Argosomes: Intracellular Transport Vehicles for Intercellular Signals?

See allHide authors and affiliations

Science's STKE  19 Mar 2002:
Vol. 2002, Issue 124, pp. pe13
DOI: 10.1126/stke.2002.124.pe13


Cell clusters in the immature tissues of developing organisms create morphogen gradients that guide cellular differentiation into specific cell fates. Although the process of simple diffusion in gradient establishment has been well studied, there are other mechanisms by which cells establish morphogen gradients. Christian discusses the recent findings that morphogens may establish gradients through the use of plasma membrane-containing exovesicles, termed "argosomes."

Morphogens are molecules that are synthesized by a small group of cells and establish a gradient of signaling activity as they move away from their source. Hedgehog (Hh), Wingless (Wg), and Decapentaplegic (Dpp) all function as morphogens in Drosophila wing and leg imaginal discs, but the mechanisms by which these secreted proteins travel to distal cells remains controversial. The best-studied, and perhaps simplest, mechanism involves the formation of long-range extracellular gradients by passive diffusion (Fig. 1A). Until recently, direct evidence for this model was lacking because of technical problems in detecting proteins in the extracellular space. In addition, the observations that Wg and Dpp are firmly associated with heparan sulfate proteoglycans (HSPGs), whereas Hh is attached to the cell surface by the addition of cholesterol, seemed to preclude simple diffusion. Both of these issues were recently resolved with the direct visualization of extracellular gradients of Wg and Dpp proteins in unfixed tissues (1-3), together with genetic evidence showing that cholesterol modification (4) and the presence of HSPGs (5, 6) promote, rather than impede, diffusion of Hh. The latter finding raises the possibility that extracellular gradients form by facilitated, rather than passive, diffusion.

Fig. 1.

Models for distributing ligands during formation of a morphogen gradient. Morphogens are secreted from a source (the leftmost cell in each panel) and might establish an extracellular gradient by passive diffusion away from the source (A) or establish an intracellular gradient by sequential endocytosis, active intracellular transport, and exocytosis (B). In either case, the morphogen activates a signal transduction cascade that is strongest in cells nearest the source and progressively weakens in more distal cells. This is translated into a gradient of expression of downstream target genes.

Morphogen gradients are also proposed to form through a type of "bucket brigade," in which ligands are shuttled from cell to cell by consecutive rounds of endocytosis, intracellular transport, and exocytosis (7) (Fig. 1B). Precedent for this mechanism exists in the immune system, where vesicle shedding is used as a means to transmit signals. In immune cells, proteins are selectively recruited from the cytoplasm into multivesicular bodies or are incorporated directly into microvesicles that are then externalized and taken up by target cells (8, 9). The initial evidence that Wg, Hh, and Dpp might use a similar intercellular route for transportation consisted of the observation that each of these proteins can be detected in vesicles inside of cells located at some distance from their site of synthesis (1-3, 10). Biochemical analyses have confirmed their presence both inside and outside of cells (2, 3). Despite this circumstantial evidence, the idea that ligand-containing vesicles are secreted and trafficked to other cells, as opposed to targeted for degradation after receptor-mediated endocytosis, remains controversial. The best evidence that endocytosis and intracellular trafficking are required for the spread of morphogens through target tissues might be that Dpp (in the wing disc) and Wg (in the embryo) are unable to signal through groups of cells in which endocytosis is blocked (1, 11). This mechanism may be tissue-specific, however, because in imaginal discs Wg can signal beyond clones of endocytosis-defective cells (2).

A recent paper by Greco et al. (12) provides new support for active intracellular transport as a means of moving morphogens. These authors identified a population of membrane exovesicles, termed "argosomes," which bud off from the basolateral membranes of Drosophila imaginal disc cells and are transported throughout the disc epithelium. Unlike the multivesicular bodies that form in immune cells, argosomes do not originate internally and appear to consist of only a subset of membrane domains, raising the possibility that proteins are sorted and selectively targeted for incorporation into argosomes. Intriguingly, time-lapse analysis shows that these vesicles traverse cells at a rate consistent with movement on cellular motors, and photobleaching experiments reveal that the rate of spread of argosomes is compatible with the speed at which morphogen gradients form. Finally, the authors show that a fraction of Wg-containing vesicles colocalize with argosomes, suggesting that these may be a vehicle for the movement of Wg protein. One potential inconsistency with this idea is that although argosomes are derived from the basolateral surface of the disc epithelium, where the extracellular Wg gradient forms, Wg- (and Dpp-) containing endocytic vesicles are enriched in the apical region of cells. Apical localization of Wg is essential for signaling, at least in the embryonic ectoderm, because apically but not basolaterally restricted Wg protein can rescue Wg-dependent segmental patterning in vivo (13). One way to reconcile these differences would be to hypothesize that the extracellular gradient of Wg is converted into an intracellular apical gradient, with argosomes moving the endocytosed ligand from cell to cell during an intermediate step.

