutralized dengue 2 virus in a plaque reduction assay. Of the neutralizing monoclonal antibodies, five passively protected mice in vivo against lethal intracerebral dengue 2 challenge. The protective monoclonal antibodies were directed against viral determinants that fell into at least three spatially separate families of epitopes on E-glycoprotein, the antigenicities of which were preserved after heat/detergent denaturation.

Selection of attenuated dengue 4 viruses by serial passage in primary kidney cells. IV. Characterization of a vaccine candidate in fetal rhesus lung cells.

A strain of primary dog kidney (PDK)-passaged dengue (DEN) 4 (H-241) virus cloned by terminal dilution (PDK 35-TD3) was propagated in large volumes in fetal rhesus lung (FRhL) cells to produce a candidate vaccine for evaluation in man. Production seed (FRhL p2) and candidate vaccine (FRhL p3) were subjected to rigorous safety tests to exclude contaminating microbial agents. There was no significant monkey neurovirulence of parental or PDK-passaged DEN-4 virus or of control fluid cultures. FRhL-passaged viruses retained the phenotypic characteristics: small (occasional medium) plaque; temperature sensitivity at 38.5 degrees C; and absence of plaque formation in African green monkey kidney cells, cytopathic effect in LLC-MK2 cells, and viral growth in human monocytes. FRhL p2 virus displayed low virulence for monkeys; only one of four animals was viremic and three of four developed low-titered antibody. FRhL p3 virus produced viremia in three monkeys and moderate to high hemagglutination-inhibition and neutralizing antibody titers in all animals. Virus at both passages in FRhL exhibited reduced neurovirulence in suckling mice as compared to parental DEN-4. Because of its safety and desirable monkey virulence attributes PDK 35-TD3 FRhL p3 is recommended for human phase I trial.

Dengue-2 vaccine: oral infection, transmission, and lack of evidence for reversion in the mosquito, Aedes aegypti.

The dengue-2 vaccine virus (S-1), and its parent virus (PR-159), were compared for their ability to infect orally, to replicate in, and subsequently to be transmitted by Aedes aegypti mosquitoes. The vaccine virus was markedly less efficient in its ability to infect mosquitoes orally. After ingesting infectious bloodmeals containing 3, 7 to 8.2 log10MID50/ml of the respective viruses, 56% (220/396) of the mosquitoes became orally infected with the parent virus contrasted with 16% (66/397) for the vaccine virus. None of the 16 infected mosquitoes transmitted the vaccine virus, while 14% (3/22) of the mosquitoes transmitted the parent virus. The vaccine virus remained temperature-sensitive (restrictive temperature 39 degrees C) after orally infecting and replicating in Ae. aegypti mosquitoes.

Lack of attenuation of a candidate dengue 1 vaccine (45AZ5) in human volunteers.

A dengue type 1, candidate live virus vaccine (45AZ5) was prepared by serial virus passage in fetal rhesus lung cells. Infected cells were treated with a mutagen, 5-azacytidine, to increase the likelihood of producing attenuated variants. The vaccine strain was selected by cloning virus that produced only small plaques in vitro and showed reduced replication at high temperatures (temperature sensitivity). Although other candidate live dengue virus vaccines selected for similar growth characteristics have been attenuated for humans, two recipients of the 45AZ5 virus developed unmodified acute dengue fever. Viremia was observed within 24 hr of inoculation and lasted 12 to 19 days. Virus isolates from the blood produced large plaques in cell culture and showed diminished temperature sensitivity. The 45AZ5 virus is unacceptable as a vaccine candidate. This experience points out the uncertain relationship between in vitro viral growth characteristics and virulence factors for humans.

Preparation of an attenuated dengue 4 (341750 Carib) virus vaccine. I. Pre-clinical studies.

Dengue 4 (DEN-4) virus strain 341750 Carib was modified by serial passage in primary canine kidney (PCK) cell cultures. By the 15th PCK passage, this virus was less infectious for monkeys and resulted in a significantly reduced viremia as compared to the parent DEN-4 virus. The 30th PCK passage of DEN-4 341750 Carib was non-infectious for monkeys. A vaccine prepared at the 20th PCK passage in DBS-FRhL-2 cells stimulated the production of both neutralizing and hemagglutination inhibition antibodies in monkeys; these animals were also protected against challenge with the homologous strain as well as a heterologous strain of DEN-4. An ID50 titration in monkeys resulted in a titer of greater than 10(4) plaque-forming units (PFU) for the vaccine virus and 0.5 PFU for the parent virus. Reduced monkey infectivity of this magnitude has been correlated with human attenuation in previous dengue vaccine candidates. The DEN-4 strain 341750 Carib PCK-20/FRhL-4 vaccine has been characterized and sufficiently tested to be considered for safety and immunogenicity trials in humans.

