Clathrin-independent endocytosis: an increasing degree of complexity

Abstract

This article aims at providing an update on the complexity of clathrin-independent endocytosis. It is now almost 30 years since we first wrote a review about its existence; at that time many people believed that with the exception of macropinocytosis, which will only be briefly mentioned in this review, all uptake could be accounted for by clathrin-dependent endocytosis. Now it is generally accepted that there are different clathrin-independent mechanisms, some of them regulated by ligands and membrane lipid composition. They can be both dynamin-dependent and -independent, meaning that the uptake cannot be accounted for by caveolae and other dynamin-dependent processes such as tubular structures that can be induced by toxins, e.g. Shiga toxin, or the fast endophilin mediated endocytosis recently described. Caveolae seem to be mostly quite stable structures with other functions than endocytosis, but evidence suggests that they may have cell-type dependent functions. Although several groups have been working on endocytic mechanisms for years, and new advanced methods have improved our ability to study mechanistic details, there are still a number of important questions we need to address, such as: How many endocytic mechanisms does a cell have? How quantitatively important are they? What about the complexity in polarized cells where clathrin-independent endocytosis is differentially regulated on the apical and basolateral poles? These questions are not easy to answer since one and the same molecule may contribute to more than one process, and manipulating one mechanism can affect another. Also, several inhibitors of endocytic processes commonly used turn out to be less specific than originally thought. We will here describe the current view of clathrin-independent endocytic processes and the challenges in studying them.

Keywords: Caveolae; Caveolin; Clathrin; Endocytosis; Endophilin; Rho proteins.

Fig. 1

An overview of endocytic mechanisms in a non-polarized and b polarized epithelial cells. The different mechanisms are described in the text: In addition to clathrin-dependent endocytosis, we have indicated the following pathways: Caveolae, now regarded as quite stable structures; flotillin, may mediate transfer of a ligand to the invagination rather than contribute to the endocytic process; the Cdc42- dependent but dynamin-independent uptake; the Rho/Rac pathway; the toxin-induced tubules with endophilin; FEME (see text); as well as macropinocytosis. In polarized cells the apical clathrin-independent uptake is regulated by a number of factors, e.g. protein kinase A, which does not affect the basolateral uptake. It should be noted that in MDCK cells all caveolae are at the basolateral pole

Fig. 2

Ruthenium red added during fixation of cells and the resulting black staining of the membrane reveal that caveolae which may appear as free vesicles in the cytosol are surface connected. Bar 100 nm. This figure is reproduced from (Sandvig et al. 2011). 10.1016/j.ceb.2011.03.007

Fig. 3

Examples from HEp-2 cells of clathrin-coated areas with different extent of invagination (a–e) and schematically drawn in f. In g, clathrin-coats in control cells and cells treated with 10 mM mβCD for 15 min were scored for invagination as classified in f (approximately 200 coated pits in each experiment). The relative frequency of the different types is shown, and reveal that mβCD prevents invagination of clathrin-coated pits. Bar 100 nm. This figure is reproduced from (Rodal et al. 1999). http://www.molbiolcell.org/content/10/4/961.full.pdf+html

Fig. 4

Cholesterol is required for macropinocytosis. A431 cells were serum-starved for 4 h, and in a and b the cholesterol was lowered by a 30 min incubation with 5 mM mβCD (CDx) before the cells were incubate for 15 min with HRP (horseradish peroxidase) (10 mg/ml) in the presence of mβCD and 1 µM TPA (12-O-tetradecanoylphorbol 13-acetate). There is no ruffling and macropinocytosis. In contrast, in c and d, a mixture of mβCD (2.5 mM) and mβCD (2.5 mM) with cholesterol (CDxCh) was used in order not to affect membrane cholesterol, and as shown, the cells are then able to form ruffles and macropinosomes. Bars 1 µm; FS free surface; IS intercellular space. This figure was reproduced from (Grimmer et al. 2002); http://jcs.biologists.org/content/joces/115/14/2953.full.pdf