Optineurin and myosin VI colocalize in NRK cells. Endogenous optineurin localizes to the Golgi complex using a pAb to optineurin (a) and an mAb to TGN38 (b). Optineurin and myosin VI are localized at the Golgi complex using a pAb to optineurin (d) and an mAb to myosin VI (e). Colocalization in vesicular structures close to the plasma membrane was demonstrated with the pAb to optineurin (g) and by transiently overexpressing myosin VI tagged with GFP (h). White boxes indicate areas enlarged in the pictures below (g′, h′, and i′). Endogenous myosin VI, visualized using the pAb against the whole tail, is recruited into aggregates containing overexpressed GFP-optineurin (j–k). The merged images are shown in c, f, i, i′, and l. Bars, 10 μm.

Knockdown of optineurin in NRK cells causes loss of myosin VI from the Golgi complex. Mock-transfected (a, b, e, and f) or siRNA-transfected (c, d, g, and h) NRK cells were used for immunofluorescence and were double labeled with antibodies to the globular tail of myosin VI (a, c, e, and g) and antibodies to TGN38 (b and d) or GM130 (f and h). Asterisks mark the position of the nucleus. Bars, 10 μm.

Myosin VI colocalizes with Rab8 at the Golgi complex and the plasma membrane. In NRK cells transiently overexpressing Rab8 tagged with GFP (b and e) myosin VI was detected with an antibody to the globular tail (a) or to the whole tail (d). Myosin VI and Rab8 are found in the same labeled structures/vesicles at the Golgi complex (c) and at the plasma membrane (f). Arrows indicate vesicles containing both proteins. Bars, 10 μm.

The transport of VSV-G to the cell surface is dramatically reduced in optineurin-depleted HeLa cells. Mock-treated or siRNA-treated HeLa cells were transfected with ts045-VSV-G-GFP to measure the rate of exocytosis . VSV-G at the cell surface was detected in indirect immunofluorescence using an mAb to the luminal domain of VSV-G and total VSV-G expressed was detected using a pAb to GFP. (A) Representative cells for the 40- and 100-min time points are shown to compare the amount of VSV-G on the cell surface in mock- and optineurin-depleted cells. (B) The ratio of cell surface over total VSV-G fluorescence was measured as described in Materials and methods to determine the rate of transport from the Golgi complex to the cell surface. Error bars: ±SEM. Bars, 10 μm.

Optineurin is a myosin VI–binding partner. (A) A carton showing the major regions in optineurin and myosin VI tail. Optineurin contains an NH₂-terminal Rab8-binding domain (aa 141–209; Hattula and Peranen, 2000) and a COOH-terminal myosin VI–binding domain (aa 412–520). Myosin VI tail has a proposed helical tail domain (HT) and a globular tail (GT), which contains the optineurin-binding site RRL in position 1107–1109. The putative phosphorylation site TINT is at position 1088–1091. (B) Using the ProQuest yeast two-hybrid screen optineurin was identified as a myosin interacting protein. The interaction was verified using the mammalian two-hybrid system. To narrow down the optineurin-binding site on the myosin VI tail the whole tail (T; column 1), the helical tail (HT; column 2) or the globular tail (GT) with or without the large insert (GT+LI or GT-LI; columns 3 and 4) were fused to the Gal4 DNA-binding domain and coexpressed with full-length optineurin fused to the VP16 activating domain in CHO cells. Luciferase activity is shown relative to control cells containing only the myosin VI construct with the empty prey vector. Deletion of myosin VI tail at residue 1117 (TΔC; column 5) did not abolish binding. Mutating three adjacent amino acids each (WKS, KVY, RRL, and REE to AAA) revealed RRL as the optineurin-binding site (columns 6–9). Two threonine residues in positions 1088 and 1091 (TINT) were mutated to either alanine (AINA) or glutamate (EINE; columns 10 and 11). (C) Optineurin binds to purified myosin VI tail. A pull down was performed using in vitro–translated optineurin and GST-myosin VI tail. ³⁵S-labeled in vitro–translated full-length optineurin (lanes 4 and 5), Dab2 as a positive control (lanes 2 and 3) or luciferase as a negative control (lanes 6 and 7) were incubated with either 5 μg GST alone (lanes 2, 4, and 6) or with 5 μg GST-myosin VI tail (lanes 3, 5, and 7). Lane 1 shows 50% of the input used for the pulldown with [³⁵S]optineurin in lane 5. The band below Dab2 is possibly caused by degradation (lane 3). (D) Co-immunoprecipitation of endogenous myosin VI and optineurin from the cytosol of A431 cells. Immunoprecipitation was performed under native conditions using no antibody, only protein A beads as a control (lane 2, blank), an antibody to full-length optineurin (lane 3, a-Optn) or a pAb to the tail of myosin VI (lane 4, a-MVI). Lane 1 is equivalent to 5% of the input used for each immunoprecipitation. Immunoprecipitates were run out on a 10% PAGE gel, blotted, and analyzed using an antibody to optineurin. (E) Constitutive secretion in Snell's waltzer fibroblasts cannot be rescued by overexpression of the mutant myosin VI lacking the optineurin-binding site (ΔRRL). Rescue experiments were performed by generating Snell's waltzer cell lines stably expressing only GFP (SV), the mutant myosin VI without the optineurin-binding site (ΔRRL), or wild-type whole myosin VI (MVI). Levels of secretion were compared with wild-type cells stably expressing only GFP (WT). To measure constitutive secretion, a construct, which expresses a SEAP was transfected into the various Snell's waltzer cell lines and the wild-type fibroblasts. The amount of alkaline phosphatase secreted into the media was measured 24, 48, and 72 h after transfection and plotted as a percentage of total maximal secretion in wild-type cells after 72 h. Transfection efficiency was normalized by cotransfection of a control plasmid expressing a red fluorescent protein. The data from two separate experiments run in triplicate are shown.

