© 2011 John Wiley & Sons A/S
Traffic 2012; 13: 355–363
doi:10.1111/j.1600-0854.2011.01286.x
Review
The Ins and Outs of HIV-1 Tat ` Solene Debaisieux, Fabienne Rayne, Hocine Yezid and Bruno Beaumelle∗ CPBS, UMR 5236 CNRS, Universite´ de Montpellier, 1919 Route de Mende, 34923 Montpellier Cedex 05, France *Corresponding author: Bruno Beaumelle,
[email protected] HIV-1 encodes for the small basic protein Tat (86–101 residues) that drastically enhances the efficiency of viral transcription. The mechanism enabling Tat nuclear import is not yet clear, but studies using reporter proteins fused to the Tat basic domain indicate that Tat could reach the nucleus by passive diffusion. Tat also uses an unusual transcellular transport pathway. The first step of this pathway involves high-affinity binding of Tat to phosphatidylinositol (4,5) bisphosphate (PI(4,5)P2 ), a phospholipid that is concentrated in the inner leaflet of the plasma membrane and enables Tat recruitment at this level. Tat then crosses the plasma membrane to reach the outside medium. Although unconventional, Tat secretion by infected cells is highly active, and export is the major destination for HIV-1 Tat. Secreted Tat can bind to a variety of cell types using several different receptors. Most of them will allow Tat endocytosis. Upon internalization, low endosomal pH triggers a conformational change in Tat that results in membrane insertion. Later steps of Tat translocation to the target-cell cytosol are assisted by Hsp90, a general cytosolic chaperone. Cytosolic Tat can trigger various cell responses. Indeed, accumulating evidence suggests that extracellular Tat acts as a viral toxin that affects the biological activity of different cell types and has a key role in acquired immune-deficiency syndrome development. This review focuses on some of the recently identified molecular details underlying the unusual transcellular transport pathway used by Tat, such as the role of the single Trp in Tat for its membrane insertion and translocation. Key words: endocytosis, exocytosis, HIV-1, membrane insertion, protein translocation, unconventional secretion, Tat, toxin Received 18 March 2011, revised and accepted for publication 19 September 2011, uncorrected manuscript published online 22 September 2011, published online 11 October 2011
The Plasticity of Tat Structure HIV-1 encodes the Tat protein, which is a regulatory protein that enhances viral transcription and replication.
Tat is a small protein whose length varies between 80 and 103 residues depending on viral isolates, with the predominant form being 101 residues long (Figure 1A). Despite HIV-1’s high mutation rate (1), Tat is a relatively well-conserved protein, especially its first 56 residues, indicating that several of these residues have important functional roles in Tat activity and/or structure (Figure 1B). Among the well-preserved residues are the cysteine residues (except for Cys31) that are involved in the co-ordination of Zn2+ binding (2), most of the proline residues, as well as the acidic N-terminal region (residues 1–9) and the basic domain (residues 49–57) (Figure 1B). Most structural studies of the full-length Tat, obtained using two-dimensional nuclear magnetic resonance (NMR) and molecular dynamic calculations, failed to identify a stable conformation in the absence of Tat ligand (3–5). Collectively, these studies indicate that the basic region of the protein (residues 49–57) is well exposed to the solvent, while the single tryptophan of the protein, Trp11, is at least partly buried at the centre of the protein. Mutagenesis and molecular dynamic studies indicated that Tat structure is stabilized by strong electrostatic interactions between the acidic Nterminal domain (especially Asp/Glu 2) and the basic domain (4,6,7). Nevertheless, Tat did not show any major secondary structural organization during circular dichroism analyses, and its secondary structure sharply relies on solvent polarity (5,8). These