Guiding Tail-Anchored Membrane Proteins to the ER in a Chaperone Cascade

Home , Protein targeting

JBC Papers in Press. Published on October 1, 2019 as Manuscript REV119.006197 The latest version is at http://www.jbc.org/cgi/doi/10.1074/jbc.REV119.006197

Guiding Tail-anchored Membrane Proteins to the ER In a Chaperone Cascade

Shu-ou Shan*

Division of Chemistry and Chemical Engineering, California Institute of Technology 1200 E. California Blvd., Pasadena, CA 91125

Running title: chaperones guide tail-anchored protein targeting

*To whom correspondence should be addressed. Phone: 626-395-3879. Fax: 626-568-9430. E- mail: [email protected] Downloaded from

Key words: chaperone, membrane proteins, tail-anchored protein, Hsp70, protein targeting, ATPases

http://www.jbc.org/ at CALIFORNIA INSTITUTE OF TECHNOLOGY on October 2, 2019

1 ABSTRACT homeostasis in the cell. Before arrival at the Newly synthesized integral membrane appropriate membrane destination, newly proteins must traverse the aqueous cytosolic synthesized membrane proteins must traverse environment before arrival at their membrane the cytosol and, in some cases, multiple other destination and are prone to aggregation, aqueous cellular compartments where misfolding, and mislocalization during this improper exposure of their transmembrane process. The biogenesis of integral domains (TMDs) will lead to rapid and membrane proteins therefore poses acute irreversible aggregation. In addition, the challenges to protein homeostasis within a degeneracy of TMD-lipid interactions poses cell and requires the action of effective challenges to the fidelity of their insertion at molecular chaperones. Chaperones that the appropriate biological membrane, Downloaded from mediate membrane protein targeting not only especially in eukaryotic cells that contain need to protect the nascent transmembrane multiple membrane-enclosed organelles. The domains from improper exposure in the proper localization and folding of membrane

cytosol, but also to accurately select client proteins therefore relies critically on http://www.jbc.org/ proteins and actively guide their clients to the molecular chaperones, which not only protect appropriate target membrane. The nascent membrane proteins from off-pathway mechanisms by which cellular chaperones interactions but also actively guide them to work together to coordinate this complex the correct biological membrane. The process are only beginning to be delineated. mechanism by which the cellular chaperone at CALIFORNIA INSTITUTE OF TECHNOLOGY on October 2, 2019 Here we summary recent advances in studies network overcomes these challenges during of the tail-anchored membrane protein (TA) membrane protein biogenesis remains an targeting pathway, which revealed a network outstanding question. of chaperones, cochaperones, and targeting In the past decade, an increasing factors that together drive and regulate this number of factors have been described that essential process. This pathway is emerging represent components of multiple, distinct as an excellent model system to decipher the protein targeting pathways that deliver mechanism by which molecular chaperones nascent membrane proteins to diverse overcome the multiple challenges during organelles such as the endoplasmic reticulum post-translational membrane protein (ER), mitochondria, and peroxisomes (1-6). biogenesis and to gain insights into the One of these pathways, the guided entry of functional organization of multi-component tail anchored protein (GET) pathway, has chaperone networks. been studied at exquisite mechanistic detail. –––––––––––––––––––––––––––––––––––– This review will summarize recent advances Generation and maintenance of a in our understanding of the GET pathway, functional proteome requires the proper with a focus on a hierarchical chaperone folding, assembly, and localization of all the network found in this pathway that suggest cellular proteins. Integral membrane proteins sophisticated solutions to the challenges of comprise over 30% of the proteins encoded membrane protein biogenesis as well as new by the genome and mediate numerous questions about the role and mechanisms of essential cellular processes including molecular chaperones during this process. molecular transport, energy generation, signaling, and cell-to-cell communication. Diverse targeting pathways accommodate Compared to soluble proteins, the biogenesis membrane proteins with distinct TMD of integral membrane proteins poses locations. particularly acute challenges to protein

