Role of Phosphorylation in Regulation of the Assembly of Endocytic Coat Complexes

See allHide authors and affiliations

Science  07 Aug 1998:
Vol. 281, Issue 5378, pp. 821-824
DOI: 10.1126/science.281.5378.821


Clathrin-mediated endocytosis involves cycles of assembly and disassembly of clathrin coat components and their accessory proteins. Dephosphorylation of rat brain extract was shown to promote the assembly of dynamin 1, synaptojanin 1, and amphiphysin into complexes that also included clathrin and AP-2. Phosphorylation of dynamin 1 and synaptojanin 1 inhibited their binding to amphiphysin, whereas phosphorylation of amphiphysin inhibited its binding to AP-2 and clathrin. Thus, phosphorylation regulates the association and dissociation cycle of the clathrin-based endocytic machinery, and calcium-dependent dephosphorylation of endocytic proteins could prepare nerve terminals for a burst of endocytosis.

Clathrin-mediated endocytosis plays a key role in the recycling of synaptic vesicles in nerve terminals, and several components of the molecular machinery involved in this process have been identified (1). These include, in addition to clathrin and the clathrin adaptors, the guanosine triphosphatase dynamin 1, the amphiphysin dimer, and synaptojanin 1. Dynamin 1 oligomerizes into collar structures at the neck of deeply invaginated clathrin-coated pits, and its conformational change is thought to be an essential step leading to vesicle fission (2). Synaptojanin 1 is a presynaptic inositol 5-phosphatase enriched on endocytic intermediates (3). The amphiphysin dimer (4–6) binds to both dynamin 1 and synaptojanin 1 through the COOH-terminal SH3 domains of its two subunits (4,7, 8). Disruption of SH3-mediated interactions of amphiphysin blocks clathrin-mediated endocytosis at the step of invaginated coated pits (9). The amphiphysin dimer also binds to clathrin (10, 11) and to the α-adaptin subunit of the plasma membrane clathrin adaptor AP-2 (9, 12) and thus may mediate recruitment of dynamin 1 and synaptojanin 1 to the site of clathrin-mediated endocytosis.

To identify potential major binding partners for the proline-rich domain (PRD) of dynamin other than amphiphysin, a total brain extract was affinity purified on glutathione S-transferase (GST) fusion proteins comprising the entire PRD or COOH-terminal truncations of the PRD (Fig. 1A) (13). A set of proteins was specifically bound by full-length PRD and by a deletion construct missing the last 16 amino acids (PRDΔC16) but not by a construct lacking an additional 16 amino acids (PRDΔC32) or a larger portion of the PRD (Fig. 1, B and C). The crucial 16 amino acids contain the amphiphysin binding site (8). The major affinity-purified protein bands were identified as amphiphysins 1 and 2, the α and β subunits of the AP-2 clathrin adapter complex (α- and β-adaptin, respectively) (14), and clathrin, based on both their electrophoretic mobilities and coenrichment during affinity purification with the corresponding immunoreactivities as determined by immunoblotting. Additional proteins were found by immunoblotting to be specifically retained by the PRD constructs containing the amphiphysin binding site (Fig. 1C) (15). These included dynamin 1, synaptojanin 1, and other components of the endocytic machinery, such as eps15 and AP180 (16). Two other SH3 domain–containing proteins, SH3p4 (endophilin 1) and SH3p8 (endophilin 2) (17), bound to all GST-PRD constructs (Fig. 1C), suggesting a localization of the binding site for these proteins upstream of the amphiphysin binding site.

Figure 1

A macromolecular complex comprising several endocytic proteins can be affinity-purified from a Triton X-100 brain extract by the PRD of dynamin (13). (A) Schematic drawing of the GST fusion proteins used for affinity chromatography (8). (B) Coomassie blue staining of the starting rat brain extract and of the material affinity purified by the four constructs shown in (A). (C) Immunoblot analysis of the material shown in (B). (D) Immunoblot analysis of the material affinity-purified by the PRD in the absence and presence of the indicated amounts of amphiphysin 1, SH3 domain of amphiphysin 1, and central fragment of amphiphysin 1 (amino acids 262 to 435).

Other than amphiphysin, none of endocytic proteins that were affinity-purified by using the PRD and PRDΔC16 constructs contain an SH3 domain. Thus these proteins may form multimeric complexes with amphiphysin. In agreement with this possibility, binding of AP-2 and clathrin was increased by the addition of full-length recombinant amphiphysin 1 and decreased by the addition of amphiphysin fragments (Fig. 1D). Furthermore, antibodies to amphiphysin could coprecipitate proteins of the complex from brain lysates (see below). If these complexes occur in vivo, their formation would be likely to undergo regulation because of the cyclic nature of the association and dissociation of the endocytic machinery. Because several proteins of the complex are known to be phosphoproteins [for example, clathrin, AP-2 (18, 19), dynamin 1, synaptojanin 1, and amphiphysin (4, 20, 21)], we used immunoprecipitation experiments to determine whether protein-protein interactions with the complexes were regulated by phosphorylation.

