PMC

Schematic representation of the putative bilobed subunit organization of the mammalian (large type) and yeast (small type) COG complexes.

Comparison of the subunit composition of the small and large Cog5-containing complexes. Partially purified COG complex (purified by ammonium sulfate precipitation, butyl-Sepharose chromotography and MonoQ chromatography) was subjected to ceramic hydroxyapatite chromatography. Samples of the fractions were separated by 10% SDS-PAGE and subjected to immunoblotting with antibodies specific to the indicated proteins as described in the legend to A. Labels are as in A.

Ultrastructure of the Golgi apparatus in wild-type, mutant, and revertant cells. Monolayers of cells were fixed in situ and epon embedded. Thin epon sections (40–50 nm) were viewed by transmission EM as described in the Materials and methods. Juxtanuclear regions with Golgi cisternae are shown: (A) ldlB, (B) ldlB[COG1] (mutation complemented by the corresponding cDNA), (C) ldlC, (D) ldlC[COG2], and (E) wild-type CHO. N, nucleus. Bars, 200 nm.

Coimmunoprecipitation of COG complex subunits. (A) Cog2 was immunoprecipitated from rat liver cytosol (lanes 2 and 3) with an anti-Cog2 antipeptide antibody in the absence (lane 2) or presence (lane 3) of the immunogenic peptide. The precipitates were fractionated by 6% SDS-PAGE and immunoblotted with the indicated antibodies. Lane 4 shows 10% of the starting cytosol sample. (B) Bovine brain COG complex, partially purified through the MonoQ chromatography step, was immunoprecipitated in the absence (lane 1) or presence (lane 2) of anti-Cog1 monoclonal antibody. The precipitates and 2.5% of the starting sample (lane 3) were fractionated by 6% SDS-PAGE and immunoblotted with antibodies to the indicated COG subunit or the Sec8 component of the mammalian Sec6/8 complex.

COG complex architecture. These anaglyph stereo images are best viewed with red/green stereo glasses (). Fixed COG (800 kD, 13S) is in the first and second rows. Unfixed COG is in the third and fourth rows. Fixed mammalian Sec6/8 complex is displayed in the fifth row (), with F-actin filament in the rightmost panel. Other calibration molecules are in the bottom row (left to right): thyroglobulin (670 kD, 19S), IgG (150 kD), dynactin (1,200 kD; ), clathrin triskelion (630 kD, 7S; ), and cytoplasmic dynein (820 kD, 20S). All panels are 375,000× when viewed as 1-inch squares.

Immunofluorescence localization of Cog1 and Cog7. (A) CHO cells (top) or HeLa cells expressing HA epitope–tagged Cog7 (bottom) were fixed and double stained with primary antibodies to Cog1 or the HA epitope tag (HA–Cog7; center) and mannosidase II (Man II; left) and secondary antibodies labeled with either Alexa^®488 (green)/Alexa^®568 (red; top) or Alexa^®488 (green)/Alexa^®546 (red; bottom). Confocal microscopic images were collected and are displayed either as a field representing the distribution of a single antibody (left and center) or as the merged images (right). Bars, 10 μm. (B) HeLa cells expressing HA epitope–tagged Cog7 were fixed and double stained with primary antibodies to the indicated COG complex subunits and the HA epitope tag (HA–Cog7) and secondary antibodies labeled with either Alexa^®488 (green) or Alexa^®546 (red). Images are presented as in A.

Steady-state levels of COG components and sizes of complexes in wild-type CHO, ldlB (Cog1), and ldlC (Cog2) cells. (A) Total cell lysates from the indicated wild-type CHO, ldlB, and ldlC cells, as well as ldlB and ldlC revertants corrected by transfection with the corresponding cDNA expression vectors (ldlB[COG1] and ldlC[COG2]) were fractionated by 7% SDS-PAGE and immunoblotted with antibodies to the indicated COG subunits or control proteins (β-COP and Sec8). (B) Cytosols from wild-type CHO (top), ldlB (middle), and ldlC (bottom) cells were fractionated by Superose 6 gel filtration chromatography. Samples of the fractions were analyzed by 7% SDS-PAGE and immunoblotted with the indicated antibodies. The position of the void volume and the elution positions of size standards (thyroglobulin, 669 kD; apoferritin, 440 kD) are shown at the top.

Identification of two Cog5-containing protein complexes in bovine brain cytosol. (A) Partially purified COG complex from bovine brain cytosol was subjected to ceramic hydroxyapatite chromatography (see Materials and methods). The eluted fractions were separated by 10% SDS-PAGE and either silver stained (top) or subjected to immunoblotting with anti-Cog5 antibody (bottom). “Load” is the sample before chromatography, and FT3 and FT7 are representative flow-through fractions; fractions 28–55 were collected during elution with a phosphate gradient. The positions of molecular weight markers (kD) are shown on the left. Peaks 1 and 2 represent Cog5-containing proteins eluting in fractions 31 and 33–36, respectively. (B) Peak 1 (9% of fraction 31) and peak 2 (2% of the pooled fractions 33–36) from the chromatogram shown in A were subjected to gel filtration chromatography on a SMART Superose 6 column followed by 10% SDS-PAGE and immunoblotting with the anti-Cog5 antibody. Load is the sample before chromatography. The void volume and the elution positions of size standards (thyroglobulin, 669 kD; catalase, 232 kD) are shown at the top. All samples loaded onto gels and immunoblots represent equivalent amounts within a chromatogram.

MiniQ purification of COG and identification of its subunits. (A) Peak 2 from the ceramic hydroxyapatite chromatogram in A was further purified (), and a portion of each fraction from the last chromatographic step (MiniQ) was subjected to 7.5% SDS-PAGE and silver staining (top; molecular weight markers [kD] shown on the left) or 10% SDS-PAGE and immunoblotting (bottom four panels) with antibodies to the indicated proteins. To resolve the individual bands, the 7.5% SDS-PAGE gel was run until proteins smaller than 60 kD ran off the gel. No additional components (or contaminants) were detected in the size range below 65 kD (verified by 15% SDS-PAGE; unpublished data). The load lane is the sample before chromatography. (B) Enlarged view of the lower portion of the lane containing fraction 19 in A. The indicated identities of the proteins were determined by mass spectrometry, as described in the text. The pairs marked by brackets were excised together and sequenced as a mixture. The protein marked with an asterisk is rabaptin-5, which is not part of the complex (A). The positions of molecular weight (kD) markers are shown on the left.