Kinases in clathrin-mediated endocytosis.

The process of clathrin-coated vesicle (CCV) formation/disassembly has been studied intensively, and numerous proteins have been identified which aid this process. We have learnt a great deal about individual components of the CCV machinery, and now the ultimate aim is to elucidate the mechanisms regulating clathrin-mediated trafficking. One of the fundamental processes governing the complicated network of interactions is phosphorylation. It has been known for some time that several proteins associated with clathrin-coated vesicles are substrates for protein kinases. These proteins include clathrin, three of the four adaptor complex subunits, dynamin 1, synaptojanin 1 and the amphiphysins. However, the identities of the kinases involved in this process remained largely unknown until recently. This short review discusses advances in our knowledge of how CCV formation/disassembly is regulated by the phosphorylation/dephosphorylation cycle and the role played by specific protein kinases in that process.

Modular phosphoinositide-binding domains--their role in signalling and membrane trafficking.

The membrane phospholipid phosphatidylinositol is the precursor of a family of lipid second-messengers, known as phosphoinositides, which differ in the phosphorylation status of their inositol group. A major advance in understanding phosphoinositide signalling has been the identification of a number of highly conserved modular protein domains whose function appears to be to bind various phosphoinositides. Such 'cut and paste' modules are found in a diverse array of multidomain proteins and recruit their host protein to specific regions in cells via interactions with phosphoinositides. Here, with particular reference to proteins involved in membrane traffic pathways, we discuss recent advances in our understanding of phosphoinositide-binding domains.

A PtdIns(3)P-specific probe cycles on and off host cell membranes during Salmonella invasion of mammalian cells.

Salmonella invade nonphagocytic cells by eliciting their own internalization; upon contact with the host cell, the bacteria induce membrane ruffles highly localized to the point of contact between the invading bacterium and the host cell. The bacterium is then internalized into an unusual cytosolic organelle, the Salmonella-containing vacuole (SCV). Early endosomal markers (including EEA1) have recently been shown to be associated with the SCV shortly after invasion. EEA1, a protein involved in early endosome fusion, is recruited to early endosomal membranes in part by the interaction between its FYVE finger and phosphatidylinositol 3-phosphate [PtdIns(3)P], a characteristic lipid of early endosomes. This suggests a possible role for PtdIns(3)P during Salmonella infection. To investigate this, we generated a highly specific probe for PtdIns(3)P that was used to follow invasion of Salmonella in nonphagocytic cells. Here, we show that PtdIns(3)P is present on the membranes of SCVs shortly after invasion and also that it is present on the membrane ruffles produced immediately prior to invasion. We also show that this specific probe cycles on and off the membranes of nascent SCVs even when PtdIns 3-kinase activity is inhibited, demonstrating that invading Salmonella influence the composition of the membranes that envelop them during invasion.

Analysis of the cat eye syndrome critical region in humans and the region of conserved synteny in mice: a search for candidate genes at or near the human chromosome 22 pericentromere.

We have sequenced a 1.1-Mb region of human chromosome 22q containing the dosage-sensitive gene(s) responsible for cat eye syndrome (CES) as well as the 450-kb homologous region on mouse chromosome 6. Fourteen putative genes were identified within or adjacent to the human CES critical region (CESCR), including three known genes (IL-17R, ATP6E, and BID) and nine novel genes, based on EST identity. Two putative genes (CECR3 and CECR9) were identified, in the absence of EST hits, by comparing segments of human and mouse genomic sequence around two solitary amplified exons, thus showing the utility of comparative genomic sequence analysis in identifying transcripts. Of the 14 genes, 10 were confirmed to be present in the mouse genomic sequence in the same order and orientation as in human. Absent from the mouse region of conserved synteny are CECR1, a promising CES candidate gene from the center of the contig, neighboring CECR4, and CECR7 and CECR8, which are located in the gene-poor proximal 400 kb of the contig. This latter proximal region, located approximately 1 Mb from the centromere, shows abundant duplicated gene fragments typical of pericentromeric DNA. The margin of this region also delineates the boundary of conserved synteny between the CESCR and mouse chromosome 6. Because the proximal CESCR appears abundant in duplicated segments and, therefore, is likely to be gene poor, we consider the putative genes identified in the distal CESCR to represent the majority of candidate genes for involvement in CES.

