Dengue virus is an enveloped positive-strand RNA virus with a genome =:11 kilobases in length. The four serotypes of dengue virus are currently the most important members of the flavivirus family in terms of geographical distribution and the incidence of infection in humans. In this communication we describe successful cloning of a stable full-length cDNA copy of dengue type 4 virus that can be used as the template for in vitro transcription of infectious RNA. Evidence is presented that dengue virus recovered from permissive cells transfected with the in vitro RNA transcripts retained a mutation that was engineered into full-length cDNA. The properties of the virus produced by cells transfected with infectious RNA transcripts of dengue cDNA resembled those of the virus from which the cDNA clone was derived. The dengue virus recombinant DNA system should prove helpful in gaining a better understanding of the molecular biology of dengue viruses and should facilitate the development of a safe and effective live vaccine for use in humans.

Dengue type 4 virus (DEN4) cDNA was used as a vector to express genes of the distantly related tick-borne encephalitis virus (TBEV). Full-length chimeric TBEV/DEN4 cDNAs were constructed by substituting TBEV genes coding for proteins such as capsid (C); pre-membrane, which is the precursor of membrane (M); envelope (E); or nonstructural protein NS1 for the corresponding DEN4 sequences. RNA transcripts prepared from cDNAs were used to tr ect permissive mian cells. Two viable chimeric viruses that contained TBEV CME or ME genes were recovered. (11), we initiated the construction of chimeric viruses containing sequences from dengue type 4 virus (DEN4) and the distantly related TBEV, which belongs to another serotype complex of the Flaviviridae (12). DEN4 and TBEV have the same genome organization and share the same strategy of gene expression, but comparison of sequences between the two viruses indicates that the homology is relatively low (13). The immediate objective of our study was to determine whether any TBEV/DEN4 gene constellation could produce a viable chimeric virus and to characterize those viable chimeric viruses that were recovered. Two chimeric viruses that contained the capsid/membrane/ envelope (CME) or ME structural protein genes of TBEV were viable and exhibited the antigenicity of TBEV. MATERIALS AND METHODSChimeric TBEV/DEN4 cDNA. Previously, subgenomic cDNA fragments ofTBEV (strain Sotjin) were cloned and the nucleotide sequence was determined (13). Plasmids pGEM2-CME, containing nucleotides (nt) 76-1977, and 10532The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.

The dengue type 4 virus (DEN4) genome contains a 384-nucleotide (nt) 3 noncoding sequence in which the last 81 nt, predicted to form a secondary structure, are thought to be essential for virus replication. Immediately upstream of the secondary structure, short RNA sequences that are conserved among mosquito-borne flaviviruses have been identified. A series of deletions that range from 30 to 262 nt were introduced into this upstream region of full-length DEN4 cDNA to create viable deletion mutants, some of which might prove to be useful for inclusion in a live attenuated virus vaccine. When studied by an infectious-center assay, most full-length RNA transcripts of the deletion constructs exhibited reduced infectivity when transfected into simian LLC-MK 2 cells compared with the full-length RNA transcripts of wild-type parental virus. Deletion mutations that extended as far as the 5 boundary of the 3 noncoding region and whose 3 boundary did not extend beyond the last 113 nt of the 3 end were viable. With the exception of mutant 3d 303-183, which contained a deletion of nt 303 to 183 from the 3 terminus, deletion mutants produced plaques that appeared late on simian LLC-MK 2 cells or exhibited a small-plaque morphology on mosquito C6/36 cells compared with the wild-type virus. These mutants also replicated less efficiently and attained a lower titer in LLC-MK 2 cells than parental wild-type virus. Significantly, mutant 3d 303-183 grew to a high titer and was least restricted in growth. Mutant 3d 303-183 and four other moderately to severely restricted mutants were selected for evaluation of infectivity and immunogenicity in rhesus monkeys. There was a suggestion that occurrence and duration of viremia were reduced for some of the deletion mutants compared with the wild-type virus. However, more convincing evidence for attenuation of some of the mutants was provided by an analysis of antibody response to infection. Mutant 3d 303-183 induced an antibody response equivalent to that stimulated by wild-type virus, whereas other mutants induced low to moderate levels of antibodies, as measured by radioimmunoprecipitation and virus neutralization. The immunogenicity of these 3 DEN4 deletion mutants in monkeys appeared to correlate with their efficiency of growth in simian LLC-MK 2 cells. One or more mutants described in this paper may prove to be useful for immunization of humans against disease caused by dengue virus.

Dengue virus contains an li-kilobase positive-strand RNA genome that codes for, in one open reading frame, three structural proteins (capsid, premembrane, and envelope), followed by seven nonstructural proteins. The structural protein genes of a full-length cDNA done of type 4 dengue virus were replaced with the corresponding genes of dengue 1 or dengue 2 to create intertypic chimeric cDNA. The RNA transcripts made from these templates were infectious when transfected into permissive cells in culture. Progeny ofchimeric cDNA produced apparently authentic dengue 1 or dengue 2 structural proteins, together with dengue 4 nonstructural proteins, and as a consequence exhibited type 1 or type 2 serological specificity. Both of the chimeras ultimately grew to the same titer as their type 1 or type 2 parent, but the type 2/type 4 chimera grew very slowly. This chimera also produced small plaques; in contrast, the type 1/type 4 chimera produced normal size plaques. The type 2/type 4 chimera retained the mouse neurovirulence of the dengue 2 virus, which was the source of its structural protein genes. Each of the mice inoculated intracerebrally with the chimera died, but survival time was prolonged. The retardation of replication of the type 2/type 4 chimeric virus suggests that this virus and possibly other intertypic dengue virus chimeras with similar properties should be examined for attenuation in primates and possible usefulness in a live dengue virus vaccine for humans.