Identification and Characterization of a Nerve Terminal-enriched Amphiphysin Isoform*

Amphiphysin is a nerve terminal-enriched protein thought to function in synaptic vesicle endocytosis, in part through Src homology 3 (SH3) domain-mediated interactions with dynamin and synaptojanin. Here, we report the characterization of a novel amphiphysin isoform (termed amphiphysin II) that was identified through a homology search of the data base of expressed sequence tags. Antibodies specific to amphiphysin II recognize a 90-kDa protein on Western blot that is brain-specific and highly enriched in nerve terminals. Like amphiphysin (now referred to as amphiphysin I), amphiphysin II binds to dynamin and synaptojanin through its SH3 domain. Further, both proteins bind directly to clathrin in an SH3 domain-independent manner. Taken together, these data suggest that amphiphysin II may participate with amphiphysin I in the regulation of synaptic vesicle endocytosis.

Amphiphysin is a nerve terminal-enriched protein thought to function in synaptic vesicle endocytosis, in part through Src homology 3 (SH3) domain-mediated interactions with dynamin and synaptojanin. Here, we report the characterization of a novel amphiphysin isoform (termed amphiphysin II) that was identified through a homology search of the data base of expressed sequence tags. Antibodies specific to amphiphysin II recognize a 90-kDa protein on Western blot that is brainspecific and highly enriched in nerve terminals. Like amphiphysin (now referred to as amphiphysin I), amphiphysin II binds to dynamin and synaptojanin through its SH3 domain. Further, both proteins bind directly to clathrin in an SH3 domain-independent manner. Taken together, these data suggest that amphiphysin II may participate with amphiphysin I in the regulation of synaptic vesicle endocytosis.
Following their exocytosis at the nerve terminal, synaptic vesicle membranes are retrieved by internalization through clathrin-coated pits and vesicles (1). A number of proteins have been identified that are thought to be involved in this process, including dynamin, which is similar to the product of the Drosophila shibire gene (2,3). A temperature-sensitive mutation in this gene leads to a block in endocytosis of synaptic vesicles (4), and recent data suggests that dynamin functions in the nerve terminal by mediating the fission of endocytic vesicles (5,6). Another protein putatively involved in synaptic vesicle endocytosis is synaptojanin, which was identified based on its ability to bind to the Src homology 3 (SH3) 1 domains of Grb2 (7). Synaptojanin is enriched in brain and is concentrated in the nerve terminal, where it precisely co-localizes with dynamin (8). Both dynamin and synaptojanin are associated with synaptophysin-positive membrane fractions on the endocytic limb of the synaptic vesicle cycle, and both proteins undergo dephosphorylation in response to synaptic vesicle mobilization (8). Cloning of synaptojanin revealed a domain homologous to a family of enzymes that dephosphorylate inositol phospholipids and inositol polyphosphates at the 5Ј-position of the inositol ring (5-phosphatases) as well as a proline-rich C terminus with consensus binding sites for SH3 domains (9). We have recently identified a major binding partner for synaptojanin as SH3P4 (10), a member of a novel family of SH3 domain-containing proteins (11).
Recently, two mammalian amphiphysin homologues have been identified, BIN1 (20) and SH3P9 (11). BIN1, which is localized to the nucleus, was identified as a Myc-binding protein (20). The Myc oncoprotein contains functionally critical N-terminal Myc box regions that have been implicated in cell growth, apoptosis, and malignancy (21,22), and BIN1 binding to these domains inhibits malignant cell transformation by Myc (20). Little information is available on SH3P9, which was discovered along with SH3P4 in a library screen isolating SH3 domain-containing proteins (11). BIN1 and SH3P9, which are very similar in sequence, are homologous to amphiphysin at their N termini as well as at their C-terminal SH3 domains but do not contain a large central domain found in amphiphysin.