Effects of elevated expression of inositol 1,4,5-trisphosphate 3-kinase B on Ca2+ homoeostasis in HeLa cells.

Ins(1,4,5)P(3) 3-kinase (IP3K) phosphorylates the Ca(2+)-mobilizing second messenger Ins(1,4,5)P(3) to yield the putative second messenger Ins(1,3,4,5)P(4). A HeLa cell line was established expressing the rat B isoform of IP3K under the control of an inducible promoter. The IP3KB-transfected cell line possessed 23-fold greater IP3K activity than untransfected cells after induction of IP3KB expression, but only 0.23-fold greater activity when IP3KB expression was not induced. Elevating IP3KB expression significantly reduced levels of Ins(1,4,5)P(3) and increased levels of Ins(1,3,4,5)P(4) after stimulation of cells with histamine, but had no effect on basal levels. Histamine- and ATP-evoked cytosolic Ca(2+) responses were dramatically reduced upon elevation of IP3KB expression. On stimulation with a supramaximal dose of histamine, 67% of cells induced to express IP3KB gave no detectable elevation in cytosolic Ca(2+), compared with 3% of uninduced cells. The quantity of Ca(2+) within thapsigargin-sensitive and -insensitive stores was unaffected by elevation of IP3KB expression, as was capacitative Ca(2+) entry. These data suggest that IP3KB may play a significant role in the regulation of Ins(1,4,5)P(3) levels, and consequently in Ca(2+) responses following stimulation of cells with Ins(1,4,5)P(3)-elevating agonists.

GAP1IP4BP contains a novel group I pleckstrin homology domain that directs constitutive plasma membrane association.

The group I family of pleckstrin homology (PH) domains are characterized by their inherent ability to specifically bind phosphatidylinositol 3,4,5-trisphosphate (PtdIns(3,4,5)P(3)) and its corresponding inositol head-group inositol 1,3,4,5-tetrakisphosphate (Ins(1,3,4,5)P(4)). In vivo this interaction results in the regulated plasma membrane recruitment of cytosolic group I PH domain-containing proteins following agonist-stimulated PtdIns(3,4,5)P(3) production. Among group I PH domain-containing proteins, the Ras GTPase-activating protein GAP1(IP4BP) is unique in being constitutively associated with the plasma membrane. Here we show that, although the GAP1(IP4BP) PH domain interacts with PtdIns(3,4, 5)P(3), it also binds, with a comparable affinity, phosphatidylinositol 4,5-bisphosphate (PtdIns(4,5)P(2)) (K(d) values of 0.5 +/- 0.2 and 0.8 +/- 0.5 microm, respectively). Intriguingly, whereas this binding site overlaps with that for Ins(1,3,4,5)P(4), consistent with the constitutive plasma membrane association of GAP1(IP4BP) resulting from its PH domain-binding PtdIns(4,5)P(2), we show that in vivo depletion of PtdIns(4,5)P(2), but not PtdIns(3,4,5)P(3), results in dissociation of GAP1(IP4BP) from this membrane. Thus, the Ins(1,3,4,5)P(4)-binding PH domain from GAP1(IP4BP) defines a novel class of group I PH domains that constitutively targets the protein to the plasma membrane and may allow GAP1(IP4BP) to be regulated in vivo by Ins(1,3,4,5)P(4) rather than PtdIns(3,4,5)P(3).

Effective formation of major histocompatibility complex class II-peptide complexes from endogenous antigen by thyroid epithelial cells.

