HOPS prevents the disassembly of trans-SNARE complexes by Sec17p/Sec18p during membrane fusion

Trans-SNARE complex is largely associated with Sec17p

To determine whether Sec17p was associating with free Vam3p or with Vam3p in a SNARE complex, we again used vacuoles from BJ3505 nyv1Δ and DKY6281 vam3(ΔN) strains. At the end of the fusion reaction, we immunodepleted Sec17p from the vacuolar detergent extracts and asked whether SNARE complexes, as assayed by Nyv1p that was bound to full-length Vam3p, were removed through their association with Sec17p (Figure 2A). When fusion is robust (Figure 2B, lane I), virtually all the newly formed SNARE complexes are removed by Sec17p immunodepletion (Figure 2C, left panel, compare lane 10 with lane 9 or 11), showing that most of the newly formed SNARE complexes interact with Sec17p.

Open in a separate window

Figure 2

Trans-SNARE complex is largely associated with Sec17p. (A) Assay scheme. Vacuoles from BJ3505 nyv1Δ (93 μg) were incubated with vacuoles (93 μg) from DKY6281 vam3(ΔN) or DKY6281 vam3(ΔN) nyv1-CCIIM under standard fusion conditions for 45 min before an aliquot was assayed for Pho8p activity (B). The rest were solubilized in detergent and divided equally into three portions, each to be incubated with agarose resin conjugated with either αPorin I (a control IgG), αSec17p, or no IgG. After 1 h incubation, supernatant was collected by centrifugation and incubated with fresh resin for the second time to ensure complete depletion of Sec17p (lanes 3, 7, 14, and 18). A small aliquot of the supernatant collected after the second incubation was set aside for western blotting, whereas the rest were incubated with immobilized αVam3-N. Proteins co-precipitated with Vam3p were subject to SDS–PAGE and immunoblot analysis (C). Note that there are more Nyv1 molecules on DKY6281 vam3(ΔN) vacuoles than on DKY6281 vam3(ΔN) nyv1-CCIIM vacuoles (compare lanes 1 and 12), as reported earlier (Jun et al, 2007), but the amount of new SNARE complex formed in reactions I and II are comparable (compare lanes 11 and 22).

At least three types of SNARE complex form during these incubations: trans-SNARE complex, cis-SNARE complex that was trans- before fusion, and cis-SNARE complex that formed de novo after fusion and had never been trans-. To determine whether Sec17p is associated with the trans-SNARE complex, we exploited vacuoles from DKY6281 vam3(ΔN) nyv1-CCIIM. The transmembrane domain of Nyv1p on these vacuoles had been replaced by the Ykt6p prenylation motif (Jun et al, 2007). Vacuoles with lipid-anchored Nyv1p fuse poorly (Figure 2B, lane II) and thus a larger percentage of the new SNARE complex will be trans-SNARE complex as compared with reactions involving vacuoles with the wild-type Nyv1p. Earlier studies using Nyv1p-CCIIM vacuoles (Figure 6E of Jun et al, 2007) established that >50% of the new SNARE complex formed is trans-SNARE complex. Remarkably, immunoadsorption of Sec17p completely depleted the new SNARE complex from detergent extracts of docked NYV1-CCIIM vacuoles (Figure 2B, compare lane 21 with lane 20 or 22), showing that the new SNARE complex, including the trans-SNARE complex, is largely associated with Sec17p. What is the significance of the interaction between Sec17p and the trans-SNARE complex?