In this paper, we have characterized a third amphiphysin homologue, which appears to be an alternatively spliced product of the gene encoding BIN1 and SH3P9. This protein, which we refer to as amphiphysin II, contains a large central insert domain that is partially homologous to the central domain of amphiphysin (now referred to as amphiphysin I) but that is absent from BIN1 and SH3P9. We generated amphiphysin II-specific antibodies and determined that the protein is brainspecific and highly enriched in nerve terminals. Like amphiphysin I, amphiphysin II binds through its SH3 domain to dynamin and synaptojanin. Further, amphiphysin I and II both bind to clathrin through a region outside of their SH3 domains. Taken together, these data suggest that amphiphysin II may function in synaptic vesicle endocytosis in the nerve terminal.

EXPERIMENTAL PROCEDURES
Identification of Amphiphysin II-A search of the data base of expressed sequence tags (dbEST) with the sequence of human amphiphysin I (13) 1 The abbreviations used are: SH3, Src homology 3; AP2, assembly protein 2; EST, expressed sequence tag; dbEST, EST data base; PCR, polymerase chain reaction; GST, glutathione S-transferase; PBS, phosphate-buffered saline; PAGE, polyacrylamide gel electrophoresis.
To obtain the amino terminus of amphiphysin II, we performed PCR on cDNA prepared from adult human cortex (23). The forward primer (5Ј-GCGGGATCCATGGCAGAGATGGGCAGTAAAG) corresponded to nucleotides 1-22 of the coding sequence of BIN1 (20) (from GenBank TM clone U68485), and the reverse primer (5Ј-GCGGAATTCGTCCAGCA-GACTGGCCTGC) corresponded to nucleotides 1083-1065 of amphiphysin II in a region of the amphiphysin II-specific insert domain. A specific PCR product was subcloned, sequenced, and found to have a 100% overlap with 564 base pairs at the 5Ј-end of the I.M.A.G.E. consortium clone and to extend the clone by an additional 519 base pairs.
Southern Blot Analysis of Amphiphysin II-Human genomic DNA (8 g/sample) was digested with BamHI or EcoRI, and the fragments were separated on an agarose gel and transferred to Hybond Nϩ membrane (Amersham Corp.). A cDNA probe from nucleotides 1396 -1671 of amphiphysin II was prepared by PCR, labeled with [ 32 P]dCTP by random priming, and hybridized to immobilized DNA under high stringency using standard protocols.
Production of Amphiphysin II-specific Antibodies-To generate an amphiphysin II antibody, the I.M.A.G.E. consortium clone was used as a template in PCR reactions with Vent Polymerase (New England Biolabs) with the forward primer, 5Ј-GCGGGATCCTCCCAGTTTGAG-GCCCCG (nucleotides 1035-1050 of amphiphysin II) and the reverse primer, 5Ј-GCGGAATTCTCATGGGACCCTCTCATGT (nucleotides 1413-1396 of amphiphysin II). The PCR product was cloned in frame into the BamHI-EcoRI sites of pGEX-2T (Pharmacia Biotech Inc.), and a GST fusion protein (GST-amphiphysin II) was expressed and purified as described (7). Polyclonal antiserum 1874 was prepared by injecting a rabbit with 200 g of GST-amphiphysin II using Titermax adjuvant (CytRx Corp) with standard protocols. Rabbit sera were tested by Western blot on crude brain homogenate. Antibodies specific to amphiphysin II were affinity-purified from the serum in a two-step procedure. First, 2 ml of 1874 serum were incubated overnight at 4°C with approximately 2.0 mg of a GST fusion protein encoding a region of the amphiphysin II-specific insert domain immobilized on polyvinylidene difluoride membrane (Bio-Rad). The fusion protein was prepared by PCR as described above using the forward primer 5Ј-GCGGGATCCTC-CCAGTTTGAGGCCCCG (nucleotides 1035-1050 of amphiphysin II) and the reverse primer, 5Ј-GCGGAATTCACACAGCAAAGGTGC-CCTCG (nucleotides 1240 -1221 of amphiphysin II). The membrane was washed, and the bound antibodies were eluted by incubation for 5 min in 3 ml of 50 mM glycine, pH 2.5. The affinity-purified antibodies were subsequently neutralized by the addition of 150 l of 2 M Tris, pH 8.0 (single affinity purification). In the second step of affinity purification, 0.5 ml of singly affinity-purified antibodies were incubated overnight at 4°C with strips of polyvinylidene difluoride membrane containing approximately 2.0 mg of a GST-amphiphysin I fusion protein prepared by PCR as described above using the forward primer 5Ј-GCGGGATCCTCCCAGAATGAAGTCCCTG (nucleotides 1122-1140 of amphiphysin I) and the reverse primer, 5Ј-GCGGAATTCCTAATCTA-AGCGTCGGGTG (nucleotides 2198 -2180 of amphiphysin I). The unbound antibodies (doubly affinity-purified) were collected and tested on Western blots.