In autoimmune thyroid disease, thyroid epithelial cells (TEC) express major histocompatibility complex (MHC) class II molecules, potentially enabling them to present thyroid self-antigens to CD4-positive T cells. However, despite this, TEC may fail to present endogenous antigen as a result of limited processing or MHC class II loading capacity, or inadequate MHC class II levels. We addressed these issues using the cloned rat TEC line, Fischer rat thyroid cell line (FRTL5), which was transfected using an adenoviral expression vector that expressed ovalbumin (OVA) as an integral membrane protein. OVA-expressing FRTL5 cells very efficiently activated a panel of OVA-specific, class II-restricted T-cell hybridomas. This response was dependent on induction of MHC class II molecules by interferon-gamma (IFN-gamma) and was blocked by anti-MHC class II antibodies. Poor responses were seen to exogenously added OVA or OVA peptides. These results provide the most direct evidence to date that TEC can form MHC class II-peptide complexes derived from self-antigen in sufficient quantities to activate T cells.

A study of the coregulation and tissue specificity of XG and MIC2 gene expression in eukaryotic cells.

CD99, the product of the MIC2 gene, exhibits an erythroid-specific quantitative polymorphism coregulated with the polymorphism of the XG blood group gene. As a preliminary study of this phenomenon, human XG and CD99 recombinant proteins were expressed in murine RAG cells and analyzed by flow cytometry. Both proteins were expressed independently and at a similar level in single and double transfectants. Immunoprecipitation and Western blot analysis, using the murine monoclonal antibodies NBL-1 and 12E7, revealed species of 26 kd (XG) and 32 kd (CD99), respectively. A putative 28-kd intracellular precursor of CD99 was also detected, as was a 26-kd species after neuraminidase treatment of CD99-expressing cells. No evidence of association or complex formation between XG and CD99 proteins could be proven, either on transfected RAG cells or on human erythrocytes. These results were confirmed using somatic hybrids between single transfectants. These findings suggest that the phenotypic relationship between XG and CD99 is mostly regulated at the transcriptional level, but they do not formally exclude some posttranscriptional effect. Studies on the tissue specificity of XG expression showed that surface expression of the XG protein could not be restored in somatic hybrids between B-lymphoblastoid cell lines from Xg(a+) persons and fibroblasts (RAG) or erythroid (MEL) cells. RT-PCR analysis of the transcripts revealed the existence of an XG mRNA in each cell line, suggesting that the tissue-specific regulation of cell surface XG expression occurs either at a quantitative transcriptional level or is a posttranscriptional event. By Northern blot analysis, XG transcripts were detected in erythroid tissues and several nonerythroid tissues. (Blood. 2000;95:1819-1826)

In vivo dynamics of the F-actin-binding protein neurabin-II.

Neurabin-II (spinophilin) is a ubiquitously expressed F-actin-binding protein containing an N-terminal actin-binding domain, a PDZ (PSD95/discs large/ZO-1) domain and a C-terminal domain predicted to form a coiled-coil structure. We have stably expressed a green fluorescent protein (GFP)-tagged version of neurabin-II in PC12 cells, and characterized the in vivo dynamics of this actin-binding protein using confocal fluorescence microscopy. We show that GFP-neurabin-II localizes to actin filaments, especially at cortical sites and areas underlying sites of active membrane remodelling. GFP-neurabin-II labels only a subset of F-actin within these cells, as indicated by rhodamine-phalloidin staining. Both actin filaments and small, highly motile structures within the cell body are seen. Photobleaching experiments show that GFP-neurabin-II also exhibits highly dynamic behaviour when bound to actin filaments. Latrunculin B treatment results in rapid relocalization of GFP-neurabin-II to the cytosol, whereas cytochalasin D treatment causes the collapse of GFP-neurabin-II fluorescence to intensely fluorescent foci of F-actin within the cell body. This collapse is reversed on cytochalasin D removal, recovery from which is greatly accelerated by stimulation of cells with epidermal growth factor (EGF). Furthermore, we show that this EGF-induced relocalization of GFP-neurabin-II is dependent on the activity of the small GTPase Rac1 but not the activity of ADP-ribosylation factor 6.