2 Diverse pathways mediate the targeting of arrow a; (11)). In addition, the SND genes are nascent membrane proteins to the ER, via synthetically lethal with the GET genes (13). which proteins enter the endomembrane These observations suggest that the SND system in eukaryotic cells. Despite being components provide a backup system for the overly simplistic, it has been useful to SRP and GET pathways to deliver membrane conceptualize the multitude of targeting proteins with relatively downstream TMDs. mechanisms in terms of the needs of Analogous diversity is observed with membrane proteins with distinct TMD translocases at the ER membrane: insertion of locations. For example, most membrane some SRP-dependent membrane proteins and proteins harboring a TMD near the N- less hydrophobic TAs are dependent on the terminus are recognized by the universally ER membrane protein complex (EMC) (Fig. Downloaded from conserved signal recognition particle (SRP) 1; (15-17)). In addition, Snd2/Snd3 as soon as their first TMD emerges from the genetically and physically interacts with exit tunnel of the translating ribosome. Sec72p(13), a component of the post-

Ribosome profiling work in yeast further translational Sec62/63/71/72 translocase http://www.jbc.org/ suggested that SRP can engage ribosomes conserved across eukaryotic even earlier, before the targeting signals on organisms(18,19). The diversity and the nascent polypeptide are translated(7). Via redundancy of targeting and translocation interaction with the SRP receptor, SRP machineries are thought to provide a robust delivers translating ribosomes to the Sec61p network that accommodates the targeting at CALIFORNIA INSTITUTE OF TECHNOLOGY on October 2, 2019 translocase at the ER membrane (or the needs of diverse membrane proteins with SecYEG translocase at the bacterial plasma different TMD location, topology, and charge membrane), often before an additional 60- distribution. 100 residues of the nascent protein is At the other extreme is the class of synthesized (Fig 1, left path; (8-12)). The tail-anchored membrane proteins (TAs) strictly co-translational nature of the SRP whose TMD is near the C-terminus (Fig. 1, pathway ensures that the nascent TMDs are right path). TAs comprise up to 5% of the effectively shielded by proteinaceous eukaryotic membrane proteome and mediate environments in either the SRP or the Sec61p diverse cellular processes including protein (or SecYEG) complex, thus minimizing translocation across organellar membranes, exposure to the aqueous cytosolic vesicle fusion, apoptosis, and protein quality environment during their biogenesis control(2,20-22). As the C-terminal TMD is Much less is known about the obscured by the ribosome during translation, targeting of membrane proteins harboring it was predicted early on that TAs undergo internal TMDs (Fig. 1, middle path). A obligatorily post-translational mechanisms of genetic screen identified three genetically targeting(22). The past decade has witnessed linked SND (for SRP-independent targeting) the discovery of several pathways that proteins, Snd1 in the cytosol and Snd2 and mediate the targeted delivery and insertion of Snd3 at the ER membrane, whose loss led to TAs, including the GET-, SND-, and EMC- mislocalization of this class of proteins(13). dependent pathways (Fig. 1; More recently, the human orthologue of yeast (2,6,13,16,20,21,23)). The GET pathway, Snd2 has been described(14). Nevertheless, which targets relatively hydrophobic TAs to localized ribosome profiling data suggested the ER, is especially well studied. This that SRP is responsible for the cotranslational pathway is also remarkably conserved among ER-localization of most membrane proteins eukaryotic cells: all the components in the containing internal TMDs (Fig. 1, dashed yeast GET pathway have orthologues or

3 functional homologs in mammalian cells. facilitates TA transfer from Sgt2 to Get3 (Fig. The readers are referred to (2,6,20,21) for 2, steps 5-6;(30)). In mammalian cytosol, the comprehensive reviews of the GET pathway C-terminal part of the BAG6 complex and the targeting of tail-anchored proteins in (comprised of BAG6, TRC35 and UBL4A) general. Here, I will focus on the works that was shown to be structurally and functionally uncovered and characterized a multi- homologous to Get4/5 and facilitates TA component chaperone system required for the loading onto TRC40 from SGTA, the biogenesis of this essential class of mammalian Sgt2 homologue (29,31,32). membrane proteins. Thus, the substrate loading mechanism via the Sgt2-to-Get3 transfer is conserved among