Trans-SNARE complex is sensitive to Sec17p/Sec18p

Earlier studies have shown that a partially lipid-anchored trans-SNARE complex formed during vacuole fusion is subject to Sec18p and Vam7p-dependent remodelling, implicating Sec17p (along with Sec18p) in disassembling the trans-SNARE complex (Jun et al, 2007). To study SNAREs with wild-type membrane anchors, we have exploited a reconstituted proteoliposome fusion system made from recombinant proteins and a set of pure lipids that closely match the composition of vacuole membranes (Mima et al, 2008). Fusion is measured by lipid mixing between proteoliposomes (Figure 3A and B). By placing Nyv1p on one set of liposomes and Vam3p and Vti1p on the other, we are able to bypass Sec17p/Sec18p-dependent priming and directly study new SNARE complex formation. In this assay, fusion (Figure 3A) and new SNARE complex formation (Figure 3C) require the addition of the soluble SNARE Vam7p. Both fusion and SNARE complex formation are stimulated by HOPS. It is noteworthy that similar rates of fusion can be achieved despite drastically different amounts of SNARE complex (compare Figure 3C, lanes 6 and 10, and the corresponding fusion in Figure 3A, filled circles and filled inverted triangles, respectively), suggesting that not all new SNARE complexes are functionally equal.

Open in a separate window

Figure 3

A truncated Vam7p (Vam7p-3Δ) arrests the fusion of proteoliposomes while allowing trans-SNARE complex formation. Proteoliposomes bearing Nyv1p were mixed with proteoliposomes carrying Vam3p and Vti1p. Wild-type Vam7p, Vam7p-3Δ, and/or HOPS (61 nM) was added as indicated (A, B). Fusion between donor and acceptor liposomes was measured by the dequenching of NBD-PE fluorescence (see Materials and methods). (C) At the end of the reaction, proteoliposomes were solubilized in detergent, and the proteins that adsorbed to immobilized αVam3 were analysed by SDS–PAGE and immunoblot.

To resolve trans-SNARE complex formation from fusion and possible formation of post-fusion cis-SNARE complexes, we used a C-terminally truncated Vam7p (Vam7p-3Δ) that efficiently drives trans-SNARE complex formation on isolated vacuoles in the absence of fusion (Schwartz and Merz, 2009). In the reconstituted fusion system, Vam7p-3Δ supports trans-SNARE complex formation (Figure 3C, lanes 12–17) without fusion (Figure 3B). The yield of trans-SNARE complex increases with increasing amounts of Vam7p-3Δ in the reaction. In addition, HOPS promotes the formation of trans-SNARE complexes in the presence of Vam7p-3Δ (Figure 3C). Vam7p-3Δ therefore permits study of trans-SNARE complex without the interference of post-fusion cis-SNARE complexes.

To determine whether trans-SNARE complexes can be disassembled by Sec17p and Sec18p, reactions were divided into two stages, assembly of trans-SNARE complexes followed by their disassembly by Sec17p/Sec18p. During the first stage incubation, 6% of the Nyv1p entered into trans-SNARE complex with Vam3p in the presence of Vam7p-3Δ (Figure 4, compare lanes 5 and 13). If the reaction was allowed to continue for another 30 min (second stage), more trans-SNARE complex was detected (Figure 4, compare lane 7 and 13). Sec17p and Sec18p added at the beginning of the second stage reduced the amount of Nyv1p associated with Vam3p to about 2% (Figure 4, lane 9), significantly below the level reached in the first stage (6%), showing that Sec17p and Sec18p disassemble the complex already formed in the first stage. Please note that under our immunoprecipitation conditions, the Sec17p/Sec18p is inactivated by EDTA/RIPA buffer (Supplementary Figure S2), so that any SNARE complex disassembly we measure has occurred on the membrane rather than in detergent.

Open in a separate window

Figure 4

Trans-SNARE complex is disassembled by Sec17p/Sec18p. Donor proteoliposomes (50 μM) bearing Nyv1p were mixed with acceptor proteoliposomes (400 μM) bearing Vam3p and Vti1p (with 1 mM MgCl₂, 1 mM ATP and an ATP regenerating system). Samples received either 1.2 μM Vam7p-3Δ, 1.2 μM Vam7p-3Δ with 12 μM FYVE, or control buffer. After 30 min at 27°C (Stage I), samples with Vam7p-3Δ were either transferred to ice or incubated at 27°C for another 30 min without Sec17p/Sec18p (lanes 7 and 10) or with the addition of Sec17p/Sec18p (0.34 μM/0.24 μM or 0.68 μM/0.24 μM; lanes 8, 9, 11, and 12). Samples from lanes 10 to 12 also received FYVE at the end of Stage I to prevent the formation of new SNARE complex. At the end of Stage II, Vam7p-3Δ (1.2 μM) was added to the sample as shown in lane 1, which had not initially received Vam7p-3Δ. All samples were assayed for trans-SNARE complex (see Materials and methods). The Nyv1p signal, normalized for the recovery of Vam3p, was averaged from three independent experiments. Error bars represent standard error of the mean (s.e.m.) calculated using KaleidaGraph. Note that lane 13 measures trans-SNARE complex formation in the first stage only. If the second stage reaction continues at 27°C without inhibitors, more trans-SNARE complex will form (compare lanes 13 and 7).

