Endocytosis is critical for controlling the protein–lipid composition of the plasma membrane, uptake of nutrients as well as pathogens, and also plays an important role in regulation of cell signalling. While a number of pathways for endocytosis have been characterized in different organisms, all of these require remodelling of the cell cortex. The importance of a dynamic actin cytoskeleton for facilitating endocytosis has been recognized for many years in budding yeast, and is increasingly supported by studies in mammalian cells. Our studies have focused on proteins that we have shown to act at the interface between the actin cytoskeleton and the endocytic machinery. In particular, we have studied interactions of Sla1p, which binds to both activators of actin dynamics, i.e. Abp1p, Las17p and Pan1p, and to cargo proteins such as the pheromone receptor Ste2p. More recently we have mapped the interaction of Sla1p with Lsb5p, a protein that has a similar structure to the GGA [Golgi-localizing, γ-adaptin ear homology domain, Arf (ADP-ribosylation factor)-binding] family of proteins with an N-terminal VHS (Vps27p/Hrs/STAM)-domain and a GAT (GGAs and TOM1) domain. We show that Lsb5p can interact with yeast Arf3p (orthologous with mammalian Arf6) and we demonstrate a requirement for Arf3p expression in order to localize Lsb5p to the cell cortex.

Introduction

There is a growing body of evidence that demonstrates a requirement for a dynamic actin cytoskeleton to facilitate the process of endocytosis. Much of the initial evidence for a role for actin in the process of endocytosis derives from studies in yeast which demonstrate that strains mutated in actin-binding proteins show concomitant defects both in cortical actin organization and in endocytosis [13]. Research on mammalian cells has been less conclusive for a role for the actin cytoskeleton in endocytosis, possibly due to further levels of complexity in the processes [4]. The growing consensus does, however, support the idea of evolutionary conservation of the endocytic machinery, and many studies using mammalian cells are now indicating the importance of cortical actin in facilitating endocytosis [57], In particular, elegant studies from the Merrifield laboratory have used real-time fluorescence microscopy to reveal the presence of actin at sites of endocytosis and also indicate the likely scenario of sequential complex assembly and disassembly of proteins from these sites during the process of endocytosis as observed in yeast [810].

Sla1p and the assembly of complexes required for endocytosis

Previous work in our laboratory has characterized the cortical protein Sla1p in the regulation of actin dynamics and in coupling this role to the functioning of the endocytic machinery [1113]. While Sla1p does not yet have a direct mammalian homologue, it does have a similar domain composition and a comparable spectrum of interactions to the protein CIN85 (Cbl-interacting protein of 85 kDa)/CD2AP (CD2-associated protein) found to be involved in endocytosis in mammalian cells [14,15]. A combination of in vitro and in vivo studies reveal that Sla1p localizes with the endocytic machinery but it functions to regulate actin dynamics through its interactions with Arp2/3 (actin-related protein 2/3) activator proteins (Abp1p, Las17p and Pan1p) [12,16].

Sla1p is proposed to be a central player in the current model for endocytosis in Saccharomyces cerevisiae. The overall model proposed by Kaksonen et al. [17] is that there is a sequential assembly and disassembly of complexes at the sites of endocytosis on the plasma membrane. Early members of the complex include Sla1p, Sla2p (HIP1-R homologue), Pan1p (Eps15 homologue) and End3p. These components are present in non-motile and slow moving patches and are possibly involved in binding coat components, membrane deformation and recruitment of the actin cytoskeleton. Actin recruitment is followed by a slow and then a rapid movement of the patch, proposed to correspond to the invagination, scission and movement of the vesicle away from the membrane. Regulation of the complex by kinases Ark1/Prk1 [homologues of mammalian AAK1 (adaptor-associated kinase 1) and GAK1 (cyclin-G-associated kinase 1)] causes Sla1p and Pan1p to be removed from the complex [18]. In strains in which ark1 and prk1 have been deleted, Sla1p and Pan1p remain in the complex and endocytosis is abrogated indicating the importance of disassembly processes as part of this dynamic event.

