![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
Acambis, Inc., Cambridge, Massachusetts 02139,1 Northbrook, Illinois 60062,2 Sierra Division, Discovery and Development Services, Charles River Laboratories, Inc., Sparks, Nevada 89431,3 Aventis Pasteur, Marcy-L'Etoile F-69280, France;,4 University of Texas Medical Branch, Galveston, Texas 77555-06095
Received 17 November 2003/ Accepted 22 December 2003
 |
ABSTRACT |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
INTRODUCTION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We have used the YF 17D virus vaccine virus as a vector for the construction of YF/DEN1 through YF/DEN4 chimeras (ChimeriVax-DEN1-4) and evaluated their safety and immunogenicity in mice and monkeys (14-16). A phase I clinical trial with monovalent ChimeriVax-DEN2 virus was completed recently, which showed an excellent safety and immunogenicity profile of the vaccine candidate (Acambis). All YF/DEN chimeras described previously were constructed using gene donors (prME) from low-passage human isolates of DEN viruses. These chimeras, which were generated by transfection of Vero cells with RNA transcripts and not subjected to plaque purifications, contained viruses with mutations in their genomes due to the quasispecies nature of RNA viruses. To manufacture vaccines for incorporation into a tetravalent formulation, a plaque purification strategy was adopted to produce a more homogeneous population of virus stocks that would be less likely to accumulate mutations during cell culture passages. When these viruses were analyzed after 20 passages in Vero cells, the uncloned viruses had accumulated more mutations than plaque purified chimeras (45).
This is the first report describing the safety of a recombinant chimeric tetravalent DEN virus vaccine in a formal monkey neurovirulence test and showing the protective efficacy of the vaccine in a monkey challenge model. Based on these results, a tetravalent dose and formulation were selected for use in a clinical trial.
|
MATERIALS AND METHODS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
M)2001 (Fig. 1). The BstBI-NarI fragment containing the C-terminal portion of the DEN1 virus E gene (normally present in YFM5.2 plasmids) was resynthesized by reverse transcription-PCR on virion RNA of DEN1 virus strain PUO359 and individually cloned in a modified low-copy-number vector, pCL1921, resulting in plasmid pCL/D1E. The absence of mutations in both plasmids was verified by sequencing. DNA template for in vitro transcription was prepared by three-fragment ligation (Fig. 1). The BstBI-NarI fragment of pCL/D1E plasmid and the Nar-AatII fragment from the YFM5.2/DEN2 plasmid (containing most of the YF virus-specific NS genes) (8, 16) were ligated with the large BstBI-AatII portion of the 5'3'DEN1(M)2001 plasmid. Ligation products were linearized with XhoI and phenol-chloroform extracted, and RNA transcripts were synthesized in vitro using SP6 RNA polymerase (Epicentre).
|
Preparation of Master Seed, Working Seed, and Vaccine Lots of ChimeriVax-DEN1-4 viruses under cGMP.
Master Seed (MS, P8) viruses were prepared at Molecular Medicine (La Jolla, Calif.) using GMP-compliant procedures. Vero cells were grown in MEME medium (supplemented with 2 mM L-glutamine, 1% minimal essential medium MEM nonessential amino acids, and 10% 0.1-µm-pore-size-filtered, gamma-irradiated fetal bovine serum (FBS) to near confluence in 10-layer Nunc cell factories (NCF). The cells were then infected with PMS (P7) virus at a multiplicity of infection of approximately 0.001 and incubated at 36 ± 2°C under 5% ± 2% CO2 at 80% ± 5% relative humidity for 3 to 4 days. Virus-containing cell culture supernatant fluids were harvested when an early cytopathic effect was observed and processed by filtration to remove cell debris. The MS virus was formulated by adding FBS to a final concentration of 50%. The formulated materials were filtered through a sterile 0.22-µm-pore-size filter and stored at 
–60°C. Working Seed (WS, P9) and Vaccine Lot (VL, P10) viruses were produced using GMP-compliant procedures by Aventis-Pasteur (Marcy L'Etoile, France). WS (P9) virus was manufactured, formulated, and filled in the same manner as MS (P8) virus. The manufacture of WS and VL viruses involved a biomass expansion phase in NCFs. For production of VL, the cells were washed twice with serum-free medium prior to virus infection and refed with serum-free medium to reduce the level of FBS in the final bulk. Downstream processing consisted of filtration to remove cell debris followed by digestion of nucleic acids by benzonase (15 UI/ml at 5 ± 3°C for 16 h), concentration by tangential-flow filtration (50 kDa), and diafiltration of the virus. VL viruses were stabilized by the addition of sorbitol to a final concentration of 10%.
Sequencing. Consensus sequencing of viruses was done as described previously (45). Briefly, virion RNAs were extracted from supernatants of infected cells by using TRI-Reagents LS (Molecular Research Center). RNAs were amplified with the Titan One-Tube reverse transcription-PCR kit (Roche), and cDNA fragments were purified using the QIAquick PCR purification kit or the QIAquick gel extraction kit (Qiagen) and sequenced using the CEQ DYE Terminator cycle-sequencing kit (Beckman). The sequencing reaction products were resolved with a CEQ2000 automated sequencer (Beckman Coulter) and analyzed by Sequencher 3.0 (GeneCodes).
