During cellular invasion, human papillomavirus type 16 (HPV16) must transfer its viral genome (vDNA) across the endosomal membrane prior to its accumulation at nuclear PML bodies for the establishment of infection. After cellular uptake, the capsid likely undergoes pH-dependent disassembly within the endo-/lysosomal compartment, thereby exposing hidden domains in L2 that facilitate membrane penetration of L2/vDNA complexes. In an effort to identify regions of L2 that might physically interact with membranes, we have subjected the L2 sequence to multiple transmembrane (TM) domain prediction algorithms. Here, we describe a conserved TM domain within L2 (residues 45 to 67) and investigate its role in HPV16 infection. In vitro, the predicted TM domain adopts an alpha-helical structure in lipid environments and can function as a real TM domain, although not as efficiently as the bona fide TM domain of PDGFR. An L2 double point mutant renders the TM domain nonfunctional and blocks HPV16 infection by preventing endosomal translocation of vDNA. The TM domain contains three highly conserved GxxxG motifs. These motifs can facilitate homotypic and heterotypic interactions between TM helices, activities that may be important for vDNA translocation. Disruption of some of these GxxxG motifs resulted in noninfectious viruses, indicating a critical role in infection. Using a ToxR-based homo-oligomerization assay, we show a propensity for this TM domain to self-associate in a GxxxG-dependent manner. These data suggest an important role for the self-associating L2 TM domain and the conserved GxxxG motifs in the transfer of vDNA across the endo-/lysosomal membrane.A ll nonenveloped viruses are faced with a cellular membrane barrier through which they must transfer their genetic material to establish a successful infection. These viruses behave as metastable particles, rigid enough to survive transmission through the extracellular milieu but structurally poised to undergo conformational rearrangements or partial disassembly upon encountering specific factors during cellular entry, including the engagement of host cell receptors, proteolytic and chaperone activity, and low pH of the intracellular endosomal compartment. In response to these factors, capsids rearrange to expose hidden and often hydrophobic membrane translocation domains (MTDs), thereby ensuring that the coordinated process of membrane penetration occurs only when the viral capsid has reached the appropriate intracellular locale. Through convergent evolution, viruses have become equipped with a variety of different MTDs, designed to accomplish the feat of membrane penetration in several ways (recently reviewed in reference 1). Amphipathic helices, myristoylated and/or hydrophobic peptides, and phospholipase activity have all been documented to facilitate membrane penetration through pore formation, local membrane disruption, or gross fragmentation of membranes although many molecular details of these MTD-driven activities remain obscure and are ongoing subjects of inve...