There is concordant evidence that HBV budding takes place at intracellular membranes rather than at the plasma membrane. In regard to our microscopy studies, the viral envelope accumulated in crescent-shaped ER-associated structures where it actively recruited the viral core/capsid. This CS structure shares hallmarks of viral interaction platforms, as an envelopment-defective core mutant (C.K96A) failed to be attracted. Our results are in accord with recent confocal microscopy and CO-IP studies that demonstrated a lack of interaction between the HBV L protein and the C.K96A mutant [64]. Nonetheless, the C.K96A mutant has been shown to warrant the formation and secretion of genome-free empty virions, another HBV particle type with relatively unknown functions [65]. Unlike complete virions, empty virion assembly requires only the S envelope protein and may thus proceed via distinct mechanisms and pathways. Our observation that the characteristic recruitment of the viral core to the L-specific CS structure and even the formation of enveloped HBV-replicating nucleocapsids preserved upon the ER exit block led us to infer that HBV assembly initiates at the ER. Therefore, ERGIC-53 and the COPII coat proteins presumably come into play. However, their temporal and spatial modes of action are yet ambiguous. In particular, we still do not know whether Sec24A may act as a direct COPII adaptor for the viral envelope or as an indirect adaptor recognizing the envelope-associated ERGIC-53 cargo receptor. A bipartite use of adaptors and receptors has been suggested to amplify the cargo capture into vesicles [66].