results indicate that Tat is a flexible and intrinsically disordered protein. But this does not prevent very tight binding, with Kd values in the nanomolar range, to partners such as the transactivation response element (TAR, a short RNA recognized by Tat) and cyclin T1 (9) or phosphatidylinositol (4,5) bisphosphate (PI(4,5)P2 ) (10). Tat probably adopts different conformations upon interaction with these different partners. For instance, the structure of only the first 49 amino acids of Tat could be resolved in a recent study examining the crystal structure of Tat in complex with cyclinT1, CDK9 and an ATP analogue. This indicates that the remainder of the protein is not involved in this interaction (2). Conversely, Tat binding to PI(4,5)P2 essentially relies on residues 49–51 in the basic domain, together with the membrane insertion of Trp11 side chain (11). Hence, Tat can assemble with various partners (proteins or lipids) to form high-affinity complexes each involving a specific conformation of Tat. The capacity of Tat to establish tight interactions, each involving a different and specific conformation of the protein, is probably favoured by Tat structural plasticity that allows its spatial adaptation to its different partners. www.traffic.dk 355
Debaisieux et al. 1000 Number of Tat variants
A
800 600 400 200 0 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103
Number of residues
B
Figure 1: Analysis of Tat primary sequences. The 1303 HIV-1 Tat sequences (2009 update) from Los Alamos National Laboratory database (http://www.hiv.lanl.gov/content/sequence/HIV/mainpage.html) were compiled for the analysis (after alignment and elimination of the gaps). A) Length of HIV-1 Tat depending on viral isolates. B) Sequence conservation. WebLogo (http://weblogo.berkeley.edu/logo.cgi) was used to obtain this representation where letter size is proportional to residue conservation.
Overview of Tat Transcellular Transport and Its Implication in AIDS Quickly after the demonstration that exogenous Tat could enter cells and reach the cytosol from the outside (12), Tat was found to be secreted by infected cells in the absence of cell lysis, presumably by an unconventional secretion mechanism because Tat is devoid of a signal sequence (13). This HIV-1 protein thus displays an unusual trafficking as it can be exported by infected cells, before being captured by various target cell types (Figure 2). This review focuses on recent studies that shed light on the mechanisms involved in Tat secretion, uptake and nuclear targeting.
production of immunoregulatory cytokines by lymphocytes, monocytes and macrophages, thereby contributing to immune suppression (16,17). Extracellular Tat can also be directly cytopathic to CD4+ T cells (18,14), neurons (19) and astrocytes (20). Tat could therefore be involved both in the decline of CD4+ T cells, which is a characteristic of the onset of acquired immune-deficiency syndrome (AIDS), and in HIV-1-associated neurological disorders, which are among the most severe sources of morbidity for AIDS patients under highly active antiretroviral therapy (21). The effects of extracellular Tat on target cells have been reviewed in detail elsewhere (5,16).
Tat Basic Domain as a Protein Vehicle Tat intercellular transport is tightly linked to HIV-1 pathogenesis. Indeed, extracellular Tat acts as a viral toxin that can affect several cell types, triggering various cell responses. For instance, secreted Tat strongly increases the number of the HIV-1 co-receptors CXCR4 and CCR5 at the surface of CD4+ T cells, favouring viral infection (10,16). Extracellular Tat also modulates the 356
The capacity for Tat to enter the cytosol from the outside has been used to import proteins into cells. It was first observed that the full-length Tat or Tat37 – 72 targeted attached proteins into the cytosol (22). The minimum effective sequence able to insure targeting was then identified to be Tat48 – 60 , i.e. Tat basic domain (23).
