Benjamin J. Nichols

Flotillin-1 defines a clathrin-independent endocytic pathway in mammalian cells.

Previous studies provide evidence for an endocytic mechanism in mammalian cells that is distinct from both clathrin-coated pits and caveolae, and is not inhibited by overexpression of GTPase-null dynamin mutants. This mechanism, however, has been defined largely in these negative terms. We applied a ferro-fluid-based purification of endosomes to identify endosomal proteins. One of the proteins identified in this way was flotillin-1 (also called reggie-2). Here, we show that flotillin-1 resides in punctate structures within the plasma membrane and in a specific population of endocytic intermediates. These intermediates accumulate both glycosylphosphatidylinositol (GPI)-linked proteins and cholera toxin B subunit. Endocytosis in flotillin-1-containing intermediates is clathrin-independent. Total internal reflection microscopy and immuno-electron microscopy revealed that flotillin-1-containing regions of the plasma membrane seem to bud into the cell, and are distinct from clathrin-coated pits and caveolin-1-positive caveolae. Flotillin-1 small interfering RNA (siRNA) inhibited both clathrin-independent uptake of cholera toxin and endocytosis of a GPI-linked protein. We propose that flotillin-1 is one determinant of a clathrin-independent endocytic pathway in mammalian cells.

Endocytosis without clathrin coats

Endocytosis is involved in an enormous variety of cellular processes. To date, most studies on endocytosis in mammalian cells have focused on pathways that start with uptake through clathrin-coated pits. Recently, new techniques and reagents have allowed a wider range of endocytic pathways to begin to be characterized. Various non-clathrin endocytic mechanisms have been identified, including uptake through caveolae, macropinosomes and via a separate constitutive pathway. Many markers for clathrin-independent endocytosis are found in detergent-resistant membrane fractions, or lipid rafts. We will discuss these emerging new findings and their implications for the nature of lipid rafts themselves, as well as for the potential roles of non-clathrin endocytic pathways in remodeling of the plasma membrane and in regulating the membrane composition of specific intracellular organelles.

Dynamics of putative raft-associated proteins at the cell surface.

Lipid rafts are conceptualized as membrane microdomains enriched in cholesterol and glycosphingolipid that serve as platforms for protein segregation and signaling. The properties of these domains in vivo are unclear. Here, we use fluorescence recovery after photobleaching to test if raft association affects a proteins ability to laterally diffuse large distances across the cell surface. The diffusion coefficients (D) of several types of putative raft and nonraft proteins were systematically measured under steady-state conditions and in response to raft perturbations. Raft proteins diffused freely over large distances (>4 μm), exhibiting Ds that varied 10-fold. This finding indicates that raft proteins do not undergo long-range diffusion as part of discrete, stable raft domains. Perturbations reported to affect lipid rafts in model membrane systems or by biochemical fractionation (cholesterol depletion, decreased temperature, and cholesterol loading) had similar effects on the diffusional mobility of raft and nonraft proteins. Thus, raft association is not the dominant factor in determining long-range protein mobility at the cell surface.

Two syntaxin homologues in the TGN/endosomal system of yeast

Intracellular membrane traffic is thought to be regulated in part by SNAREs, integral membrane proteins on transport vesicles (v‐SNAREs) and target organelles (t‐SNAREs) that bind to each other and mediate bilayer fusion. All known SNARE‐mediated fusion events involve a member of the syntaxin family of t‐SNAREs. Sequence comparisons identify eight such proteins encoded in the yeast genome, of which six have been characterized. We describe here the remaining two, Tlg1p and Tlg2p. These have the expected biochemical properties of t‐SNAREs, and are located in separable compartments which correspond to a putative early endosome and the yeast equivalent of the TGN, respectively. They co‐precipitate with the v‐SNARE Vti1p, which is implicated in Golgi–endosome traffic and, remarkably, binds to five different syntaxins. Tlg1p also binds the plasma membrane v‐SNARE Snc1p. Both Tlg1p and Tlg2p are required for efficient endocytosis and to maintain normal levels of TGN proteins. However, neither is required for intra‐Golgi traffic. Since no further syntaxins have been identified in yeast, this implies that the Golgi apparatus can function with a single syntaxin, Sed5p.

