Supplemental Material

This article contains the following supporting material:

• Figure 1S. Clathrin is recruited to the reforming medial-Golgi apparatus in NRK cells. - NRK cells were treated with 1% 1-BtOH for 30 min followed by replacement with control medium for the indicated times. Cells were double labeled using antibodies specific to clathrin (red) and ManII (green). In untreated cells clathrin and ManII displayed minimal co-staining (A-C). In response to BtOH treatment, clathrin staining became dispersed (D) and ManII was highly punctate (E) and localized to clathrin negative structures (F, inset). ManII assumed an intermediate tight morphology (arrowheads) following alcohol removal and co-localized with clathrin (G-I and J-L). Panels M-O: ManII regained its normal co-localization with clathrin after 90min washout (inset). All micrographs are projected Z-series images taken using a Leica confocal microscope (Methods). Overlap of single sections yielded identical results. Scale bar, 10ƒÝm. Panel P: Quantitation of Golgi volumes in untreated, BtOH treated and washout NRK cells (Materials and Methods). C: untreated; 30B: 30min BtOH treatment; 20R: 20min washout (Recovery); 40R: 40min washout (Recovery).

  In contrast to control untreated cells, at early times during BtOH washout the tight ManII-immunoreactive structures colocalized with clathrin. By 90min of washout, ManII and TGN38 (data not shown) staining resumed the characteristic control lace-like appearance, showing that the alcohol effect is completely reversible. Significantly, the normal segregation between clathrin and ManII was restored.

• Figure 2S. The presence of the AP180 clathrin binding domain does not interfere with medial-Golgi morphology, but inhibits its recovery from BtOH treatment. - (A, B) NRK cells that were transiently transfected with cDNA encoding the Flag-tagged clathrin binding domain (CBD) of AP180. Transfected cells were detected using a polyclonal anti-Flag antibody and the Golgi apparatus was visualized with an anti-ManII antibody. (A) Untreated: the characteristic lace-like distribution and juxtanuclear localization of ManII was not disrupted by the presence of the CBD. (B) At 90min following BtOH washout, cells containing the CBD consistently displayed a tight Golgi morphology. (C) Quantitation of normal, tight and fragmented Golgi morphologies in untransfected or CBD transfected NRK cells (see Figure 3 for details). (D) Kinetic analysis of Golgi recovery in the absence and presence of CBD. Cells were treated as in Figure 3, and at the indicated times after alcohol removal the Golgi morphologies were quantified in untransfected cells (-CBD) and in those expressing the CBD.
• Figure 3S. Knockdown of the clathrin heavy chain delays Golgi apparatus reformation. - NRK cells were transfected with non-silencing siRNA or CHC siRNA (rchc2) for 3 days, after which they were either untreated or treated with 1% 1-BtOH for 30min followed by washout in normal medium for 60min (60R). Asterisk, clathrin knockdown cells and corresponding Golgi structures. Note that as determined by the X22 antibody, clathrin knockdown appeared complete. Scale bar, 20μm.
• Figure 4S. Golgi reformation is normal in mock transfected NRK cells. - NRK cells were mock transfected for 3.5 days and either untreated or treated with 5μg/ml BFA for 40min (40min BFA), followed by 90min washout (90R) after which Golgi morphology was determined using either (A) a mouse monoclonal anti-ManII and rabbit anti-clathrin antibody or (B) a rabbit anti-GRASP65 antibody in conjunction with the monoclonal anti-CHC, X22 antibody. ManII redistributes into the ER (f) while GRASP65 redistributes independently of the ER (h) and maintains a loose juxtanuclear distribution in response to BFA treatment. Both ManII and GRASP65 recover their normal juxtanuclear distribution following BFA washout (i-l). Scale bar, 20μm.
• Figure 5S. Characterization of isolated clathrin coated vesicles. - Ten μg of protein from GH3 total homogenates (TH) or the CCV fraction (CCV) were resolved by SDS-PAGE, transferred to PVDF membranes and probed with antibodies against CHC, Man6PR and β-COP. CHC and Man6PR are enriched in the CCV preparation, while β-COP is absent. The cross-reactive lower molecular band in the CCV fraction is non-specific. Man6PR: mannose 6 phosphate receptor; β-COP: β subunit of coatomer I.
• Figure 6S. Limited clathrin recruitment to the reforming Golgi apparatus following nocodazole or BFA washout. - (A) NRK cells were treated with 30 μM nocodozole for 90min followed by washout for 20 and 40min. Cells were stained with rabbit anti-CHC antibody (red) and mouse monoclonal 53FC3 anti-ManII (green) followed by appropriate secondary antibodies. Nuclei were stained with Hoechst 33342 (blue). a-c: untreated cells; d-f: nocodazole treated cells; g-l: nocodazole washout cells. ManII-immunoreactive Golgi ministacks generated by nocodazole treatment do not contain clathrin immunoreactive material (d-f). Following nocodazole removal, ManII positive structures coalesce in the juxtanuclear region and clathrin is recruited to only a subset of the population (g-l, arrowheads). (B) NRK cells were treated with 5 μg BFA/ml for 40min, after which the drug was removed for the indicated times. a-c: untreated cells; d-f: BFA treated cells; g-l: cells following BFA removal. Upon BFA treatment ManII looses its Golgi localization and redistributes to the ER (d-f). During BFA washout ManII regains a Golgi-like distribution via large vesicle clusters that lack significant clathrin-immunoreactive staining (g-l). Overlap of single sections yielded identical results. Scale bar, 10μm.

  Following nocodazole washout the low level of co-segregation between clathrin and ManII is consistent with a minimal role of clathrin/ CCVs in reorganization of ministacks. This confirms our CHC siRNA data which show that diminished CHC has a negligible effect on normal Golgi reassembly (Figure 7). In addition, the lack of colocalization between ManII and clathrin following BFA washout, supports our model whereby clathrin is not required for ManII exit/ traffic between ER and Golgi, but rather for the formation of the Golgi cisternae (Figure 6).

• Figure 7S. Clathrin heavy chain knockdown inhibits transferrin endocytosis. - (A) HeLa cells were either mock transfected or treated with a control, non-silencing siRNA (C siRNA) or with a specific CHC siRNA (hchc2) for 3days; hchc2 is the human equivalent the rchc2 CHC siRNA used in NRK cells. Ten ug of total homogenates were loaded onto SDS-PAGE followed by transfer to PVDF and blotting with the indicated antibodies. The bands were quantified using Quantity One$reg;. (B) HeLa cells were transfected with hchc2 for 3days, after which they were incubated with either transferrin-FITC or fluorescein-labeled dextran for 20min and 3 hours, respectively, (Materials and Methods), followed by processing for immunofluorescence microscopy with a polyclonal anti-clathrin antibody (red). Scale bar, 20μm.