A chaperone cascade guides TAs to the ER. eukaryotic cells. Downloaded from Components of the GET pathway Despite these advances, how newly were initially identified through biochemical synthesized TAs are captured by Sgt2 reconstitutions and genetic interaction remained a long standing puzzle. Purified

analyses of the secretory pathway in yeast. Sgt2 is ineffective in capturing TAs in the http://www.jbc.org/ Work in rabbit reticulocyte lysate identified a soluble form, and attempts to directly load 40 kDa ATPase, TRC40, which crosslinks TA onto Sgt2 led to extensively aggregated efficiently to the C-terminal TMD of model complexes(23). For many years, generation TAs and allows insertion of the bound TA of soluble, functional Sgt2•TA or SGTA•TA into ER microsomes(24,25). The yeast complexes has relied on cell lysates that at CALIFORNIA INSTITUTE OF TECHNOLOGY on October 2, 2019 homologue of TRC40, Get3, was contain endogenous chaperone(30,33) or epistatically linked to two ER-localized super-physiological Sgt2/SGTA membrane proteins, Get1 and Get2(26), concentrations(32). Importantly, Sgt2 which were subsequently shown to act as contains a conserved tetratricopeptide repeat both a receptor complex for TA-loaded Get3 (TPR) domain that associates with multiple and a translocase that mediates TA insertion heat shock proteins including Hsp70, Hsp90, into the ER membrane (Fig. 2, step 7; and Hsp104(30,34,35). This observation led (27,28)). As the central targeting factor in the to the hypothesis that heatshock proteins are GET pathway, the structure, dynamics, and further required to facilitate TA loading on activity of Get3 have been extensively Sgt2(35). Experimental evidence for this studied, providing a high-resolution model emerged recently through the work of mechanistic model for how this targeting Cho et al, who demonstrated that the major factor couples its ATPase cycle to the ER cytosolic Hsp70 in yeast, Ssa1, is highly targeting of TAs (Section “Get3: an ATP- effective in capturing newly synthesized TAs driven protean clamp”). and efficiently transfers the bound TAs to Nevertheless, it soon became clear that TA Sgt2, in a manner dependent on its interaction capture by Get3 (or TRC40) is a facilitated with the Sgt2 TPR motif(23). In vivo, process in the crowded cytosolic transient inactivation of Ssa1 severely environment. Nascent TA released from the disrupted TA insertion into the ER, ribosome was poorly captured by partially analogous to observations with GET gene purified TRC40(29). Using biochemical deletions(23). Together, Hsp70, Sgt2, Get4/5 reconstitutions, Wang et al showed that the and Get3 form the minimal components that products of two additional genes epistatically allow reconstitution of the molecular events linked to Get3, Get4 and Get5, form a required to generate a soluble, translocation- scaffold complex that bridges between Get3 competent targeting complex in the and an upstream cochaperone, Sgt2, and cytosol(23).

4 Collectively, these works demonstrate that reverse direction is unfavorable under even a compositionally simple integral physiological conditions. Measurements of membrane protein, such as the TA, is the kinetic stabilities of Sgt2•TA and sequentially funneled through a multi- Get3•TA complexes supported this model, component Hsp70-cochaperone cascade showing that their half-times for spontaneous (Figure 2). Newly synthesized TAs released dissociation are ~40 min(23) and ~4 from the ribosome is captured by Ssa1, which hr(33,36), respectively. Thus, successive effectively shields the TA-TMD from substrate transfers in the Hsp70-Sgt2-Get3 aggregation in the aqueous cytosol (steps 1- triad is thermodynamically driven, with TAs 2). Ssa1 assembles the first transfer complex engaging in increasingly stable interactions via interaction of its C-terminus with the TPR with chaperones as they progress through the Downloaded from domain of Sgt2, in which TA is rapidly pathway. transferred (steps 3-4). A second client If the downstream chaperones bind TAs transfer complex is assembled via the Get4/5 more tightly, why is participation of Hsp70