Molecular mechanisms of clathrin-independent endocytosis

There is good evidence that, in addition to the canonical clathrin-associated endocytic machinery, mammalian cells possess multiple sets of proteins that are capable of mediating the formation of endocytic vesicles. The identity, mechanistic properties and function of these clathrin-independent endocytic pathways are currently under investigation. This Commentary briefly recounts how the field of clathrin-independent endocytosis has developed to date. It then highlights recent progress in identifying key proteins that might define alternative types of endocytosis. These proteins include CtBP (also known as BARS), flotillins (also known as reggies) and GRAF1. We argue that a combination of information about pathway-specific proteins and the ultrastructure of endocytic invaginations provides a means of beginning to classify endocytic pathways.

Lipid raft proteins have a random distribution during localized activation of the T-cell receptor

The extent to which lipid raft proteins are organized in functional clusters within the plasma membrane is central to the debate on structure and function of rafts. Glycosylphosphatidylinositol (GPI)-linked proteins are characteristic components of biochemically defined rafts. Several studies report a function for rafts in T-cell stimulation, but it is unclear whether molecules involved in T-cell receptor (TCR) signalling are recruited to (or excluded from) T-cell synapses through asymmetric distribution of raft microdomains or through specific protein–protein interactions. Here we used FRET analysis in live cells to determine whether GPI-linked proteins are clustered in the plasma membrane of unstimulated cells, and at regions where TCR signalling has been activated using antibody-coated beads. Multiple criteria suggested that FRET between different GPI-linked fluorescent proteins in COS-7 or unstimulated Jurkat T-cells is generated by a random, un-clustered distribution. Stimulation of TCR signalling in Jurkat cells resulted in localized increases in fluorescence of GPI-linked fluorescent proteins and cholera toxin B-subunit (CTB). However, measurements of FRET and ratio imaging showed that there was no detectable clustering and no overall enrichment of GPI-linked proteins or CTB in these regions.

Coassembly of Flotillins Induces Formation of Membrane Microdomains, Membrane Curvature, and Vesicle Budding

Endocytosis has a crucial role in many cellular processes. The best-characterized mechanism for endocytosis involves clathrin-coated pits [1], but evidence has accumulated for additional endocytic pathways in mammalian cells [2]. One such pathway involves caveolae, plasma-membrane invaginations defined by caveolin proteins. Plasma-membrane microdomains referred to as lipid rafts have also been associated with clathrin-independent endocytosis by biochemical and pharmacological criteria [3]. The mechanisms, however, of nonclathrin, noncaveolin endocytosis are not clear [4, 5]. Here we show that coassembly of two similar membrane proteins, flotillin1 and flotillin2 [6-8], is sufficient to generate de novo membrane microdomains with some of the predicted properties of lipid rafts [9]. These microdomains are distinct from caveolin1-positive caveolae, are dynamic, and bud into the cell. Coassembly of flotillin1 and flotillin2 into microdomains induces membrane curvature, the formation of plasma-membrane invaginations morphologically similar to caveolae, and the accumulation of intracellular vesicles. We propose that flotillin proteins are defining structural components of the machinery that mediates a clathrin-independent endocytic pathway. Key attributes of this machinery are the dependence on coassembly of both flotillins and the inference that flotillin microdomains can exist in either flat or invaginated states.