Lsb5p: a GGA-like protein at the plasma membrane

Two hybrid approaches led us to identify Lsb5p as an Sla1p interactor. Lsb5p has also been listed in a report of proteins interacting with the yeast WASP (Wiskott–Aldrich syndrome protein) homologue Las17p [19]. Sequence analysis reveals that Lsb5p contains a number of recognizable motifs reminiscent of membrane trafficking proteins (Figure 1). At its N-terminus is a region with homology to both VHS (Vps27p/Hrs/STAM) and ENTH (epsin N-terminal homology) domains. It also contains a putative Arf-interacting GAT {GGAs [Golgi-localizing, γ-adaptin ear homology domain, Arf (ADP-ribosylation factor)-binding] and TOM1} domain and a motif termed NPF (Asn-Pro-Phe), which, in other proteins, has been demonstrated to interact with EH (EPS15 homology) domains [20]. VHS and GAT domains are found as adjacent domains in a family of proteins termed the GGA proteins [21,22]. However, this family of proteins also contains a γ-adaptin ear domain and a clathrin-binding motif which is not in Lsb5p. Furthermore, the GGA proteins do not contain NPF motifs. The GGA proteins have been shown to function largely within the endoplasmic reticulum and Golgi apparatus [22,23].

Proteins at the interface of actin regulation and endocytosis

Figure 1
Proteins at the interface of actin regulation and endocytosis

(A) Interaction sites between several cortical patch proteins involved in endocytosis. Dotted lines represent regions of the proteins that have been shown to interact either biochemically or through two-hybrid approaches [19,24,25]. Domains indicated are: SH3, Src homology-3; SHD1, sla1 homology domain-1; SHD2, sla1 homology domain-2; C-terminal rpts, repeats containing a threonine residue that is phosphorylated by the Ark1/Prk1 kinases; VHS; GAT; WH1, WASP homology-1; pro, proline-rich; WH2, WASP homology-2. (B) Yeast GFP–Lsb5p expressed in mammalan COS7 cells and co-stained with anti-tubulin. Lsb5 localizes to the periphery of cells (arrowheads) and also to the mid-body (arrow), and both sites are observed to contain mammalian Arf6.

Figure 1
Proteins at the interface of actin regulation and endocytosis

(A) Interaction sites between several cortical patch proteins involved in endocytosis. Dotted lines represent regions of the proteins that have been shown to interact either biochemically or through two-hybrid approaches [19,24,25]. Domains indicated are: SH3, Src homology-3; SHD1, sla1 homology domain-1; SHD2, sla1 homology domain-2; C-terminal rpts, repeats containing a threonine residue that is phosphorylated by the Ark1/Prk1 kinases; VHS; GAT; WH1, WASP homology-1; pro, proline-rich; WH2, WASP homology-2. (B) Yeast GFP–Lsb5p expressed in mammalan COS7 cells and co-stained with anti-tubulin. Lsb5 localizes to the periphery of cells (arrowheads) and also to the mid-body (arrow), and both sites are observed to contain mammalian Arf6.

In our characterization of Lsb5p, we were able to demonstrate localization of Lsb5p to the plasma membrane in an actin-independent manner and that deletion of LSB5 in a yeast strain already lacking a particular actin-associated protein, Ysc84p, led to an almost complete block in endocytosis [24]. From these results we proposed that yeast has at least two definable endocytic pathways and that Lsb5p is an important adaptor-like protein in one of these. More recently, we have mapped the interaction with Las17p to the N-terminal domain of Lsb5p and the polyproline region of Las17p. We have also shown that Sla1p interacts with the C-terminal NPF motif of Lsb5p. Most interestingly, we observed an interaction with yeast Arf3p, which in previous studies was reported not to play a role in endocytosis [25,26].

Does Arf3p have a role in endocytosis in yeast?