Traffic 2012; 13: 355–363
The Ins and Outs of HIV-1 Tat Infected T-cell
Target cell
Nucleus
Nucleus
TAT TAT Viral genes
Cellular genes
TAT
TAT
Endosome
In TAT
Out Seric [Tat] = 0.2 - 4 nM TAT Figure 2: Schematic representation of Tat transcellular transport. Infected cells produce Tat that, in addition to its transcriptional activity (resulting from its binding to viral RNA in the nucleus as depicted), is actively secreted directly through the plasma membrane using an unconventional secretion mechanism that relies on the tight interaction between Tat and PI(4,5)P2 . This phosphoinositide is specifically concentrated at the plasma membrane and enables Tat recruitment at this level. It is the first membrane translocation step of Tat (‘Out’). Tat is thereby released in the outside medium where its concentration, measured both in vitro and in vivo, is within the nanomolar range (10,14,15). Circulating Tat can bind to various target cell types, endothelial cells, lymphocytes, monocytes, macrophages, neurons and so on. It can then be endocytosed (not depicted). Once in the endosome, Tat crosses the membrane and reaches the cytosol. This is the second step of Tat transmembrane transport (‘In’). From the cytosol, Tat can, directly or indirectly, modulate the expression of various cellular genes such as several cytokines in monocytes. It can also trigger more deleterious cell responses such as cell death, in T-cells and neurons, for instance (16).
When used to deliver proteins to the cytosol, this domain is often called Tat protein transduction domain (PTD) or Tat cell-penetrating peptide (CPP). Other CPPs have been used successfully, such as penetratin, a 16-residue basic peptide derived from the antennapedia homeodomain, and oligoarginines. The mechanism of cytosolic delivery of proteins attached to Tat-PTD/CPP may involve translocation of the chimera through the plasma membrane or the endosomal membrane (reviewed in Refs. 24 and 25). Some studies indicate that, as shown for Tat (26,27), endosome acidification is required for TatPTD to reach the cytosol of treated cells (28, 29). These results thus suggest that Tat and Tat-PTD could use similar mechanisms to translocate from the endosome lumen to the cytosol. Nevertheless, it should be noted that Tat translocation to the cytosol requires not only the basic domain but also Trp11. Hence, although they show some limited similarities, the molecular mechanisms responsible for Tat and Tat-PTD access to the cytosol appear to be different. Indeed, in the case of the fulllength Tat, an elaborate low pH detection system controls the initiation of translocation. The low pH sensor triggers molecule remodelling upon exposure to endosomal pH, thereby allowing the Trp side chain to insert into
Traffic 2012; 13: 355–363
the membrane, a prerequisite for translocation (7) (see below).
Tat Cytosolic Partners It recently became evident that Tat interacts with several cytosolic partners and that these interactions affect both Tat and cell activities. Tat binds tubulin through its 36–39 residues. This interaction has important functional consequences because Tat alters microtubule dynamics and thereby activates a mitochondria-dependent apoptotic pathway (30). Tat–tubulin interactions also affect the activity of the mitotic spindle, resulting in defects in chromosome congression and sister chromatid segregation (31). This delay in polymerization of microtubules also induces mislocalization of several proteins in Drosophila melanogaster oocyte during development (32). Tat can also interact with protein phosphatase-1 (PP1) via Tat V36 and F38, which are in the tubulin-binding domain, indicating that Tat binding to tubulin or PP1 is mutually exclusive. Tat–PP1 interaction enables nuclear targeting of the phosphatase and its delivery to specific nuclear substrates critical for HIV-1 357
Debaisieux et al.
transcription. Hence, Tat can address PP1 to the nucleus to enhance its transcriptional activity (33). The human inhibitor of the myoD family domain-containing protein (I-mfa) HIC is one of the major Tat cytosolic partners. It sequesters a significant fraction of cellular Tat in the cytosol, thereby controls Tat activity and regulates HIV-1 transcription and replication. HIC was found to interact with several Tat domains: the cysteine and core and basic domains (34). IκB−α is the best characterized and ubiquitously expressed member of the IκB family of nuclear factor-κB (NF-κB)-negative regulators. IκB−α binds to the Tat basic domain. This interaction promotes cytoplasmic sequestration of Tat and inhibition of HIV-1 transcription (in an NF-κB-independent manner) (35). It is therefore clear from these studies that Tat transcriptional activity can be regulated by cytosolic partners that can retain Tat in the cytosol and prevent it from reaching the nucleus.
