The adenovirus E4orf4 protein expressed at high levels kills cancer cells but not normal human primary cells. Previous studies suggested that disruption of processes that regulate mitosis may underlie E4orf4 toxicity. Here we have used live imaging to show that E4orf4 induces a slowed defective transit through mitosis, exhibiting a delay or often failure in cytokinesis that may account for an accumulation of G 1 tetraploids in the population of dying E4orf4-expressing cells.
E4orf4 is a human adenovirus early protein that functions during infection to facilitate progression to the late phase (1-7); however, when overexpressed alone in cancer cells, it induces some accumulation of 4n (tetraploid) cells (8) and slow cell death, perhaps involving mitotic catastrophe (9-14). We showed previously that E4orf4-expressing tumor cells with both 2n and 4n DNA content are arrested in G 1 and die (8). This arrest in G 1 seems due to an inability to initiate new rounds of DNA synthesis (8). E4orf4's effects require an interaction with the B55 class of protein phosphatase 2A regulatory subunits (9, 14-17), and we have found that at high levels, E4orf4 inhibits PP2A B55 (18) to induce effects on anaphase-promoting complex (APC) and perhaps other processes that may disrupt normal passage through mitosis (18,19).In this report, we have specifically addressed the defect in mitosis by performing a series of live-cell imaging studies by timelapse microscopy on asynchronously growing as well as synchronized cells. H1299 cells expressing mCherry-tagged histone 2B (H2B) (H1299 cells that had previously been stably transfected with H2B-mCherry in pcDNA3 [catalog number 20972; Addgene, MA, USA]) (20) were infected at a multiplicity of infection (MOI) of 50 with adenoviral vectors expressing either E4orf4 or the reverse tetracycline-controlled transactivator (rtTA) protein. Construction of these viral vectors has been described previously (9,11). Following an 18-h expression period, the cells were plated on 6-well plates in medium containing 20% serum, and time-lapse microscopy was performed using an AxioVision 3 microscope equipped with AxioCam HR (Zeiss, Thornwood, NY) digital camera. Pictures were taken at 15-min intervals over 18 h at a ϫ10 magnification using both phase-contrast microscopy (to observe whole-cell morphology) and fluorescence microscopy (to observe histone 2B and hence the cell nuclei). For each condition, 50 cells were randomly chosen, monitored through the entire time course, and scored for their phenotypes. Cells were scored for the time required to round up (considered to be the start of mitosis), the time required from the first rounded phenotype to the appearance of two nuclei (telophase), the time required for the doubly nucleated cells to separate (cytokinesis), and the time required for the daughter cells to return to their flat morphology (considered returned to G 1 ).As summarized in Fig. 1A, asynchronously growing cells infected with the control rtTA viral vector mostly proceeded rapidly through M phase, a...