Flaviviruses are a group of medically relevant pathogens, known to cause serious disease in animals and humans. The previously defined roles of the flavivirus non-structural protein 2A (NS2A) in RNA replication, and modulation of the host antiviral response, has recently been extended to include virus assembly and secretion. In West Nile virus subtype Kunjin (KUN), an Isoleucine (I)-to-Asparagine (N) substitution at position 59 of the NS2A protein blocked the production of secreted virus particles in cells electroporated with viral RNA carrying this mutation. In this study, prolonged incubation of mutant KUN NS2A-I59N replicon RNA, in an inducible BHK-derived packaging cell line (expressing KUN structural proteins C, prM, and E), generated escape mutants that rescued the secretion of infectious virus-like particles. Sequencing identified three groups of revertants that included (i) reversions to wild-type, hydrophobic Ile, (ii) pseudorevertants to more hydrophobic residues (Ser, Thr, and Tyr) at codon 59, and (iii) pseudorevertants retaining Asn at NS2A codon 59, but containing a compensatory mutation (Thr-to-Pro) at NS2A codon 149. Engineering hydrophobic residues at NS2A position 59, or the compensatory T149P mutation into NS2A-I59N replicon RNA, restored the assembly of secreted virus-like particles in packaging cells. T149P mutation also rescued virus production when introduced into the full-length KUN RNA containing an NS2A-I59N mutation. Immunofluorescence and electron microscopy analyses of NS2A-I59N replicon-expressing cells showed a distinct lack of virus-induced membranes normally present in cells expressing wild-type replicon RNA. The compensatory mutation NS2A-T149P restored the induction of membrane structures to a level similar to those observed during wild-type replication. These results further confirm the role of NS2A in virus assembly, demonstrate the importance of hydrophobic residues at codon 59 in this process, implicate the involvement of NS2A in the biogenesis of virus-induced membranes, and suggest a vital role for these induced membranes in virus assembly. To further our understanding of how mutations within NS2A are able to affect the induction of virus-induced membranes, leading to a block in virus assembly, the membrane topology of KUN NS2A was investigated. Using a plasmid encoding NS1 and NS2A proteins with C-terminal c-myc and FLAG epitopes, NS2A proteins containing N-linked acceptor sites and C-terminal truncations were generated. Assays were performed to identify the subcellular localization of specific sequences within NS2A by Western blot and immunofluorescence analyses. While the membrane topology could not be determined experimentally, the findings of this study support the assertion that cleavage at the NS1/NS2A junction requires the majority, if not all of the NS2A protein for proper processing to occur, and suggests that the interaction between hydrophilic loops and -helical transmembrane segments plays an important role in the formation and stability of the flavivirus NS2A protein topology. Based on the knowledge of polyprotein processing events, and utilizing a range of software packages, a topology model of NS2A was predicted. The likelihood of additional sequences within NS2A affecting the ability to induce virus-specific membranes, and facilitate virion assembly, has led to the development of an invasive bacterial screening system, as a delivery vehicle to screen libraries of mutated KUN replicon clones. Using these invasive bacteria to deliver mutated KUN replicons into BHK-derived packaging cells, mutations causing a deficiency in either RNA replication or encapsidation can be identified by performing -gal assays on cells maintained in the presence, or absence of Doxycycline (suppressing the expression of structural proteins), respectively. Furthermore, this system was adapted for use in a 96-well plate format, allowing for high-throughput screening. Thus, KUN replicon clones capable of RNA replication, but unable to assemble and secrete virus-like particles can be identified and further analyzed, in the hope of mapping amino acid residues and motifs involved in encapsidation of flavivirus RNA. Finally, a range of hypotheses are discussed, explaining the possible mechanisms by which NS2A is involved in flavivirus assembly. A number of future directions and applications are also presented.