Cells and viruses. LS-10 Vero cells were obtained from a qualified working cell bank at passage 137 from Aventis Pasteur and grown in MEME containing 10% FBS (HyClone, Ogden, Utah). In addition to ChimeriVax-DEN1-4 PMS, MS, WS, and VL, the following viruses were used: stocks of vaccine viruses (P10) for ChimeriVax-DEN1-4 were produced from MS virus stocks (P8) by two subsequent passages in Vero cells under non-GMP-compliant conditions; wild-type (WT) DEN virus stocks (DEN1 [West Pacific 74], DEN2 [S16803 PDK-10], DEN3 [CH53489], and DEN4 [341750]) were kindly provided by Kenneth Eckels (Walter Reed Army Institute of Research (WRAIR), Silver Spring, Md.); and YF-VAX (YF vaccube strain 17D) was purchased from Aventis Pasteur. All WT DEN viruses and YF-VAX were used without any further passages.
Potency and identity tests. Three different plaque assay techniques using Vero cells were used to measure the potency and identity of DEN viruses. The first was a standard plaque assay using agarose double overlay, where plaques were visualized by addition of Neutral Red to the plates. This technique was generally used to measure the potency of chimeras that produce visible plaques (16, 17); titers were expressed as PFU per milliliter. The second was an immunofocus-forming assay (IFF), using methyl cellulose instead of agarose, where plaques were visualized by the addition of DEN virus serotype-specific monoclonal antibodies and goat anti-mouse immunoglobulin G conjugated to alkaline phosphatase. This technique was generally used for determination of the potency (Fig. 2) and identity of individual ChimeriVax-DEN serotype viruses in a mixture and for detection of WT DEN viruses that produce extremely small plaques in Vero cells (14). Titers were expressed as focus-forming units (FFU) per milliliter. The third was a modified IFF assay, where 6- or 12-well plates were replaced with 96-well plates. The titers in this assay, which is performed exclusively by Aventis Pasteur, were expressed as 50% tissue culture infectious dose per milliliter. Using selected samples, these three techniques produced comparable determinations of DEN virus potency.
|
50 to 100 PFU) is neutralized by different serum dilutions and titers are expressed as the highest dilution of serum inhibiting 50% of the plaques (PRNT50) (14-16, 36). For measurement of neutralizing antibody to YF 17D virus, a constant-serum varying-virus test was performed and the results were expressed as the log10 neutralization index (LNI) (35). Animal studies. All studies were carried out using Institutional Animal Care and Use Committee-approved protocols and in accordance with the U.S. Department of Agriculture Animal Welfare Act (9 CFR parts 1 to 3), as described in the Guide for Care and Use of Laboratory Animals (41).
(i) Mice. The neurovirulence phenotypes of chimeras were assessed in suckling mice. Pregnant ICR (CD1) mice were purchased from Harlan (Indianapolis, Ind.). Suckling mice were pooled at the age of 2 to 3 days and randomly distributed to dams (9 to 13 mice/dam). The mice were inoculated at the age of 3 to 4 days by the intracerebral (i.c.) route with 0.02 ml of various passages of ChimeriVax-DEN1-4 viruses and observed for 21 days, and deaths were recorded. The virus dose administered to each group was determined by back titer determination.
(ii) Monkeys. Two studies were conducted with monkeys (Sierra Division of Discovery and Developments Services, Charles River Laboratories, Inc., Sparks, Nev.) to assess the safety, immunogenicity, and protective activity of ChimeriVax-DEN1-4 chimeras. Safety (neurovirulence) was assessed and compared to that of YF-VAX following i.c. inoculation, as prescribed by the WHO requirement for testing of YF virus vaccines (50) (experiment 1). Viremia and immunogenicity/protection were measured following subcutaneous (s.c.) inoculation with four different tetravalent formulations and challenge with WT DEN viruses (experiment 2).
(iii) Experiment 1. Twenty-two (11 male and 11 female) experimentally naive, flavivirus-seronegative cynomolgus monkeys (2.0 to 4.2 years of age and weighing 3.2 to 3.8 kg for males; 2.1 to 4.6 years of age and weighing 2.2 to 3.0 kg for females [day 1 values]) were assigned to two treatment groups (n = 11 for each group). The vaccine formulation (0.25 ml), consisting of approximately 5 log10 PFU of each cGMP ChimeriVax-DEN1-4 viruses (5.6 log10 PFU total, stabilized in 10% sorbitol solution) or reference article, YF-VAX (4.7 log10 PFU), was inoculated into the frontal lobe of the cerebral cortex. Animals were observed for 30 days postdose and then euthanized and necropsied. During the observation period, the monkeys were evaluated for clinical signs (twice daily), changes in food consumption (once daily), body weight (approximately weekly), serum chemistry, and hematology parameters (predose on day 1 and then on days 3, 5, 7, 15, and 31). Clinical signs were scored as specified by the WHO protocol (36, 50). Blood samples were collected on days 1 (predose) and 2 to 11 for viremia analysis and on days 1 and 31 for DEN1 through DEN4 virus antibody titer determination. At necropsy, gross pathological findings were recorded and tissues were preserved for histopathological examination. A selected subset of tissues was evaluated by a board-certified pathologist (liver, spleen, heart, kidneys, and adrenal glands) or Inessa Levenbook (brain and spinal cord). Lesions in the meninges and the brain/spinal cord matter were scored using a scale of 0 to 2, on the basis of the following observations: grade 0, no visible lesions; grade 1 (minimal), one to three small and/or one large infiltrate, mostly perivascular; a few small foci of more diffuse infiltration, unconnected with blood vessels; grade 2 (mild), more than three small and/or two or more large perivascular infiltrates; several foci of cellular infiltration, unconnected with blood vessels (some neurons may be involved in these foci of inflammation). The degree of neurovirulence was estimated for the target and discriminator areas, as described previously (26, 36, 50). For cynomolgus monkeys, the substantia nigra and cervical and lumbar enlargements of the spinal cord represent the target formations whereas basal ganglia and thalamic nuclei are considered discriminator areas. Individual and group mean lesion scores for the target and discriminator areas were calculated both separately and as a combined score. Statistical analyses of lesion scores were performed by using a nonparametric Kruskal-Wallis test for comparison between groups.
