Cdc42p functions at the docking stage of yeast vacuole membrane fusion

Abstract

Membrane fusion reactions have been considered to be primarily regulated by Rab GTPases. In the model system of homotypic vacuole fusion in the yeast Saccharomyces cerevisiae, we show that Cdc42p, a member of the Rho family of GTPases, has a direct role in membrane fusion. Genetic evidence suggested a relationship between Cdc42p and Vtc1p/Nrf1p, a central part of the vacuolar membrane fusion machinery. Vacuoles from cdc42 temperature-sensitive mutants are deficient for fusion at the restrictive temperature. Specific amino acid changes on the Cdc42p protein surface in these mutants define the putative interaction domain that is crucial for its function in membrane fusion. Affinity-purified antibodies to this domain inhibited the in vitro fusion reaction. Using these antibodies in kinetic analyses and assays for subreactions of the priming, docking and post-docking phase of the reaction, we show that Cdc42p action follows Ypt7p-dependent tethering, but precedes the formation of trans-SNARE complexes. Thus, our data define an effector binding domain of Cdc42p by which it regulates the docking reaction of vacuole fusion.

Fig. 1. Cdc42p is associated with the vacuolar membrane. (A) 70 µg of isolated vacuoles (Vac), cytosol (Cyt) or a whole-cell extract (Cell) from strain BJ3505 were analyzed by SDS–PAGE and western blotting for their content of Cdc42p, Ypt7p and Pgk1p. (B) Longer exposure of the blots for Cdc42p and Ypt7p in (A) with the bands visible in the whole-cell extract (lane 3).

Fig. 2. In vitro fusion of vacuoles from cdc42^ts mutants. Standard fusion reactions with cytosol were incubated for 70 min at 23 or 30°C. The indicated tester strain combinations were used: OMY14/OMY15(wt), OMY16/OMY17 (cdc42^ts-123) and OMY18/OMY19 (cdc42^ts-124). Growth of cells and preparation of vacuoles was at 23°C instead of the 30°C used in the standard protocol. Four experiments were averaged. The 100% fusion activities of the wild type at 23°C were between 2.15 and 2.38 U.

Fig. 3. Cdc42p functions up to the docking stage. (A) Antibody inhibition. The indicated amounts of affinity-purified antibodies to Cdc42p were added to standard fusion reactions without cytosol. After pre-incubation for 10 min on ice, ATP was added and fusion was measured after 70 min at 27°C. (B) Kinetic analysis. Standard fusion reactions without cytosol were started. After the indicated times at 27°C inhibitors or control buffer were added. The samples were left on ice for 10 min. Then they were transferred to 27°C or left on ice for the remainder of the 70 min reaction period. Finally, fusion activity was assayed. Inhibitors used were: anti-Sec18p, 2 µM; Gdi1p, 5 µM; anti-Vam3p, 2 µM; Mastoparan, 20 µM; anti-Cdc42p, 8 µM.

Fig. 4. Sec17p/α-SNAP release is not compromised by anti-Cdc42p. Standard fusion reactions (3-fold volume) without cytosol were incubated for 70 min without ATP on ice (0 min) or with ATP at 27°C for 15 or 70 min. Where indicated, 6 µM anti-Cdc42p was added. Reactions were chilled on ice and centrifuged (10 000 g, 10 min, 4°C). The supernatants (S) were recovered and the pellets (P) resuspended in 90 µl of PS buffer. All samples were trichloroacetic acid precipitated and processed for non-reducing SDS–PAGE and western blotting with the indicated antibodies. Alkaline phosphatase (Pho8p) serves as a membrane integral vacuolar marker.

Fig. 5. Kinetic analysis of Cdc42p action. (A and B) Standard fusion reactions (10-fold volume) without cytosol were incubated at 27°C in the presence of (A) 2 mM GTPγS or (B) 2 mM BAPTA. After 30 min, reactions were chilled on ice and vacuoles were re-isolated (6800 g, 2 min, 4°C). Vacuoles were resuspended in fresh fusion buffer with cytosol. In the case of the BAPTA block, 200 µM CaCl₂ was added. Aliquots were supplemented with the indicated inhibitors and incubated for another 60 min at 27°C. Finally, fusion activity was measured. Fusion without re-isolation was 3.94 U. Inhibitors used were: anti-Sec18p, 2 µM; Gdi1p, 3.7 µM; anti-Cdc42p, 6 µM; anti-Vam3p, 2 µM; BAPTA, 2 mM; GTPγS, 2 mM. (C) Ca²⁺ release from vacuoles in the course of the fusion reaction was assayed as described in Materials and methods. Inhibitors used were: Gdi1p, 4 µM; anti-Sec18p, 1 µM; anti-Cdc42p, 6 µM.

Fig. 6. Cdc42p acts before trans-SNARE pairing. Standard fusion reactions (5-fold volume) with cytosol containing vacuoles from strains BJ3505Δvam3 and BJ3505Δnyv1 were incubated without ATP on ice or with ATP at 27°C. One of the samples at 27°C received buffer only (None) and another one 6 µM affinity-purified antibodies to Cdc42p (Ab). After 45 min, trans-SNAREs were assayed as described in Materials and methods by the amount of Nyv1p that co-immuno precipitates with Vam3p. As can be seen by comparison of co-precipitated Nyv1p with the corresponding input, formation of trans-SNARE pairs in the control reaction is in the range of literature values (Ungermann et al., 1998a).