Fig. 2. Localization of EB1 and its C-terminal binding partner APC to centrosomes is independent of intact cytoplasmic MTs. (A) MDCK cells (a-c) and MDCK cells treated with nocodazole to disrupt the MT cytoskeleton (a'-c') were immunolabelled for -tubulin (a,a'), APC (b,b') or EB1 (c,c'). Filamentous MTs (a), MT-dependent cortical APC clusters (arrowheads in b) and MT plus-end localization of EB1 (c) are disrupted by nocodazole (a', black arrowheads in b',c'). Localization of APC to cell-cell contact sites is independent of intact MTs and seems to be enhanced in nocodazole-treated MDCK cells (arrows in b,b'). Bar, 50 µm. (B) Sections of nocodazole-treated MDCK cultures shown in (A) were stained for centrosome markers -tubulin (green in a-c) or pericentrin (red in d-f) and co-stained for APC (red in a-c) or EB1 (green in d-f). APC and EB1 remain localized around the centrosome in nocodazole-treated cells, whereas their MT-dependent localizations are disrupted (see A). Images in d-f show an example in which EB1 localizes to both centrosomes of a cell. Bar, 2 µm. (C) Section of a nocodazole-treated U-2 OS cell expressing GFP-APC (green in a-c) stained with the centriolar marker centrin (red in a-c). GFP-APC preferentially localizes to one of the two centrioles in this cell [Movie 2 (http://jcs.biologists.org/supplemental)]. Bar, 2 µm.