(iv) Experiment 2. Twenty-four male and female monkeys were assigned to four treatment groups (six monkeys/group). Monkeys in each group received a single 0.5-ml s.c. dose of each of four tetravalent formulations (5,5,5,5 [group 1], 3,5,5,3 [group 2], 5,5,5,3 [group 3], and 3,3,3,3 [group 4] log10 TCID50 of ChimeriVax-DEN1, ChimeriVax-DEN2, ChimeriVax-DEN3, and ChimeriVax-DEN4 viruses, respectively). Monkeys were evaluated for changes in clinical signs and body weight. Sera for virus and antibody titer determinations were collected at various time points. Viremia was measured on days 2 to 11. First, as a safety test, total viremia was measured to determine whether viremia levels were within acceptable limits according to the WHO guidelines (36, 50) (i.e., <500 mouse 50% i.c. lethal dose [LD] units for any individual monkey, estimated to equal 4.3 log10 Vero cell PFU/ml for YF-VAX). Next, titers of individual ChimeriVax-DEN viruses in sera were measured using a serotype-specific immunofocus assay on Vero cells (14). Neutralizing-antibody titers (PRNT50) were measured in heat-inactivated sera (without addition of complement) obtained on days 1 (preimmune), 31, 121, 180 (prechallenge), and 210 (1 month postchallenge) against homologous or heterologous (WT DEN1 to DEN4 viruses, days 31, 180, and 210) viruses as described previously (14). To assess protective immunity, the monkeys were rerandomized into four new groups and challenged with WT DEN virus strains 6 months postimmunization. The control group consisted of 16 animals (4 monkeys per serotype) shown to be flavivirus naive by prescreening for presence of JE and DEN1 through DEN4 virus neutralizing antibodies by PRNT50. Since the suitability of the challenge viruses had not been established previously in cynomolgus monkeys, 8 of 16 control monkeys (2 per serotype) were inoculated in advance with WT challenge strains to ensure that these viruses would induce acceptable levels of viremia in these monkeys. Viremia and neutralizing-antibody levels were measured postchallenge (5 log10 PFU/dose) to determine protection against each DEN virus serotype.
Statistical analysis. Differences in responses among groups or between two groups of animals were analyzed for significance using Fisher's exact test, two tailed for dichotomous variables and analysis of variance (ANOVA) for continuous variables; all analyses were performed using JMP software version 5.0. Statistical analyses of neuropathological lesion scores in experiment 1 were performed using a nonparametric Kruskal-Wallis test for comparison between groups.
|
RESULTS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
1 mm) in Vero cells at P7, which increased to 2 to 3 mm during three further passages, as a result of mutations in the envelope genes (Table 1). In contrast, no change in the genome sequence of the DEN4 chimera was observed up to P20, despite a modest increase in its plaque size (from 1 mm at P7 to 2.0 mm at P20). To further search for the presence of virus mutants not detected by consensus sequencing, vaccine viruses were plaqued in Vero cells. The plaque morphology was determined after staining with DEN serotype-specific monoclonal antibodies. As shown in Fig. 2, only the DEN3 vaccine strain produced heterogeneous plaque sizes. Direct sequencing of small and large plaques from this vaccine revealed a minority population of mutant viruses (E202 K to R) that had not been identified by consensus sequencing, which has a limit of detection of about 10%. However, when a non-GMP version of DEN3 virus at P10 (vaccine passage level) that was produced by passage of the MS (P8) virus was subjected to consensus sequencing, both parent and mutant viruses could easily be detected (Table 1). Both small-plaque (E202K) and large-plaque (E202R) viruses retained their attenuated phenotype in the 4-day-old suckling-mouse model (data not shown). Mouse neurovirulence. The neurovirulence of the DEN1 chimera for infant mice was significantly reduced from P8 to P10 (P < 0001 for all doses measured by the log-rank method). An increase in the 50% lethal dose from <2.1 (P7) to >5.1 log10 (P10) PFU/ml correlated with the appearance of the E204 K-to-R mutation (Table 1, experiment 1). Another clone of the DEN1 chimera (clone E) with a single E204 K-to-R mutation was also significantly less virulent than its parent WT chimera (Guirakhoo, unpublished). For the DEN2 and DEN3 chimeras, accumulations of mutations (prM24V and E251F in DEN2 or heterogeneity in the E202 residue of DEN3) did not significantly alter their neurovirulence for infant mice. However, extensive passage in Vero cells (P20) seemed to decrease the neurovirulence of these viruses for mice, sometimes without an apparent genetic change (Table 1, experiment 2). All passages of the DEN1 through DEN4 virus chimeras were significantly less virulent (P < 0001, log-rank method) than their vector, YF 17D virus.
To demonstrate the absence of interference between chimeras administered as a tetravalent formulation, groups of 4-day-old infant mice were inoculated by the i.c. route with each of four chimeras as monovalent (2 log10 PFU) or tetravalent (2 log10 PFU of each serotype) formulations. A control group was inoculated with 2 log10 PFU of YF-VAX. As expected, all nine mice in the YF-VAX group died, with an average survival time (AST) of 7.7 days. The mortality rate with monovalent chimeras was 44, 22, 0, and 33% with an AST of 13.7, 14, >21, and 13.3 days for DEN1, DEN2, DEN3, and DEN4 virus chimeras, respectively. The mortality rate (33%) and AST (14.7 days) in the tetravalent group did not differ significantly from that of the monovalent chimeras with the highest neurovirulence (i.e., DEN1, DEN2, and DEN4 viruses). This experiment demonstrated that the attenuated DEN3 virus chimera in a tetravalent mixture did not interfere with the replication of other chimeras in the brains of suckling mice.