Yeast Arf3p is the homologue of mammalian Arf6, which regulates the membrane trafficking between the plasma membrane and endosome and is also involved in the reorganization of the actin cytoskeleton and cytokinesis [27]. However, deletion of yeast ARF3 was previously reported not to affect endocytosis and only mild effects are seen on polarity [26]. The interaction between Lsb5p and Arf3p was demonstrated biochemically. We tested whether the interaction was enhanced in the presence of ubiquitin, which has been reported to stabilize the interaction between Arfs and the GAT domain of GGA proteins within the Golgi [28,29]. However, the presence of ubiquitin did not affect the Arf3p–Lsb5p interaction. We also determined whether deletion of either the ARF3 or LSB5 gene affected localization of the other protein. While deletion of LSB5 did not affect Arf3p localization to the plasma membrane, deletion of ARF3 did lead to a loss of Lsb5p localization. This leads to the suggestion that Arf3p may play a role in endocytosis through its recruitment of Lsb5p to the plasma membrane [25]. The reason for a lack of phenotype in the arf3 null strain previously noted [26] is possibly because that study reported effects of the deletion only in a wild-type strain. It might be that Arf3p functions in a minor or redundant pathway in a similar way to Lsb5p [25], making phenotypic effects clear only in the absence of other proteins.

Despite the absence of a clear mammalian Lsb5 homologue we postulated that if yeast Lsb5p can interact with Arf3p, it might also be able to interact with its mammalian homologue, Arf6. We were able to demonstrate an interaction using the two-hybrid approach with active GTP-Arf6 and but not with GDP-Arf6. We also expressed GFP (green fluorescent protein)-tagged yeast Lsb5p in mammalian COS7 cells. If the Arf interaction is conserved one might predict that it could localize to the sites of its interacting Arf. Interestingly, we observed localization to the cell periphery and also to the mid-body during cytokinesis [25]. Arf6 has been observed in both of these locations, suggesting that domains important in the Arf6 interaction may have been highly conserved.

Summary

Recent advances, especially from real-time imaging approaches, have greatly increased our understanding of the stages involved in the endocytic process in yeast. The role of actin in endocytosis appears to be conserved across evolution, though the generality of actin's role in mammalian endocytosis is still not completely clear. Our studies have now progressed to investigating additional components of the endocytic machinery to determine the importance of recruitment of coat proteins and the role of a plasma membrane Arf protein.

Cell Architecture: from Structure to Function: A Focus Topic at BioScience2005, held at SECC Glasgow, U.K., 17–21 July 2005. Edited by S. Cockroft (University College London, U.K.), Y. Goda (University College London, U.K.), R. Insall (Birmingham, U.K.) and M. Wakelam (Birmingham, U.K.).

Abbreviations

     
  • Arf

    ADP-ribosylation factor

  •  
  • GAT

    GGAs and TOM1

  •  
  • GFP

    green fluorescent protein

  •  
  • GGA

    Golgi-localizing, γ-adaptin ear homology domain, Arf-binding

  •  
  • NPF

    Asn-Pro-Phe

  •  
  • VHS

    Vps27p/Hrs/STAM

  •  
  • WASP

    Wiskott Aldrich syndrome protein

K.R.A. is supported by a Medical Research Council (MRC) Senior Non-clinical Research Fellowship (G117/394). R.C. is supported by a Biotechnology and Biological Sciences Research Council grant (BB/C510091/1) to K.R.A.

