Role of Caveolae in Endocytosis of Cholera Toxin

In addition to the recent studies on internalization of virus mentioned above, studies on cholera toxin (CT) uptake have also been important for the ongoing discussion of the role of caveolae in endocytosis. CT belongs to a large group of protein toxins with enzymatically active A-moieties, and B-moieties that act as ligands and bind to cell- surface components. These toxins must be endocytosed before they exert their toxic effect(s). In the case of CT, the B-moieties bind to the ganglioside GM1 and, following internalization and transport via the Golgi complex to the ER, the A-moiety is translocated to the cytosol where it affects the activity of adenylyl cyclase [11].

Although clathrin-mediated and clathrin- and caveolin-independent endocytosis can play a role in the uptake of CT [65-69], this does not in any way exclude a role of caveolae as well, and it is widely believed that caveolae or caveolae-like structures are the major vehicle for CT internalization and the subsequent delivery to the TGN [70-74]. The main reason for this is that CT binds to GM1, which is in part present in caveolae [68,75]. However, the consequences of caveolae-mediated CT/ GM1 internalization for the cell's steady-state balance of membrane-associated caveolin and caveolae have generally been ignored. In principle, two possibilities for CT-stimulated internalization of caveolae exist. Binding of CT to GM1 may induce a wave of caveolae internalization as described above for SV40 and okadaic acid, leading to down-regulation of caveolae at the plasma membrane. Alternatively, CT could induce a constitutive internalization mechanism for caveolae including an efficient recycling of caveolae or caveolin back to the plasma membrane. This would not lead to a depletion of the plasma membrane pool of cav-eolae, but requires a rapid replacement of those CT-containing caveolae that become internalized with new, empty caveolae.

The first report of a major role for caveolae in CT endocytosis was an ultrastructural study published in 1982 by Montesano et al. [72]. Here, the authors used CT-B chain (CT-B)-gold and found that the conjugate preferentially bound to caveolae at 2 °C. Interestingly, they noticed an increase in CT-B-gold in caveolae relative to CT-B-gold bound to noncaveolar plasma membrane over time at 37 °C. These observations were originally interpreted as the result of an up-concentration of CT in endocytic caveolae. However, an alternative explanation could be selective internalization of CT bound to the noncaveolar plasma membrane. In fact, CT/ GM1 in caveolae could represent an immobilized fraction of the ligand-receptor complex. In the study by Montesano et al., CT-B-gold was only found very rarely, if ever, in clathrin-coated pits. However, other studies have documented small, but significant amounts of CT-B-gold [68], as well as biotin-GM1 detected by antibiotin antibodies conjugated to gold [75] in clathrin-coated pits. CT-B-HRP has also been localized to clathrin-coated pits [66]. Parton [68] was in fact very careful not to draw too far-reaching conclusions based on the preferential localization of CT-B to caveolae, and stressed that at least some GM1 must be internalized by clathrin-coated pits. We would like to emphasize that the finding that only minimal CT-B can be observed at any given time point in clathrin-coated pits does not exclude a significant role for the coated pits in CT-B uptake. Thus, clathrin-coated pits are highly dynamic structures with a half-life of minutes at the cell surface [8], and CT-B may continuously move into newly formed clathrin-coated pits to become efficiently internalized. This is in agreement with the findings that clathrin-dependent endocytosis can account for about 50% of CT internalization [66,76].