Rat brain extract (22) was depleted of nucleotides and incubated in the presence or absence of adenosine triphosphate (ATP) and either a protein kinase inhibitor or a protein phosphatase inhibitor mixture. This material was subjected to immunoprecipitation with monoclonal antibodies specifically directed against the NH2-terminal region of amphiphysin 1, which do not recognize amphiphysin 2 (22). Amphiphysin 1 had a slower mobility after incubation in the presence of both ATP and phosphatase inhibitors (Fig. 2A), which confirms the effectiveness of the phoshorylation reaction under these conditions (21). A similar shift (Fig. 2A) was exhibited by amphiphysin 2 (23). In addition to amphiphysin 1, the antibodies coprecipitated amphiphysin 2, dynamin 1, synaptojanin 1, AP-2 (α- and β-adaptin), and clathrin. Preincubation of the extract with ATP and phosphatase inhibitors did not affect coprecipitation of amphiphysin 2 but did cause a significant decrease in coprecipitation of other components of the complex, which suggests that phosphorylation affects these interactions (Fig. 2A).

Figure 2

Interaction of amphiphysin with other endocytic proteins is regulated by phosphorylation. Desalted rat brain extracts were used in immunoprecipitation and affinity-purification experiments after preincubation under the conditions indicated in (22). (A) Protein blots of immunoprecipitates generated from a total rat brain Triton X-100 extract (22) by monoclonal antibodies directed against the NH2 region of amphiphysin 1. The presence of ATP and phosphatase inhibitors affects coprecipitation of dynamin 1, synaptojanin 1, AP-2 adaptor components, and clathrin but not of amphiphysin 2. (B) Monoclonal antibodies directed against amphiphysin 1 were used to generate immunoprecipitates from Triton X-100 extracts of a total brain homogenate (10). Immunoprecipitates were reacted by immunoblotting with the antibody CD9, which recognizes both amphiphysins 1 and 2 (5) or by an overlay assay (8) with GST or a GST fusion protein comprising amino acids 1 to 150 of amphiphysin 1 (6). The amphiphysin 1 fragment binds both amphiphysins 1 and 2 regardless of their state of phosphorylation as revealed by the upper mobility shift. (C) Immunoblot for dynamin 1 and synaptojanin 1 of the starting Triton X-100 brain extract and of the material affinity-purified by a GST fusion protein comprising full-length amphiphysin 1 (7). (D) The AP-2 complex and clathrin bind directly to distinct sites in the central part of amphiphysin 1. Coomassie blue staining of material affinity purified from a Triton X-100 brain extract (22) by GST fusion proteins (100 μg/ml) comprising indicated fragments of amphiphysin 1. (E) Triton X-100 rat brain extracts preincubated in the presence or absence of ATP and phosphatase inhibitors were affinity-purified on a GST fusion comprising amino acids 262 to 435 of amphiphysin 1. Eluates were analyzed by protein blotting. (F and G) Brain cytosol was affinity-purified on a GST fusion protein comprising the appendage domain of α-adaptin (25). The affinity-purified material was reacted by protein blotting for amphiphysin 1 and dynamin 1 (F) and band intensity was quantified by a Phosphor- Imager (G). Bars represent the mean ± SD of two independent experiments.

The coprecipitation of amphiphysin 2 from both phosphorylated and dephosphorylated brain extracts supports the presence of amphiphysin heterodimers (4) and indicates that amphiphysin dephosphorylation does not affect heterodimer stability. A GST fusion protein comprising the first 150 amino acids of amphiphysin 1 bound to both amphiphysins 1 and 2 in an overlay assay (Fig. 2B), indicating that dimerization is mediated by this coiled-coil region of amphiphysin; this suggests the possibility that both heterodimers and homodimers are present. Binding was not affected by phosphorylation (Fig. 2B).

To determine whether the phosphorylation site or sites that affect binding of dynamin 1 and synaptojanin 1 to amphiphysin were on these proteins or on amphiphysin, we incubated brain extracts in the presence of ATP and the phosphatase inhibitor mixture and then loaded them on a GST-amphiphysin 1 fusion protein column (7) after we terminated the kinase reactions by adding EDTA. In both cases, binding of dynamin 1 and synaptojanin 1 to the SH3 domain of amphiphysin 1 was significantly reduced by previous exposure of the cytosol to ATP (Fig. 2C) (24). Thus, phosphorylation of dynamin 1 and synaptojanin 1 regulates their interaction with amphiphysin.

We next tested whether clathrin and AP-2 from control and ATP-pretreated cytosol (21) bound differently to recombinant amphiphysin. In preliminary affinity-chromatography experiments the AP-2 binding site was localized to a region (amino acids 322 to 375 of human amphiphysin 1) distinct from, but adjacent to, the clathrin binding site (amino acids 347 to 405 of human amphiphysin 1) (11) (Fig. 2D). Therefore, we used a GST fusion protein of an amphiphysin 1 fragment comprising both regions (amino acids 262 to 435 of human amphiphysin 1) for these experiments (Fig. 2E). Binding of clathrin and α- and β-adaptin was very similar in the two conditions (Fig. 2F), arguing against an effect of clathrin and AP-2 phosphorylation on their binding to amphiphysin. The phosphorylation of amphiphysin, however, was found to affect its binding to AP-2. When a rat brain cytosolic extract was affinity-purified on a GST fusion protein comprising the appendage domain of α-adaptin (25)—that is, the amphiphysin binding portion of AP-2—the phosphorylated forms of both amphiphysin 1 and 2 were retained less efficiently than the corresponding dephosphorylated forms (Fig. 2, F and G).

