The p11/S100A10 light chain of annexin A2 is dispensable for annexin A2 association to endosomes and functions in endosomal transport

Abstract

Background: Annexin A2 is a peripheral membrane protein that belongs to the annexin family of Ca(2+) and phospholipid-binding proteins. This protein, which plays a role in membrane organization and dynamics in particular along the endocytic pathway, exists as a heterotetrameric complex, consisting of two annexin A2 molecules bound via their N-termini to a dimer of p11/S100A10 light chains. The light chain, and thus presumably formation of the heterotetramer, was reported to control annexin A2 association to the plasma membrane and to cortical actin, as well as the distribution of recycling endosomes. However, the specific role of the light chain and the functions of monomeric versus heterotetrameric annexin A2 have remained elusive in the endocytic pathway.

Methodology/principal findings: Here, we have investigated whether p11 plays a role in the endosomal functions of annexin A2. Using morphological and biochemical approaches, we found that p11, unlike annexin A2, was not present on early endosomes. Neither was the heterotetramer detected on purified early endosomes, while it was clearly present in total cell lysates. Moreover, knockdown of p11 with siRNAs did not affect annexin A2 targeting to early endosomes, and, conversely, binding of annexin A2 to purified endosomes or liposomes occurred without p11 in vitro. Finally, while we could confirm that annexin A2 knockdown inhibits transport beyond early endosomes, p11 knockdown had no such effects on early-to-late endosome transport.

Conclusions/significance: Our data show that the binding of annexin A2 to endosomal membranes and its role in endosomal trafficking are independent of the p11/S100A10 light chain. We thus conclude that annexin A2 functions are fully supported by the monomeric form of the protein, at least the endocytic pathway leading to lysosomes.

Figure 1. Subcellular localization of p11 compared to F-actin and annexin A2.

The distribution in HeLa cells of p11 (H21 monoclonal antibody), AnxA2 (HH7 monoclonal antibody, and F-actin (phalloidin) was analyzed by immunofluorescence using a conventional saponin-based permeabilization protocol (A) or a protocol optimal for AnxA2 and p11 detection after cytosol wash-out (B–D), as follows: (A) p11 and F-actin double fluorescence after saponin-based permeabilization; (B) AnxA2 and F-actin double fluorescence after cytosol wash-out; (C) p11 and F-actin double fluorescence analysis after cytosol wash-out; (D) p11 and AnxA2-GFP double fluorescence analysis after cytosol wash-out. Arrowheads in B and C indicate concentrated AnxA2 and/or p11 bundles at cell surface, with strong colocalization with F-actin. Arrows in D indicate p11/AnxA2 colocalization. Bar: 10 µm.

Figure 2. p11 does not colocalise with endosomal markers.

The distribution of p11 was analyzed as in Fig 1 (washout protocol), using antibodies against EEA1 (A) and Lamp1 (B). Since AnxA2 is present on early endosomes, we used wide field, and not confocal, microscopy in (A) to ensure that structures that may contain both p11 and the early endosomal marker EEA1 were not missed in this analysis (hence some background nuclear staining). Cropped images are shown on both right panels. Bar: 10 µm.

Figure 3. Biochemical analysis of p11 distribution and annexin A2/p11 interaction detection.

(A), Total cell lysates and early endosomes (EE) were prepared from BHK cells, and then analyzed by SDS gels (equal proteins amounts loaded in each lane) and western blotting using the indicated antibodies. The p11 and AnxA2 signals were scanned and the right panel shows the ratio of the p11 over AnxA2 signals in the corresponding fractions (A.U: arbitrary units). (B) Experiments and quantification (right panel) were as in (A) except that total membranes and cytosol obtained after high speed centrifugation were analyzed. (C) Hela cells were transfected with (AnxA2^H28-GFP), and cell lysates were prepared. Then, AnxA2-GFP was immunoprecipitated from the lysates with anti-GFP antibodies (IP: immunoprecipitate; Ab- : control without the specific antibody; load: input fraction before immunoprecipitation). Samples were then analyzed by SDS gels and western blotting with the indicated antibodies. The H28 antibody only recognizes AnxA2^H28, while, in our hands, the HH7 antibody recognizes WT endogenous AnxA2. HC IgG, heavy chain of anti-GFP antibody used for immunoprecipitation. (D) The experiments were as in (C), except that purified early endosomes were used as starting materials. The left panel shows western blots with anti-GFP antibody, while the right panels show blots with anti-AnxA2 antibody. For comparison, the small right panel (EE load) indicates the mobility of (untagged) AnxA2 in a gel of the early endosome (EE) starting material (load). HC IgG, heavy chain of anti-GFP antibody used for immunoprecipitation, LC IgG, light chain of anti-GFP antibody.

Figure 4. Recombinant annexin A2 binding to endosome and liposome membranes.

(A) Early endosomes (left panel) or PA/PE/cholesterol liposomes (right panel) were prepared and incubated in vitro for 30min at 37°C with 5 µg purified recombinant AnxA2 (AnxA2^H28). Membrane-bound and free AnxA2 were then separately recovered after high speed centrifugation, and analyzed by SDS gels and western blotting, as indicated. The faster migrating form of AnxA2 in panel A corresponds to the core domain of the protein lacking the N-terminus, and results from AnxA2 proteolytic cleavage. Note that the core domain does not become efficiently membrane-associated, as expected , (B) The recombinant mutant AnxA2^H28 is selectively detected with the monoclonal antibody H28 (see 4A), which does not detect WT AnxA2 , while the HH7 monoclonal antibody recognizes WT AnxA2 , as shown with blots of HeLa and BHK lysates.

Figure 5. Endosomal targeting of annexin A2 is not affected when p11 is downregulated.

(A) p11 was downregulated in HeLa cells using two siRNA duplexes (RNAi1 and RNAi2), and lysates were prepared. The control (ctrl RNAi) siRNA used was designed against viral stomatitis virus (VSV-G) G-protein, which is not expressed in HeLa cells; equal amount of proteins were loaded in each lane and tubulin was used as an equal loading marker. (B) AnxA2 (diced AnxA2 RNAi) or p11 (RNAi2) was downregulated in HeLa cells, and total lysates were analyzed by western blotting as indicated (C) p11 was knocked down (as in A–B), and then a post-nuclear supernatant was prepared (left panel) and used as to further fractionate total endosomes (right panel). The samples were then analyzed by western blotting using the indicated antibodies. (D) AnxA2-GFP (AnxA2^H28-GFP) was expressed in mock-treated (left panel) or p11-siRNA2-treated HeLa cells (right panel) and then cells were processed for immunofluorescence as in Fig 1 (using the special permeabilizaiton protocol). Arrows show AnxA2-GFP/EEA1 colocalizations and cropped images are shown under both conditions. Bar: 10 µm.

Figure 6. p11 downregulation does not alter transport from early to late endosomes.

Rhodamin-dextran was endocytosed for 10 min at 37°C and then chased for 40 min in marker-free medium, after p11 knockdown as in Fig 5 (left panels, with the corresponding mock-treatment on the left) or AnxA2 knockdown as in Fig 5 (right panels, with the corresponding mock-treatment on the left). Cells were then processed for immunofluorescence using the indicated antibodies as in Fig 1. Bar: 10 µm.