scaffold complex, which bridges between necessary? An intriguing observation is that http://www.jbc.org/ Sgt2 and Get3 to facilitate TA transfer onto stepwise substrate loading via Ssa1 Get3 (step 5). The TA is then delivered to the significantly enhances the conformational ER membrane via the interaction of Get3 quality of TA substrates: while direct loading with the Get1/2 receptors (steps 6-7). of TAs on Sgt2 is inefficient and resulted in Although the complexity of the GET pathway largely aggregated, inactive complexes, at CALIFORNIA INSTITUTE OF TECHNOLOGY on October 2, 2019 is counter-intuitive, many observations in this Sgt2•TA complexes generated via transfer pathway suggest potential chemical and from Ssa1 are not only soluble, but also biological rationales for the evolution of this functionally competent in undergoing elaborate chaperone cascade, and their subsequent steps in the GET pathway(23). further investigation could provide valuable Although the client interactions of Ssa1 and insights into the roles and organization Sgt2 remain to be studied at higher resolution, principles of chaperone networks in general. kinetic determinants are likely responsible Below, I highlight and discuss the for these observations. The aggregation of implications of some of these observations, single-pass membrane proteins in aqueous with the hope to stimulate additional studies environments tends to be rapid (τ < 10 sec into this and conceptually analogous multi- (23)) and, without external energy input, component chaperone systems. irreversible. Although Sgt2 and Get3 bind TAs with high kinetic stability, their substrate Improved Client Conformational Quality binding kinetics at physiological via Stepwise Loading. concentrations (0.5 – 1 µM; (37,38)) are What drives the directional substrate probably too slow to compete with TA transfers in the GET pathway? Quantitative aggregation. In contrast, Hsp70 binds client measurements suggested that both the Ssa1- proteins rapidly in the ATP state (~106 M-1s-1 to-Sgt2 and Sgt2-to-Get3 TA transfers are (39,40)) and is far more abundant in the energetically downhill, with the transfer cytosol (~15 µM; (37,38)) compared to Sgt2 equilibrium ~100-fold and ~20-fold in favor and Get3. These factors enable cytosolic of the downstream chaperone in the Hsp70s to more effectively compete with off- respective transfer complexes(23,33). This pathway misfolding and aggregation implies that the downstream chaperones bind processes, allowing nascent TAs to be TAs much more strongly than their respective captured in a soluble, functionally competent upstream chaperones, and TA transfer in the conformation. The conformational quality of

5 TAs appears to be effectively preserved These observations strongly suggest that TAs during both the Ssa1-to-Sgt2 and Sgt2-to- are also physically shielded from alternative Get3 handovers, likely through concerted chaperones in the cytosol during their transfer substrate transfer mechanisms (section from Sgt2 to Get3. “Client Privilege In the Chaperone Cascade” A recent study further highlights that below). Thus, the sequential substrate conserved molecular mechanisms have loading and transfers in the GET pathway are evolved to ensure client privilege during the governed by a combination of Sgt2-to-Get3 transfer. A conserved helix 8 thermodynamic forces and kinetic constraints, (termed α8) lining the substrate-binding which together ensure that these hydrophobic groove of Get3, which was unresolved in proteins are maintained in a soluble, most crystal structures, was found to Downloaded from translocation-competent state en route to the specifically promote rapid and privileged TA ER membrane. transfer from Sgt2 to Get3(36). Mutations of α8 slowed TA transfer from Sgt2 to Get3