Preparation of Tissue Fractions-Various tissues were homogenized at 1:10 (w:v) in 20 mM HEPES-OH, pH 7.4, containing 0.83 mM benzamidine, 0.23 mM phenylmethylsulfonyl fluoride, 0.5 g/ml aprotinin, and 0.5 g/ml leupeptin followed by centrifugation for 5 min at 800 ϫ g max . Postnuclear supernatants were separated on SDS-PAGE, transferred to nitrocellulose, and processed for Western blot analysis. Subcellular fractionation of brain homogenates to generate synaptic fractions was performed as described (8).
Immunofluorescence Microscopy-Rats were anesthetized and perfused through the ascending aorta with 50 ml of 0.1 M NaPO4, pH 7.4 (phosphate buffer) followed by 400 ml of 4% paraformaldehyde in phosphate buffer. The brains were dissected out, postfixed, and cryoprotected in 30% sucrose, and 15-m sections were cut on a freezing microtome. Sections were permeabilized with phosphate-buffered saline (PBS) (20 mM NaPO 4 , pH 7.4, 150 mM NaCl) containing 0.3% Triton X-100 for 10 min and then blocked for 30 min with PBS containing 5% bovine serum albumin and 5% normal goat serum. The sections were then incubated in primary antibodies for 2 h in PBS containing 1% bovine serum albumin. The immunolabeling was revealed with CY-3conjugated goat anti-rabbit IgG in PBS for 1.5 h. All steps were performed at room temperature with extensive washes in PBS.
Purification of Clathrin-Clathrin-coated vesicles were purified from rat brain as described (24). Clathrin was extracted by incubating the purified vesicles with 0.5 M Tris-Cl, pH 7.4, for 30 min at 4°C followed by removal of insoluble material by centrifugation in a Beckman TLA 100.1 rotor at 50,000 rpm. for 30 min (25). The supernatant fraction was then incubated overnight at 4°C with ϳ25 g each of various amphiphysin fusion proteins bound to glutathione-Sepharose. Samples were washed in 0.5 M Tris, pH 7.4, and bound proteins were eluted with SDS-PAGE sample buffer.

RESULTS
Amino Acid Sequence of Amphiphysin II-A partial amino acid sequence for amphiphysin II was determined from an I.M.A.G.E. consortium clone identified through an amphiphysin I homology search of the dbEST. Analysis of the protein suggested that it was an alternatively spliced version of the amphiphysin homologue BIN1 (20). To obtain the amino terminus of amphiphysin II, we performed PCR on cDNA prepared from adult human cortex (23) using a forward primer encoding the N terminus of BIN1 (from GenBank TM clone U68485) and a reverse primer in a region of the I.M.A.G.E. consortium clone unique to amphiphysin II. A specific PCR product was generated that overlapped the I.M.A.G.E. consortium clone by 188 amino acids and extended the amino terminus of amphiphysin II by 173 amino acids. The amino acid sequence of amphiphysin II is shown aligned to amphiphysin I, BIN1, and SH3P9 (Fig. 1A). Amphiphysin II, BIN1, and SH3P9 are essentially identical at their N termini and their SH3 domain-containing C termini. Minor differences between SH3P9 and the other two sequences are due to interspecies variations (20) (amphiphysin II and BIN1 are from humans, whereas SH3P9 is from mice). The central domain of amphiphysin II contains a 123-amino acid insert that is not found in SH3P9 or BIN1. The amphiphysin II insert domain is homologous to amphiphysin I through its N-terminal half, but the homology is decreased in the C terminus of the insert domain.