Monoclonal antibodies for dengue virus prM glycoprotein protect mice against lethal dengue infection.

Five murine monoclonal antibodies (Mabs) reactive against the prM glycoproteins of DEN-3 and -4 were used to passively protect mice in vivo against lethal challenge with homologous and heterologous dengue virus serotypes. Four of the 5 prM-reactive monoclonals cross-protected mice against heterologous challenge, whereas 1 protected against challenge with only the homologous serotype. Although in vitro binding to virions was readily demonstrated, only 2 of the prM Mabs had detectable neutralizing activity. The neutralizing activity could not be enhanced by anti-mouse immunoglobulin or complement. However, 4 of the 5 prM Mabs fixed complement. This is the first report of prM-specific Mabs that are protective in mice.

Evaluation of recombinant dengue viral envelope B domain protein antigens for the detection of dengue complex-specific antibodies.

To increase the specificity of dengue (DEN) diagnosis based on antibody detection, we have evaluated recombinant proteins as antigens that incorporate most of the B domain of the DEN virus envelope protein fused to the trpE protein of Escherichia coli (trpE-DEN). A pooled antigen consisting of trpE-DEN proteins representing all four serotypes of DEN virus was used in an indirect ELISA for the detection of IgG or IgM antibody. This assay was compared with a standard IgG indirect ELISA and an IgM-capture ELISA using DEN virus-infected cell culture pooled antigens. The results indicated that the trpE-DEN antigens and the cell culture antigens were equally sensitive for detecting IgM and IgG antibodies in convalescent sera from Peru and Indonesia representing virus isolation-confirmed primary and secondary DEN infections, respectively. Fourteen day postinfection IgG antibody-positive sera obtained from individuals infected with DEN-1 virus who had been vaccinated with other flaviviruses were more strongly reactive with the cell culture antigen than with the recombinant antigen, but by day 21 postinfection, a strong antibody response to the trpE-DEN antigens was present. These results suggested that the early antibody response was directed predominantly towards shared flavivirus group antigens that were not detected with the trpE-DEN antigens. Comparison of the trpE-DEN-1 recombinant antigen with a DEN-1 virus-infected cell lysate antigen for the detection of IgG antibody in sera from a cohort of 55 individuals from Peru who seroconverted over a one-year period indicated greater specificity for the recombinant antigens. Also, sera from individuals with no known DEN infections that had been sequentially vaccinated with yellow fever and Japanese encephalitis reacted with the DEN virus cell culture antigen in the IgG ELISA, but did not react with the trpE-DEN pooled antigens. Similarly, YF IgM antibody positive samples that showed cross-reactivity with the DEN virus cell culture antigens, did not react with the trpE-DEN pooled antigens. These results indicated that the trpE-DEN pooled antigen provided a more specific diagnosis of dengue infections than DEN virus-infected cell culture antigen and avoided the biohazards associated with handling live virus during the preparation of diagnostic reagents. The trpE-DEN pooled antigen should permit a better approach to distinguish between past DEN and other flavivirus infections in epidemiologic surveys, and also increase the specificity of serologic diagnosis of acute DEN infections.

Selection of attenuated dengue 4 viruses by serial passage in primary kidney cells. V. Human response to immunization with a candidate vaccine prepared in fetal rhesus lung cells.

A dengue 4 (strain H241, PDK35-TD3 FRhL p3) vaccine attenuated by passage in primary dog kidney cells followed by passage and final vaccine preparation in DBS-FRhL-2 cells was tested in five yellow fever-immune volunteers. Only two volunteers seroconverted by producing hemagglutination-inhibiting and neutralizing antibodies. Mild illness, compatible with dengue infection was found only in the individuals who later developed antibodies. Both volunteers developed a rash by the 8th day following vaccination, coinciding with a slight elevation in temperature and leukopenia. Additionally, several serum enzymes were elevated during the observation period. Dengue 4 virus was isolated from the blood of the two infected volunteers starting as early as day 5 post vaccination. During the viremic period, which lasted 5 days, phenotypically-changed virus was recovered, indicating genetic instability of the vaccine virus. The clinical disease and immune response in the two infected individuals was probably related to replication of the variant virus. Further testing of this vaccine in its present form is not indicated.

Dengue-2 vaccine: infection of Aedes aegypti mosquitoes by feeding on viremic recipients.

Colonized Aedes aegypti mosquitoes were fed on voluntary recipients of an experimental, live, attenuated, dengue type 2 (PR 159/S-1) vaccine to estimate the frequency of vector infection and the stability of the virus in mosquitoes. Two volunteers were viremic at the time of mosquito feeding, but only two of 114 mosquitoes that took a viremic blood meal became infected with the vaccine virus. Strains of virus recovered from the bodies of the mosquitoes and the volunteer's blood retained the temperature sensitivity and small plaque growth characteristics of the vaccine virus. Dengue viral antigen was not detectable in any of the mosquito heads by direct immunofluorescence and in vitro virus transmission by droplet feeding was not observed. This experiment showed that vector mosquitoes can be infected with vaccine virus by feeding on viremic vaccinees. Furthermore, the virus is sufficiently stable to retain the in vitro growth characteristics associated with the vaccine virus.