Tat Nuclear Import Tat is required for HIV-1 transcription and is therefore able to reach the nucleus. When the Tat basic domain was mutated (36, 37), or deleted (38), Tat was not delivered to the nucleus, and transactivation was concomitantly inhibited. Moreover, when the Tat basic domain was attached to beta-galactosidase (β−Gal), the resulting chimeras were efficiently targeted to the nucleus (39, 40). These results indicated that Tat basic domain is both necessary and sufficient for Tat nuclear import, and most mechanistic studies of Tat nuclear import use the Tat basic region attached to β−Gal (41), glutathione S-transferase (GST) (42) or green fluorescent protein (GFP) (43–45) in different experimental systems. These differences might explain why contrasting results were obtained. The first in vitro study, using mechanically perforated cells and Tat48 – 60 -β−Gal, showed that nuclear import of this 476 kDa chimera was dependent on ATP or GTP hydrolysis but was inhibited by the presence of cytosol. Moreover, nuclear accumulation of Tat48 – 60 -β−Gal was not affected by the presence of 200-fold excess of a β−Gal chimera prepared using the nuclear localization signal (NLS) of SV40 large tumour antigen (NLSSV40 -β−Gal), indicating that Tat nuclear import does not rely on importin α (41). The second study confirmed these points by showing that Tat49 – 60 binds directly to importin β. The latter proved to be both necessary and sufficient for Tat49 – 60 -GST (27 kDa) nuclear import when assayed using permeabilized cells and purified import factors, but no cytosol (42). A third group performed fluorescence recovery after photobleaching (FRAP) experiments on live cells and found that Tat47 – 57 -EGFP (27 kDa) entered the nucleus by passive diffusion and not using an energydependent pathway as it was the case for NLSSV40 -EGFP. The size of the construct was also found to be critical for import because Tat47 – 57 -EGFP4 (110 kDa) was excluded from the nucleus (43). This live cell study indicates that 358
both cytosolic factors and protein size regulate Tat47 – 57 nuclear import. Although Tat47 – 57 -EGFP bound to either GST-importin α or GST-importin β when purified proteins were used, this interaction was not observed within cells according to fluorescence lifetime imaging (FLIM) experiments (45). This absence of significant Tat47 – 57 interaction with importins in cellulo is consistent with a passive diffusion mechanism for Tat47 – 57 access to the nucleus. So far, this conclusion is valid for the Tat basic domain only and, to our knowledge, there has not yet been a study of Tat nuclear import using the full-length protein. Because cytosolic or nuclear partners of Tat are known to regulate Tat nuclear targeting and Tat nucleocytoplasmic distribution (34, 35), results obtained using the Tat basic domain should be extrapolated with care to the full-length Tat.
Tat Role in HIV-1 Transcription Tat has an essential role in viral transcription. In the absence of Tat, a short RNA, TAR, is produced from the HIV-1 long terminal repeat (LTR). Tat, cyclin T1 and CDK9 form a ternary complex that binds TAR and then induces the phosphorylation of RNA polymerase II, releasing pausing and enabling the production of full-length HIV1 transcripts. As detailed elsewhere, Tat additionally interacts with several other transcription factors and co-activators to ensure robust transcription from HIV-1 LTR (2,46).