Client Privilege In the Chaperone Cascade. ~100-fold and largely abolished the role of http://www.jbc.org/ Client handover from Hsp70 to the Get4/5 complex. Moreover, Get3 lost its downstream chaperones, such as GroEL/ES privilege to capture TAs from Sgt2 upon and Hsp90, is integral for the folding of mutation of α8, and the TA substrate was numerous proteins. Despite the importance of instead lost to external chaperones such as these transfer events, their detailed molecular CaM(36). These defects in vitro are at CALIFORNIA INSTITUTE OF TECHNOLOGY on October 2, 2019 mechanisms are poorly understood. For corroborated by the enhanced stress example, current models largely assume a sensitivity and TA insertion defects of yeast passive mechanism in which recalcitrant cells harboring Get3(α8) mutations (36,44). substrates released from Hsp70 simply Coupled with the observation that α8 can diffuse to the GroEL chaperonin to complete crosslink to TAs(44), it was proposed that the their folding (41,42), and bacterial outer flexible α8 motif mediates the earliest membrane proteins are assumed to associate contacts of Get3 with the TA during its with and dissociate from multiple transfer from Sgt2, helping to guide the TA periplasmic chaperones before insertion into into the substrate binding groove of Get3 the membrane (43). Intriguingly, studies in while also shielding the TA from off-pathway the GET pathway provided convincing chaperones during this process (Fig. 3;(36)). evidence for a strongly facilitated, highly Importantly, privileged client transfer privileged client handover mechanism during provides an effective mechanism to not only the Sgt2-to-Get3 (or SGTA-to-TRC40) TA protect nascent membrane proteins from re- transfer. First, the transfer is kinetically facile, exposure to the aqueous cytosolic with a halftime of 10-20 seconds(32,33). This 2 environment, but also ensure that substrates is >10 -fold faster than spontaneous TA are retained within a dedicated biogenesis dissociation from Sgt2(23), suggesting that pathway en route to the target membrane. the transfer occurs via a more active The molecular mechanisms that underlie mechanism than simple TA release and the active and privileged substrate transfers in diffusion from Sgt2 to Get3. Further, the the GET pathway remain to be determined. transfer is impervious to the presence of off- On one hand, many individual domains and pathway chaperones that act as a TA trap, interactions in the Sgt2•Get4/5•Get3 transfer such as calmodulin (CaM), whereas the complex have been extensively studied. Sgt2 isolated SGTA•TA or Sgt2•TA complex (and SGTA) is characterized by a subclass of quickly loses the bound TA to CaM(32,36). TPR domains frequently found in HSC

6 cochaperones, including Hsp organizing biophysical properties impact their substrate protein (HOP in human and Sti1 in yeast) and recognition, are likely key to understanding C-terminus of HSC interacting protein (CHIP) the kinetic acceleration and privilege of the (35,45-48). Five conserved residues in this TA during its transfers in the GET pathway. TPR domain form a dicarboxylate clamp that recognizes a C-terminal EEVD motif in Client Selection and Triage. Hsp70, Hsp90, and Hsp100 (Fig. 3, left inset), While the roles of Hsp70 and Get3 in linking Sgt2 to multiple protein folding the GET pathway (client capture and pathways. The N-terminus of Sgt2 mediates targeting to the ER, respectively) are easier to its homodimerization and forms an understand, the precise roles of Sgt2 (and interaction platform for the ubiquitin-like SGTA) are less clear. An interesting Downloaded from (UBL) domain of Get5 (Fig. 3, lower inset), hypothesis is that this cochaperone provides linking Sgt2 to downstream components of a mechanism to reject suboptimal substrates the GET pathway(49-51). At the other end of from the GET pathway. Co-