To control for SNARE complex assembly during the second stage of the reaction, we also performed this stage of the reaction with purified recombinant FYVE domain. FYVE domain inhibits vacuole fusion by binding to PI(3)P (Fratti et al, 2004), a regulatory lipid that is essential for trans-SNARE complex formation (Collins and Wickner, 2007). As FYVE domain inhibits the trans-SNARE complex formation on proteoliposomes (Figure 4, lane 6 versus 7), adding FYVE domain at the start of the second stage of the reaction allows direct evaluation of the effect of Sec17p and Sec18p on existing complexes. As shown in Figure 4 (lanes 10–12), Sec17p/Sec18p disassembles trans-SNARE complex in a concentration-dependent manner. This explains why Sec17p and Sec18p block the fusion of proteoliposomes in the absence of HOPS (Mima et al, 2008).

HOPS inhibits the Sec17/18-dependent disassembly of the trans-SNARE complex

To allow SNARE-dependent membrane fusion in the presence of Sec17p and Sec18p, the disassembly of trans-SNARE complexes must be inhibited. HOPS and Sec17p/Sec18p synergistically promote proteoliposome fusion (Mima et al, 2008), suggesting that HOPS might regulate the Sec17p/Sec18p-mediated disassembly of trans-SNARE complex. As mentioned above, we allowed trans-SNARE complex to form in an initial incubation in the presence or absence of HOPS. To arrest further trans-SNARE complex assembly while examining HOPS function, we could not use FYVE domain as in Figure 4, as FYVE binds to phosphoinositides that are part of the HOPS membrane receptor (Burd and Emr, 1998; Kutateladze et al, 1999). We therefore added either CaCl₂ (10 mM) or KCl (250 mM), which effectively prevents fusion (data not shown) and trans-SNARE complex formation (Supplementary Figures S3 and S4). Without HOPS, trans-SNARE complexes are disassembled by Sec17p and Sec18p in the second stage of the reaction (lanes 9–12 in Figure 5A and B). The disassembly rate is slower in the presence of 250 mM KCl, presumably due to salt inhibition of Sec17p/Sec18p. Strikingly, trans-SNARE complexes are quite resistant to Sec17p/Sec18p-mediated disassembly in the presence of HOPS (lanes 5–8 in Figure 5A and B). There is a slight decrease of trans-SNARE complex at high concentrations of Sec17p/Sec18p (compare lane 8 with lane 5), suggesting that the HOPS inhibition of trans-SNARE complex disassembly is not absolute.

Open in a separate window

Figure 5

HOPS inhibits the disassembly of trans-SNARE complex by Sec17p/Sec18p. Donor proteoliposomes (50 μM) bearing Nyv1p, acceptor proteoliposomes (400 μM) bearing Vam3p and Vti1p, and 1.2 μM Vam7p-3Δ were mixed with 61 nM HOPS or its buffer alone. Trans-SNARE complex was allowed to form at 27°C for 30 min. Samples then received either 10 mM CaCl₂ (A) or 250 mM KCl (B) to stop the formation of new SNARE complex. Samples were then divided into four portions and received either Sec17p/Sec18p (0.68 μM/0.24 μM or 1.36 μM/0.48 μM) (lanes 7, 8, 11, and 12) or control buffer (lanes 5, 6, 9, and 10). After 30 min of further incubation at 27°C (except samples in lanes 6 and 10, which were placed on ice), trans-SNARE complex was assayed. Data from three experiments were analysed as in Figure 4. In both (A, B), the P-values between lanes 5 and 8 or 5 and 7 are larger than 0.25.