Tat Targeting to the Plasma Membrane Tat was recently found to bind tightly to membraneembedded PI(4,5)P2 (10). To establish this interaction, Tat recognizes PI(4,5)P2 headgroups through a basic residue triplet (residue 49–51). This is reminiscent of other PI(4,5)P2 –protein interactions (47), but PI(4,5)P2 –Tat interaction is unusual in that phosphoinositide binding triggers conformational changes that enable the membrane insertion of Tat Trp11. In fact, Tat does not significantly bind PI(4,5)P2 in the absence of a phospholipid membrane. Moreover, the study of the interaction of Tat mutants with PI(4,5)P2 -containing membranes showed that membrane insertion of the tryptophan side chain is strictly required for high-affinity binding of Tat to PI(4,5)P2 . For instance, the conservative mutation W11Y that preserved the native conformation of the protein (7) decreases 30-fold Tat affinity for PI(4,5)P2 liposomes (10). Moreover, a Tat peptide (Tat42 – 61 ) encompassing the basic domain (Tat49 – 57 ) binds PI(4,5)P2 liposomes very weakly (Figure 3). Tat shows a Kd < 0.3 nM for immobilized PI(4,5)P2 liposomes according to surface plasmon resonance (SPR) measurements (10). This is approximately 20-fold lower than that of the PH domain of phospholipase Cδ (PHPLCδ ), although the latter is one of the strongest cellular ligands for PI(4,5)P2
Traffic 2012; 13: 355–363
The Ins and Outs of HIV-1 Tat Epsin ENTH AP180 ENTH PH-PLCd HIV-1 Tat Tat(42-61) 0.1
1
10 Kd (nM)
100
1000
Figure 3: Affinity of Tat and cellular proteins for PI(4,5)P2 . Liposomes contained 3–5% PI(4,5)P2 and were immobilized on Biacore chips for surface plasmon resonance (SPR) experiments. Results from the studies mentioned were used for Tat, Tat42 – 61 (10), PHPLCδ (48), AP180 N-terminal homology (ANTH) and epsin N-terminal homology (ENTH) (49). Changes in PI(4,5)P2 concentration in the 3–10% range only marginally affected protein affinities (49).
(Figure 3). This superior affinity enables Tat to displace PHPLCδ -EGFP from the cell membrane (10) and suggests that Tat could interfere with cell activities relying on a machinery that involves proteins recruited at the plasma membrane in a PI(4,5)P2 -dependent manner. This includes key activities of immune cells such as AP-2/clathrinmediated endocytosis, phagocytosis, polarized secretion and cytokinesis (50–52). For instance, cytosolic Tat could prevent further Tat endocytosis, thereby regulating its own uptake by the AP-2/clathrin pathway. Another HIV-1 protein, Gag, binds PI(4,5)P2 through its N-terminal matrix domain (53). This interaction enables Gag recruitment at the plasma membrane and is required for HIV-1 particle assembly and budding at this level (54). Gag and Tat likely compete for PI(4,5)P2 in the infected cell, but it is not clear whether Tat could regulate budding or if, conversely, Gag can affect Tat secretion.
Is Tat Essentially Nuclear or Membrane-Bound? The answer is ‘it depends on the cell type’. In primary T cells, either HIV-1-infected or Tat-transfected Tat was found to concentrate at the plasma membrane using both immunofluorescence and cell fractionation techniques. It was also detected in the nuclear fraction. The use of Tat mutants unable to bind PI(4,5)P2 (i.e. Tat(49–51)A or Tat-W11Y), as well as different treatments that mask or deplete PI(4,5)P2 showed that this phosphoinositide is responsible for Tat targeting at the plasma membrane (10). Such a strong Tat affinity for the plasma membrane and a concomitant weak nuclear concentration in primary T-cells were clearly unexpected. Nevertheless, at least two types of evidence indicate that Tat accumulation into the nucleus is not required for its transcriptional activity. First, the affinity constants involved in the assembly of Tat-TAR-cyclinT1 transcriptional complex in solution are subnanomolar (9)
Traffic 2012; 13: 355–363
and therefore one or two log lower than the Kd of Tat–PI(4,5)P2 interaction (16 nM) obtained under equivalent conditions, i.e. using free liposomes and isothermal titration calorimetry (10). Hence, the fact that Tat accumulates at the plasma membrane of HIV-1-infected primary T cells indicates that enough Tat is present within HIV-1 transcriptional complexes, enabling robust transactivation and viral multiplication. Second, preventing Tat accumulation within the nucleus of Jurkat T cells by attaching a Rev nuclear export signal poorly affected transactivation (10). Tat is also membrane bound in other cell types such as macrophages (unpublished data). It is not yet clear why Tat concentrates in the nuclei of transformed cell lines such as Jurkat T cells (10, 55) or HeLa cells (35, 56). It seems likely that transformation, which results in a dramatic enhancement of the expression of various importins (57, 58), is involved in the nuclear accumulation of Tat.