Several uptake mechanisms have been described for CT and these may - to some extent - be cell type-dependent. Thus, CT can be internalized both by cla-thrin-dependent and -independent mechanisms, and the effect of cholesterol-binding drugs (nystatin, filipin, cyclodextrin) on CT-internalization varies [6567,71,77,78]. In studies with dominant negative mutants of epsin and eps15 (molecules required for clathrin-mediated endocytosis), Nichols et al. [67] found that a proportion of CT-B is internalized by clathrin-coated pits. Moreover, Shogomori and Futerman [65] reported that although CT-B is present in detergent-insoluble rafts or microdomains, it becomes internalized by a raft-independent mechanism, presumably via clathrin-coated pits. The same is true for the epidermal growth factor receptor [79]. Recently, Hansen et al. [69] showed that CT bound to GM1 localized to a detergent-insoluble, raft-like fraction of the enterocyte brush border membrane, and was subsequently internalized by a cholesterol-independent, but clathrin-dependent mechanism. It could be argued that uptake mediated by cla-thrin-coated pits would not lead to delivery of cargo to the TGN. However, Shiga toxin, which binds to globotriasylceremide (Gb3), is internalized via clathrin-coated pits and vesicles by an as-yet unknown mechanism [80] and subsequently transported to the TGN and the ER [81]. In accordance with this, expression of antisense clathrin heavy chain inhibits the toxicity of Shiga toxin [82]. Interestingly, Shiga toxin-HRP, like CT-B-HRP, also accumulates in caveolae at 37 °C (our unpublished data). It should be noted that ligands taken up from caveolae and cla-thrin-coated pits can ultimately be located in the same endosome (see Section 4.7).

In our attempts to characterize the endocytic mechanisms of CT uptake [66], Caco-2 cells were transfected with caveolin, leading to the formation of basolateral caveolae in the otherwise caveolae/caveolin-negative, highly polarized cell line [20]. This did not change the uptake or effect of CT. Furthermore, treatment of the cav-1-expressing Caco-2 cells with the cholesterol-binding drug filipin only reduced internalization of CT slightly (<20%), suggesting that although caveolae may be involved in CT endocytosis, they do not play a major role. In contrast, cholesterol-depletion with mbCD, a treatment that both removes caveolae and inhibits the formation of deeply invaginated clathrin-coated pits and subsequently coated vesicle formation [37,83], resulted in a 30-40% reduction in CT uptake [66], indicating that coated pits/vesicles also play a role, though not exclusive, in CT uptake. Similarly, in baby hamster kidney (BHK) cells expressing antisense clathrin heavy chain [84], endocytosis of CT was reduced by about 50%. Moreover, in HeLa cells expressing the dominant-negative dynamin mutant K44A, a 50-70% reduction in CT uptake was seen [66]. In these cells both endocytosis mediated by caveolae and by clathrin-coated pits/vesicles are inhibited, as well as possible caveolin- and cla-thrin-independent mechanisms that may depend on dynamin. In BHK and HeLa cells, in which either antisense clathrin or the dominant-negative dynamin mutant was expressed, CT is still internalized after cholesterol depletion with mbCD [85]. In a recent report it was similarly shown that the overexpression of dominant mutants that inhibit clathrin-, caveolin- or Arf6-dependent endocytic pathways does not prevent CT uptake and trafficking to the Golgi and ER, or CT cytotoxicity.

Interestingly, even under conditions where all three endocytic pathways were inhibited simultaneously, CT has a toxic effect [86]. Importantly, it must be borne in mind that the blocking of one or more pathways may lead to up-regulation of other pathways that are normally quiescent [87], or it may induce new pathways [88]. Taken together, these results suggest that CT can be internalized by caveolae and clathrin-coated pits/vesicles as well as by one or more mechanism not involving these structures. However, caveolae do not seem to play a major role in the uptake.