A 15-amino acid insert, which is unique to the BIN1 sequence, encodes a nuclear localization signal ( Fig. 1A; Ref. 20). A 31amino acid insert specific to amphiphysin II is located between amino acids 174 and 204. The pattern of homology seen for amphiphysin II with BIN1 and SH3P9, in which regions of 100% identity are interrupted by regions unique to each protein, suggests that the three proteins arise from alternative splicing of the same gene. To confirm this, we performed Southern blot analysis with human genomic DNA digested with BamHI or EcoRI and hybridized at high stringency with a cDNA probe encoding a region of the amphiphysin II SH3 domain conserved in BIN1 and SH3P9. For both digests, a single hybridizing band was observed (Fig. 1B), further suggesting that amphiphysin II, BIN1, and SH3P9 are encoded by the same gene.
Detection of Amphiphysin II in Brain Homogenate-To begin to characterize the amphiphysin II protein, a polyclonal antiserum (antiserum 1874) was produced against a GST fusion protein encoding the C-terminal 217 amino acids of amphiphysin II. To generate an antibody reactive with amphiphysin II but not BIN1 or SH3P9, the serum was affinity-purified against a fusion protein encoding a region of the amphiphysin II-specific insert domain (amino acids 376 -444). When tested against strips of crude rat brain homogenate, the affinitypurified antibody was found to react strongly with bands at 90 and 125 kDa (singly affinity-purified 1874; Fig. 2). The band at 125 kDa co-migrated with amphiphysin I (CD5; Fig. 2), suggesting that the antibody is cross-reacting with amphiphysin I. We therefore incubated the singly affinity-purified 1874 antibody with the amphiphysin I fusion protein immobilized on polyvinylidene difluoride membranes, and the unbound material reacted specifically with the 90-kDa band (doubly affinitypurified 1874; Fig. 2). The predicted molecular weight for amphiphysin II is 64,703, suggesting that amphiphysin II runs aberrantly on SDS-PAGE, as is the case for amphiphysin I and BIN1, which have predicted molecular masses of approximately 76 and 50 kDa and run on SDS-PAGE at 128 and 70 kDa, respectively (13,20). Expression of amphiphysin II constructs in bacteria confirm the aberrant migration of the amphiphysin II protein (data not shown).
Tissue and Subcellular Distribution of Amphiphysin II-Western blot analysis of tissue extracts with the doubly affinity-purified amphiphysin II antibody revealed expression of amphiphysin II in brain only (Fig. 3A). In contrast, amphiphysin I, although enriched in brain, was detected in testis extracts as well (Fig. 3A). This is in contrast to BIN1, which is present in a wide variety of tissues and most highly expressed in skeletal muscle (20). Upon subcellular fractionation of brain extracts, amphiphysin I and II demonstrate a similar distribution pattern, with both proteins present in soluble and membrane-associated fractions (Fig. 3B). The subcellular localization of amphiphysin II was then studied using immunofluorescence microscopy. In all brain regions examined, amphiphysin II immunoreactivity had a punctate appearance, suggesting a concentration of the antigen in synaptic terminals. Thus, a rim of bright puncta were seen surrounding cell bodies and primary dendrites, which themselves were not immunolabeled (Fig. 4, A and B). In the cerebellum, the molecular layer, which is abundant in synaptic terminals, was brightly stained, while the granule cell layer, which is densely packed with granule cell bodies, exhibited immunoreactivity only in discrete clusters corresponding to the mossy fiber terminals. Axons in the white matter were not immunolabeled. Similarly, in the spinal cord, immunonegative cell bodies and dendrites of motor neurons were surrounded by brightly stained puncta, while axonal bundles did not exhibit any immunoreactivity. We never observed labeling of nuclei that are enriched in BIN1 (20). The staining pattern for amphiphysin II was practically identical with the immunolabeling observed with an antibody against synaptophysin, a synaptic terminal protein (Fig. 4, E and F). Amphiphysin I immunolabeling (Fig.  4, C and D) also demonstrated a similar punctate synaptic-like distribution (14) although labeling was also present at moderate levels in the cell body cytoplasm.