this transfer complex, Get4 binds with immunoprecipitation experiments by Wang http://www.jbc.org/ nanomolar affinity to ATP-bound Get3 and et al first showed that Sgt2 can distinguish bridges the Get3 dimer interface (Fig. 3, right between TAs destined to the ER versus inset)(52,53). As detailed later (section “Get3: mitochondria(30). By systematically varying an ATP-driven protean clamp”), the the TMD in model TAs, biochemical interactions of Get4/5 not only bring Sgt2 and analyses showed that Sgt2 preferentially at CALIFORNIA INSTITUTE OF TECHNOLOGY on October 2, 2019 Get3 into close proximity but also optimize binds TMDs that have higher hydrophobicity the conformation and nucleotide state of Get3 and helical content, features that distinguish for TA capture. On the other hand, due to the GET substrates from mitochondrial TAs(33). multiple flexible elements in Sgt2 and Get4/5, Another study, which examined a large set of the organization and architecture of this TAs, further showed that Sgt2 does not transfer complex is largely unknown. In efficiently capture TAs with low addition, the C-terminal domain of Sgt2, rich hydrophobicity TMDs, which can be inserted in glutamine and methionine, forms the into the ER via the alternative EMC binding site for hydrophobic TMDs on TA pathway(16). Although the preferences of substrates (Fig. 3, SBD), but the molecular Sgt2 for more hydrophobic TAs are basis of substrate recognition by Sgt2 paralleled by Get3(33), the high kinetic remains unclear (30). Whether the upstream stability of the Get3•TA complex renders it Hsp70-to-Sgt2 TA transfer is also privileged less effective at rejecting suboptimal TAs. and the molecular mechanisms that give rise The lifetime of Get3 bound to a model GET to privileged client transfer remain substrate is 1–4 hrs, whereas TA insertion outstanding questions. Finally, the client into the ER occurs within 5–10 interaction of Hsp70 is extensively regulated min(33,36,54). Thus, most TAs that have by its own ATPase cycles and by been loaded on Get3 are committed to cochaperones, such as Hsp40, that tune the insertion into the ER, and suboptimal TAs conformation and nucleotide state of Hsp70; that bound Get3 less tightly do not efficiently this further raises questions as to whether and dissociate before the insertion. The upstream how additional Hsp70 cochaperones are chaperone in the pathway, Ssa1, is known to involved in the targeting pathway and the promiscuously associate with diverse nascent client transfer process. Deciphering the proteins(55). Sgt2 was therefore proposed to conformation and dynamics of substrate- provide a key selection filter that rejects TAs bound Hsp70 and Sgt2, and how these and other membrane proteins destined to

7 alternative organelles or targeting pathways the core targeting factor in the GET pathway (Fig. 2, dashed arrows b-e). In the that binds TA substrates with extraordinarily mammalian system, the BAG6 complex that high stability, Get3 provides a valuable replaces Get4/5 contains an additional UBL opportunity to elucidate the client interaction domain that recruits the ubiquitin ligase of a membrane protein chaperone at high RNF126 and can mediate poly-ubiquitylation resolution. As a member of the SIMIBI (after of substrates loaded on SGTA(32,56), signal recognition particle (SRP), MinD, and potentially generating an additional branch BioD) class of nucleotide hydrolases, studies that triages mislocalized membrane proteins of Get3 also provided insights into the to quality control pathways. regulatory mechanism of an emerging family

These observations suggest a modular of dimerization-activated GTPase and Downloaded from organization of the GET pathway, in which ATPases whose mode of action differs each chaperone/cochaperone fulfills a significantly from that of the classic signaling distinct function that together enable the GTPases (64-67).