Optineurin is depleted from NRK and HeLa cells using siRNA. HeLa (A) or NRK (B) cells were either mock transfected with water or transfected twice at 48-h intervals with siRNA specific to optineurin. After 4 d cells were blotted and probed with antibodies to myosin VI, optineurin, or actin (A, a and B, a). In a parallel experiment mock-transfected, siRNA-transfected NRK, or HeLa cells were used for immunofluorescence and double labeled with antibodies to optineurin (A, b and d; B, b and d) and GM130 (A, c and e; B, c and e). Bars, 10 μm.

The loss of optineurin results in Golgi fragmentation. Mock (a–c) or optineurin siRNA-treated HeLa cells (d–f and d′–f′) were double labeled in immunofluorescence experiments with TGN46 (b, e, and e′) and GM130 (a, d, and d′). White boxes indicate areas enlarged in the pictures below (d′, e′, and f′). The Golgi fragments contain both marker proteins as indicated by the yellow color in the merged images (c, f, and f′). The arrows highlight the overlap between the marker proteins in (d′–f′). Bars, 10 μm. TEM analysis of mock (g) or siRNA-transfected (h) HeLa cells. Arrows indicate the position of Golgi stacks that can be found in both mock-transfected and siRNA-transfected cells. Bars, 200 nm.

The recruitment of myosin VI to tubular vesicular structures formed after overexpression of the constitutively active form of Rab8 is lost in optineurin-depleted cells. The constitutively active form of Rab8 (GFP-Rab8-Q67L) overexpressed in NRK cells localizes with myosin VI at the Golgi complex (a and b) and recruits myosin VI onto tubular and vesicular structures emerging from the Golgi complex (e and f). In optineurin-depleted NRK cells, which are overexpressing GFP-Rab8-Q67L myosin VI no longer colocalizes with Rab8-Q67L at the Golgi complex (c and d) and is not recruited to tubular and vesicular structures positive for Rab8-Q67L (g and h). White boxes highlight the areas of the cells enlarged in the pictures below (e′,f′, g′, and h′). Endogenous myosin VI was detected with either a pAb to the globular tail (a, c, g, and g′) or to the whole tail (e and e′). Bars, 10 μm.

A cartoon illustrating the possible interaction of optineurin with motor protein complexes at the Golgi complex. Optineurin may play a central role in coordinating actin-based and microtubule-based motor function for maintaining Golgi morphology. The optineurin-binding partner huntingtin has been shown to interact with HAP1, which in turn was reported to form a complex with the dynactin subunit p150^glued and modulate/regulate the dynein–dynactin complex. Loss of minus end–directed dynein motor activity causes in fragmentation of the Golgi ribbon structure. On the other hand loss of myosin VI results in a reduction in the size of the Golgi complex, because now the dynein complex is still active and able to retract the Golgi complex toward the MTOC. However, if optineurin is depleted, both motor complexes are nonfunctional and the Golgi complex is fragmented.