Unconventional Protein Secretion Most secreted proteins contain a conventional signal peptide to target them to the endoplasmic reticulum (ER)/Golgi secretion apparatus. However, some proteins that are devoid of signal peptide are secreted, indicating that these proteins use alternative, unconventional or non-classical secretion pathways (59). Examples of such proteins include interleukin 1β that transits via late endosomes–lysosomes on its way out (60), FGF-2 that is exported directly through the plasma membrane in a PI(4,5)P2 -dependent manner (61), HTLV-1 Tax that is secreted via the secretory pathway (62) and engrailed homeodomain that is associated with vesicular compartments for secretion (63). Hence, unconventional secretion does not necessarily mean direct export through the plasma membrane. Other proteins secreted using unconventional pathways include galectin-1, FGF-1 and HMGB-1 (59). Interestingly, the secretion of many of these proteins is regulated by caspase-1, indicating that cells can regulate the unconventional secretion pathways (64). This has also been shown for engrailed and FGF-2 secretion, which are regulated by phosphorylation (63,65). Unconventional protein secretion is the focus of recent reviews (59, 66).
Tat Unconventional Secretion The capacity of Tat to be secreted by infected cells in the absence of cell lysis was first observed in 1990 (13). These pioneer results showed that Tat is secreted through an unconventional pathway (66). Tat concentration in infected T-cell supernatants reached the level observed in sera of HIV-1-infected individuals, i.e. the nanomolar range (14, 15) (Figure 2). These measurements are consistent with the reported efficiency of Tat secretion. Indeed, despite being unconventional, Tat secretion is highly active, and 359
Debaisieux et al.
around 2/3 of cellular Tat is exported during the 2.2 day life span of an infected primary T cell (10). Moreover, secreted Tat is biologically active in a variety of assays such as the growth of Kaposi sarcoma cells (13) and LTR transcellular activation (10, 67). Exogenous Tat can also affect the physiology of several target cell types as depicted in Figure 2 and reviewed elsewhere (5,16). Altogether, these data indicate that Tat secretion, although observed in vitro, is a biologically relevant process. The Tat secretion mechanism was first studied using pharmacological approaches in transfected COS-1 cells. Tat export was found to be temperature-dependent and insensitive to weak bases or brefeldin A, indicating that secreted Tat does not transit through endosomes (similar to interleukin 1β)) (60), or the ER/Golgi apparatus (similar to HTLV-1 Tax) (68). Comparable data were obtained later using similar inhibitors in Jurkat T cells. The use of a glycosylatable Tat variant confirmed that secreted Tat does not transit through the ER during export, and the insensitivity of Tat secretion to cytoskeletal poisons such as cytochalasin D, nocodazole or taxol indicated that Tat release does not rely on vesicular transport (11). Tat–PI(4,5)P2 interaction is essential for Tat recruitment at the plasma membrane and secretion. Indeed, mutants such as Tat-W11Y and Tat-(49-51)A that do not bind PI(4,5)P2 remain cytosolic and are weakly secreted (10). The Tat secretion/translocation process has not been characterized further than the initial membrane insertion step enabled by Trp11. FGF-2 secretion also relies on PI(4,5)P2 binding. Nevertheless, FGF-2 does not require a membrane to bind PI(4,5)P2 and, compared to Tat, FGF-2 shows a lower specificity and affinity for PI(4,5)P2 (61). Hence, the mechanisms of PI(4,5)P2 recognition and membrane translocation are probably different for Tat and FGF-2.