Preparation of an attenuated dengue 4 (341750 Carib) virus vaccine. II. Safety and immunogenicity in humans.

To determine safety and immunogenicity, a single 0.5 ml dose of a monovalent live-attenuated dengue (DEN) 4 (341750 Carib) vaccine was given sc to 3 groups of flavivirus nonimmune volunteers in increasing concentrations. Two recipients received 10(3) plaque forming units (PFU)/dose (1:100 dilution of stock vaccine). One remained asymptomatic, but became viremic between days 12 and 15, experienced a mild elevation of temperature (37.4 degrees C), and developed DEN-4 specific antibody. Neither recipient of the 10(4) PFU became infected. Eight volunteers then received undiluted vaccine (10(5) PFU). Viremia and antibody (neutralizing, hemagglutination inhibition, and IgM) developed in 5 of the 8 (63%). These 5 volunteers also developed a scarcely noticeable macular, blanching rash and minimal temperature elevations (37.3, 38.1, 37, 37.9, and 37.9 degrees C). Clinically insignificant decreases in total white blood cell, lymphocyte, and polymorphonuclear cell counts and an elevation in mononuclear cell counts occurred in association with viremia. This vaccine is safe, reasonably immunogenic, and suitable for further evaluation.

Immunization of monkeys with baculovirus-dengue type-4 recombinants containing envelope and nonstructural proteins: evidence of priming and partial protection.

Groups of rhesus monkeys were immunized with baculovirus-dengue type-4 (DEN-4) recombinant-infected cell extracts. One recombinant contained all of the DEN-4 structural proteins and two nonstructural (NS) proteins (C-M-E-NS1-NS2a), while the other was a fusion protein containing a portion of the respiratory syncytial virus G glycoprotein and DEN-4 envelope glycoprotein (RSVG-E). Both preparations were immunogenic; all monkeys receiving either immunogen responded with the production of antivirion antibodies in enzyme immunoassays. All except one monkey receiving the recombinant b(C-M-E-NS1-NS2a) made antibodies to NS1. One monkey that received b(RSVG-E) showed the production of low levels of neutralizing antibodies. Following challenge with unmodified DEN-4 virus, seven of nine monkeys in the immunized group became infected and were viremic for a mean of 4.1 days. The control, sham-inoculated monkeys were also viremic; the mean number of days of viremia in this group was 4.7 days. The remaining monkeys in the immunized group (n = 7), although not protected, had evidence of priming. Hemagglutination inhibition antibody responses following challenge indicated an anamnestic response in this group of animals. Based on these results, it was concluded that future immunization schedules should be altered to optimize immune responses and that immunization with more potent and purified immunogens would probably result in higher seroconversion rates and antibody levels in monkeys.

Phase 1 studies of Walter Reed Army Institute of Research candidate attenuated dengue vaccines: selection of safe and immunogenic monovalent vaccines.

We describe the results of initial safety testing of 10 live-attenuated dengue virus (DENV) vaccine candidates modified by serial passage in primary dog kidney (PDK) cells at the Walter Reed Army Institute of Research. The Phase 1 studies, conducted in 65 volunteers, were designed to select an attenuated vaccine candidate for each DENV serotype. No recipient of the DENV candidate vaccines sustained serious injury or required treatment. Three vaccine candidates were associated with transient idiosyncratic reactions in one volunteer each, resulting in their withdrawal from further clinical development. Increasing PDK cell passage of DENV-1, DENV-2, and DENV-3 candidate vaccines increased attenuation for volunteers, yet also decreased infectivity and immunogenicity. This effect was less clear for DENV-4 candidate vaccines following 15 and 20 PDK cell passages. Only one passage level each of the tested DENV-2, -3, and -4 vaccine candidates was judged acceptably reactogenic and suitable for expanded clinical study. Subsequent studies with more recipients will further establish safety and immunogenicity of the four selected vaccine candidates: DENV-1 45AZ5 PDK 20, DENV-2 S16803 PDK 50, DENV-3 CH53489 PDK 20, and DENV-4 341750 PDK 20.

Large-scale purification of inactivated hepatitis A virus by centrifugation in non-ionic gradients.