Monkey studies. Two experiments were performed to address the safety (experiment 1) and efficacy (experiment 2) of tetravalent vaccine formulations in monkeys.
(i) Experiment 1. Evaluation of safety and neurovirulence following i.c. inoculation of tetravalent vaccine in monkeys. Twenty-two monkeys were divided into two groups (n = 11 for each group). The vaccine formulation (0.25 ml) consisting of approximately 5 log10 PFU of each cGMP ChimeriVax-DEN1-4 viruses (5.6 log10 PFU total) or reference article, YF-VAX (4.7 log10 PFU), was inoculated into the frontal lobe of the cerebral cortex. Following inoculation, there were no vaccine-related changes in body weight, serum chemistry, or hematology parameters. Neither vaccine produced gross pathological findings. Possible test and reference article-related clinical signs were limited to pupil dilation in 5 of 11 monkeys inoculated with YF-VAX (group 1) and in 2 of 11 monkeys inoculated with ChimeriVax-DEN tetravalent vaccine (group 2) between days 16 and 31.
(a) Viremia. The magnitude and duration of viremia are shown in Table 2. Of 11 monkeys inoculated with YF-VAX, 10 (91%) became viremic. The duration of viremia was 1-3 days, with peak titers ranging from 1.3 to 2.9 log10 PFU/ml. All 11 monkeys inoculated with tetravalent ChimeriVax-DEN became viremic. The duration of the viremia, as measured by total virus in the serum, was 4 to 9 days, with peak titers ranging from 1.7 to 3.3 log10 PFU/ml. The mean duration of viremia for tetravalent ChimeriVax-DEN vaccine (5.5 days) was significantly longer than that of YF-VAX (2.09 days) (P = 0.0094, ANOVA). Although the mean peak total viremia (i.e., the combined titers of all four chimeras) titer for tetravalent ChimeriVax-DEN (2.9 log10PFU/ml) was higher than the mean peak titer for the reference YF-VAX (2.3 log10 PFU/mL) (P < 0.0001, ANOVA) virus, the titers for both vaccines remained within acceptable group and individual monkey specifications, based on WHO requirements for the YF 17D vaccine (50).
|
(b) Neutralizing-antibody response.
All monkeys seroconverted following i.c. inoculation with YF-VAX. On day 31, YF virus-specific neutralizing-antibody titers ranged from 2.07 to 3.88 in the LNI assay (an LNI of 
0.7 log10 is generally considered seropositive). A small proportion (2 of 11, 2 of 11, 1 of 11, and 1 of 11) of the YF-VAX-treated monkeys showed cross-reactivity with heterologous DEN1, DEN2, DEN3, or DEN4, respectively, in a PRNT50 assay on day 31. Such antibody cross-reactivity is not unexpected among flaviviruses. However, remote exposure of these monkeys to a heterologous flavivirus not detected by prestudy antibody screening cannot be ruled out.
All monkeys treated with tetravalent ChimeriVax-DEN vaccine seroconverted to each DEN serotype. Neutralizing-antibody titers ranged from 160 to 20,480, 40 to 2,560, 20 to 640, and 80 to 5,120 for DEN1, DEN2, DEN3, and DEN4 viruses, respectively, in the PRNT assay; no monkeys had cross-reacting antibodies to YF 17D virus in the LNI assay. The geometric mean titer (GMT) was 1,202, 640, 85, and 726 for DEN1, DEN2, DEN3, and DEN4 virus chimeras, respectively.
(c) Histopathology. Neither vaccine produced histopathological findings in nonneural tissues. In neural tissues, the neurovirulence of the tetravalent ChimeriVax-DEN vaccine was minimal. The target, discriminator, and combined scores for the group given tetravalent ChimeriVax-DEN vaccine were significantly lower than the scores for the group that received YF-VAX (P < 0.05). Vaccine-related lesions in the meninges and in the brain or spinal cord of monkeys from both groups were limited to minimal to mild inflammation. These lesions were observed in 10 (91%) of 11 monkeys treated with YF-VAX and in 6 (55%) of 11 monkeys inoculated with tetravalent ChimeriVax-DEN vaccine. Perivascular infiltrates consisted of mononuclear cells. Rare foci of microglial cell infiltration of blood vessels were noted but considered unrelated to vaccine administration. The lesions in monkeys vaccinated with tetravalent ChimeriVax-DEN vaccine were significantly less severe than those in monkeys vaccinated with YF-VAX vaccine. Lesions induced by tetravalent ChimeriVax-DEN vaccine did not exceed grade 1, but lesions caused by YF-VAX were scored 1 or 2 (Table 3).
|
(a) Viremia.
