The Sar1 GTPase coordinates biosynthetic cargo selection with endoplasmic reticulum export site assembly - PubMed (original) (raw)

The Sar1 GTPase coordinates biosynthetic cargo selection with endoplasmic reticulum export site assembly

M Aridor et al. J Cell Biol. 2001.

Abstract

Cargo selection and export from the endoplasmic reticulum is mediated by the COPII coat machinery that includes the small GTPase Sar1 and the Sec23/24 and Sec13/31 complexes. We have analyzed the sequential events regulated by purified Sar1 and COPII coat complexes during synchronized export of cargo from the ER in vitro. We find that activation of Sar1 alone, in the absence of other cytosolic components, leads to the formation of ER-derived tubular domains that resemble ER transitional elements that initiate cargo selection. These Sar1-generated tubular domains were shown to be transient, functional intermediates in ER to Golgi transport in vitro. By following cargo export in live cells, we show that ER export in vivo is also characterized by the formation of dynamic tubular structures. Our results demonstrate an unanticipated and novel role for Sar1 in linking cargo selection with ER morphogenesis through the generation of transitional tubular ER export sites.

PubMed Disclaimer

Figures

Figure 2

Sar1-GTP triggers the formation of VSV-G-containing tubules. Permeabilized NRK cells (Plutner et al. 1992) were incubated either in the absence (A) or presence (B) of cytosol, with cytosol and 10 μg Sar1-GTP (C), with Sar1-GTP only (D), with Sar1-GDP only (E), or with activated ARF1 (ARF1-GTP) only (F), in 220 μl for 30 min at 32°C as described in Materials and Methods. The distribution of VSV-G was determined using indirect immunofluorescence with specific antibody (Plutner et al. 1992). The arrowheads in B and C denote punctate VTCs. The arrows in d denote tubules containing accumulated VSV-G; the arrowheads denote foci of tubule formation. All figures are representative of at least five independent experiments. Scale bar: 5 μm.

Figure 1

Sar1 interacts with the VSV-G cytoplasmic tail and assembles COPII coats in an export signal-dependent manner. (A, a–e) VSV-G containing ER microsomes were incubated as described (Rowe et al. 1996) in the absence (a) or presence (b–d) of GST-Sar1-GTP, on ice (d) or at 32°C (a–c) for 30 min, before transfer to ice and incubation in the absence (c) or presence (a, b, and d) of the reversible cross linker DTSSP. Subsequently, membranes were solubilized in detergent buffer and the lysate was incubated with anti–GST antibody. The amount of VSV-G (a–e) or ribophorin (f and g) recovered in the immunoprecipitate was determined by immunoblotting with specific antibody (Aridor et al. 1998). e shows 25% of the total VSV-G present in the incubation. h and i show a silver-stained gel used to determine the stoichiometry of the GST-Sar1/VSV-G complex. ER microsomes were incubated as described above in the absence (h) or presence of GST-Sar1-GTP (i). The cross-linked complex was isolated by immunoprecipitation with anti–GST antibodies, denatured in the presence of 0.1% SDS at 55°C, and reisolated with VSV-G-specific antibody and analyzed by SDS-PAGE. The location of the antibody heavy chain remaining associated with denatured GST-Sar1-VSV-G complex is indicated (HC). *Nonspecific band recovered under both incubation conditions. VSV-G and Sar1 were identified in a duplicate gel analyzed by immunoblotting (not shown). (B) GST (a and b) or the GST-VSV-G (GST-tail) fusion protein were incubated with purified Sar1-GTP (a–d) or Sar1-GDP (f–h) in the absence (a, c, f, and g) or presence (b, d, e, and h) of purified Sec23/24 complex as described. The amount of Sar1 (top) or Sec23 (bottom) recovered on GS beads was determined by immunoblotting with specific antibody. (C) GST or GST-tail fusion proteins containing either the wild-type 29 amino acid tail of VSV-G (wt) (YTEIDM at positions 19–24), a mutant harboring a D21A and E23A substitution (A19A21), or Ala substitutions for the entire export signal (A19-24) were incubated with Sar1-GTP and purified Sec23/24 proteins as indicated, and the recovery of Sar1 or Sec23 was determined as described above. Figures are representative of at least four independent experiments.

Figure 3

Electron microscopy analysis of Sar1-dependent tubular domains. (A–C) Permeabilized cells were incubated as described in Fig. 2 in the presence of Sar1-GTP, and prepared for EM as described (Bannykh et al. 1996). In A, a typical image of Sar1-dependent tubules (arrowheads) forming at a focal point on the nuclear envelope (ne) at early time points. In B, Sar1-dependent tubules are immunolabeled with specific antibody for Sar1 (arrowheads) and detected with 10 nm gold-labeled secondary antibodies. In C, a cluster of tubules (large arrowhead) is immunolabeled with 6 nm gold for VSV-G (arrowheads). (D and E) Quick-freeze, deep-etch micrographs of permeabilized cells incubated in the presence of Sar1-GTP as described in Fig. 2. Arrowheads show 40–80-nm tubular elements that are formed in response to Sar1 activation. Scale bar: 100 nm.

