INTRODUCTION

The Rab small G protein family consists of nearly 30 members in mammal and nearly 10 members in yeast and is implicated in intracellular vesicle trafficking, including budding of a vesicle from a donor membrane, docking, and fusion of the vesicle with the acceptor membrane (for reviews, see Refs. 1-5). Some Rab family members constitute a subfamily with highly homologous members. The Rab family members cycle between the GDP-bound inactive and GTP-bound active forms and between the cytosol and membrane fractions, and these two types of cycling are essential for their action in vesicle trafficking (1-5). The GDP-bound form is complexed with Rab GDI¹ and stays in the cytosol. This form is converted to the GTP-bound form by the action of Rab GEP. In the budding process, the GTP-bound form interacts with the donor membrane to induce budding of a vesicle. Before, during, or after the budding of the vesicle, the GTP-bound form is converted to the GDP-bound form by the action of Rab GAP. In the targeting/docking process, the GTP-bound form interacts with the vesicle to translocate it to the acceptor membrane. Before, during, or after the fusion of the vesicle, the GTP-bound form is converted to the GDP-bound form by the action of Rab GAP. Once the GDP-bound form is produced, it is complexed with Rab GDI and returns to the cytosol fraction. Of these three types of regulators of the Rab family members, Rab GDI is active on all the Rab family members thus far examined (for reviews, see Refs. 5 and 6). One GAP active on yeast Ypt6 and -7, but not on Ypt1 and Sec4, has been cloned and characterized (7), but no GEP specific for each Rab family member or Rab subfamily has been identified in yeast or mammal, and no GAP specific for each Rab family member or Rab subfamily has been identified in mammal.

The Rab3 subfamily consists of four members, Rab3A, -B, -C, and -D (for reviews, see Refs. 5 and 8). Of these members, Rab3A and -C are implicated in Ca²⁺-dependent exocytosis, particularly in neurotransmitter release from nerve terminal. A GEP and a GAP active on Rab3A have been partially purified from rat brain (9, 10), but neither their primary structures nor their precise properties have been studied. Therefore, we have attempted here to purify a GAP specific for Rab3A or the Rab3 subfamily.

EXPERIMENTAL PROCEDURES

Lipid-modified and -unmodified Rab3As were purified from the membrane and soluble fractions, respectively, of Sf9 cells expressing the Rab3A cDNA (11). The lipid-modified form of Rab2, -3B, -3C, -3D, -5A, and -11 was purified from the membrane fraction of Sf9 cells expressing each cDNA.

Lipid-modified Rab3A (3 pmol) was incubated for 10 min at 30 °C in a reaction mixture (10 µl) containing 25 mM Tris/HCl at pH 8.0, 10 mM EDTA, 5 mM MgCl₂, 0.5 mM DTT, 0.3% CHAPS, and 1.5 µM [-³²P]GTP (1 × 10⁴ cpm/pmol). The reaction was stopped by adding 2.5 µl of 80 mM MgCl₂. To this mixture (12.5 µl), the sample to be assayed was added in a total volume of 50 µl and further incubated for 5 min at 30 °C. The mixture was applied to a nitrocellulose filter, and the radioactivity retained on the filter was determined by Cerenkov counting.

All the purification procedures were performed at 0-4 °C. The synaptic soluble (SS) fraction was prepared from 200 rat brains (12). One-fifth of the SS fraction (450 ml, 315 mg of protein) was directly applied to a Q-Sepharose FF column (2.6 × 23 cm) equilibrated with Buffer A (20 mM Tris/HCl at pH 7.5, 0.5 mM EGTA, 0.5 mM EDTA, and 1 mM DTT). After the column was washed with 600 ml of Buffer A, elution was performed with a 600-ml linear gradient of NaCl (0-0.5 M) in Buffer A, followed by 120 ml of 0.5 M NaCl in Buffer A at a flow rate of 5 ml/min. Fractions of 8 ml each were collected. One peak of the Rab3 GAP activity appeared in Fractions 62-70. These fractions (72 ml, 36 mg of protein) were collected. The rest of the SS fraction was subjected to the same Q-Sepharose column chromatography four times in a similar manner. The samples of the five Q-Sepharose column chromatographies were pooled and diluted with 720 ml of Buffer B (20 mM potassium phosphate at pH 7.5, 0.5 mM EDTA, and 1 mM DTT). The sample was applied to a hydroxyapatite column (2.6 × 6.6 cm) equilibrated with Buffer B. After the column was washed with 350 ml of the same buffer, elution was performed with a 500-ml linear gradient of potassium phosphate (20-212 mM) in Buffer B, followed by a 150-ml linear gradient (212-500 mM) and 150 ml of 500 mM potassium phosphate in Buffer B at a flow rate of 1.25 ml/min. Fractions of 10 ml each were collected. One peak of the Rab3 GAP activity appeared in Fractions 29-40. These fractions (120 ml, 18 mg of protein) were collected and diluted with 240 ml of Buffer A. The sample was applied to a heparin-Sepharose CL-6B column (0.5 × 5 cm) equilibrated with Buffer A. After the column was washed with 20 ml of the same buffer, elution was performed with 0.5 M NaCl in Buffer A at a flow rate of 0.5 ml/min. Fractions of 1 ml each were collected. One peak of the Rab3 GAP activity appeared in Fractions 2-6. These fractions (5 ml, 4 mg of protein) were collected, and one-fifth of the sample was diluted with 2 ml of Buffer A and applied to a Mono Q PC 1.6/5 column equilibrated with 280 mM NaCl in Buffer A. After the column was washed with 2 ml of the same buffer, elution was performed with a 3-ml linear gradient of NaCl (280-500 mM) in Buffer A, followed by a 0.5-ml linear gradient of NaCl (0.5-1 M) and 0.5 ml of 1 M NaCl in Buffer A at a flow rate of 0.1 ml/min. Fractions of 0.1 ml each were collected. One peak of the Rab3 GAP activity appeared in Fractions 10 and 11 (see Fig. 1A). These fractions (0.2 ml, 14 µg of protein) were collected. The rest of the heparin-Sepharose sample was subjected to the same Mono Q column chromatography four times in a similar manner. The samples of the five Mono Q column chromatographies were pooled and stored at 80 °C until use.