Formalin-inactivated hepatitis A virus (HAV) can be purified for vaccine preparation by centrifugation in Renografin-76 (diatrizoate meglumine and diatrizoate sodium) gradients. Both continuous-flow rate-zonal and isopycnic methods were used for the separation of a major antigen component from minor antigen and host protein. The major antigen component, which appeared to contain complete virions by electron microscopy, could be recovered from gradients and accounted for approximately one third of the total antigen in the starting material. The HAV-specific purified antigen could be enriched 200-300-fold by either centrifugation procedure. The purified HAV antigen, when adsorbed to alum and inoculated into mice, was found to be highly immunogenic.

Preparation of noninfectious hepatitis A virus hemagglutinin for detecting hemagglutination inhibition antibodies.

Hepatitis A virus (HAV) harvested from infected MRC-5 cells can hemagglutinate various species of erythrocytes at acid pH (Eckels et al., 1989). Further studies revealed that the majority of the hemagglutinin (HA) in MRC-5 and BS-C-1 cells was cell-associated. A simplified procedure for preparing HAV-HA consisted of collecting infected cells in phosphate-buffered saline followed by three cycles of freeze-thawing and sonication. The fluids were clarified and stored at 4 degrees C. The analysis of HA by rate-zonal sucrose gradient centrifugation indicated that the majority of HA co-migrated with infectious virus. Complete inactivation of infectious HAV with 0.03% beta-propiolactone (BPL) did not affect HA activity, while inactivation with 0.05% formalin caused a 16-fold reduction in titer. There was no difference in HAI antibody titers when BPL-treated HA was compared to untreated HA in the hemagglutination inhibition (HAI) test.

Nucleotide sequence of envelope protein of Japanese encephalitis virus SA14-14-2 adapted to vero cells.

Live attenuated Japanese encephalitis (JE) virus SA(14)-14-2 produced in primary dog kidney cells (PDK) was adapted to Vero cells. In an effort to gain insight into the molecular basis of the biological characteristics of the SA14-14-2(Vero) strain, the 1500 nucleotide sequence encoding the envelope (E) gene which possesses major neutralizing epitopes was determined and compared with the sequences of two other attenuated JE virus strains, SA14-14-2(PHK) and SA14-14-2(PDK). The amino acid sequence of the C-terminal region (a.a. 280-500) was found to be identical for all three strains, while the N-terminal region (a.a. 1-279) shows sequence variation. The distribution of mutations in the N-terminal region was nearly the same among the three attenuated strains, suggesting that the N-terminal sequences might be related with virus-host cell specificity. However, it was found that Lys and Val (a.a. 138 and 176, respectively), known to be responsible for attenuation, are still conserved in SA(14)-14-2(Vero). Animal testing showed that SA(14)-14-2(Vero) has an attenuation phenotype similar to that of the parent SA(14)-14-2(PDK) strain in mice.

Evaluation of dengue virus strains for human challenge studies.

Discordance between the measured levels of dengue virus neutralizing antibody and clinical outcomes in the first-ever efficacy study of a dengue tetravalent vaccine (Lancet, Nov 2012) suggests a need to re-evaluate the process of pre-screening dengue vaccine candidates to better predict clinical benefit prior to large-scale vaccine trials. In the absence of a reliable animal model and established correlates of protection for dengue, a human dengue virus challenge model may provide an approach to down-select vaccine candidates based on their ability to reduce risk of illness following dengue virus challenge. We report here the challenge of flavivirus-naïve adults with cell culture-passaged dengue viruses (DENV) in a controlled setting that resulted in uncomplicated dengue fever (DF). This sets the stage for proof-of-concept efficacy studies that allow the evaluation of dengue vaccine candidates in healthy adult volunteers using qualified DENV challenge strains well before they reach field efficacy trials involving children. Fifteen flavivirus-naïve adult volunteers received 1 of 7 DENV challenge strains (n=12) or placebo (n=3). Of the twelve volunteers who received challenge strains, five (two DENV-1 45AZ5 and three DENV-3 CH53489 cl24/28 recipients) developed DF, prospectively defined as ≥2 typical symptoms, ≥48h of sustained fever (>100.4°F) and concurrent viremia. Based on our study and historical data, we conclude that the DENV-1 and DENV-3 strains can be advanced as human challenge strains. Both of the DENV-2 strains and one DENV-4 strain failed to meet the protocol case definition of DF. The other two DENV-4 strains require additional testing as the illness approximated but did not satisfy the case definition of DF. Three volunteers exhibited effusions (1 pleural/ascites, 2 pericardial) and 1 volunteer exhibited features of dengue (rash, lymphadenopathy, neutropenia and thrombocytopenia), though in the absence of fever and symptoms. The occurrence of effusions in milder DENV infections counters the long-held belief that plasma leakage syndromes are restricted to dengue hemorrhagic fever/dengue shock syndromes (DHF/DSS). Hence, the human dengue challenge model may be useful not only for predicting the efficacy of vaccine and therapeutic candidates in small adult cohorts, but also for contributing to our further understanding of the mechanisms behind protection and virulence.