All 24 monkeys became viremic (Table 4). The onset of viremia in monkeys inoculated with the high-dose formulation (group 1) occurred as early as day 2, and all six animals were viremic on day 3. No viremia was detected in any group 1 monkey after day 8 (level of detection, 1.7 log10 PFU/ml). The mean peak titer and duration of viremia for this group were 2.6 log10 PFU/ml and 4.0 days, respectively. In low-dose group animals (group 4), viremia was delayed until day 3. Only one of six animals showed borderline viremia of 1.7 log10 PFU/ml on this day. Six (100%) of six monkeys were viremic from day 6 until the last day of serum collection (day 11). The mean peak titer and duration of viremia were 2.5 log10 PFU/ml and 5.5 days, respectively. Viremia in groups 2 and 3 (where the doses of DEN1 and/or DEN4 virus chimeras were reduced) was observed from days 2 to 11, due to circulation of different serotypes on each day (Table 5). The lowest mean peak viremia (2.1 log10 PFU/ml) was identified in group 3, where animals had received a reduced dose of ChimeriVax-DEN4 virus. The shortest duration of viremia (2.5 days) was observed in group 2, where animals had received reduced doses of DEN1 and DEN4 virus chimeras.
|
|
In group 1, all four serotypes were detected in four (67%) of six monkeys on day 3. All animals became viremic to DEN3 and DEN4 virus chimeras. DEN3 and DEN4 viruses were found in the majority of monkeys up to day 7, but DEN2 and DEN1 virus chimeras were not detected beyond days 4 and 5, respectively, in any animal. DEN1 and DEN2 virus chimeras and DEN2 virus chimera could not be found in monkeys F20967 and F21339M, respectively. The highest peak viremia, which lasted for 6 days, was induced by the DEN4 virus chimera (in monkey F21786F). The mean peak viremia titers were 1.2, 1.1 (1.5 and 1.6 when only viremic animals were considered, respectively), 2.3, and 1.9 log10 PFU/ml for DEN1 to DEN4 viruses, respectively. The mean duration of viremia was 1.5, 0.8 (1.8 and 1.2 when only viremic animals were considered, respectively), 2.3, and 3.7 days for DEN1 to DEN4 viruses, respectively (Table 4).
In group 2, monkeys were immunized with 100-fold-lower doses of ChimeriVax-DEN1 and ChimeriVax-DEN4. No DEN1 virus chimera could be detected in any of the monkeys. Other chimeras were detected from days 2 to 11 (Table 5). Four (67%) of six monkeys became viremic to DEN2 and DEN3 viruses. Only two (33%) of six monkeys became viremic to DEN4 virus. The highest viremia (3.4 log10 PFU/ml), which lasted for 3 days, was caused by DEN3 virus in monkey F21565F. The longest duration of viremia was 8 days with DEN4 virus in monkey F21355M. The mean peak viremia titers were 0, 1.5, 1.8, and 0.9 (0, 2.3, 2.6, and 2.7 when only viremic animals were considered, respectively) log10PFU/ml for DEN1 to DEN4 viruses, respectively. The mean durations of viremia were 0, 1.3, 3, 1.5 (2, 4.5, and 4.5 days when only viremic animals were considered, respectively) for DEN1 to DEN4 viruses, respectively (Table 4).
In group 3, most of the monkeys showed early viremia with DEN1 to DEN3 viruses and a delay in viremia (replication) with the DEN4 virus chimera. Only DEN3 and DEN4 viruses could be detected in monkey F21544M, and only the DEN2 virus chimera could be detected in monkey F21149M (for 1 day [Table 5]). The highest peak viremia (3.1 log10 PFU/ml) was induced by DEN3 virus (monkey F21342M) and DEN4 virus (monkey F21384F). The duration of viremia in these animals was 3 and 7 days, respectively. The mean peak viremia titers were 1.2, 1.4, 1.8, 1.4 (1.8, 1.7, 2.2, and 2.1 when only viremic animals were considered, respectively) log10 PFU/ml for DEN1 to DEN4 viruses, respectively. The mean duration of viremia was 1.5, 2, 2.2, 2.3 (2.2, 2.4, 2.6, and 2.8 days when only viremic animals were considered, respectively) days for DEN1 to DEN4 viruses, respectively (Table 4).
In group 4, chimeric viruses could generally be detected before day 3 (the exception was monkey F21522M, in which the DEN2 virus chimera could be detected on day 2). In contrast to group 1, in which most viruses were detected on days 3 to 4 and no virus was detected beyond day 7, the majority of viruses in group 4 were detected between days 5 and 7 and the DEN3 and DEN4 virus chimeras could be detected through day 11 (Table 5). All monkeys became viremic to DEN4 virus, and four of six became viremic to DEN1, DEN2, and DEN3 viruses (DEN1 virus could not be found in monkeys F21522M and F21570F). The highest peak viremia (3.1 log10 PFU/ml) was induced by the DEN4 virus chimera (monkey F209108F) and lasted for 4 days. The longest duration of viremia (8 days) was also related to the DEN4 virus chimera (monkey F21311M), with a peak titer of 2.8 log10 PFU/ml. The mean peak viremia titers were 1.2 (1.9 when only viremic animals were considered), 1.9, 1.6 (2.0 when only viremic animals were considered), and 2.5 log10 PFU/ml for DEN1 to DEN4 viruses, respectively. The mean duration of viremia was 2 (3 when only viremic animals were considered), 3.8, 3.5 (4.2 when only viremic animals were considered), and 5.3 days for DEN1 to DEN4 viruses, respectively (Table 4).