Figure 4

A selective domain containing cargo is formed from the ER after Sar1 activation. (A) Permeabilized NRK cells were incubated in the presence of Sar1-GTP as described above, and the distribution of VSV-G (a–e) was compared with that of Sar1 (f), Sec22 (g), p63 (h), BIP (i), and Man II (j) using indirect immunofluorescence. VSV-G colocalized with Sar1 (a and f) and the cargo SNARE protein Sec22 (b and g), but not with the resident ER membrane protein p63 (c and h), the resident soluble ER protein BIP (d and i), or Man II (e and j). Figures are representative of at least four independent experiments. Scale bar: 15 μm. (B) Sar1-dependent ER tubular domains are detergent resistant. After incubation of permeabilized cells in the presence of Sar1-GTP, the cells were extracted with ice-cold Triton X-100 as described in Materials and Methods before fixation and analysis using indirect immunofluorescence. Sar1-coated tubules (d) resist the detergent extraction and retain the cargo proteins VSV-G (a–c) and Sec22 (e), while the resident ER membrane protein p63 is largely extracted (f) (compare with the nonextracted condition shown in A, h). Shown are deconvolved images. Figures are representative of at least four independent experiments. Scale bar: 15 μm.

Figure 4

Figure 5

Formation and elongation of Sar1-dependent tubules. Permeabilized NRK cells were labeled with DiOC6, washed, and incubated with Sar1-GTP. DiOC6 fluorescence (A and B) or video-enhanced DIC (C, see below) were used to follow Sar1-dependent tubule formation (Allan and Vale 1994). Deconvolution microscopy (A) was used to generate a three-dimensional reconstruction of a DiOC6-labeled cell from 36 sections, 0.2-μm each (scale bar: 5 μm). More than 120 tubules are visible within the cell (for full three-dimensional reconstruction, see Movie 1 in the supplemental material). The faint outline delimits the cell boundary. Tubule elongation in DiOC6-labeled cells was monitored in real time. B1–B8 show continuous images taken at 4-s intervals to follow elongation of a single tubule. Elongation rates observed were B1–B2, 0.55 μm/s; B2–B3, 0.3 μm/s; B3–B4, 0.3 μm/s; B4–B5, 0 μm/s; B5–B6, 0.45 μm/s; B6–B7, 0.38 μm/s; B7–B8, 0 μm/s. (C) DIC was used to follow Sar1-induced tubule elongation in unlabeled cells. C1–C8 depict 1.25-s interval recordings of the elongation of a single tubule. Observed rates are C1–C2, 0.264 μm/s; C2–C3, 0.824 μm/s; C3–C4, 1.48 μm/s; C4–C5, 0.056 μm/s; C5–C6, 0.396 μm/s; C6–C7, 1.3 μm/s; C7–C8, −0.18 μm/s. Scale bar: 2 μm for B and C (shown in B5).

Figure 6

Sar1-dependent recruitment of microtubule motors. Permeabilized NRK cells were incubated with (B–C and E–F) or without (A and D) cytosol in the presence (A, C, D, and F) or absence (B and E) of Sar1-GTP for 30 min at 32°C. Cells were fixed and stained for dynein (using a 70.1 anti-intermediate chain antibody; A1, B1, and C1), kinesin (using the H1 anti-kinesin heavy chain antibody; D1, E1, and F1), Sar1 (A2 and D2), and Sec23 (B2, C2, E2, and F2). A–F3 show merged results of separate corresponding images. A1–A3 are compiled from 16 0.2-μm optical sections; B1–B3 are compiled from 8 0.2-μm sections; C1–C3 are compiled from 9 0.2-μm sections; D1–D3 are compiled from 11 0.2-μm sections; E1–E3 are compiled from 12 0.2-μm sections; and F1–F3 are compiled from 20 0.2-μm sections. Note that addition of activated Sar1 (Sar1-GTP) in the presence of cytosol leads to robust recruitment of Sec23 (C2 and F2) and specific mobilization of kinesin-to-ER export domains coated by Sec23 (F3). Scale bars: 5 μm.

