We previously studied the processing of dengue virus nonstructural protein NS1 by in vitro transcription/translation, and by in vivo expression using vaccinia virus as a vector. In vitro, the full length NS1-NS2A precursor was made, and was translocated into dog pancreas microsomal membranes and glycosylated, but NS1/NS2A cleavage did not occur. In vivo, brefeldin A blocked secretion of NS1 but did not inhibit NS1/NS2A cleavage. Taken together, these results suggested that NS1/NS2A cleavage occurs in the Golgi, or in a compartment between the ER and the Golgi. We have now shown that two independent methods that block ER to Golgi transport (the drug CCCP and incubation at 14degrees C) block NS1/NS2A cleavage. This confirms that NS1/NS2A cleavage does not occur until after NS1-NS2A exits the ER. Previously, we had also analyzed a recombinant virus expressing chimeric 72%prM-NS1(8)-NS2A protein, containing only the 8 C-terminal residues of NS1. The chimera was cleaved at the NS1/NS2A junction, indicating that only the last 8 amino acids of NS1 are required for cleavage. More recently mutants were constructed which lacked the C-terminal 31% of NS2A or the signal sequence. Both of these changes blocked cleavage of the chimeric protein, as expected from studies of authentic NS1/NS2A cleavage.