(b) Neutralizing-antibody response. Sera obtained on day 1 (preimmunization) were negative (PRNT50 < 10) against four DEN virus serotypes. As shown in Table 6, all monkeys (six of six) in group 1 seroconverted to all four chimeric DEN viruses by day 31 postimmunization and remained seropositive until challenged with WT viruses 6 months later. The level of DEN virus-specific neutralizing antibodies (day 31) varied from 40 to 320 for DEN1, 80 to 5,120 for DEN2, 40 to 1,280 for DEN3, and 640 to >2,560 for DEN4 viruses. The GMTs on day 31 were 113, 718, 285, and 1,140 for DEN1 to DEN4 viruses, respectively. All monkeys in group 4 also seroconverted to all four chimeric DEN viruses by day 31 postimmunization. The level of DEN virus-specific neutralizing antibodies varied from 20 to 320 for DEN1 virus, 10 to 320 for DEN2 virus, 160 to 2,560 for DEN3 virus, and 640 to >2,560 for DEN4 virus. The GMTs on day 31 were 80, 57, 403, and 1,280 for the DEN1 to DEN4 viruses, respectively. Only three (50%) of six monkeys in the dose-adjusted formulation group 3 seroconverted to all four chimeric DEN viruses by day 31 postimmunization. Three monkeys (F212117F, F20977F, and F21534M) did not seroconvert to DEN4 virus. Interestingly, monkeys F20977F and F21534M, which were later challenged with WT DEN4 virus, had developed a low PRNT50 of 20 and 40, respectively, prior to challenge (see below). The level of DEN-specific neutralizing antibodies varied from 10 to 640 for DEN1 virus, <10 to >10,240 for DEN2 virus, 10 to 2,560 for DEN3 virus, and <10 to >1,280 for DEN4 virus. The GMTs on day 31 were 80, 218, 320, and 32 for DEN1 to DEN4 viruses, respectively. Four of six monkeys in group 4 seroconverted to all four chimeric DEN viruses by day 31 postimmunization. One monkey (F21544M) did not seroconvert to DEN2 and DEN4 viruses, and one monkey (F21149 M) did not seroconvert to DEN3 and DEN4 viruses. The level of DEN virus-specific neutralizing antibodies varied from 10 to 2,560 for DEN1 virus, <10 to 320 for DEN2 virus, <10 to 10,240 for DEN3 virus, and <10 to >1,280 for DEN4 virus. The GMTs on day 31 were 142, 34, 489, and 66 for DEN1 to DEN4 viruses, respectively.
|
|
|
|
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The appearance of mutations in DEN virus chimeras resulted sometimes in a change in plaque size or neurovirulence of the virus for infant mice. The E204 K-to-R substitution in the DEN1 virus envelope, for example, increased the plaque size and reduced its neurovirulence for infant mice and viscerotropism and viremia for nonhuman primates (Guirakhoo, unpublished), whereas mutations in DEN2 and DEN3 virus genomes did not appear to have a major impact on the neurovirulence of the vaccine viruses. Because the vaccine lot of DEN1 virus chimera had acquired an attenuating mutation, it was necessary to further evaluate the impact of this mutation in additional relevant animal models, such as monkeys and mosquitoes. The viscerotropism and neurovirulence of mutant and WT DEN1 virus chimeras were compared in monkeys inoculated by the s.c. or i.c. routes. These viruses were also evaluated for their replication in mosquito vectors by infection of Aedes aegypti, the principal mosquito vector for DEN viruses, by the intrathoracical route (Guirakhoo, unpublished). The results provided reassurance that ChimeriVax-DEN1 mutant vaccine will remain safe in the human host and will not replicate in mosquitoes, even if it reverts to the WT sequence in a vaccinated individual.
The suckling-mouse neurovirulence test was adopted as a quality control release test for all vaccine lots because it had been shown to detect minor changes in the genome of ChimeriVax-JE (37) or ChimeriVax-DEN chimeras that are associated with neurovirulence. The neurovirulence of the DEN1 virus chimera was significantly reduced from P7 to P10 (vaccine level), and the virus remained attenuated up to P20 in Vero cells. No change in neurovirulence was observed for the other chimeras (DEN2 to DEN4 viruses) between PMS (P7) and bulk vaccines (P10). Extensive passaging to P20, however, produced viruses which were generally less neurovirulent than their P7 parents. In the case of the DEN1 virus chimera, this was obviously due to the E204 mutation. For other chimeras, however, slight reduction in neurovirulence for infant mice could not be supported by a change in the virus genome (e.g., DEN4 virus did not accumulate any mutations based on consensus sequencing). It is possible that subpopulations of mutants (estimated to be 10% of the total virus population) with reduced neurovirulence for infant mice are present, which cannot be detected by consensus sequencing. Alternatively, viruses could accumulate cellular proteins by in vitro passages, which may reduce their replication in suckling-mouse brains.
Our data are in agreement with other studies which showed that chimerization using attenuated backbones such as YF 17D (5, 8, 14-16, 48), DEN2 (4) or DEN4 (49) virus is generally sufficient to produce vaccine candidates for dengue (nonencephalitic). To produce chimeric vaccine viruses for encephalitic flaviviruses, it seems that further attenuations, usually within but not limited to the E genes, may be necessary. Successful vaccine candidates were produced using attenuated E genes of Japanese encephalitis (JE) (SA14-14-2-strain) (7), West Nile (1) and St. Louis encephalitis (K. Pugachev, submitted for publication) viruses within the YF 17D backbone or West Nile and Langat viruses within the DEN4 virus backbone (43, 44).
Monkey neurovirulence studies.
The YF 17D vaccine was first used for human immunization in 1936. Over the past 60 years, 400 million people have received the vaccine, with a remarkable record of safety and efficacy. However, recent reports estimate a very low incidence (1:250,000) of YF 17D vaccine-associated viscerotropic disease and encephalitis (6, 31). In dengue, involvement of the central nervous system is an equally rare event. Reports of patients with a diagnosis of DEN virus encephalitis indicate that the virus very rarely crosses the blood-brain barrier and targets neuronal cells (21, 29). Most neurological adverse events appear to be due to encephalopathy in DHF/DSS caused by edema or electrolyte imbalance. Because ChimeriVax-DEN viruses contain the replicative machinery of YF 17D together with E genes of DEN viruses, it is possible that similar viscerotropic and encephalitis events could occur with this vaccine. Since these events are rare, they cannot be studied in early clinical trials. However, animal models can be used in preclinical studies to predict safety of the vaccine candidates for humans. The neurovirulence and viscerotropism of ChimeriVax-DEN1-4 tetravalent vaccine were determined in a formal neurovirulence test performed as specified by the WHO protocol (50). Prior to this test, it was necessary to demonstrate that these viruses can independently replicate in vivo and that neurovirulence of the "hotter" viruses (DEN1, DEN2, and DEN4 viruses) is not masked by that of the "colder" (DEN3 virus) one. Inoculation of a tetravalent mixture of four chimeric viruses into brains of suckling mice demonstrated that neurovirulence and AST of a tetravalent mix is similar to that of the "hotter" monovalent components. The independence of replication of the four chimeras was also demonstrated in mosquito vectors (22).