The Mono Q sample of Rab3 GAP (13 µg of protein) was subjected to SDS-PAGE. The protein band corresponding to Rab3 GAP was cut out from the gel and digested with lysyl endopeptidase, and the digested peptides were separated by C18 reverse phase high pressure liquid column chromatography as described (13). The amino acid sequences of the peptides were determined with a peptide sequencer (Shimadzu PSQ-1-gas phase sequencer). To determine the N-terminal amino acid sequence, the Mono Q sample of Rab3 GAP (2.6 µg of protein) was subjected to SDS-PAGE and transferred to a polyvinylidene difluoride membrane. The protein band was cut out from the membrane and directly subjected to the peptide sequencer.

The hybridization procedures for screening a human brain cDNA library were carried out (14). The cDNA clones obtained in a gt10 phage vector were recloned using a pBluescript plasmid, and the nucleotide sequence was determined (15).

The 2946-bp fragment containing the complete Rab3 GAP cDNA coding region was synthesized by polymerase chain reaction and inserted into pRSET. pRSET-Rab3 GAP was transformed into Escherichia coli DE3, and His₆-tagged Rab3 GAP was purified by Ni²⁺-nitrilotriacetic acid-agarose column chromatography (16). Deletion mutant cDNAs (1-909, 910-1800, and 1801-2946 bp) were synthesized by polymerase chain reaction and inserted into pGEX-2T. pGEX-2T-Rab3 GAP deletion mutants were transformed into E. coli DH5, and glutathione S-transferase-tagged Rab3 GAP deletion mutants were purified by glutathione-Sepharose beads (17).

Proteins were separated by SDS-PAGE (8% polyacrylamide gel) as described (18). Protein concentrations were determined with bovine serum albumin as a reference protein (19).

RESULTS

Rab3 GAP was purified with lipid-modified (geranylgeranylated and methylated) Rab3A as a substrate from the SS fraction of rat brain by column chromatographies, including Q-Sepharose, hydroxyapatite, heparin-Sepharose, and Mono Q column chromatographies. On these column chromatographies, Rab3 GAP appeared in a single peak. On the final Mono Q column chromatography, the GAP activity well coincided with two proteins with molecular masses of about 150 kDa and 130 kDa (Fig. 1, A and B). When this sample was subjected to sucrose density gradient ultracentrifugation followed by re-Mono Q column chromatography, the GAP activity well coincided again with these two proteins (Fig. 1, C and D). The mass estimated by sucrose density gradient ultracentrifugation was about 290 kDa. Of these two proteins, the 130-kDa protein showed the GAP activity as estimated by an overlay method (Fig. 2A). It is likely that Rab3 GAP is a heterodimer with 150-kDa and 130-kDa proteins, but further study is necessary for this conclusion.

The Rab3 GAP sample from the first Mono Q column chromatography indeed showed GAP activity as estimated by thin layer chromatography (Fig. 2B). It was active on lipid-modified Rab3A, but not on the unmodified form (Fig. 3A). Of the many Rab family members, Rab3 GAP was active on the Rab3 subfamily members, including Rab3A, -B, -C, and -D, and was inactive on other Rab family members, including Rab2, -5A, and -11 (Fig. 3B). Of the Rab3 subfamily members, Rab3 GAP was most active on Rab3A, -C, and -D, but weakly active on Rab3B.