Sec17p and Sec18p can disassemble a complex of four vacuolar SNARE soluble domains and this disassembly is not affected by HOPS (Mima et al, 2008), suggesting that membrane integrity might be required for the HOPS-dependent preservation of the trans-SNARE complex. To directly address this, we modified the trans-SNARE complex disassembly assay by also preparing a parallel set of samples, which received Triton X-100 during the second stage of the reaction (Figure 6A, lanes 7–12). Triton X-100 solubilization of donor and acceptor membranes allowed Vam3p (from acceptor proteoliposome) and Nyv1p (from donor proteoliposomes) that had not been in trans-SNARE complex to form a complex in detergent despite the addition of excess soluble GST-Nyv1p to compete with Nyv1p. This led to the detection of a modest increase of Vam3p–Nyv1p association (compare Figure 6A, lanes 1 and 7, lanes 4 and 10). Strikingly, almost all these complexes were effectively disassembled by Sec17p/Sec18p regardless of the presence of HOPS (compare Figure 6A, lanes 7 and 9, lanes 10 and 12). Trans-SNARE complexes formed on apposed liposomes that were not subsequently exposed to detergent maintained their resistance to Sec17p/Sec18p in the presence of HOPS (Figure 6A, lanes 1–6). Thus, the HOPS-dependent preservation of trans-SNARE complexes appears to rely on the integrity of the two apposed membranes.

Open in a separate window

Figure 6

HOPS preservation of trans-SNARE complex requires membrane integrity (A) and HOPS:Vam7p interaction (B). (A) trans-SNARE complex was formed at 27°C for 30 min as described in Figure 5. After receiving 10 mM CaCl₂, aliquots received various amounts of Sec17p/Sec18p (0/0, 0.68 μM/0.24 μM, or 1.36 μM/0.48 μM) with (lanes 7–12) or without (lanes 1–6) 1% Triton X-100. After 30 min of further incubation at 27°C, trans-SNARE complex was assayed. Data from five independent experiments were analysed. Error bars represent s.e.m. For samples in lanes 7–12, GST-Nyv1p was added before the addition of Triton X-100 to reduce new SNARE complex formation in detergent. (B) Proteoliposomes bearing Nyv1p (50 μM) and proteoliposomes (400 μM) bearing Vam3p and Vti1p were mixed with 1.2 μM Vam7p-3Δ at 27°C for 30 min to allow trans-SNARE complex formation. After receiving 10 mM CaCl₂, samples were aliquoted into tubes containing HOPS, HOPS with Y42A-PX, or control buffers. The final concentration of Y42A-PX is 10.8 μM, and thus the molar ratio of Y42A-PX:Vam7p-3Δ=9:1. Each aliquot was further divided into three portions to receive Sec17p/Sec18p (0/0, 0.68 μM/0.24 μM, or 1.36 μM/0.48 μM). After 30 min of additional incubation at 27°C, trans-SNARE complex was assayed. Data from four independent experiments were analysed. Error bars represent s.e.m.

To address whether SNARE recognition is required for this new function of HOPS, we examined the effect of saturating HOPS for its capacity to bind SNAREs. We have previously shown that HOPS interacts preferentially with Vam7p among the vacuolar SNAREs, and that this interaction is specific for the N-terminal PX domain of Vam7p (Stroupe et al, 2006). We therefore tested whether the addition of an excess of this PX domain would block HOPS preservation of trans-SNARE complexes. To avoid complications arising from the affinity of this PX domain for vacuolar PI(3)P, we used the Y42A mutant PX domain, which specifically lacks affinity for phosphoinositides (Cheever et al, 2001). Proteoliposomes bearing Vam3p and Vti1p were incubated with those bearing Nyv1p in the presence of Vam7p-3Δ to allow trans-SNARE complex assembly, then mixed with CaCl₂ to prevent further trans-SNARE complex assembly. After adding combinations of HOPS and/or Y42A-PX, a second incubation was performed that included Sec17p and Sec18p where indicated. The trans-SNARE complexes formed during the initial incubation were readily disassembled by added Sec17p/Sec18p (Figure 6B, lanes 4–6), and HOPS inhibited this disassembly (Figure 6B, lanes 1–3). Adding Y42A-PX alone had no effect on the Sec17p/Sec18p-mediated disassembly (Figure 6B, lanes 10–12), but it prevented HOPS from preserving the trans-SNARE complex (Figure 6B, lanes 7–9). Thus, HOPS appears to rely on both membrane bilayer integrity (Figure 6A) and on its interaction with SNAREs (Figure 6B) to preserve trans-SNARE complexes.