In order to further analyze the importance of caveolae in CT-uptake, and not least to evaluate the internalization efficiency of CT bound to GM1 inside and outside caveolae, we incubated HeLa cells with red fluorescent CT-B on ice and subsequently at 37 °C for various periods of time (between 0 and 120 min), followed by fixation and immunocytochemical detection of caveolae using an antibody against endogenous caveolin-1. A partial co-localization of CT-B and caveolin on the cell surface following labeling at 0 °C was seen. However, upon heating to 37 °C, CT-B was internalized and gradually accumulated in the perinuclear region. After 30 min and more, some - but not all - of the internalized CT-B was co-localized with TGN38, a marker of the TGN (Fig. 4.2A), while a small fraction of CT was still seen at the plasma membrane where it apparently co-localized with caveolin. Interestingly, during this CT-B uptake, no change in the distribution of endogenous caveolin was detected, irrespective of the time of CT-B incubation. At all time points (0-120 min) the peripheral rim of caveolae remained stable. In blind experiments, where the microscope operator only examined caveolin in the confocal microscope and was unaware of the incubation time with CT-B at 37 °C, no difference between the time points could be established. Similarly, no differences in the distribution of caveolae could be observed when cells were incubated with the holotoxin (CT A+B chain) (our unpublished data).

Since the individual cells bound CT-B to varying degrees, and some were unlabeled, such cells could serve as internal controls for a changed pattern in the distribution of endogenous caveolin stimulated by CT-B binding. Importantly, at all incubation times the caveolar localization was the same, irrespective of the GM1 expression level of the cell and consequently the degree of CT-B binding (Fig. 4.2B). Taken together, these data strongly indicate that CT-B uptake does not cause a synchronized wave of caveolae internalization. To further analyze the effect of CT-B uptake on the distribution of caveolae, we expressed GFP-tagged caveolin-1 fusion proteins [31] in HeLa cells and human skin fibroblasts. When these cells were incubated with CT-B on ice and subsequently incubated at 37 °C for various periods (0-120 min), and then fixed and examined in the confocal microscope, results similar to those with endogenous caveolin were obtained: CT-B became internalized and accumulated in the TGN over time, while the pattern of caveolar expression was unchanged - no wave of incoming caveolae or caveolin stimulated by CT-B was observed (Fig. 4.2C and D). When the pattern of CT-B internalization in cells expressing GFP-tagged caveolin was compared to that of nontransfected, neighboring cells in the same culture, no differences were seen (Fig. 4.2C). Thus, the expression of GFP-tagged caveolin does not influence CT-B uptake and transport.

-t -p.

'•• .A

Cav-1 _






_ Cav-1



_ Cav-GFP



D 5 min




Merge _

15 min

30 min

Fig. 4.2 (A) HeLa cells were labeled with Alexa 594 CT-B (red) on ice for 1 h, washed, and incubated at 37 °C for 2 h to allow internalization of CT-B. The cells were subsequently fixed and immunolabeled for endogenous caveolin-1 (white) and TGN-38 (green). After 2 h, the majority of CT-B has been internalized to the TGN. (B) HeLa cells were allowed to internalize CT-B for 2 h, as described above. The cells were subsequently fixed and immunolabeled for endogenous caveolin-1 (green). The panel shows two cells expressing GM1 that have internalized CT-B, and adjacent cells that have not internalized CT-B. The cellular distribution of caveolin-1 is the same, irrespective of whether the cells have internalized CT-B. (C) HeLa cells trans-fected with Cav-GFP were allowed to internalize CT-B for 2 h, as described above. The panel shows two transfected cells and adjacent, nontransfected cells. The pattern of internalized CT-B is the same in transfected and nontransfected cells. (D) HeLa cells transfected with Cav-GFP and incubated with CT-B for 5 to 30 min at 37 °C. While CT-B moves from the cell surface to the perinuclear region (the TGN), the distribution of Cav-GFP-labeled caveolae is not changed. Scale bars = 20 mm.

