The imaging of events in living cells offers a way to test models of cell behaviour and develop new hypotheses. The invaginating ‘pits’ on the surface of cells are the latest subject of this approach.
Writing last month in Cell, Ehrlich and colleagues1 reported progress in a 40-year line of investigation. They have studied the ‘coated pits’ that form on the surface of cells and which take up material from the surface and the extracellular environment.
Seminal work by Roth and Porter2 first showed that there are regions on the surface of mosquito eggs that are specialized to allow these cells to take up yolk proteins from their environment. Roth and Porter called these areas coated pits because the surface facing the inside of the cell was coated with an electron-dense material. That material was later shown to be made up of a protein named clathrin3. Over the past 20 years, many other components of coated pits have been identified. These include the AP2 adaptor protein, which links transmembrane cargo molecules to the clathrin lattice, and dynamin, an enzyme required to transform coated pits into small, spherical vesicles inside the cell4 (although its precise role remains controversial).
Clathrin-coated pits have become firmly established as the major pathway for the uptake of nutrients and signalling molecules in almost all non-bacterial cells. This pathway is also essential in tissue-specific processes such as immune surveillance, and it is frequently hijacked by pathogens seeking entry into the cell4. A combination of biochemistry, electron microscopy and cell biology has provided a series of snapshots of the coated-vesicle cycle. These snapshots define a temporal order of events (Fig. 1) that includes the assembly of clathrin and other coat proteins just under the cell surface, the selection of cargo molecules, an increase in membrane curvature, and scission of the invaginating pit. The vesicle thereby produced then rapidly loses its coat, allowing it to fuse with another intracellular compartment, the early endosome.
By labelling clathrin, AP2 and cargo molecules with fluorescent ‘tags’, Ehrlich et al.1 have now gone a step further: rather than taking static snapshots, they have produced movies of the vesicle cycle in living cells. This is not the first example of live-cell imaging of coated-pit formation5. But it is the first such study to probe the connection between coat proteins, such as clathrin and AP2, and cargo molecules (such as low-density lipoprotein and its receptor) that are known to be internalized by this route. They thus provide compelling evidence that the fluorescent particles observed indeed represent functional coated pits. The authors introduced the fluorescently tagged molecules into cells cultured under conditions in which the molecules were stably expressed, and subsequently identified growing coated pits through an increase in the size of fluorescent spots on the plasma membrane with time.
It has been suggested6 that there are ‘hotspots’ on the inside of the cell surface at which coated pits assemble, implying that initiator components are specifically localized to such hotspots. But Ehrlich et al. observe that the initiation of new pits occurs randomly, suggesting that the proteins or lipids responsible must also be distributed randomly on the plasma membrane.
The authors also find that the fluorescently labelled cargo is recruited to newly forming pits. Surprisingly, though, a significant proportion of the developing pits (around 25%) rapidly disassemble without incorporating cargo. So Ehrlich et al. suggest that the capture of cargo might stabilize nascent pits and commit them to forming vesicles. In practice, this would prevent cells from assembling vesicles in the absence of cargo, thus efficiently coupling function to assembly. The idea that cargo recruitment is linked to successful vesicle formation is consistent with the observation that cargoes such as growth-factor receptors can modulate the vesicle machinery to facilitate their own internalization7. Recruitment of such cargo early in pit formation would provide a window of opportunity where the rate at which pits commit to forming vesicles might be enhanced.
Ongoing debate surrounds the role of dynamin, which has been implicated in the final pinching-off of coated pits. Ehrlich and colleagues' observation that dynamin is recruited to newly forming coated pits at early time points is consistent with studies that show a role for this protein in pit invagination as well as scission8. However, as dynamin was also detected on the pits that disassemble prematurely, it is unlikely to be a major player in the stabilization of nascent pits. If there is instead a link between cargo recruitment and commitment to vesicle formation, coat components such as epsin, which has been implicated both in the generation of membrane curvature9 and as a possible cargo-binding protein10, might mediate this coupling.
Although this is a pleasing interpretation, other possibilities exist, not least because the coated pits that abort might be expected to be able to capture, and be stabilized by, other, unlabelled cargo molecules, which are relatively abundant on the cell surface. One alternative possibility is that a ‘commitment’ step directly precedes, and is necessary for, subsequent cargo recruitment. In this model, cargo recruitment and a commitment to vesicle formation would still be linked, but the commitment step would come first.
The authors also find that relatively large cargo molecules, such as reovirus, take longer than smaller cargoes to become included in a fully formed coated pit, hinting that the type of cargo might help to determine pit size. Indeed, hen egg cells have been found to contain large coated pits that are specialized for the uptake of yolk proteins. But generally, coated pits contain more than one type of cargo molecule and, when visualized by electron microscopy, have a fairly uniform diameter (100–200 nm), which should be large enough to accommodate a reovirus particle. To determine whether clathrin assembles at the same rate around cargoes of different sizes (thus leading to coated vesicles of different sizes), or whether different cargoes affect the rate of coat assembly, pit formation rates in the presence of large and small cargoes will need to be compared under the same conditions.
Ehrlich et al. also examined the dynamics of other fluorescently tagged components of coated pits. They find that AP2 and clathrin assemble into coated pits and disassemble from coated vesicles at apparently the same rates as each other. In vitro data have shown that assembly and disassembly of the two molecules have very different biochemical requirements; in fact, they have even been shown to be uncoupled in cells in which clathrin is knocked out11. Ehrlich and colleagues' findings show that these processes are nonetheless coordinated with precision in vivo.
The exciting aspect of this study1 is that it allows cell biologists to re-evaluate their models of the coated-vesicle cycle, which are based on a huge body of biochemical and cell-biological data. Furthermore, Ehrlich and colleagues' approach should make it possible to examine the behaviour of other coated-pit components — and the technique could be adapted to reveal what happens when key proteins are mutated or ablated in live cells.
Ehrlich, M. et al. Cell 118, 591–605 (2004).
Roth, T. F. & Porter, K. R. J. Cell Biol. 20, 313–323 (1964).
Pearse, B. M. Proc. Natl Acad. Sci. USA 73, 1255–1259 (1976).
Conner, S. D. & Schmid, S. L. Nature 422, 37–42 (2003).
Rappoport, J., Simon, S. & Benmerah, A. Traffic 5, 327–337 (2004).
Gaidarov, I., Santini, F., Warren, R. A. & Keen, J. H. Nature Cell Biol. 1, 1–7 (1999).
Clague, M. J. & Urbe, S. J. Cell Sci. 114, 3075–3081 (2001).
Hill, E. et al. J. Cell Biol. 152, 309–323 (2001).
Ford, M. G. J. et al. Nature 419, 361–366 (2002).
Traub, L. M. J. Cell Biol. 163, 203–208 (2003).
Wettey, F. R. et al. Science 297, 1521–1525 (2002).