Thus, complex formation of a multimeric complex between various endocytic proteins is inhibited by phosphorylation (26). We explored the effect of protein kinase inhibitors on coprecipitation with amphiphysin 1 of dynamin 1 and AP-2 (27). The general protein kinase inhibitor K252a (28) strongly inhibited the effect of ATP on both coprecipitation and amphiphysin mobility (Fig. 3, A and B). The kinase or kinases responsible for these effects remain to be identified.

Figure 3

Effect of protein kinase and protein phosphatase inhibitors on coprecipitation of dynamin 1 and AP-2 (β-adaptin subunit) with amphiphysin 1. Monoclonal antibodies directed against amphiphysin 1 were used to generate immunoprecipitates from rat brain cytosol (22), which had been preincubated with ATP, a phosphatase inhibitor mixture, and the compounds indicated. Immunoprecipitates were reacted by protein blotting (A) and band intensity was quantified by a PhosphorImager (B). (C) Monoclonal antibodies directed against amphiphysin 1 were used to generate immunoprecipitates from rat brain cytosol incubated in the presence of ATP and phosphatase inhibitors as indicated. Immunoprecipitates were reacted by protein blotting with antibody to amphiphysin 2 (upper) and the relative amount of coprecipitated dynamin 1 was quantified with a PhosphorImager (lower). Note upper shift of the amphiphysin 2 band correlating with the presence of phosphatase inhibitors. Similar results were obtained for amphiphysin 1.

Dynamin 1, synaptojanin 1, and the amphiphysins undergo stimulation-dependent dephosphorylation in nerve terminals, and dephosphorylation is blocked by inhibitors of the Ca2+/calmodulin-dependent phosphatase calcineurin (4, 20, 21). In the absence of phosphatase inhibitors, ATP was not sufficient to produce a significant mobility shift of the amphiphysins in SDS–polyacrylamide gel electrophoresis (SDS-PAGE) and a corresponding inhibition of the binding of dynamin to amphiphysin (Fig. 3C). The calcineurin inhibitor cyclosporin A, however, enhanced the shift and decreased the coprecipitation of dynamin with amphiphysin. An even greater effect on both parameters was observed if two other phosphatase inhibitors, okadaic acid and vanadate, were added to cyclosporin A, which suggests an involvement of other phosphatases in addition to calcineurin (Fig. 3C).

Thus, amphiphysin appears to play a key role as a regulated linker connecting AP-2/clathrin to dynamin 1 and synaptojanin 1. High-level expression of the SH3 domain of amphiphysin, which binds dynamin and synaptojanin 1, has dominant negative effects on clathrin-mediated endocytosis (9). A similar effect would be expected for overexpression of the amphiphysin region that contains the AP-2 and clathrin binding sites. Accordingly, transfection of an amphiphysin 1 construct comprising amino acids 250 to 588 (B and C domains) (31) in Chinese hamster ovary cells blocked receptor-mediated uptake of transferrin (Fig. 4, A and B). Furthermore, expression of this construct produced a change of the clathrin immunostaining from the typical punctate to a diffuse pattern (Fig. 4, C and D), consistent with a disruption of clathrin assembly.

Figure 4

The amphiphysin 1 region containing the AP-2 and clathrin binding sites has a dominant-negative effect on clathrin-mediated endocytosis. Double immunofluorescence of Chinese hamster ovary cells transfected with an amphiphysin 1 fragment (amino acids 250 to 588) comprising the clathrin and AP-2 binding site (31) [B + C region as defined in (6)]. Cells were transiently transfected with the B + C region of amphiphysin 1 and then incubated with CY3-labeled transferrin for 20 min. After fixation cells were processed for amphiphysin 1 and clathrin immunofluorescence. Magnification, ×400.

Phosphorylation and dephosphorylation reactions play an important role in regulation of the endocytic machinery. Ca2+-dependent dephosphorylation of endocytic proteins (4, 20, 21) after nerve terminal depolarization may prime the nerve terminal for efficient compensatory endocytosis after a burst of exocytosis. Ca2+-dependent dephosphorylation may underlie some of the reported positive effects of Ca2+ on synaptic vesicle endocytosis (32) and a dephosphorylation-dependent assembly of cytosolic endocytic coat proteins may explain the increased number of clathrin cages and clathrin-coated pits observed in ATP-depleted cells (33). In nonneuronal cells AP-2 assembly into clathrin coats correlates with its dephosphorylation (19). It is possible that a general property of proteins involved in endocytosis is to undergo constitutive phosphorylation and to assemble in the dephosphorylated state.

  • * To whom correspondence should be addressed. E-mail: pietro.decamilli{at}


View Abstract

Navigate This Article