Interaction of Amphiphysin II with Dynamin and Synaptojanin-Amphiphysin I binds through its SH3 domain to proline-rich sequences in synaptojanin and dynamin. To examine the binding properties of amphiphysin II, we used GST-amphiphysin II in overlay assays on rat brain postnuclear supernatant fractions (7). The fusion protein binds specifically to bands of 145 and 100 kDa (Fig. 5A), a pattern identical to that seen for Grb2 binding (7,8), suggesting that the bands are synaptojanin and dynamin. To confirm the identity of these proteins and to determine if the interactions were SH3 domainmediated, we performed affinity purifications from rat brain extracts using fusion proteins encoding the SH3 domains of amphiphysin I and II. As a control, we also used fusion proteins encoding a non-SH3 domain region of amphiphysin I and II. The SH3 domain fusion proteins from both amphiphysin I and II strongly bind to dynamin and synaptojanin, whereas the non-SH3 domain amphiphysin I and II fusion proteins do not bind (Fig. 5B).
Amphiphysin Binding to Clathrin-While examining the binding of dynamin and synaptojanin to amphiphysin II, we FIG. 2. Specificity of amphiphysin antibodies. Proteins from rat brain homogenates were separated on SDS-PAGE, transferred to nitrocellulose, and processed for Western blots with antibodies raised against amphiphysin I (CD5) or amphiphysin II (1874). The 1874 serum was affinity-purified using a GST fusion protein encoding the amphiphysin II insert domain (single). The singly affinity-purified antibody was then incubated with a fusion protein against amphiphysin I to remove cross-reactive antibodies (double).

FIG. 3. Tissue and subcellular distribution of amphiphysin II.
A, postnuclear extracts from a variety of tissues as indicated were separated on SDS-PAGE, transferred to nitrocellulose, and subjected to Western blot analysis with antibodies specific to amphiphysin I and amphiphysin II as indicated. The arrows on the right indicate the migratory positions of the proteins detected on the blots. B, proteins of brain subcellular fractions were separated by SDS-PAGE, transferred to nitrocellulose, and subjected to Western blots as described in A. Subcellular fractions were prepared as described (8). H, homogenate; P, pellet; S, supernatant; LP, lysed pellet; LS, lysed supernatant. The arrows on the right indicate the migratory positions of the proteins detected on the blots.
noticed the presence of a 170-kDa protein that bound to the non-SH3 domain fusion protein. In Fig. 6A, we demonstrate that the 170-kDa band is clathrin and that it binds to the non-SH3 domain fusion proteins of amphiphysin I and II but does not bind to the SH3 domain of either of these proteins. To further study the clathrin binding, we generated clathrincoated vesicles from rat brain and extracted the clathrin using 0.5 M Tris (25). Clathrin prepared in this manner bound to the non-SH3 domain fusion proteins from amphiphysin I and II but did not bind to the fusion proteins encoding the SH3 domains (Fig. 6B). It has previously been demonstrated that amphiphysin I binds to the ␣-subunit of AP2 (14,16). Therefore, we examined the binding of AP2 to our amphiphysin fusion proteins. The two ␣-adaptin subunits (␣ a and ␣ c ) of AP2 bound weakly to the SH3 domain of amphiphysin II but did not bind to the SH3 domain of amphiphysin I or to the non-SH3 domain fusion proteins (Fig. 6A). To further explore this issue, we generated a GST fusion protein against the insert domain of amphiphysin II and a corresponding fusion protein from amphiphysin I. Interestingly, the amphiphysin I but not the amphiphysin II fusion protein bound ␣-adaptin (Fig. 6C). DISCUSSION Amphiphysin I is a nerve terminal-enriched protein that is thought to function in the endocytosis of synaptic vesicles. Here, we report the amino sequence of amphiphysin II, which we identified based on the sequence of a cDNA clone from the dbEST. Recently, the amino acid sequence of two other amphiphysin I homologues, BIN1 (20) and SH3P9 (11), were reported. BIN1, SH3P9, and amphiphysin II align with long blocks of virtually identical amino acid sequence interrupted by stretches of sequence specific to the individual proteins. Southern blot analysis of human genomic DNA suggests that the three proteins are generated from a single gene. Thus, it seems likely that the three proteins are encoded by alternative splicing leading to the generation of unique sequences that impart distinct properties upon each protein. For example, the only apparent difference between BIN1 and SH3P9 is the presence of a small insert sequence (RKKSKLFSRLRRKKN) in BIN1, which encodes a nuclear localization signal. This is consistent with the function of BIN1 as a Myc-binding protein localized in the nucleus (20) and would suggest that SH3P9 may share a common function with BIN1 but be localized outside of the nucleus. A striking feature of amphiphysin II is that it contains a large central domain of 123 amino acids that is not present in either of the other two molecules. This region is strongly homologous to amphiphysin I throughout its N-terminal half, including the presence of a highly conserved proline-rich sequence that forms a potential SH3 domain-binding site (14,29). The insert domain also contains a binding site for clathrin as discussed below.