The use of yeast two-hybrid screens in studies of protein:protein interactions involved in trafficking.

The yeast two-hybrid system has provided a convenient means to both screen for proteins that interact with a protein of interest and to characterise the known interaction between two proteins. Several groups with an interest in the molecular mechanisms that underlie discrete steps along trafficking pathways have exploited the yeast two-hybrid system. Here, we provide a brief background to the technology, attempt to point out some of the pitfalls and benefits of the different systems that can be employed, and mention some of the areas (within the trafficking field) where yeast two-hybrid interaction assays have been particularly informative.

Direct interaction of the trans-Golgi network membrane protein, TGN38, with the F-actin binding protein, neurabin.

TGN38 is a type I integral membrane protein that constitutively cycles between the trans-Golgi network (TGN) and plasma membrane. The cytosolic domain of TGN38 interacts with AP2 clathrin adaptor complexes via the tyrosine-containing motif (-SDYQRL-) to direct internalization from the plasma membrane. This motif has previously been shown to direct both internalization and subsequent TGN targeting of TGN38. We have used the cytosolic domain of TGN38 in a two-hybrid screen, and we have identified the brain-specific F-actin binding protein neurabin-I as a TGN38-binding protein. We demonstrate a direct interaction between TGN38 and the ubiquitous homologue of neurabin-I, neurabin-II (also called spinophilin). We have used a combination of yeast two-hybrid and in vitro protein interaction assays to show that this interaction is dependent on the serine (but not tyrosine) residue of the known TGN38 trafficking motif. We show that TGN38 interacts with the coiled coil region of neurabin in vitro and binds preferentially with the dimeric form of neurabin. TGN38 and neurabin also interact in vivo as demonstrated by coimmunoprecipitation from stably transfected PC12 cells. These data suggest that neurabin provides a direct physical link between TGN38-containing membranes and the actin cytoskeleton.

Phosphorylation of the medium chain subunit of the AP-2 adaptor complex does not influence its interaction with the tyrosine based internalisation motif of TGN38.

Tyrosine based motifs conforming to the consensus YXXphi (where phi represents a bulky hydrophobic residue) have been shown to interact with the medium chain subunit of clathrin adaptor complexes. These medium chains are targets for phosphorylation by a kinase activity associated with clathrin coated vesicles. We have used the clathrin coated vesicle associated kinase activity to specifically phosphorylate a soluble recombinant fusion protein of mu2, the medium chain subunit of the plasma membrane associated adaptor protein complex AP-2. We have tested whether this phosphorylation has any effect on the interaction of mu2 with the tyrosine based motif containing protein, TGN38, that has previously been shown to interact with mu2. Phosphorylation of mu2 was shown to have no significant effect on the in vitro interaction of mu2 with the cytosolic domain of TGN38, indicating that reversible phosphorylation of mu2 does not play a role in regulating its direct interaction with tyrosine based internalisation motifs. In addition, although a casein kinase II-like activity has been shown to be associated with clathrin coated vesicles, we show that mu2 is not phosphorylated by casein kinase II implying that another kinase activity is present in clathrin coated vesicles. Furthermore the kinase activity associated with clathrin coated vesicles was shown to be capable of phosphorylating dynamin 1. Phosphorylation of dynamin 1 has previously been shown to regulate its interaction with other proteins involved in clathrin mediated endocytosis.

TGN38 cycles via the basolateral membrane of polarized Caco-2 cells.

