A Novel Sequence in AP180 and CALM Promotes Efficient Clathrin Binding and Assembly

Abstract

The clathrin heavy chain N-terminal domain interacts with endocytic adapter proteins via clathrin binding motifs to assemble clathrin triskelia into cages. However, the precise mechanism of clathrin assembly is not yet known. Clathrin assembly protein AP180 has more clathrin binding motifs than any other endocytic protein and has a major role in the assembly of the clathrin coat during synaptic vesicle biogenesis. We now demonstrate that some of the previously identified binding motifs in AP180 may be non-functional and that a non-conventional clathrin binding sequence has a major influence on AP180 function. The related protein, clathrin assembly lymphoid myeloid leukemia protein (CALM), has fewer clathrin binding motifs and functions ubiquitously in clathrin-mediated endocytosis. The C-terminal ~16 kDa sub-domain in AP180, which has relatively high similarity with CALM, was shown in earlier work to have an unexplained role in clathrin binding. We identified the specific sequences in this sub-domain that bind to clathrin. Evidence for a role for these sequences in promoting clathrin binding was examined using in vitro and ex vivo experiments that compared the clathrin binding ability of site mutants with the wild type sequence. A sequence conserved in both AP180 and CALM (LDSSLA[S/N]LVGNLGI) was found to be the major interaction site and mutation caused a deficit in clathrin assembly, which is the first example of a mutation having this effect. In contrast, single or double mutation of DL(L/F) motifs in full length AP180 had no significant effect on clathrin binding, despite higher clathrin affinity for isolated peptides containing these motifs. We conclude that the novel clathrin interaction sites identified here in CALM and AP180 have a major role in how these proteins interface with clathrin. This work advances the case that AP180 and CALM are required to use a combination of standard clathrin N-terminal domain binding motifs and the sequence identified here for optimal binding and assembling clathrin.

Fig 1. Binding of clathrin to 15mer peptides from the AP180 CLAP domain and AP180 and CALM ADΔCLAP domains.

(A) Diagram indicating AP180 and CALM domains, the location of existing clathrin heavy chain (CHC) and adapter protein complex 2 (AP2) binding motifs [7] and the clathrin binding sites identified in this work. Note CALM isoform 3 was used to obtain results in this figure and Fig 2. CALM isoform 1 was used in subsequent figures. (B) The relative amount of purified clathrin bound to AP180 15mers containing known CBMs in an overlay assay. Numbers correspond to numbered binding motifs in (A). The dark blue columns and purple columns correspond to the overlay experiment with the AP180 and CALM ADΔCLAP 15mers, respectively. The intensity for each peptide was divided by the intensity of clathrin binding to peptide 6. (C) The amount of clathrin binding to overlapping 15mer peptides from the ADΔCLAP of AP180 isoform 2 and CALM isoform 3 (see Figure A in S1 File for full list of aligned peptides), relative to peptide 6 in (B). Same colour scheme as (B). Sequences and spots for Site 1 are shown. Bold residues in sequence alignment indicate identity between AP180 and CALM. The respective C-termini from AP180 and CALM are both referred to as Site 2. (D) Clathrin binding for AP180 and CALM Site 2 15mers compared to 15mer peptides from the C-termini of other clathrin adaptors (normalized to peptide 6 in (B), n = 1).

Fig 2. Substitution of amino acid residues in peptides from Site 1 and Site 2 to identify those important for clathrin binding.

Hydrophobic amino acids (Leu, Ile, Met, and Phe), Asp residues and conserved residues in the extreme C-terminus were substituted with Ala (red residues) in peptide array overlay assays with purified clathrin. Bold residues indicate conserved amino acids. (A) CALM Site 1 substitutions and their effect on clathrin binding to 15mers (n = 1). The intensity for each 15mer has been divided by the intensity of clathrin binding to 15mer E (B) AP180 Site 2 substitutions and their effect on clathrin binding to 15mers (n = 1). The intensity for each 15mer has been divided by the intensity of clathrin binding to 15mer I (C) CALM Site 2 substitutions and their effect on clathrin binding to 15mers (n = 1). The intensity for each 15mer was divided by the intensity of clathrin binding to 15mer M.

