A Capsid-Encoded PPxY-Motif Facilitates Adenovirus Entry

Wodrich, Harald; Henaff, Daniel; Jammart, Baptist; Segura-Morales, Carolina; Seelmeir, Sigrid; Coux, Olivier; Ruzsics, Zsolt; Wiethoff, Christopher M.; Kremer, Eric J.

introduced into Mason-Pfizer monkey virus, was previously shown to abolish late domain functions with no apparent structural changes, which would impair virus assembly [35, 36] . To control for unintended mutations introduced during the cloning, we reverted the PGAA sequence back to PPSY (Ad5-VI-wt). Because the ,360 copies of protein VI appears to be in contact with several proteins in the mature capsid, modifications that disrupt the tertiary structure could also affect the capsid composition. In large-scale preparations, mutant and wt virus banded at identical densities and gave similar yields of particles as determined by genome and protein quantification. A biochemical analysis of the capsid composition of purified viral particles showed no apparent differences between wt and mutant viruses (Figure 2A and data not shown). To confirm that viral capsid integrity between mutant and Figure 1 . Timecourse of protein VI release during Ad entry. Ad5-488 was pre-bound to hTERT-RPE1 cells at 4uC (top row) and shifted to 37uC for 5 min (second row), 15 min (third row) and 45 min (bottom row) as indicated. Protein VI was detected using anti-protein VI antibodies (left column), adenoviral particles by detecting the Alexa-488 fluorescent signal (second column) and the MTOC by staining the primary cilia (third column). A composite of all three signals including the nucleus (in grayscale) is shown to the left. The inset shows an inverted magnification of representative virus, protein VI and the primary cilia signals from the small dashed box. In the composite protein VI signals are shown in red, Ad is shown in green, the primary cilia in grey and the nucleus in blue. Colocalization of protein VI and Ad appears as yellow. The scale bar is 10 mm. doi:10.1371/journal.ppat.1000808.g001 Figure 2B capsid integrity and morphology of the mutant virus was indistinguishable from the wt virus. In contrast, the infectious versus physical particle ratio of Ad5-VI-M1 was ,20-fold lower than Ad5-VI-wt as assayed by plaque formation on monolayers of 911 cells ( Figure 2C ). Because infectious vs. physical particles can vary between preparations, we assayed plaque size, which is more informative measurement of propagation rate. Plaques were significantly smaller for Ad5-VI-M1 versus Ad5-VI-wt (see below), suggesting that the altered PPxY domain affects some stages of virus propagation.