efficient, selective, and unidirectional Earlier work has provided beautiful http://www.jbc.org/ targeting of nascent TAs. As nascent proteins structural illustrations for how Get3 are funneled through this cascade, they undergoes ATP- and interaction partner- engage more specialized chaperones with induced conformational rearrangements that increasingly high affinity and become more can be coupled to substrate binding and committed to insertion into the ER. release (Fig. 4A, TA-loading phase on the left; at CALIFORNIA INSTITUTE OF TECHNOLOGY on October 2, 2019 Analogous functional specialization of see references (2,20,21) for more downstream cochaperones that collaborate comprehensive reviews on Get3 structure and with Hsp70 are well documented. For function). Get3 is an obligate homodimer in example, TPR-containing co-chaperones, which the ATPase domains directly bridge Sti1/HOP, mediate the handover of kinase the dimer interface and are structurally and substrates from Hsp70 to Hsp90, enabling functionally coupled to a helical domain (Fig. Hsp90 to complete the folding of numerous 4B). Early crystallographic work showed that members of the kinase superfamily(57-59). non-hydrolyzable ATP analogues induce Upon encounter with protein aggregates, adjustments at the dimer interface, which are Hsp70 could recruit and collaborate with the amplified into larger movements of the Hsp100 family of chaperones to refold the helical domains that bring them close to one aggregated proteins(60-63). More broadly, another (Figs. 4A, “closed” Get3 and Fig. 4B, client handover from Hsp70 to downstream left panel; (44,68-70)). Importantly, “closing” chaperones or cochaperones could provide a brings together conserved hydrophobic versatile triaging mechanism via which residues in the helical domains of Get3 to distinct classes of client proteins are sorted to form a contiguous hydrophobic groove, their dedicated biogenesis pathways. This which provides the docking site for the TA- hypothesis and the detailed molecular TMD(71). Get4/5 selectively binds to and mechanisms of client triage in these stabilizes ATP-bound closed Get3 and chaperone/cochaperone systems remain to be further inhibits its ATPase activity (Fig. 4A, studied. “occluded”; (53,72,73)). Presumably, these ATP- and Get4/5-induced rearrangements Get3: An ATP-driven Protean Clamp. optimizes Get3 for capture of the TA At the end of the chaperone cascade, substrate from Sgt2 (Fig. 4A, “TA Loading”). Get3 receives the TA substrates from Sgt2 At the other extreme, the cytosolic domain of and delivers them to the ER membrane. As Get1 (Get1-CD) form a coiled-coil that

8 inserts like a wedge into the Get3 dimer of the pathway (Figs. 4A, “Targeting”, and interface, inducing a wide open conformation Fig. 4C). Biochemical analyses demonstrated of Get3 (Fig. 4A, “open” and Fig. 4B, right that these substrate-induced dynamic motions panel; (74-76)). Get1-CD not only disrupts led to adjustments at the Get3-Get4 interface the TA binding groove of Get3, but also that enable more facile dissociation of Get3 induces both the switch I and switch II loops from the Get4/5 complex (Fig. 4A, step 5). at the Get3 ATPase site into a conformation Once released from Get4/5, the TA substrate incompatible with ATP binding(73-77). activates ATP hydrolysis on Get3 (Fig. 4A, These Get1-induced rearrangements are step 6). These biochemical changes are believed to be responsible for triggering the coupled with increasing dynamics of the release of TA from Get3 at the ER for Get3•TA complex to more extensively Downloaded from membrane insertion. sample the open conformation (Fig. 4C)(54). A dilemma posed by this early model is This renders dissociation from Get4/5 that an exclusively closed Get3•TA complex irreversible and primes Get3•TA for

would preclude downstream events in the interaction with and remodeling by the http://www.jbc.org/ pathway that require Get3 to dissociate from Get1/2 receptors (Fig. 4A, step 7), driving the Get4/5. These include the interaction of Get3 relocalization of the targeting complex from with the Get1/2 receptor, whose binding sites the cytosol to the ER membrane. on Get3 heavily overlap with that of While earlier models of Get3-TA Get4/5(53), and ATP hydrolysis by Get3, interaction invoked a ‘lock-and-key’ type at CALIFORNIA INSTITUTE OF TECHNOLOGY on October 2, 2019 which is inhibited by Get4/5(72). More mechanism in which the TA substrates fit generally, chaperones that engage and deliver into a pre-organized hydrophobic groove on membrane proteins not only need to Get3, observations from the single molecule effectively capture substrates, but also to study suggest a distinct model in which Get3 promptly release the bound substrates at the forms a rapidly fluctuating ‘protean clamp’ target membrane. The mechanisms that that stably traps substrates. Analogous enable membrane protein chaperones to conformational dynamics of chaperone-client transition from the substrate-loading mode to interactions have been observed with the substrate-releasing mode are not well multiple ATP-independent chaperones that understood. mediate outer membrane protein biogenesis The resolution to this dilemma was in the bacterial periplasm(78-80). In addition, provided by a more recent single molecule although earlier work based on peptide spectroscopy study, which uncovered substrates associated a lid-closed unusual substrate-induced dynamic motions conformation of Hsp70 with the high affinity in this ATPase and elucidated how these client-binding state, more recent NMR, EPR dynamics drive the targeting phase of the and single molecule experiments revealed GET pathway (Fig. 4A, right). In this work, remarkable conformational heterogeneity the open-to-closed conformational and dynamics in Hsp70 when it engages full rearrangements of Get3 was directly length protein substrates(81-84). It is monitored using a pair of Förster resonance conceivable that rapidly fluctuating energy transfer (FRET) dyes incorporated at chaperones and dynamic chaperone-client its helical domains (Fig. 4B, green and red interactions operate in many systems to retain stars)(54). Contrary to the accepted models, substrates with high affinity, while also this study found that the TA substrate providing functional switches to propel the initiates sub-millisecond timescale opening progression of vectorial pathways. motions in Get3 that drive the targeting phase