The potential neurovirulence and acute toxicity of the ChimeriVax-DEN1-4 tetravalent vaccine was compared with those of YF-VAX in the monkey neurovirulence test. i.c. injection of tetravalent DEN virus vaccine and YF-VAX was well tolerated at doses of 4.1, 4.3, 4.9, and 5.0 log10 PFU for ChimeriVax-DEN1 to ChimeriVax-DEN4, respectively, within the tetravalent vaccine, and 4.7 log10 PFU for YF-VAX. Vaccine-related clinical signs of pupil dilation observed between days 16 and 31 were more frequent in the YF-VAX-treated group, especially on days 17 and 18, when clinical scores were significantly different between groups. Vaccine-related gross anatomic and histological findings from nonneural tissues were secondary to vaccine-induced immunostimulation. Peak total viremia titers and number of viremic days were greater in monkeys treated with the tetravalent DEN virus vaccine compared to those in YF-VAX-treated monkeys. Nevertheless, viremia levels for both groups remained within acceptable limits for individual monkey and group titers established under the WHO requirements for the YF 17D vaccine (50). All monkeys seroconverted following i.c. inoculation of tetravalent vaccine, and none had a detectable cross-reactivity with YF 17D virus in the LNI assay. Antibody levels against the DEN1 to DEN4 virus serotypes did not significantly vary within the tetravalent vaccine (P > 0.15 by ANOVA), despite higher viremia levels detected for DEN3 and DEN4 virus serotypes. Viremia, which reflects the release of viruses from infected cells into the bloodstream and rapid clearance by the reticuloendothelial system (40), was measured for determination of the viscerotropism of viruses. However, since the tissue tropism of ChimeriVax-DEN viruses is unknown at present, daily measurements of viremia may not precisely reflect the dynamics of virus replication in the host. The neurovirulence of the tetravalent DEN virus vaccine preparation was minimal, and the target region scores were significantly lower than the scores for the group treated with YF-VAX (P < 0.05). Clinical and histological lesion scores induced by the YF 17D vaccine were similar to those reported for the YF 17DD substrain (30). These data are also consistent with our previous observations with ChimeriVax-JE (39) or ChimeriVax-DEN2 vaccines (data not shown), in which the mean lesion scores for monkeys treated with chimeras were significantly lower than the scores for the group given the YF 17D vaccine. Thus, the tetravalent formulation met or exceeded the safety test requirements for the YF 17D vaccine. It would be interesting to see if other chimeric flavivirus vaccines (based on the YF 17D, DEN2, or DEN4 virus backbone) have similar or better safety profiles when inoculated into monkeys by the i.c. route.
Immunogenicity and protection. Neutralizing antibodies directed against the E protein are considered the principal mediator of immunity against flavivirus reinfection, and the demonstration of neutralizing antibodies is a surrogate of protective immunity. Studies with animal models, principally mice, have demonstrated that the transfer of immune serum raised against E proteins of DEN2, YF 17D (3), and tick-borne encephalitis (42) viruses and monoclonal antibodies against JE virus (52) conferred protection against challenge with WT viruses. The level of protection in immunized mice was directly proportional to the passive titer of neutralizing antibodies. Adoptive transfer of T cells is also effective in preventing illness in recipient mice, indicating that effector T cells can limit the severity of disease by eliminating infected cells. Thus, an effective vaccine should elicit durable neutralizing-antibody responses, as well as strong memory B- and T-cell responses. A live vaccine against DEN virus would therefore be more efficient than inactivated or subunit vaccines, which require multiple doses, do not provide strong memory or durable immunity, and do not efficiently elicit major histocompatibility complex class I restricted T-cell responses.
All monkeys became viremic on inoculation of a single dose of one of four tetravalent formulations. All four serotypes were detected in all four groups, with the exception of serotype 1, which was not detected in any animals of group 2. The magnitude of viremia in the high-dose group formulation (group 1) was similar to that of the low-dose group (group 4) (2.6 and 2.5 log units, respectively). However, the duration of viremia in the low-dose group was longer than in the high-dose group (5.5 and 4 days, respectively). A similar inverse relationship between dose and duration of viremia has been reported previously in monkeys inoculated with YF 17D (10) or in humans vaccinated with ChimeriVax-JE virus (35, 38). The onset of viremia was, however, dose dependent (i.e., early in the high-dose group [days 2 to 7] and late in the low-dose group [days 3 to 11]). It is possible that higher doses of vaccine induce a stronger innate immune response that limits the replication of the virus. When the concentration of one or two components in a tetravalent formulation was reduced, the corresponding viruses were either undetectable (e.g., DEN1 virus in group 2) or detected very late (e.g., DEN4 virus in groups 2 and 3). As observed in i.c.-inoculated monkeys, the DEN3 and DEN4 virus chimeras induced viremia with a higher magnitude and duration than did the equivalent concentrations of the DEN1 and DEN2 virus chimeras. Interestingly, the DEN3 and DEN4 virus chimeras maintained the WT envelope genes during vaccine manufacture whereas the DEN1 and DEN2 virus chimeras acquired one or two mutations within their envelope genes that may have contributed to their lower viremia in monkeys. There is a theoretical risk that ChimeriVax-DEN1-4 viruses could infect insects (mosquitoes, biting flies) or ticks feeding on the blood of vaccinated subjects and be transmitted to animals or other humans on refeeding. If secondary transmission were to occur, there is an associated risk of virus mutation and reversion to virulence. This theoretical risk is considered minimal according to the results of vector studies with mosquitoes (22) and the findings in both monkeys and humans of low viremia titers which were invariably below the threshold for oral infection of mosquito vectors, which is generally in the range of 3.5 to 5.0 log10 PFU/ml (19). It is thus highly unlikely that viremia caused by the tetravalent vaccine in humans would result in mosquito infection.
