If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
To whom correspondence should be addressed: University Medical Center Utrecht, Dept. of Cell Biology, AZU G02.525, Heidelberglaan 100, 3584 CX Utrecht, The Netherlands. Tel.: 31-30-2507577 or 31-30-2506551; Fax: 31-30-2541797;
* This work was supported by a grant from the Netherlands Organization for Scientific Research.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Recycling of endocytosed membrane proteins involves passage through early endosomes and recycling endosomes. Previously, we demonstrated a role for clathrin-coated vesicles in transferrin receptor recycling. These clathrin-coated vesicles are formed from recycling endosomes in a process that was inhibited in dynamin-1G273D-overexpressing cells. Here we show a second transferrin recycling pathway, which requires phosphatidylinositol 3-kinase activity. Two unrelated phosphatidylinositol 3-kinase inhibitors, LY294002 and wortmannin, retained endocytosed transferrin in early endosomes but did not affect transfer through recycling endosomes. The inhibitory effects of LY294002 and dynamin-1G273D on transferrin recycling were additive. In combination with brefeldin A, a drug that prevents the formation of clathrin-coated buds at recycling endosomes, LY294002 inhibited transferrin recycling synergistically. Collectively, these data indicate two distinct recycling pathways. One pathway involves transfer from early endosomes to recycling endosomes, from where clathrin/dynamin-coated vesicles provide for further transport, whereas the other route bypasses recycling endosomes and requires phosphatidylinositol 3-kinase activity.
Membrane proteins that are endocytosed via clathrin-coated vesicles are first targeted to early endosomes (EE).
). Transferrin (Tf) and the transferrin receptor (TfR) have extensively been used as prototype markers for the recycling pathway(s). Tf remains associated with its receptor after endocytosis and dissociates only after recycling of the complex to the plasma membrane. Endocytosed Tf·TfR is diverged from lysosome-targeted proteins in EE with a t½ of <2–3 min (
and references therein). The recycling pathway involves recycling endosomes (RE), a network of tubular membranes, which in some cell types, but not all, is concentrated in the perinuclear area as the so-called recycling compartment (
). For nonpolarized cells, we recently demonstrated that a significant fraction of endocytosed Tf is recycled to the plasma membrane by clathrin-coated vesicles that bud from RE in a dynamin-dependent manner (
). Cells overexpressing dynamin-1G273D (dynts), a temperature-sensitive dynamin mutant, accumulated dynts-carrying clathrin-coated buds on RE and retained ∼30% of endocytosed Tf in this compartment at the nonpermissive temperature. Whether TfR is actively or passively incorporated into RE-derived clathrin-coated vesicles remains, however, unclear (
The distinction between EE and RE is mainly based on their morphology, the flow of cargo molecules, and their spatial distribution within cells. The organizational complexity of the endosomal system is illustrated further by the distribution of Rab GTPases that regulate endocytosis and recycling. Rab5 regulates transfer from the plasma membrane to EE, whereas Rab4 (
), indicating that EE and RE cannot simply be viewed as preexisting distinct compartments. This idea is also supported by the recent observation that Rab11-containing RE seem to be generated directly from peripheral EE (
). Collectively, these data indicate that PI 3-kinase is rate-limiting in a constitutive recycling pathway. In the current study, we provide evidence for two co-existing TfR recycling pathways. One of these involves EE and RE and depends on clathrin-coated vesicles and dynamin function, whereas the other pathway bypasses RE and requires PI 3-kinase activity.
In our previous study, we demonstrated that dynts-overexpressing cells failed to recycle ∼30% of endocytosed Tf at the nonpermissive temperature due to interference with the formation of clathrin-coated vesicles from RE (
). Experimental limitations did not allow us at that time to determine the fraction of TfR that recycled via this pathway. A simple interpretation of these data was that TfR might recycle via a single pathway and that interference by dynts was inefficient. Based on the current data, however, we now conclude that there are two recycling pathways. To our knowledge, this is the first study in which two independent TfR recycling pathways have been characterized by interference (Fig. 8). A considerable fraction of endocytosed TfR is, after entry into EE, regurgitated via RE in a process that involves the budding of dynamin-dependent clathrin-coated vesicles from this compartment. Another fraction is recycled directly from EE in a PI 3-kinase dependent manner. Several arguments are in favor of two recycling pathways rather than a single route. First, PI 3-kinase and dynamin function at distinct sites. dynts cells accumulated endocytosed Tf in RE, whereas LY-treated cells retained Tf in EE (
) (Figs. Figure 1, Figure 2, Figure 3 and 6). LY did not interfere with Tf transport from EE to RE (Figs. Figure 1, Figure 2, Figure 3 and 6,C and D) or with transport from RE to the plasma membrane (Figs. 3 and 6F). Conversely, dynts did not affect the clearance of Tf from EE (Fig. 2, H andI). Second, in dynts cells, RE were increased in length and accumulated clathrin-coated buds, whereas EE were unaffected (
). On the other hand, LY did not touch the phenotype of RE, whereas the diameter of EE was significantly increased (Figs. 4 and 5). Third, the inhibitory effect of LY on Tf recycling by dynwt cells was reversed when the drug was removed directly after loading or after 10 or 20 min of chase (Fig. 6, C and E), indicating that Tf was not misdirected to compartments other than EE as a consequence of the LY treatment. LY also did not increase the amount of Tf trapped at the dynts block, arguing against the possibility that LY might have delayed arrival of Tf at RE, thereby buying time for the dynts to become more effective once the cells were shifted to 38 °C (Fig. 6, D and E). Finally, despite a complete inhibition of clathrin-coated bud formation on RE (
), BFA had only a moderate effect on the rate and almost no effect on the extent of Tf recycling. The most plausible explanation is that when the recycling pathway from RE is blocked by BFA, nearly all Tf is shunted into the PI 3-kinase-dependent recycling pathway. In this scenario, BFA and LY should have strong synergistic inhibitory effects on recycling, and this was indeed observed (Fig. 7).