HOPS does not inhibit cis-SNARE complex disassembly

Is this effect of HOPS specific for trans-SNARE complexes or does HOPS also inhibit cis-SNARE complex disassembly by Sec17p/Sec18p? Proteoliposomes bearing Vam3p, Vam7p, Vti1p, and Nyv1p were incubated at low concentration to eliminate tethering and trans-SNARE complex formation (see Materials and methods). The disassembly of their cis-SNARE complexes was measured by the loss of association between Nyv1p and Vam3p. Approximately 60% of the Nyv1p on these proteoliposomes was initially associated with Vam3p (Figure 7A). Incubation with Sec17p and Sec18p reduced the level of Nyv1p associated with Vam3p to 25%. The complexes that were insensitive to Sec17p and Sec18p were likely oriented toward the liposome lumen and thus inaccessible to the SNARE chaperones. HOPS did not inhibit cis-SNARE complex disassembly at any Sec17p/Sec18p concentration tested (Figure 7A).

Open in a separate window

Figure 7

HOPS does not inhibit cis-SNARE complex disassembly. (A) 4-SNARE proteoliposomes (100 μM lipids), 61 nM HOPS or its buffer, and various amounts of Sec17p/Se18p (0/0, 0.17 μM/0.06 μM, 0.34 μM/0.12 μM, or 0.68 μM/0.24 μM) were incubated at 27°C for 30 min. The Nyv1p bound to Vam3p was assayed as described (see Materials and methods). Data from three independent experiments were analysed as in Figure 4. (B, C) To control for the effect of HOPS pre-incubation with proteoliposomes and to use the same amount SNARE complex as in the trans-SNARE complex disassembly assay, reduced amounts of 4-SNARE proteoliposomes (3.3 μM lipids) bearing either Vam7p (B) or Vam7p-3Δ (C) were incubated with 61 nM HOPS or its buffer alone at 27°C for 30 min. Aliquots received control buffer or various amounts of Sec17p/Sec18p as in Figure 7A. After 30 min of further incubation at 27°C, cis-SNARE complex was assayed. The amount of cis-SNARE complex detected in the absence of Sec17p/Sec18p was set at 100% and was used to normalize the cis-SNARE complexes detected in the presence of Sec17p/Sec18p. Data from four independent experiments were analysed. Error bars represent s.e.m. At each Sec17p/Sec18p concentration, the differences between samples with or without HOPS were not statistically significant.

To more closely mimic the conditions used in the trans-SNARE complex disassembly assays (Figure 5) and, in particular, to account for the fact that the proportion of the SNAREs engaged in trans complexes (4–8% of total; Figures 5 and ​and6)6) is smaller than that of the cis-SNARE complex (60% Figure 7A) in our reactions, we lowered the amount of 4-SNARE proteoliposomes in the cis-SNARE disassembly assay so that the concentrations of the SNARE complexes in the two were equivalent (see Material and methods). 4-SNARE proteoliposomes that bear either Vam7p-3Δ or Vam7p were pre-incubated with HOPS (or control buffer) for 30 min before the addition of Sec17p and Sec18p at the same concentrations used in the trans-SNARE complex disassembly assay. Sec17p and Sec18p effectively disassembled cis-SNARE complexes in the presence or absence of HOPS (Figure 7B and C). Thus, the HOPS-mediated preservation of trans-SNARE complexes, which requires the integrity of the two apposed membranes, is specific to the trans conformation of the SNARE complex.

