RAL-1 controls multivesicular body biogenesis and exosome secretion

Abstract

Exosomes are secreted vesicles arising from the fusion of multivesicular bodies (MVBs) with the plasma membrane. Despite their importance in various processes, the molecular mechanisms controlling their formation and release remain unclear. Using nematodes and mammary tumor cells, we show that Ral GTPases are involved in exosome biogenesis. In Caenorhabditis elegans, RAL-1 localizes at the surface of secretory MVBs. A quantitative electron microscopy analysis of RAL-1-deficient animals revealed that RAL-1 is involved in both MVB formation and their fusion with the plasma membrane. These functions do not involve the exocyst complex, a common Ral guanosine triphosphatase (GTPase) effector. Furthermore, we show that the target membrane SNARE protein SYX-5 colocalizes with a constitutively active form of RAL-1 at the plasma membrane, and MVBs accumulate under the plasma membrane when SYX-5 is absent. In mammals, RalA and RalB are both required for the secretion of exosome-like vesicles in cultured cells. Therefore, Ral GTPases represent new regulators of MVB formation and exosome release.

Figure 1.

RAL-1 GTPase or exocyst deficiency induces alae defects. (A) C. elegans epidermal cells contain MVBs, which can fuse with the apical plasma membrane and liberate exosomes. These exosomes are integrated in the cuticle and contribute to the formation of the alae. (B) An RNAi-based screen identified 73 genes required for alae formation. (C) Disruption of ral-1 by RNAi, or by the null allele ral-1(tm5205), leads to alae defects. The number of animals is shown at the top of the graph. **, P < 0.05; ***, P < 0.01. (D) Schematic representation of the exocyst complex involved in plasma membrane attachment of secretory vesicles. Subunits found in the screen are in black. (E) Alae defects observed after disruption of several members of the exocyst complex in mutants or by RNAi.

Figure 2.

RAL-1 localizes at the surface of MVBs. (A–D) WT (RAL-1(WT); A) and dominant negative (RAL-1(DN); C) versions of RAL-1, but not the constitutively active version (RAL-1(CA); B), colocalize fully with VHA-5 in the epidermis at the time of alae formation, as shown in the quantification (D). In D, the numbers inside the bars indicate the number of puncta (number of animals). (E–H) APEX::RAL-1(DN) shows DAB staining both at apical membrane stacks (E and F, arrowheads) and at the external surface of MVBs (E and G–G’’, arrows). Animals expressing no APEX tag treated with DAB show no staining (H). The star indicates a MVB. APM, apical plasma membrane.

Figure 3.

RAL-1 affects different steps of exosome secretion. (A) Quantitative electron microscopy analysis of MVBs in epidermal cells: (1 and 2) density (1) and diameter (2) of MVBs, (3) number of ILVs per MVB, (4) ILV diameter, and (5) distance between MVB and the APM. (B) MVB density is decreased in ral-1(tm5205) compared with the WT and is increased in ral-1(RNAi) compared with control(RNAi). (C) The density of VHA-5::RFP puncta is decreased in ral-1(tm5205) compared with the WT but is unaffected in ral-1(RNAi). (D–F) In ral-1(tm5205) mutants, MVBs have an abnormal size (E) and ILV content (F). (G) In ral-1(RNAi) animals, 57% of MVBs are within 50 nm of the apical plasma membrane, compared with 20% in control animals. (H and I) Two MVBs in proximity of the apical plasma membrane from control (H) and ral-1(RNAi) (I) animals. MVBs from ral-1(RNAi) animals can form a hemifusion diaphragm (I) with the apical plasma membrane. (J and J’) EVs purified from 4T1 mammalian cells and observed by electron microscopy. (K) Depletion of either RalA or RalB by shRNA leads to a decrease in the number of EVs observed by electron microscopy compared with control shRNA (P < 0.0001 between sh control and either sh RalA or sh RalB; pool of four independent purifications, Mann-Whitney test). (L and M) Western blot of cell lysates and secreted EVs. One representative experiment (L) and pooled quantification (M) of four independent purifications (P < 0.03 between sh control and either sh RalA or sh RalB for each marker, Mann-Whitney test). Numbers in or above the bars indicate the number of animals (C), MVBs (E, F, and M) or fields (K) analyzed. APM, apical plasma membrane; CL, cell lysate; Cu, cuticle; Cy, cytoplasm; MVBm, MVB outer membrane. Errors bars, SEM.

Figure 4.

The exocyst regulates alae formation independently of RAL-1. (A) Quantitative electron microscopy analysis reveals that the MVB density is similar in sec-5(pk2358) and sec-8(ok2187) null mutants and WT controls. MVB density is increased after sec-8(RNAi) but not sec-15(RNAi) compared with control(RNAi) (P values, Mann-Whitney test). (B) Example of an MVB attached to the apical plasma membrane in sec-5(pk2358) (enlarged in B’). (C) Small vesicles (<100 nm diameter; black arrows) accumulating under the plasma membrane in sec-8(ok2187) mutants. (D and E) The exocyst subunits SEC-8 (D) and SEC-15 (E) do not colocalize with VHA-5 in the epidermis. The scale bars in D apply to E. APM, apical plasma membrane; Cu, cuticle; Cy, cytoplasm; MVBm, MVB outer membrane.

Figure 5.

SYX-5 controls MVB fusion with the apical plasma membrane. (A) In the epidermis, SYX-5 localizes in both large cytoplasmic puncta and smaller apical puncta. (B) SYX-5 displays some colocalization with VHA-5 in the epidermis at the time of alae formation. (C–E) SYX-5 colocalizes with RAL-1(CA) (C), but not with RAL-1(DN) (D), as revealed by quantification (E). The scale bars in B apply to C and D. The numbers inside the bars indicate the number of puncta (number of animals). (F) Generation of two mutant alleles for syx-5 using CRISPR/Cas9 technology. (G) Homozygote mutants for syx-5(mc51) show alae defects at the L1 stage. (H) Electron micrographs of two MVBs showing attachment to the apical plasma membrane in syx-5(mc51) mutant larva. (I) Model for the role of RAL-1 and SYX-5 in exosome biogenesis (see text). Errors bars, SEM.