SDPR induces membrane curvature and functions in the formation of caveolae

Caveolae are plasma membrane invaginations with a characteristic flask-shaped morphology. They function in diverse cellular processes, including endocytosis. The mechanism by which caveolae are generated is not fully understood, but both caveolin proteins and PTRF (polymerase I and transcript release factor, also known as cavin) are important. Here we show that loss of SDPR (serum deprivation protein response) causes loss of caveolae. SDPR binds directly to PTRF and recruits PTRF to caveolar membranes. Overexpression of SDPR, unlike PTRF, induces deformation of caveolae and extensive tubulation of the plasma membrane. The B-subunit of Shiga toxin (STB) also induces membrane tubulation and these membrane tubes also originate from caveolae. STB colocalizes extensively with both SDPR and caveolin 1. Loss of caveolae reduces the propensity of STB to induce membrane tubulation. We conclude that SDPR is a membrane-curvature-inducing component of caveolae, and that STB-induced membrane tubulation is facilitated by caveolae.

The Wnt signalling effector Dishevelled forms dynamic protein assemblies rather than stable associations with cytoplasmic vesicles

Dishevelled is a crucial effector upstream in the Wnt signalling pathway, but the molecular mechanism by which it transduces the Wnt signal remains elusive. Dishevelled is a cytoplasmic protein with a strong tendency to form puncta, which correlates with its potent activity in stimulating Wnt signal transduction when overexpressed. These puncta are thought to reflect cytoplasmic vesicles. However, we show here that the mammalian Dishevelled protein Dvl2 does not colocalise with known vesicle markers for clathrin-mediated or clathrin-independent endocytic pathways. Furthermore, Dvl2 puncta do not stain with lipid dyes, indicating that these puncta do not contain membranes. Instead, our evidence from live imaging by TIRF microscopy of Dvl2 tagged with green fluorescent protein (GFP-Dvl2) revealed that these puncta move in and out of the evanescent field near the plasma membrane in an undirected fashion, and that they can grow by collision and fusion. Furthermore, high-resolution confocal microscopy and photobleaching experiments indicate that the GFP-Dvl2 puncta are protein assemblies; there is a constant exchange of GFP-Dvl2 between puncta and a diffuse cytoplasmic pool, which, therefore, are in a dynamic equilibrium with each other. The same is true for the DIX domain of Dvl2 itself and also for Axin-GFP, which equilibrates between the punctate and cytosolic pools. Our evidence indicates that Dvl2 and Axin puncta are dynamic protein assemblies rather than cytoplasmic vesicles.

SNAREs and membrane fusion in the Golgi apparatus

Soluble factors, NSF and SNAPs, are required at many membrane fusion events within the cell. They interact with a class of type II integral membrane proteins termed SNAP receptors, or SNAREs. Interaction between cognate SNAREs on opposing membranes is a prerequisite for NSF dependent membrane fusion. NSF is an ATPase which will disrupt complexes composed of different SNAREs. However, there is increasingly abundant evidence that the SNARE complex recognised by NSF does not bridge the two fusing membranes, but rather is composed of SNAREs in the same membrane. The essential role of NSF may be to prime SNAREs for a direct role during fusion. The best characterised SNAREs in the Golgi are Sed5p in yeast and its mammalian homologue syntaxin 5, both of which are predominantly localised to the cis Golgi. The SNARE-SNARE interactions in which these two proteins are involved are strikingly similar. Sed5p and syntaxin 5 may mediate three distinct pathways for membrane flow into the cis Golgi, one from the ER, one from later Golgi cisternae, and possibly a third from endosomes. Syntaxin 5 is itself likely to cycle through the ER, and thus may be involved in homotypic fusion of ER derived transport vesicles. In all well characterised SNARE dependent membrane fusion events one of the interacting SNAREs is a syntaxin homologue. There are only eight members of the syntaxin family in yeast. Besides Sed5p two others, Tlg1p and Tlg2p, are found in the Golgi complex. They are present in a late Golgi compartment, but neither is required for transit of secreted proteins through the Golgi. We suggest that these observations are most compatible with a model for transit through the Golgi in which anterograde cargo is carried in cisternae, the enzymatic composition of which changes with time as Golgi resident enzymes are delivered in retrograde transport vesicles.