Proteins, antibodies, and cross-linking

Vam7p and Vam7p-3Δ were purified according to Schwartz and Merz (2009) with the following modifications: (1) Escherichia coli Rosetta 2 (DE3) (Novagen) was used for protein expression; (2) induced bacteria were lysed by French Press (3 passes, 900 pounds per square inch); and (3) after affinity purification on Ni-NTA agarose (Qiagen), his₆-Vam7 proteins were exchanged into RB500 (20 mM HEPES-NaOH, pH 7.4, 500 mM NaCl, 10% (v/v) glycerol) for storage. Recombinant Vam3p, Vti1p, and Nyv1p were purified from Rosetta 2 (Novagen) (Mima et al, 2008). HOPS complex containing Vps33-GST was purified from yeast CHY31 (Starai et al, 2008). Sec17p was purified from Rosetta 2 (DE3) as described (Schwartz and Merz, 2009), except that the tagless Sec17p was exchanged into RB150 (20 mM HEPES-NaOH, pH 7.4, 150 mM NaCl, 10% (v/v) glycerol) for storage. His₆-Sec18 (Haas and Wickner, 1996), I₂^B (Slusarewicz et al, 1997), GST-2xFYVE domain (Gillooly et al, 2000), his₆-Gyp1-46 (Wang et al, 2003), Gdi1p (Starai et al, 2007), Y42A-PX domain (Boeddinghaus et al, 2002), αYpt7, and its antigenic peptides (Eitzen et al, 2001), αVPS33 (Seals et al, 2000), and αSec17 (Haas and Wickner, 1996) were prepared as described. αPorin I was a kind gift from Naomi Thorngren. MARCKS effector domain (Wang et al, 2001) was synthesized by the WM Keck Biotechnology Resource Center (New Haven, CT).

αVam3 and αVam3-N (specific for N-terminal domain) were affinity purified with immobilized GST-Vam3(ΔTM) and GST-Vam3-N(ΔTM), respectively. Proteins (10 mg) were cross-linked to 2.5 ml of Sulfolink coupling gel (Thermo Scientific, Rockford, IL) following the manufacturer's instructions. Serum (10 ml) from rabbits immunized with his₆-Vam3(ΔTM) was used for antibody purification (Jun et al, 2007).

For immunoprecipitation assays, antibodies (αVam3, αVam3-N, αSec17, or αPorin I) were cross-linked to protein A Sepharose CL-4B (GE) as described (Harlow and Lane, 1998), except that 2 mg of IgG (1 OD₂₈₀=0.8 mg/ml) were used per 4 ml of protein A resin. Protein A resin cross-linked to antibodies was stored at 4°C in PBS containing 0.02% Na₃N. Immediately before use, resins were exchanged into elution buffer (0.2 M glycine, pH 2.2, 0.2 M NaCl), then equilibrated with solubilization buffer (see below).

Proteoliposome reconstitution and lipid mixing assay

SNARE proteoliposomes with POPC (42 or 44% for donor or acceptor), POPE (18%), soy PI (18%), POPS (4.4%), POPA (2.0%), CL (1.6%), ERG (8.0%), DAG (1.0%), PI(3)P (1.0%), PI(4,5)P₂ (1.0%), and fluorescent lipids (1.5% NBD-PE and 1.5% Rh-PE for donor; 1.0% dansyl-PE for acceptor) were prepared as described (Mima et al, 2008).

Fusion reactions (20 μl) contained RB150 (20 mM HEPES-NaOH, pH 7.4, 150 mM NaCl, 10% (v/v) glycerol), 1 mM MgCl₂, 1 mM ATP, an ATP regenerating system, donor proteoliposomes (50 μM lipids), and acceptor proteoliposomes (400 μM lipids). Vam7p or Vam7-3Δ (0.2–1.8 μM), HOPS (61 nM), Sec17p (0.17–2.7 μM), and Sec18p (0.06–0.96 μM) were added where indicated. To monitor lipid mixing, reaction mixtures (prepared on ice) were transferred to a 396-well plate (pre-incubated on ice) and the NBD fluorescent signal was measured (λex=460 nm, λem=538 nm, λex=515 nm) in a SpectraMAX Geminin XPS plate reader (Molecular Devices) at 27°C for 30 min (Mima et al, 2008). Fusion is presented as the ratio of the fluorescence at any time to that at 1 min.