One appealing consequence of argosome-mediated transport is that it constrains ligand movement to the plane of the epithelial sheet. As pointed out by Teleman and Cohen (3), the epithelium of the leg and the wing discs is highly folded and, as a result, the distance that a freely diffusible molecule needs to travel to reach proximal cells can be greater than that required to reach more distally located cells. Premature reception of the signal by distal cells would create discontinuities in the concentration gradient. By contrast, intracellular transport requires intimate cell-cell contact and would ensure that the gradient is transmitted in a linear fashion from proximal to distal cells.

Vesicle-mediated transport of ligands could provide a mechanism for regulating signal duration and the distance over which signals are transmitted. Preferential trafficking of argosomes toward either recycling or lysosomal compartments in different tissues could control the half-life of ligands, as well as the range over which they might signal. Consistent with this idea, the distance over which Dpp and Wg can signal depends on the rates of endocytic trafficking and degradation (1, 14). The activity and signaling range of morphogens could also be influenced by routing instructions that determine whether they are preferentially transported by an intracellular or extracellular route. This possibility is supported by evidence that proteins are selectively recruited into argosomes only if they are associated with specific subdomains of the plasma membrane (12). Specifically, whereas glycophosphatidylinositol (GPI)-anchored green fluorescent protein (GFP) or prenylated Rho-GFP is recruited to argosomes, myristoylated GFP is not. Hh, Wg, and Dpp all bind to and require HSPGs, including specific GPI-linked proteoglycans (6, 15-17), for full-signaling activity. Tissue-specific expression of HSPGs could target ligands for argosome-mediated transport in some, but not all, cells. The idea that endogenous HSPGs are involved in Wg transport is supported by the finding that heparinase treatment of imaginal discs prevents accumulation of Wg in producing cells and reduces the level of Wg found in vesicles within receiving cells (12). As previously mentioned, Hh can undergo posttranslational addition of cholesterol and palmitoyl groups, and these modifications could potentially influence whether Hh is a candidate for argosome-mediated transport in a given cell type. To further complicate the picture, genetic studies have identified several proteins, including the putative acyltransferase Porcupine (18, 19) and the multipass transmembrane protein Dispatched (20), that are required for secretion or release of Wg and Hh, respectively. The relationship, if any, between these proteins and argosomes is unknown.

The identification of argosomes as putative transporters sets the stage for further experiments to determine how these vesicles are moving, whether their intracellular trafficking is regulated, what their cargo is, and whether they are required for signal propagation. One prediction of the model of Greco et al. (12) is that Wg protein tethered to the membrane by direct addition of a GPI linkage should show the same long-range signaling activity as native Wg. This is in contrast to a form of Wg that is tethered to the membrane by the addition of a transmembrane domain and can signal only to adjacent cells in the disc (21). The idea that argosomes may be transported on cellular motors is interesting, given that kinesins are linked to membranous cargo by scaffolding proteins that can assemble cytoplasmic components of entire signal transduction cascades (22). This raises the possibilities that kinesins play a role in localizing signal transduction cascades or that signals are transmitted inside of receiving cells during transit of the ligand-receptor complex through the cell. Endocytosis of ligand-activated receptors has historically been considered a process that diminishes signaling, but growing evidence suggests that endocytic trafficking regulates the intensity of signaling, as well as colocalization of activated receptors and their downstream signaling components (23). The availability of molecular tools to selectively inhibit different trafficking events inside of cells should make it possible to dissect out the specific proteins that are required for argosome movement and to test whether this movement is required for the formation of morphogen gradients.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
View Abstract

Stay Connected to Science Signaling

Navigate This Article