An involvement of RE in Tf recycling has clearly been demonstrated (
), but kinetic evidence for a second pathway that bypasses RE has been generated too. In nonpolarized cells, a relative fast recycling pathway has been marked most clearly with fluorescent lipid analogs (
) recently fitted Tf recycling data from CHO cells in a mathematical model that describes two recycling pathways. In this model, ∼25% of endocytosed Tf recycled via RE, and the other ∼75% recycled directly from EE with a relatively high rate. The kinetics of Tf transport through polarized Madin Darby canine kidney cells were also fitted into a model with two distinct recycling pathways (
). Our data are consistent with these models except that in HeLa cells the kinetics of the two pathways were rather comparable. dyntscells retained 30% of endocytosed Tf, but the recycling rate of the other 70% was unchanged (
). This can be accounted for when the rates of Tf recycling through dynamin-dependent and LY-sensitive pathways are similar. This was indeed observed, except that Tf recycling by dynwt cells was delayed by ∼5 min in the presence of LY and thus did not occur with first order kinetics. The relative rapid clearance of Tf from RE in HeLa cells is also consistent with a much less pronounced perinuclear location of RE in HeLa cells compared with CHO and Madin Darby canine kidney cells. Tubular RE are usually not connected to EE and, in contrast to EE, predominantly concentrated in the perinuclear area. The degree at which RE accumulate in the perinuclear area is, however, highly variable between cell types, and the relation between EE and RE is still poorly understood. One view is that EE and RE are distinct compartments that are connected by vesicular transport intermediates (
). Alternatively, RE may be formed as tubular extensions of vacuolar EE, from which they detach as a tubule and then migrate to the perinuclear area. Consistent with this idea, it has recently been demonstrated that RE can be formed de novo in the cell periphery (
). Using whole mount electron microscopy, we detected TfR-containing endosomal tubules that were connected to EEA1-labeled vacuoles, but most peripheral tubular endosomes were not continuous with vacuolar EE.2 This is consistent with a model in which RE are formed directly from EE by the detachment of tubular extensions without the requirement of vesicular intermediates. The appearance of tubular endosomes in the cell periphery was not changed by LY (Figs.1L and 4B), in accordance with the observation that transfer of Tf to RE was not affected (Figs. Figure 1, Figure 2, Figure 3 and6D).
The effects of WM and LY were indistinguishable in our study. WM and LY are unrelated compounds that mechanistically interfere distinctly with PI 3-kinase kinase activity. The advantage of LY compared with WM is, however, that its inhibitory effect is more specific for PI 3-kinases and reversed upon removal of the drug (
). Further evidence for the requirement of PI 3-kinase activity in TfR recycling was obtained in a study in which microinjected inhibitory antibodies directed against the p110α subunit of class I PI 3-kinase decreased Tf recycling, whereas antibodies against hVPS34, a class III PI 3-kinase, had minor inhibitory effects (
) and associates with EE through the interaction of its FYVE domain with phosphatidylinositol triphosphate. Although in some studies PI 3-kinase inhibitors disrupted the association of EEA1 with endosomes (
). The formation of vesicles within multivesicular bodies is dependent on hVPS34 activity, and interference with the formation of multivesicular body internal vesicles by PI 3-kinase inhibitors is one of the causes of EE swelling (
). Retention of TfR in EE and swelling of this compartment in response to WM or LY, as demonstrated in this study, are consistent with these observations.
The only TfR trafficking step that was severely affected by LY was direct transport from EE to the plasma membrane. Despite accumulating evidence for a direct pathway from EE to the plasma membrane, the nature of this pathway remains unresolved. Although transport vesicles that shuttle directly from EE to the plasma membrane could be involved, such vesicles have not been identified. Alternatively, recycling may occur through direct fusion of EE with the plasma membrane. In a similar way, multivesicular bodies have been demonstrated to fuse directly with the plasma membrane, resulting in the release of their internal vesicles, called exosomes, into the extracellular milieu (
). At this point, we can only speculate on the nature of the PI 3-kinase-dependent protein(s) that are required for direct TfR recycling. Potential players that require PI 3-kinase products for their recruitment to endosomes include FYVE, phox homology, or pleckstrin homology domain-containing proteins (
). Alternatively, inositol 3-phosphates could serve as precursors for other inositides that may be essential for this pathway. Molecular mechanisms that drive endocytic recycling are still relative poorly understood, possibly due to the complexity and plasticity of the system in which parallel pathways play an important role. The demonstration that two TfR recycling pathways can be manipulated independently provides new possibilities to investigate these routes further at the molecular level.
We thank René Scriwanek for photographic assistance and Dr. S. L. Schmid for providing the dynamin-1-expressing cell lines. We are grateful to Richard Wubbolts for critical reading of the manuscript.