Trans-SNARE complex assay

Proteoliposome samples (450 μM lipids; 20 μl) were chilled (ice, 5 min). Each received 1.4 μg of GST-Nyv1(ΔTM) and, after an additional 10 min, 800 μl of RIPA buffer (25 mM Tris–Cl, pH 7.5, 150 mM NaCl, 1% NP40-Alternative, 1% deoxycholate, 0.1% SDS) containing 10 mM EDTA, protease inhibitor cocktail (0.46 μg/ml leupeptin, 3.5 μg/ml pepstatin A, 24 μg/ml pefablock-SC), and 1 mM PMSF. GST-Nyv1(ΔTM) (70 μg/ml) was added to block potential binding of native Nyv1p to Vam3p in detergent. Samples were nutated at 4°C for 20 min, then centrifuged (16 000 g, 4°C, 5 min). Supernatant (700 μl) was incubated with 15 μl of immobilized αVam3 at 4°C for 1 h with nutation. After centrifugation (3000 g, 4°C, 2 min), the αVam3 resin was resuspended in 600 μl of RIPA buffer (with 10 mM EDTA and protease inhibitor cocktail), centrifuged (3000 g, 4°C, 2 min), and the supernatant was removed by centrifugation. After repeating this process three times, proteins were eluted with 2 × sample buffer (Sambrook et al, 1989) and subjected to SDS–PAGE (12%) and western blot analysis. The blots were scanned and analysed using software UN-SCAN-IT gel 5.3 (Silk Scientific).

To assay for vacuolar trans-SNARE complex, vacuoles from a 300-μl reaction were centrifuged (11 750 g, 10 min, 4°C) and resuspended in 400 μl of solubilization buffer (20 mM Tris–Cl, pH 7.5, 150 mM NaCl, 1 mM MgCl₂, 10% glycerol, 0.5% NP40-Alternative) containing protease inhibitor cocktail and 1 mM PMSF. After 20 min at 4°C, samples were centrifuged (16 000 g, 20 min, 4°C) and supernatant (350 μl) was incubated with 30 μl αVam3-N resin at 4°C overnight. Resin was resuspended in 600 μl of solubilization buffer with protease inhibitor cocktail, centrifuged (3000 g, 4°C, 2 min), and the supernatant was removed. This process was repeated three times before proteins were eluted from the resin and assayed as above. For immunoadsorption experiments, vacuolar detergent extracts were incubated with resin bearing either αSec17 or αPorin I (30 μl) at 4°C for 1 h. Supernatant was collected after centrifugation (3000 g, 4°C, 2 min) and incubated (4°C, 1 h) with fresh resin and again centrifuged. Supernatant was then used for overnight incubation with αVam3-N resin as described above. For experiments in Figure 1, a larger fusion reaction (4.83 ml) containing BJ3505 nyv1Δ vacuoles (483 μg), DKY6281 vam3(ΔN) vacuoles (483 μg), and 40 nM Vam7p was incubated at 27°C for 45 min with either no inhibitors, Gyp1-46 (2.5 μM)/Gdi1p (0.6 μM) or MED (10 μM). Aliquots (30 μl) were set aside for Pho8p activity assay. The remainder was resuspended in 6.4 ml solubilization buffer (described above). Detergent extracts were incubated with 480 μl of immobilized αVam3(N) at 4°C overnight. Proteins were eluted with 6 ml of 0.2 M glycine (pH 2.2), 0.2 M NaCl and divided into two equal aliquots. Ice-cold TCA (100%) was added to each to 12.5%. After vigorous mixing (5 s), samples were incubated on ice for 30 min and centrifuged (16 000 g, 4°C, 20 min). The pellet in each aliquot was resuspended in 3 ml of acetone (−20°C) and centrifuged (16 000 g, 4°C, 10 min). Supernatants were removed and pellets were dried in air for 5 min. One aliquot was subjected to SDS–PAGE and Sypro Ruby staining and the other to mass spectrometry.