To begin to characterize amphiphysin II, we generated a polyclonal antiserum against an amphiphysin II-GST fusion protein. To avoid reactivity against BIN1 and SH3P9, we affinity-purified the serum using a fusion protein encoding a region of amphiphysin II restricted to the insert domain. The affinity-purified antibodies reacted with both amphiphysin I and II. Therefore, antibodies against amphiphysin I were absorbed out using an immobilized amphiphysin I fusion protein generating an amphiphysin II-specific antibody. This antibody was used to determine that amphiphysin II is highly enriched in the brain and is concentrated in the presynaptic nerve ter- FIG. 5. Amphiphysin II binding assays. A, overlay assays were performed on a rat brain postnuclear supernatant fraction using GSTamphiphysin II (amphiphysin II overlay) or GST alone (GST overlay). B, Triton X-100-soluble proteins from brain were incubated with GST fusion proteins (GST-amphiphysin I non-SH3 domain, GST-amphiphysin I SH3 domain, GST-amphiphysin II non-SH3 domain, and GSTamphiphysin II SH3 domain as indicated) which were conjugated to glutathione-Sepharose, and material specifically bound to the beads was eluted with SDS-PAGE sample buffer. The panels show immunoblots with antibodies against synaptojanin and dynamin as indicated.
minal. This is in contrast to BIN1, which has its highest expression levels in skeletal muscle and is enriched in nuclei (20). The tissue distribution and subcellular localization of SH3P9 is unknown. A number of proteins involved in synaptic vesicle recycling have been shown to be concentrated in the nerve terminal including clathrin (5), AP2 (14), dynamin and synaptojanin (8), and amphiphysin I (14). Further, recent morphological data suggest that dynamin, synaptojanin, and amphiphysin I are present with AP2 at clathrin coats in the nerve terminal and in nonneuronal cells (5,30,31).
Synaptojanin and dynamin are the major brain proteins that interact with the SH3 domains of amphiphysin I (14). Based on the conservation of the SH3 domains of amphiphysin I and II, we decided to test the ability of amphiphysin II to interact with these proteins. By overlay and column chromatography assays, dynamin and synaptojanin appear to be the major brain proteins that interact with the SH3 domain of amphiphysin II. As a control for the experiments, we used a fusion protein to a non-SH3 domain region of amphiphysin II that encoded a portion of the insert domain as well as the corresponding fusion protein from amphiphysin I. Interestingly, while these fusion proteins did not bind to synaptojanin or dynamin, they were both able to specifically precipitate clathrin from a crude Triton X-100 brain extract. To further study the interaction of clathrin with amphiphysin, we purified clathrin-coated vesicles from rat brain and extracted the clathrin with 0.5 M Tris-Cl, pH 7.4. This procedure releases clathrin as soluble triskelia and leads to an enrichment in clathrin relative to the clathrin adaptor protein AP2 (25). Under these conditions, clathrin retains its strong binding to the amphiphysin isoforms. Thus, both amphiphysin I and II bind to synaptojanin and dynamin through their SH3 domains and bind to clathrin in an SH3 domainindependent manner.