References

References
1
Benedetti
H.
Raths
S.
Crausaz
F.
Riezman
H.
Mol. Biol. Cell
1994
, vol. 
5
 (pg. 
1023
-
1037
)
2
Geli
M.I.
Riezman
H.
J. Cell Sci.
1998
, vol. 
111
 (pg. 
1031
-
1037
)
3
Raths
S.
Rohrer
J.
Crausaz
F.
Riezman
H.
J. Cell Biol.
1993
, vol. 
120
 (pg. 
55
-
65
)
4
Engvist-Goldstein
A.E.
Drubin
D.G.
Annu. Rev. Cell Dev. Biol.
2003
, vol. 
19
 (pg. 
287
-
332
)
5
Shurety
W.
Bright
N.A.
Luzio
J.P.
J. Cell Sci.
1996
, vol. 
109
 (pg. 
2927
-
2935
)
6
Engqvist-Goldstein
A.E.Y.
Warren
R.A.
Kessels
M.M.
Keen
J.H.
Heuser
J.
Drubin
D.G.
J. Cell Biol.
2001
, vol. 
154
 (pg. 
1209
-
1223
)
7
Kessels
M.M.
Engqvist-Goldstein
A.E.Y.
Drubin
D.G.
Qualmann
B.
J. Cell Biol.
2001
, vol. 
153
 (pg. 
351
-
366
)
8
Merrifield
C.J.
Trends Cell Biol.
2004
, vol. 
14
 (pg. 
352
-
358
)
9
Merrifield
C.J.
Feldman
M.E.
Wan
L.
Almers
W.
Nat. Cell Biol.
2002
, vol. 
4
 (pg. 
691
-
698
)
10
Merrifield
C.J.
Perrais
D.
Zenisek
D.
Cell (Cambridge, Mass.)
2005
, vol. 
121
 (pg. 
593
-
606
)
11
Ayscough
K.R.
Eby
J.J.
Lila
T.
Dewar
H.
Kozminski
K.G.
Drubin
D.G.
Mol. Biol. Cell
1999
, vol. 
10
 (pg. 
1061
-
1075
)
12
Gourlay
C.W.
Dewar
H.
Warren
D.T.
Costa
R.
Satish
N.
Ayscough
K.R.
J. Cell Sci.
2003
, vol. 
116
 (pg. 
2551
-
2564
)
13
Warren
D.T.
Andrews
P.D.
Gourlay
C.G.
Ayscough
K.R.
J. Cell Sci.
2002
, vol. 
115
 (pg. 
1703
-
1715
)
14
Brett
T.
Traub
L.
Fremont
D.
Structure
2002
, vol. 
10
 (pg. 
797
-
809
)
15
Lehtonen
S.
Zhao
F.
Lehtonen
E.
Am. J. Physiol. Renal Physiol.
2002
, vol. 
283
 (pg. 
F734
-
F743
)
16
Tang
H.Y.
Xu
J.
Cai
M.J.
Mol. Cell. Biol.
2000
, vol. 
20
 (pg. 
12
-
25
)
17
Kaksonen
M.
Sun
Y.
Drubin
D.G.
Cell (Cambridge, Mass.)
2003
, vol. 
115
 (pg. 
475
-
487
)
18
Zeng
G.H.
Yu
X.W.
Cai
M.J.
Mol. Biol. Cell
2001
, vol. 
12
 (pg. 
3759
-
3772
)
19
Madania
A.
Dumoulin
P.
Grava
S.
Kitamoto
H.
Scharer-Brodbeck
C.
Soulard
A.
Moreau
V.
Winsor
B.
Mol. Biol. Cell
1999
, vol. 
10
 (pg. 
3521
-
3538
)
20
Confalonieri
S.
Di Fiore
P.
FEBS Lett.
2002
, vol. 
513
 (pg. 
24
-
29
)
21
Boman
A.L.
J. Cell Sci.
2001
, vol. 
114
 (pg. 
3413
-
3418
)
22
Hirst
J.
Lindsay
M.R.
Robinson
M.S.
Mol. Biol. Cell
2001
, vol. 
12
 (pg. 
3573
-
3588
)
23
Zhdankina
O.
Strand
N.L.
Redmond
J.M.
Boman
A.L.
Yeast
2001
, vol. 
18
 (pg. 
1
-
18
)
24
Dewar
H.
Warren
D.T.
Gardiner
F.C.
Gourlay
C.G.
Satish
N.
Richardson
M.R.
Andrews
P.D.
Ayscough
K.R.
Mol. Biol. Cell
2002
, vol. 
13
 (pg. 
3646
-
3661
)
25
Costa
R.
Warren
D.T.
Ayscough
K.R.
Biochem. J.
2005
, vol. 
387
 (pg. 
649
-
658
)
26
Huang
C.F.
Liu
Y.W.
Tung
L.
Lin
C.H.
Lee
F.J.
Mol. Biol. Cell
2003
, vol. 
14
 (pg. 
3834
-
3847
)
27
Donaldson
J.G.
J. Biol. Chem.
2003
, vol. 
278
 (pg. 
41573
-
41576
)
28
Puertollano
R.
Bonifacino
J.S.
Nat. Cell Biol.
2004
, vol. 
6
 (pg. 
244
-
251
)
29
Scott
P.M.
Bilodeau
P.S.
Zhdankina
O.
Winistorfer
S.C.
Hauglund
M.J.
Allaman
M.M.
Kearney
W.R.
Robertson
A.D.
Boman
A.L.
Piper
R.C.
Nat. Cell Biol.
2004
, vol. 
6
 (pg. 
252
-
259
)