Multidimensional protein identification technology

The protein digest was pressure loaded onto a fused silica capillary desalting column containing 5 cm of 5 μm Polaris C18-A material (Metachem, Ventura, CA) packed into a 250-μm internal diameter (i.d.) capillary with a 2-μm filtered union (UpChurch Scientific, Oak Harbor, WA). The desalting column was washed with buffer containing 95% water, 5% acetonitrile, and 0.1% formic acid. After desalting, a 100-μm i.d. capillary with a 5-μm pulled tip packed with 10 cm 3-μm Aqua C18 material (Phenomenex, Ventura, CA) followed by 3 cm 5-μm Partisphere strong cation exchanger (Whatman, Clifton, NJ) was attached to the filter union and the entire split column (desalting column–filter union–analytical column) was placed inline with an Agilent 1100 quaternary HPLC (Palo Alto, CA) and analysed using a modified five-step separation described earlier (Washburn et al, 2001). The buffer solutions used were 5% acetonitrile/0.1% formic acid (buffer A), 80% acetonitrile/0.1% formic acid (buffer B), and 500 mM ammonium acetate/5% acetonitrile/0.1% formic acid (buffer C). Step 1 consisted of a 90-min gradient from 0 to 100% buffer B. Steps 2–4 had the following profile: 3 min of 100% buffer A, 2 min of X% buffer C, a 10-min gradient from 0 to 15% buffer B, and a 97-min gradient from 15 to 45% buffer B. The 2-min buffer C percentages (X) were 20, 50, 80%, respectively, for the five-step analysis. The final step, the gradient contained: 3 min of 100% buffer A, 20 min of 100% buffer C, a 10-min gradient from 0 to 15% buffer B, and a 107-min gradient from 15 to 70% buffer B.

As peptides eluted from the microcapillary column, they were electrosprayed directly into an LTQ two-dimensional ion trap mass spectrometer (ThermoFinnigan, Palo Alto, CA) with the application of a distal 2.4 kV spray voltage. A cycle of one full-scan mass spectrum (400–1400 m/z) followed by eight data-dependent MS/MS spectra at a 35% normalized collision energy was repeated continuously throughout each step of the multidimensional separation. Application of mass spectrometer scan functions and HPLC solvent gradients were controlled by the Xcalibur data system.

Analysis of tandem mass spectra

MS/MS spectra were analysed using the following software analysis protocol. Poor quality spectra were removed from the data set using an automated spectral quality assessment algorithm (Bern et al, 2004). MS/MS spectra remaining after filtering were searched with the SEQUEST^TM algorithm (Eng et al, 1994) against the Saccharomyces Genome Database (December 2005) concatenated to a decoy database in which the sequence for each entry in the original database was reversed (Peng et al, 2003). All searches were parallelized and performed on a Beowulf computer cluster consisting of 100 1.2 GHz Athlon CPUs (Sadygov et al, 2002). No enzyme specificity was considered for any search. SEQUEST results were assembled and filtered using the DTASelect (version 2.0) program (Tabb et al, 2002; Cociorva et al, 2007). DTASelect 2.0 uses a linear discriminant analysis to dynamically set XCorr and DeltaCN thresholds for the entire data set to achieve a user-specified false-positive rate (5% in this analysis). The false-positive rates are estimated by the program from the number and quality of spectral matches to the decoy database. All the data sets can be found under VAM3 in the Yeast Resource Center Public Data Repository (Riffle et al, 2005).

We thank Alex Merz (U Washington) for sharing Vam7p and Vam7p-3Δ constructs before publication and Amy Orr and Holly Jakubowski for excellent technical support. This work was supported by NIH grant GM23377. Mass spectrometry analysis was supported by NIH grant P41 RR11823. Hao Xu and William Wickner designed the experiments and wrote the manuscript. Youngsoo Jun conceived the trans-SNARE complex assay and engineered the required yeast strains. James Thompson and John Yates performed mass spectrometry analysis and wrote the corresponding methods. William Wickner isolated SNARE proteins and the lyticase used for vacuole isolation. Hao Xu performed the rest of the experiments.