Time‐lapse recordings and analyses

In the CD47 blocking experiments, cells were cultured with either 1 mg/ml CD47 antibody (clone B6H12; BD, Franklin Lakes, NJ) or with 1 mg/ml mouse IgG1 isotype control antibody (BD) as previously described (Hobolt‐Pedersen et al., 2014). In the syncytin‐1 blocking experiments, cells were cultured with 5 μg/ml syncytin‐1 blocking peptide or with the corresponding concentration of scrambled syncytin‐1‐peptide (KJ Ross‐Petersen, Klampenborg, Denmark) with daily medium change as described in (Chang et al., 2004; Soe et al., 2011).

The chambered cover‐glass was subsequently placed in the incubation chamber of a confocal Olympus Fluoview FV10i microscope (Olympus Corporation, Shinjuku, Tokyo, Japan) with 5% CO₂ and 37°C for 4 days. Using the software of the microscope, three random sites for each of the eight wells were marked (total of 24 sites), and time‐lapse images were made every 21 min for 23 h using phase contrast. This procedure was repeated for 4 continuous days, and for each new recording three new sites was chosen for each well.

Subsequently, the time‐lapse recordings were analyzed by two observers and the recordings were carefully investigated for fusion events using the FV10‐ASW 4.1/4.2 Viewer software (Olympus). When a fusion event was identified it was characterized by the number of nuclei for each fusion partner and the type of fusion (phagocytic cup, broad contact surface, filopodia/tube or from the top) according to the definitions presented in (Soe et al., 2015). The first observer went through the video material, identified the individual fusion events, marked them on the video, and documented the above‐mentioned details for each fusion event. The second observer verified the marked fusion events, inspected the videos for further events that may have been missed by the first observer and documented the above‐mentioned details separately, without having access to the first observer's documentation. Hereafter, the first observer compared the two observer's data and made the final categorization.

Data was collected from six separate experiments, performed with cells isolated from six different donors. Three experiments were performed for each antagonist. For each data set all four wells per condition for each day were considered as a single data point. Thus, n = 16 for each data set used in the statistical analyses. There were no significant differences in the effects of inhibition between days. For the CD47 blocking experiment, a total of 251 videos reflecting a total of 4,664 h were analyzed; a total of 831 fusion events (control: 483; CD47 antagonist: 348) were observed. For the syncytin‐1 blocking experiment, a total of 235 videos reflecting a total of 5,317 h were analyzed; a total of 987 fusion events (control: 482; syncytin‐1 antagonist: 505) were observed. In order to allow a direct comparison between all experiments, the raw data were converted into the number of fusion events/mm²/h and subsequently normalized to the respective control condition.