UBC-13, UEV-1, and CHN-1 mediate K63-linked poly-ubiquitination of VPS-34. (A) Ubiquitination of recombinant FLAG–VPS-34 and FLAG–BEC-1 was examined in vitro. FLAG IP was performed followed by detection of ubiquitination with anti-HA antibodies. “Long” and “short” designate long exposure time and short exposure time, respectively. Ubiquitination of VPS-34 but not BEC-1 was observed (compare lanes 1 and 5 with 6). VPS-34 ubiquitination was completely abolished when CHN-1 (lane 2) or UBC-13 (lane 4) was removed. (B) VPS-34 but not VPS-34 (255–901), which lacks the C2 domain that binds CHN-1, was ubiquitinated in vitro by UBC-13–UEV-1–CHN-1. (C) Ubiquitination of VPS-34 was examined in vitro using different forms of HA-ubiquitin (WT, K48R, and K63R). Inclusion of K63R- but not K48R-ubiquitin disrupted poly-ubiquitination of VPS-34 (compare lanes 1 and 2 with 3). “Long” and “short” designate long exposure time and short exposure time, respectively. (D) Ubiquitination of VPS-34 was examined in C. elegans. Schematic illustration of the ubiquitination assay is shown at the left. Worm lysates were prepared from strains generated by CRISPR-Cas9 that carry both flag::vps-34 and ha::ubq-2 or only ha::ubq-2. FLAG IP was performed followed by detection of ubiquitination with anti-HA antibodies. The arrow indicates FLAG-VPS34. The band indicated by the asterisk may represent nonseparated heavy and light chains of IgG. WB, Western blot. (E) Ubiquitination of VPS-34 was examined in WT, ubc-13(tm3546), and chn-1(by155) worms carrying both flag::vps-34 and ha::ubq-2. The graph shows quantification of the level of ubiquitination using ImageJ (version 1.48v). The ratio of ubiquitin versus FLAG-VPS-34 was determined and normalized to 1-fold in WT. At least four independent experiments were performed and data are shown as mean ± SD. The unpaired t test was performed to compare mutant datasets with WT. **, P < 0.0001.