It has previously been demonstrated that amphiphysin I can interact with AP2 (14,16). Thus, the interaction of amphiphysin I and II with clathrin observed here could be indirect and mediated through AP2. However, as determined by immunoblot with an antibody against the ␣-adaptin subunit (28), AP2 was not present in the non-SH3 domain fusion protein precipitates in which clathrin was highly enriched. In contrast, AP2 did bind to an amphiphysin I fusion protein encoding a region of amphiphysin I from amino acids 291-445. The non-SH3 domain amphiphysin I fusion protein, which does not bind clathrin, encodes a region of amphiphysin I from amino acids 338 -565, suggesting the presence of a binding site for ␣-adaptin between amino acids 291 and 338 of amphiphysin I. In addition, AP2 was found to bind to the SH3 domain of amphiphysin II but not amphiphysin I. It was previously demonstrated that AP2 could interact with the SH3 domains of Grb2 (14). The interaction of the SH3 domain of amphiphysin II with the ␣-adaptin subunit of AP2 reported here could be direct, through proline-rich sequences in the ␣-adaptin subunit (32).
A model has been proposed (14) that suggests that amphiphysin I, which is concentrated on clathrin coats in the were purified from rat brain, and the clathrin was extracted with 0.5 M Tris-Cl, pH 7.4. The clathrin extract (extract) was incubated with amphiphysin I and II fusion proteins as indicated, and the material specifically bound to the beads was eluted with SDS-PAGE sample buffer. The panel in B shows a Coomassie-stained SDS-gel with the migratory position of clathrin noted. C, Triton X-100-soluble proteins from brain were incubated with GST fusion proteins (GST-amphiphysin I insert domain and GST-amphiphysin II insert domain as indicated) which were conjugated to glutathione-Sepharose, and material specifically bound to the beads was eluted with SDS-PAGE sample buffer. The panel in C shows an immunoblot of equal aliquots of starting material (SM), unbound material (void), and bead fraction (beads) stained with an antibody against ␣-adaptin.
FIG. 6. Clathrin binding. A, Triton X-100-soluble proteins from brain were incubated with GST fusion proteins (GST-amphiphysin I non-SH3 domain, GST-amphiphysin I SH3 domain, GST-amphiphysin II non-SH3 domain, and GST-amphiphysin II SH3 domain as indicated) which were conjugated to glutathione-Sepharose, and material specifically bound to the beads was eluted with SDS-PAGE sample buffer. The panels show immunoblots with an antibody against clathrin or ␣-adaptin isoforms (␣ a and ␣ c ) as indicated. B, clathrin-coated vesicles (CCVs) nerve terminal (31), is recruited to these sites through its SH3 domain-independent interactions with AP2. The SH3 domain of amphiphysin I then serves to target dynamin to the endocytic site (14,30,33). Our data would suggest an additional role for amphiphysin II in the recruitment of dynamin and synaptojanin through SH3 domain-dependent interactions leading to the observed concentration of both of these proteins on clathrin-coated structures (5,30,31). Further, interactions of the SH3-independent domains of both proteins with clathrin could provide an additional targeting substrate for the amphiphysin isoforms. Cycles of phosphorylation and dephosphorylation of dynamin (34) and synaptojanin (8), which are regulated by the polarization state of the nerve terminal, may play an important role in regulating these targeting events. Further, SH3P4, which is a major SH3 domain-containing synaptojanin-binding protein in brain (10), may play a role in regulating the association of synaptojanin with the amphiphysin isoforms. Once at the endocytic site, dynamin functions in fission of the clathrincoated vesicles (5,6), and synaptojanin may function to regulate local concentrations of inositol polyphosphates or inositol phospholipids, leading to changes in the function of adaptins (35) or dynamin (36,37). Regardless, the data presented here suggest an important role for amphiphysin II in synaptic vesicle endocytosis.