Figure 6

Figure 7

Sar1 tubules are consumed by COPII vesicle budding. (A) Permeabilized NRK cells were incubated for 30 min in the presence of Sar1-GTP to form VSV-G-containing tubules (a–c). Subsequently, the incubation cocktail was replaced with a cocktail containing purified Sec23/24 and Sec13/31 (Aridor et al. 1998) and incubated for an additional 10 min at 32°C (d–f). Cells were labeled for Sar1 (a and d) or VSV-G (b and e) and imaged using a deconvolution microscope. c and f are a merge of a and b, and d and e, respectively. Large arrows in a–c illustrate the overlap of Sar1 with VSV-G. Small arrows in e and f denote location of VSV-G-containing VTCs. Arrowheads in f highlight selected regions (1–4) that demonstrate the appearance of VSV-G-containing punctate VTCs during disassembly of Sar1-containing tubules. The deconvolved image depicts three consecutive 0.1-μm sections (out of 25 total sections) from the middle of the cell to illustrate overlap. Bar in d, 5 μm. (B, 1–4) Time-lapse video (5-min intervals) showing dissolution of Sar1-dependent tubules to form punctate, pre-Golgi–like intermediates containing VSV-G (arrowheads). (C) Permeabilized cells incubated as described above were fixed and processed for electron microscopy as described in Materials and Methods. In a, numerous vesicles (arrows) coated with Sar1 (gold labeled) derived from a region dense in tubules (large arrowhead). (b) Vesicles on tubules and (c) VTCs are shown to contain Sec23. In d, VTCs are shown to contain VSV-G. Bars, 100 nm.

Figure 8

Sar1 tubules are functional intermediates in ER-to-Golgi transport. (A) Microsome membranes containing VSV-G were incubated in the presence (stage 1: Sar1-GTP, bottom) or absence (stage 1: control, top) of Sar1-GTP for 30 min at 32°C. Membranes were collected by centrifugation, washed, and further incubated for additional 30 min at 32°C (stage 2) in the absence (a–c) or presence (d) of Sar1-GTP, in the presence of Sar1-GDP (c), and in the absence (a) or presence (b–d) of purified COPII components Sec23/24 and Sec13/31 (COPII). The fraction containing COPII-coated vesicles was prepared as described (Rowe et al. 1996), and the level of VSV-G recovered determined by immunoblotting with specific antibody (shown). (B) Permeabilized cells were also pulse labeled with [35S]-methionine as described (Davidson and Balch 1993) and chased for 30 min at 32°C in the presence of wild-type Sar1 and GTP (stage 1) (a). In b, cells were subsequently incubated for 90 min at 32°C in the presence of cytosol. In c, Sar1 and GTP were not included in stage 1 before incubation in the presence of cytosol for 90 min at 32°C. Fractions of VSV-G processed to endo H–resistant forms were reported as a percentage of total VSV-G. (1 and 2) Permeabilized cells were washed and incubated in the presence of wild-type Sar1 and GTP (Stage 1) for 30 min at 32°C. Note the presence of VSV-G in Sar1-dependent tubules (1, rhodamine), but its absence in Man II containing Golgi compartments (1, fluorescein). After Stage 1, cells were further incubated in Stage 2 in the presence of cytosol for 30 min at 32°C. VSV-G shows both punctate localization to VTCs and overlap with Man II-containing compartments (inset 2). Scale bars in 1 and 2: 5 μm.

Figure 9

Export of GFP-VSV-G from the ER in living cells. (A) GFP-VSVG expressing COS cells were incubated at 40°C for 12 h to accumulate VSV-G in the reticular ER (A), and then shifted to 32°C to initiate ER export (B). Fluorescence images were taken using time-lapse video microscopy as described in Materials and Methods, and export sites of VSV-G were identified by the increased GFP fluorescence. Scale bar: 5 μm. (B) The boxed area in A is enlarged 2× and the image series (B, 1–4) (1/12-s intervals) is showing the initiation (B1, arrows a and b), formation, and elongation of tubular elements that accumulate and export VSV-G (B, arrows a in 2–4 and c in 3 and 4) from the ER. (C) A detailed time course of tubular elements containing VSV-G repeatedly initiated from the same ER export site at the indicated time points. After detachment, the tubular elements are transported on MTs and are delivered to the Golgi region (not shown) (Presley et al. 1997). In C, 6, the tubule is 14 μm in length. For full video sequence, see Movie 2 in the supplemental material.

References

1. Allan B.B., Balch W.E. Protein sorting by directed maturation of Golgi compartments. Science. 1999;285:63–66. - PubMed
1. Allan B.B., Moyer B.D., Balch W.E. Rab1 recruitment of p115 into a cis-SNARE complexprogramming budding COPII vesicles for fusion. Science. 2000;289:444–448. - PubMed
1. Allan V., Vale R. Movement of membrane tubules along microtubules in vitroevidence for specialised sites of motor attachment. J. Cell Sci. 1994;107:1885–1897. - PubMed
1. Aniento F., Emans N., Griffiths G., Gruenberg J. Cytoplasmic dynein-dependent vesicular transport from early to late endosomes. J. Cell Biol. 1993;123:1373–1387. - PMC - PubMed
1. Aridor M., Balch W.E. Principles of selective transportcoat complexes hold the key. Trends Cell Biol. 1996;6:315–320. - PubMed

The Sar1 GTPase coordinates biosynthetic cargo selection with endoplasmic reticulum export site assembly - PubMed (original) (raw)