Gm1 Caveolae

Fig.4.3 Electron micrographs of BHK cells incubated with a CT-B-HRP conjugate. (A) Cells were incubated at 0 °C with the conjugate, which is seen all over the cell surface including some caveolae. (B,C) Cells have been incubated with the conjugate at 0 °C for 1 h, then washed and further incubated for 30 min at 37 °C. Very little or no labeling is seen at the cell surface, except in caveolae, which are heavily labeled. Such images sug

Fig.4.3 Electron micrographs of BHK cells incubated with a CT-B-HRP conjugate. (A) Cells were incubated at 0 °C with the conjugate, which is seen all over the cell surface including some caveolae. (B,C) Cells have been incubated with the conjugate at 0 °C for 1 h, then washed and further incubated for 30 min at 37 °C. Very little or no labeling is seen at the cell surface, except in caveolae, which are heavily labeled. Such images sug gest that the CT-B-HRP has become internalized by a caveolae-independent mechanism. However, on very rare occasions unlabeled caveolae can be seen (arrow in C). These caveolae most likely are formed after the surface-associated CT-B-HRP conjugate was internalized, and may represent a substitution for a rare caveola that has been involved in conjugate uptake. Scale bar = 200 nm.

Next, the internalization of CT-B in relation to caveolae was followed over time in HeLa cells and human skin fibroblasts by photobleaching techniques (our unpublished results). For this purpose, cells expressing GFP-tagged caveolin were preincubated for 1 h at 4 °C with CT-B, washed and further incubated for 5-10 min in CT-B-free medium at 37 °C on stage to stabilize the culture. The cells were then exposed to a modified FRAP protocol where the entire cytoplasm was bleached in both the red and the green channels, leaving only the most peripheral rim of fluorescence corresponding to the plasma membrane (see also Ref. [31]). Fluorescence recovery in the cytoplasm was then followed over time. The rationale of this type of experiment is that if CT-B binding stimulates caveolae to become con-stitutively internalized and replaced by a recycling mechanism, then the GFP-tagged caveolae at the plasma membrane should disappear and be replaced by caveolae which are not labeled - that is, the peripheral fluorescence should decrease and the intracellular fluorescence increase. It was, however, striking that although CT-B appeared in vesicles in the bleached interior of the cells, GFP-tagged caveolin did not, but rather remained at the cell periphery. These findings strongly suggest that (GFP-tagged) caveolae are not efficiently internalized and

4.6 A Small Fraction of Caveolae may become Constitutively Internalized 181

recycled in response to CT-binding but are more likely stable, plasma membrane-associated structures.

These findings were further supported by EM observations on HeLa cells showing that even after 30-120 min of CT-B-HRP internalization, caveolae were often distinctly labeled by CT-B-HRP although the remaining plasma membrane was unlabeled (Fig. 4.3). This emphasizes that at time points where the main, mobile GM1/CT-B fraction has been internalized by a caveolae-independent mechanism^), a nonmobile fraction is still present on the cell surface, trapped in the caveolae. It was also possible by using EM to detect the internalized CT-B-HRP in the TGN [17]. The number of caveolae at the cell surface of cells pulse-chased with CT-B-HRP for 30 min at 37 °C was the same as in control cells not exposed to CT-B-HRP (our unpublished results).

Furthermore, since it has been reported that expression of the K44A dominantnegative dynamin-1 mutant inhibits caveolae-mediated internalization of CT-B in endothelial cells [28], we also quantified the number of caveolae in HeLa K44A cells expressing the mutant - that is, in cells grown without tetracycline. Interestingly, although expression of the dynamin mutant leads to a two- to three-fold increase in the number of clathrin-coated pits at the cell surface as expected [66], the number of caveolae at the cell surface was unchanged, indicating that caveolae are not constitutively internalized by a dynamin-dependent mechanism. In K44A cells grown both in the presence or absence of tetracycline and incubated for 30 min at 37 °C with CT-B-HRP, practically all caveolae were labeled (our unpublished observations).

Was this article helpful?

0 0
Essentials of Human Physiology

Essentials of Human Physiology

This ebook provides an introductory explanation of the workings of the human body, with an effort to draw connections between the body systems and explain their interdependencies. A framework for the book is homeostasis and how the body maintains balance within each system. This is intended as a first introduction to physiology for a college-level course.

Get My Free Ebook

Post a comment