Fig 3. Extraction of clathrin from synaptosome lysate by GST-tagged WT, Site 1 and 2 mutated AP180 and CALM.

(A) AP180 and (B) CALM mutated at Site 1, Site 2 and Site 1&2 (double mutant) were used in 1 h pull-downs with rat synaptosome lysate and the amount of clathrin bound was determined by Western blot with anti-clathrin heavy chain (CHC) using 40% of the sample (representative blot shown from two 1 h pull-down experiments). Ten percent of the sample was resolved by SDS-PAGE and stained with Coomassie for comparison of bait levels. The experiment was performed twice more, except the pull-down was for 10 min instead of 1 h and the result was very similar (Figure B in S1 File). (C) and (D) Comparison of clathrin binding from the combined densitometry of 10 min and 1 h pull-downs with the AP180 mutants normalised to AP180 WT and the CALM mutants to CALM WT. Data is expressed as average relative amount of clathrin bound as a fraction of the WT pull-down ± SEM (n = 4; ***, P < 0.001 compared to WT; ns = not significant).

Fig 4. Comparison of clathrin assembly by WT, Site 1 and Site 2 mutated AP180.

(A) A Coomassie stained SDS-PAGE gel of WT AP180, Site 1 (S1), Site 2 (S2) and Site 1&2 (S1&2) mutants, which were bacterially expressed and the GST-tag removed. (B) Light scattering was used to determine the clathrin assembly rate for fixed concentrations of purified AP180 and clathrin. Data shown is the mean ± SEM (n = 5 for experiment which used clathrin, n = 2 for the remaining controls). Only every 5^th error bar is shown. The S2 + clathrin error bars are staggered by 10 s to improve visual clarity. Data was normalised to baseline assembly (the sum of scattered light intensity from clathrin alone plus WT AP180 alone = 1). Curves were fitted to calculate initial rates of clathrin assembly (lines). (C) Initial clathrin assembly rates determined from curve fitting of assembly data in (B). Errors for these parameters are SEM (n = 5).

Fig 5. Comparison of WT, Site 1 and Site 2 mutated AP180 and CALM in a transferrin uptake assay.

COS-7 cells were transfected with GFP-tagged WT, Site 1, Site 2 and Site 1&2 mutants for (A) AP180 and (B) CALM. Alexa Fluor 594 labelled Tfn (red) was added and the amount of uptake was measured as an assay of endocytosis function and clathrin sequestration. Note that in (A) and (B), respectively, GFP-tagged WT AP180 and CALM exhibit a dominant negative effect due to sequestration of clathrin. Data is expressed as a fraction of Tfn uptake in untransfected control cells ± SEM (n = 4; ***, P < 0.001 compared to WT; ns = not significant). The approximately equal expression level of AP180 and CALM fusions with GFP is shown in panels (C) and (D), respectively, which was produced from COS-7 cell lysis, SDS-PAGE and Western blotting with anti-GFP or anti-actin (control for protein content).

Fig 6. Extraction of clathrin from synaptosome lysate by GST-tagged AP180 WT, Site 1 and CBM mutants.

(A) AP180 WT and the following mutant sequences, Site 1, 666-DLL-668 to AAA, 638-DLF-640 to AAA, both 666-DLL-668 and 638-DLF-640 to AAA (AAA+AAA) and 756-SLV-758 to AAA, were used in 1 h pull-downs with rat synaptosome lysate and the amount of clathrin bound was determined by Western blot with anti-clathrin heavy chain (CHC) using 25% of the sample. Five percent of the sample was resolved by SDS-PAGE and stained with Coomassie for comparison of bait levels. Representative Western blot and SDS-PAGE gel is shown from three 1 h pull-down experiments. (B) Comparison of clathrin binding from the densitometry of each 1 h pull-down with the AP180 mutants normalised to AP180 WT. Data is expressed as average relative amount of clathrin bound as a fraction of the WT pull-down ± SEM (n = 3; **, P < 0.01 compared to WT; ns = not significant).