TGN38 is a heavily glycosylated, type I integral membrane protein which is predominantly localized to the trans Golgi network (TGN), but which constitutively traffics between the TGN and the cell surface. The trafficking of TGN38 has been extensively studied in non-polarized cells, and a short, tyrosine-based, peptide motif within the cytosolic domain of the protein has been shown to be necessary and sufficient for its rapid internalization from the cell surface and efficient delivery to the TGN. Such tyrosine-based motifs have also been shown to act as basolateral targeting signals, whilst N-linked glycans (as occur on the extracytosolic domain of TGN38) can act as apical targeting signals. TGN38 has previously been shown to be sorted to the basolateral surface of polarized canine MDCK cells; a polarized cell line in which biosynthetic sorting decisions concerning the eventual destination of apical or basolateral targeted plasma membrane proteins are made at the TGN. We now show that TGN38 is targeted exclusively to the basolateral domain of polarized human Caco-2 cells, a cell line in which newly synthesized membrane proteins destined for either the apical or basolateral plasma membrane may be sorted for delivery to their final destination either at the TGN or at the cell surface. These data also demonstrate that the heavily glycosylated, extracytosolic domain of TGN38 does not contain a dominant apical targeting signal.

The steady state distribution of humTGN46 is not significantly altered in cells defective in clathrin-mediated endocytosis.

It has been shown previously that whilst the rat type I integral membrane protein TGN38 (ratTGN38) is predominantly localised to the trans-Golgi network this protein does reach the cell surface from where it is internalised and delivered back to the trans-Golgi network. This protein thus provides a suitable tool for the investigation of trafficking pathways between the trans-Golgi network and the cell surface and back again. The human orthologue of ratTGN38, humTGN46, behaves in a similar fashion. These proteins are internalised from the cell surface via clathrin mediated endocytosis, a process which is dependent upon the GTPase activity of dynamin. We thus reasoned that humTGN46 would accumulate at the surface of cells rendered defective in clathrin mediated endocytosis by virtue of the fact that they express a GTPase defective mutant of dynamin I. It did not. In fact, expression of a dominant negative GTPase defective mutant of dynamin I had no detectable effect on the steady state distribution of humTGN46. One explanation for this observation is that humTGN46 does not travel directly to the cell surface from the trans-Golgi network. Further studies on cells expressing the dominant negative GTPase defective mutant of dynamin I indicate that the major recycling pathway for humTGN46 is in fact between the trans-Golgi network and the early endosome.

Specificity of interaction between adaptor-complex medium chains and the tyrosine-based sorting motifs of TGN38 and lgp120.

Multiple sorting steps within eukaryotic cells are mediated by tyrosine-based sorting motifs. Motifs conforming to the consensus -YXXO- (where O indicates a bulky hydrophobic residue) have been shown to specify high-efficiency internalization from the plasma membrane, targeting from the plasma membrane to the trans-Golgi network and targeting to lysosomal compartments as well as being involved in basolateral sorting in polarized cells. These motifs are recognized by the medium-chain subunits of heterotetrameric adaptor complexes. Whereas these motifs have been shown to be sufficient to mediate interaction with the mu-chains, we and others have shown that their context is important in determining the affinity of interaction. In this study we have investigated the interaction between the tyrosine motifs of the type-1 integral membrane proteins TGN38 and lgp120 with medium-chain subunits using the yeast two-hybrid system. Whereas the wild-type version of the cytosolic domain of TGN38 interacts with highest affinity with mu2, we show that the cytosolic domain of lgp120 interacts almost exclusively with mu3A. The specificity of binding of tyrosine-based sorting motifs to mu-chains is shown to be highly sensitive to the context in which the motif lies. For example, the -YQTI- motif of lgp120 is effectively non-functional with regard to mu-chain binding when placed in the context of the TGN38 cytosolic domain. Deletion of four amino acids (NLKL) at the extreme C-terminus of TGN38, leaving the YXXO motif as the C-terminus, greatly enhances the affinity of interaction with mu2. Furthermore, addition of these same residues to the extreme C-terminus of lgp120 effectively abolishes the interaction of the cytosolic domain of lgp120 with mu-chains. We also show that the newly identified mu-adaptin-related protein 2 (mu4) only interacts weakly with tyrosine-based sorting motifs.

Inhibition of the interaction between tyrosine-based motifs and the medium chain subunit of the AP-2 adaptor complex by specific tyrphostins.