The 130-kDa protein was accumulated from 200 rat brains by the same series of column chromatographies as described above, and most of the Mono Q sample was subjected to SDS-PAGE. The protein band corresponding to this protein was cut out from the gel and digested completely with lysyl endopeptidase, and the digested peptides were subjected to C18 reverse phase high pressure liquid column chromatography. Over 30 peptides peaks were observed, and the amino acid sequences of the three peptides were determined. These sequences were PVPARRQRRLFDDTREAEK, LTEPAPVPIHK, and DMAPLKPEGRLHQHGK. The N-terminal amino acid sequence of the 130-kDa protein was further determined to be AADSEPESEV. On the basis of these amino acid sequences, one cDNA was cloned from a human brain cDNA library and its primary structure was determined. On the other hand, computer homology search revealed that this sequence was identical to one gene deposited in the GenBank^TM Nucleotide Sequence Data base with the accession number D31886[GenBank], of which function is unknown. The cDNA encoded a protein with a M_r of 110,521 and 981 amino acids. All the amino acid sequences of the peptides except 6 amino acids were included in this sequence. The slight differences of the amino acid sequences might be due to the differences in species. The cDNA encoding Rab3 GAP showed no homology to any known protein.

The recombinant 130-kDa protein was prepared as a His₆-tagged protein from E. coli. This recombinant protein indeed showed the GAP activities toward Rab3A and other Rab3 subfamily members (Fig. 3C). The recombinant protein was nearly inactive on other Rab family members, including Rab2, -5A, and -11. However, the specific activities of the recombinant protein toward the Rab3 subfamily members were weaker than those of the Mono Q sample. The reason for the weaker specific activities of the recombinant protein is not known, but may be due to the His₆-tagged recombinant protein produced in E. coli or to the absence of the 150-kDa protein. The catalytic domain analysis using recombinant proteins of deletion mutants indicated that the catalytic domain resided in at least 601-981-amino acid residues (data not shown). Northern blot analysis indicated that the 130-kDa protein mRNA was ubiquitously expressed in human tissues (Fig. 4).

DISCUSSION

We have purified here a GAP specific for the Rab3 subfamily of small G proteins from rat brain, isolated its cDNA, determined its primary structure, and characterized it. A Rab3A GAP has previously been reported (9), but it has partially been purified from rat brain and briefly characterized. In yeast, one GAP active on yeast Ypt6 and -7, but not on Ypt1 and Sec4, has been cloned and characterized (7), but any GAP specific for each Rab family member or Rab subfamily has not been identified in mammal. Therefore, this paper describes the first example of the isolation and characterization of a Rab GAP specific for one Rab subfamily or Rab family member in mammal. GAPs for other small G proteins including the Ras, Rho, and Arf family members have been isolated and characterized. GAPs for the Ras and Rho family members share the common catalytic domains specific for each family (for a review, see Ref. 22). Our Rab3 GAP does not have any homologous region to these GAPs. Our results that Rab3 GAP is very specific for the Rab3 subfamily members suggest that each Rab family member or Rab subfamily has their own specific GAP. It is important to isolate GAPs specific for each Rab family member or Rab subfamily and to know whether they have a common catalytic domain or their own specific catalytic domain.

We have previously shown that Rab GDI, which keeps all the Rab family members including Rab3A in the GDP-bound form and in the cytosol fraction, requires the lipid modifications of their substrate small G proteins for their interactions and the Rab GDI action (5, 6). We have recently isolated and characterized a GEP specific for the Rab3 subfamily and found that this GEP also requires the lipid modifications of Rab3A for the Rab3 GEP action.² We have shown here that Rab3 GAP requires the lipid modifications of Rab3A for the Rab3 GAP action. These results indicate that the lipid modifications of the Rab3 subfamily are essential for the actions of all the three regulators. In contrast, rabphilin3, a downstream target molecule of the Rab3 subfamily, does not require their lipid modifications of Rab3A for their interactions (5, 23).

It is currently unclear how GTP-Rab3A is associated with synaptic vesicles, but our assumption is that GTP-Rab3A is associated with the vesicles through both the protein-protein (Rab3A-rabphilin3) and lipid-lipid (geranylgeranyl moiety-vesicle phospholipid) interactions (5). If this is the case, Rab3 GAP should act on this Rab3A. We have previously shown that rabphilin3 inhibits the GAP activity of the partially purified sample of Rab3A GAP (24). We have confirmed this result by use of the Mono Q sample of Rab3 GAP (data not shown). It is unknown how Rab3 GAP is activated and what is the function of the lipid moieties of Rab3A for the Rab3 GAP action. It is unknown either when Rab3 GAP is activated, before, during, or after the fusion of synaptic vesicles with the presynaptic plasma membrane. It is important to clarify these issues to establish not only the mode of action of the Rab3 subfamily members in Ca²⁺-dependent exocytosis but also that of other Rab family members in general vesicle trafficking.