All monkeys developed neutralizing-antibody responses to all four DEN WT virus serotypes in the high- and low-dose groups (groups 1 and 4, respectively). The GMTs in the high-dose group animals, measured on days 31 and 121 (data not shown) postimmunization, were higher than those in the low-dose group. When the concentration of the DEN4 (but not DEN3) virus component was reduced, three monkeys in group 2 and two monkeys in group 3 did not seroconvert to the DEN4 virus. Since both DEN3 and DEN4 virus chimeras contain authentic WT envelopes, it is possible that in a tetravalent formulation, similar doses of both viruses need to be present to avoid any dominance or interference in terms of immunogenicity. Interestingly, when empirically derived live attenuated DEN virus vaccines (LAV) were inoculated into humans, the dominant virus in terms of viremia and neutralizing-antibody response was the DEN3 virus component (23, 47). It is likely that the DEN3 LAV was either underattenuated or accumulated quasispecies of viruses with a WT envelope.
To assess the protective efficacy of tetravalent vaccine formulations, monkeys were randomly reassigned to four new groups at 6 months postimmunization, and each group was challenged with one WT heterologous serotype. All vaccinated monkeys were protected (as shown by the lack of viremia postchallenge), with the exception of two animals (one in the DEN1 and one in the DEN4 virus challenge group), whereas all control group animals (naive animals) became viremic. At the time of challenge, unprotected animals had PRNT50 of 20 and <10 against the DEN1 and DEN4 viruses, respectively. The unprotected monkey (F21149M) in the DEN1 virus group had not become viremic to DEN1, DEN3, or DEN4 virus chimeras and showed only a single viremia day (day 10) for DEN2 virus after vaccination. Consequently, this monkey was seronegative (PRNT50, <10) for the DEN3 and DEN4 virus chimeras and had developed a very low titer (PRNT50, 1:10) for the DEN1 virus by day 31 postimmunization. Similarly, monkey F20977, which was only partially protected against DEN4 challenge, had not become viremic for the DEN1 or DEN4 virus chimeras postvaccination. It was seronegative for DEN4 and borderline positive (PRNT50, 10) for DEN1 virus 1 month postimmunization. However, both DEN2 and DEN3 virus serotypes had been detected for several days in this monkey, which resulted in the development of high levels of neutralizing antibodies (PRNT50, of 2,560 and 320 against DEN2 and DEN3 viruses, respectively). These data suggest that in general, a serum PRNT50 of >20 can protect the host against a high dose (5 log10 PFU) of a challenge virus. The actual protective titer may even be lower than 1:20, considering that in natural infections only 3 log units of virus may be inoculated into the host by a mosquito vector. Moreover, there might be cases in which an individual host can be protected with a serotype specific PRNT50 of 20 (i.e., monkey F212117F, F21544M, and F21534M). This may indicate that other factors, such as the presence of DEN virus cross-neutralizing antibodies (25) or T-cell memory, may contribute to the protection of tetravalent vaccine-inoculated humans against dengue disease. It is possible that the challenge virus that enters the cells but escapes neutralization by antibodies is eliminated by a second line of defense provided by cytotoxic T lymphocytes. The E proteins of DEN virus (27), as well as capsid (11) and nonstructural (NS) proteins (28, 32) expressed by infected cells, serve as targets for cell killing by cytotoxic T lymphocytes. It is also important to mention that there was no enhancement of infection in the two unprotected animals. This could be an important safeguard in the application of a tetravalent vaccine, where a low level of antibody to one serotype is thought to be implicated in developing DHF/DSS by antibody-dependent infection enhancement phenomena. The levels of neutralizing antibodies postchallenge were significantly higher than those prechallenge or in control groups, indicating an immunological memory provided by B and T cells having high-affinity antigen receptors in immunized monkeys.
The ChimeriVax technology offers a good probability of successful DEN virus vaccine development. The vaccine attributes include the potential for single-dose application, absent or minimal reactogenicity, durable immunity, reduced potential for interference between the individual components in a tetravalent formulation, and low cost of manufacture. Based on our data, the ideal tetravalent formulation(s) for humans may contain equal amounts of each of four ChimeriVax-DEN virus serotypes (e.g. 5,5,5,5 or 4,4,4,4 log10 PFU) and could be administered twice, e.g., at a 6-month interval, similar to LAV vaccine (47), to ensure 100% seroconversion.
|
|
|
ACKNOWLEDGMENTS |
|---|
We thank R. Tesh and the late R. Shope, University of Texas Medical Branch, Galveston, Tex., for prescreening of monkey sera, as well as K. Eckels and W. Sun, WRAIR, Silver Spring, Md., for providing wild-type DEN challenge viruses. We also thank N. Tobin for animal care and P. Papastathis (both at Acambis, Inc., Cambridge, Mass.) for cell culture support.
|
FOOTNOTES |
|---|
|
REFERENCES |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