Several intracellular membrane trafficking events are mediated by tyrosine-containing motifs found within the cytosolic domains of certain integral membrane proteins. Many of these tyrosine motifs conform to the consensus YXXPhi (where Phi represents a bulky hydrophobic residue). This YXXPhi motif has been shown to interact with the medium chain subunits of adaptor complexes that generally link relevant integral membrane protein cytosolic domains to the clathrin coat involved in vesicle formation. The motif YXXPhi is also very similar to motifs that are targets for phosphorylation by tyrosine kinases. Tyrosine kinase inhibitors known as tyrphostins are structural analogues of tyrosine, and so it is possible that tyrphostins could also inhibit interactions between medium chains and YXXPhi motifs. TGN38 is a type I integral membrane protein containing a tyrosine motif, YQRL, within the cytosolic domain. We have previously shown that this motif interacts directly with the medium chain subunit of the plasma membrane localized AP-2 adaptor complex (mu2). We have investigated a range of tyrphostins and demonstrated a specific inhibition of the interaction between mu2 and the TGN38 cytosolic domain by tyrphostin A23 through in vitro analysis and the yeast two-hybrid system. These data raise the exciting possibility that different membrane traffic events could be inhibited by specific tyrphostins.

Efficient trafficking of TGN38 from the endosome to the trans-Golgi network requires a free hydroxyl group at position 331 in the cytosolic domain.

TGN38 is one of the few known resident integral membrane proteins of the trans-Golgi network (TGN). Since it cycles constitutively between the TGN and the plasma membrane, TGN38 is ideally suited as a model protein for the identification of post-Golgi trafficking motifs. Several studies, employing chimeric constructs to detect such motifs within the cytosolic domain of TGN38, have identified the sequence 333YQRL336 as an autonomous signal capable of localizing reporter proteins to the TGN. In addition, one group has found that an upstream serine residue, S331, may also play a role in TGN38 localization. However, the nature and degree of participation of S331 in the localization of TGN38 remain uncertain, and the effect has been studied in chimeric constructs only. Here we investigate the role of S331 in the context of full-length TGN38. Mutations that abolish the hydroxyl moiety at position 331 (A, D, and E) lead to missorting of endocytosed TGN38 to the lysosome. Conversely, mutation of S331 to T has little effect on the endocytic trafficking of TGN38. Together, these findings indicate that the S331 hydroxyl group has a direct or indirect effect on the ability of the cytosolic tail of TGN38 to interact with trafficking and/or sorting machinery at the level of the early endosome. In addition, mutation of S331 to either A or D results in increased levels of TGN38 at the cell surface. The results confirm that S331 plays a critical role in the intracellular trafficking of TGN38 and further reveal that TGN38 undergoes a signal-mediated trafficking step at the level of the endosome.

Lumenal and transmembrane domains play a role in sorting type I membrane proteins on endocytic pathways.

Previous studies have shown that when the cytosolic domains of the type I membrane proteins TGN38 and lysosomal glycoprotein 120 (lgp120) are added to a variety of reporter molecules, the resultant chimeric molecules are localized to the trans-Golgi network (TGN) and to lysosomes, respectively. In the present study we expressed chimeric constructs of rat TGN38 and rat lgp120 in HeLa cells. We found that targeting information in the cytosolic domain of TGN38 could be overridden by the presence of the lumenal and transmembrane domains of lgp120. In contrast, the presence of the transmembrane and cytosolic domains of TGN38 was sufficient to deliver the lumenal domain of lgp120 to the trans-Golgi network. On the basis of steady-state localization of the various chimeras and antibody uptake experiments, we propose that there is a hierarchy of targeting information in each molecule contributing to sorting within the endocytic pathway. The lumenal and cytosolic domains of lgp120 contribute to sorting and delivery to lysosomes, whereas the transmembrane and cytosolic domains of TGN38 contribute to sorting and delivery to the trans-Golgi network.