Tat Receptors and Uptake Early papers showed that extracellular Tat can enter various cell types and reach their nucleus to transactivate reporter genes (12, 69) (Figure 2). It was then shown that Tat is a promiscuous ligand that has a number of binding sites at the cell surface, among which are receptors such as dipeptidyl aminopeptidase IV (CD26) (70), the lipoprotein receptor-related protein (LRP) (71), CXCR4 (15, 72) and heparan sulphate proteoglycans (HSPGs) (73). Tat can also bind the integrin receptors α5 β1 and αV β3 through its Arg-Gly-Asp (RGD) motif (residues 78–80), but this sequence is not present in all HIV-1 isolates (Figure 1B). This large panel of receptors for such a small protein again reflects the molecular flexibility of Tat and its capacity to undergo ligand-induced fit. Although their expression level can vary depending on cell types, Tat receptors such as CD26, the LRP and HSPGs are ubiquitous. They therefore enable Tat to bind to and affect several cell 360
types. Moreover, because most Tat receptors such as the LRP, CXCR4 and HSPGs are endocytic receptors (71, 73, 74), they will allow Tat internalization. Two endocytic pathways have been reported for Tat. Both studies used fluorescence microscopy and transactivation assays to monitor Tat uptake. The first group observed that Tat fusion proteins such as Tat-GFP entered HeLa and CHO cells using a caveolar pathway (75, 76). A study using the native protein then found that Tat used an AP-2/clathrin/dynamin 2 dependent pathway to enter T cells (26). This observation is consistent with data showing that the LRP and CXCR4 are internalized via coated pits (74,77). Nevertheless, it is still not clear whether Tat uses the clathrin or caveolar pathway or both. The productive pathway can be dependent on cell type because, for instance, T cells do not express any caveolin and are therefore devoid of caveolae (78). Alternatively, because GFP is three times bigger than Tat, the exclusion of Tat fusion proteins from coated pits could be due to size-dependent sorting as it is the case for some lipid-anchored proteins (79). Both Tat endocytosis studies nevertheless agree that endocytic uptake is required for Tat cytosolic delivery because drugs or molecular effectors that prevent Tat endocytosis inhibit transactivation of LTRdriven transfected reporter genes (26,75).
From Endosome to the Cytosol Following uptake, Tat follows the rab5-dependent default pathway to late endosomes/lysosomes (26). Within the endocytic network, Tat is exposed to low pH (pH < 6.0) that triggers Tat membrane insertion, the first step of Tat translocation to the cytosol. This transport requires endosomal acidification and drugs such as ammonium chloride, chloroquine or bafilomycin A1 that neutralize endosomal pH prevent Tat delivery to the cytosol (26, 27). The Tat insertion process can be recapitulated using model membranes such as liposomes or phospholipid monolayers (7). Just as the PI(4,5)P2 -dependent insertion process into the plasma membrane for secretion, low pH-driven Tat insertion into the endosome membrane relies on Trp11 (7) (Figure 4). Low pH is detected by a sensor made of the first acidic residue (Asp/Glu2) of Tat that interacts via an H-bond with the end of the basic domain (residues 55–57). Upon acidification, the carboxyl end of the Glu/Asp side chain becomes protonated, resulting in the destabilization of this electrostatic interaction leading to Trp11 side-chain unmasking, then membrane insertion (7). Interestingly, these molecular devices (low pH sensor and key Trp residue) that govern Tat insertion into the endosome membrane before translocation to the cytosol are amazingly similar to those used by a bacterial toxin, Pseudomonas exotoxin A, to ensure the same function (80). Tat transport from the endosome lumen to the cytosol is catalysed by the cytosolic chaperone Hsp90 that probably helps Tat to refold on the trans side of the membrane (26).
Traffic 2012; 13: 355–363
The Ins and Outs of HIV-1 Tat
is therefore not supposed to directly destabilize Tat structure. Nevertheless, it seems to initiate a rather slow (i.e. within a couple of minutes timescale in cell-free systems) molecular remodelling that culminates with Trp11 insertion (10). Whether cytosolic chaperones can assist Tat in this unfolding step intracellularly is not known.
Membrane pH 7 +++ + Endosome
--
pH