9 Additional Hsp70-cochaperone pairs in target membranes remain to be elucidated. In membrane protein targeting. the example of the GET pathway, the Cytosolic Hsp70s participate in interaction of the Hsp70 EEVD motif with almost every stage of the protein life cycle, the TPR domain of Sgt2 provides the from de novo folding, protein aggregate mechanism to direct hydrophobic TAs into remodeling, to protein quality control(42). this pathway. Notably, Hsp70 also bind to the Recent works further highlighted essential TPR motifs on the mitochondria import roles of Hsp70s in an increasing number of receptor Tom70, and this interaction is membrane protein targeting pathways. The required for the import of a subset of participation of Hsp70 in protein transport mitochondrial precursor and beta-barrel was initially recognized in studies of membrane proteins(88,90). Additional TPR- Downloaded from secretory and mitochondrial precursor containing receptors have been found on the proteins(85-89). However, the extent of membrane of other organelles, including Hsp70 participation in protein targeting had Pex5 (Peroxisomal Biogenesis Factor 5)

been unclear in the earlier work: among nine involved in the biogenesis of peroxisomal http://www.jbc.org/ preprotein substrates examined, the targeting matrix proteins(93,94), and Sec71/72 that of only two ER- and one mitochondria- associates with the Sec62/63 translocase to destined proteins were affected by Ssa1 assist the post-translational translocation of inactivation (86). Investigations of the GET precursor proteins across the ER(95). pathway added an essential class of integral Another class of mechanisms involves the at CALIFORNIA INSTITUTE OF TECHNOLOGY on October 2, 2019 membrane proteins to the list of clients whose Hsp40’s, which not only bind client proteins proper cellular localization is directly in cooperation with Hsp70 but also interact mediated by Hsp70. Additional recent studies with receptors on specific organelles. For showed that cytosolic Hsp70 and its example, the major cytosolic Hsp40s, Ydj1 associated Hsp40s, Ydj1 and Sis1, interact and Sis1, preferentially bind to the with and are required for the efficient mitochondrial import receptor Tom20, targeting of beta-barrel membrane providing a redundant import route that proteins(90) and a subset of TAs(3) to works in parallel with the Tom70-mediated mitochondria and peroxisomes. Two less pathway. Another Hsp40 involved in Mim1 abundant Hsp40s, Xdj1 and Djp1, were also biogenesis, Xdj1, is a specific interaction found to preferentially bind subsets of partner of Tom22, the central receptor of the mitochondrial membrane proteins, such as translocase of the mitochondrial outer Mim1 and Tom22, and promote their membrane(92). These observations suggest biogenesis (91,92). In the case of Xdj1, the J- that the functional coupling of Hsp70 with domain that binds and regulates Hsp70 was cochaperones that are either organelle- required for substrate import, indicating that specific, or mediate organelle-specific it cooperates with Hsp70 to carry out this interactions, could provide a general process. Collectively, these recent mechanism to direct the localization of observations suggest that cytosolic Hsp70s nascent membrane proteins to distinct provide a hub that rapidly captures nascent cellular organelles. How diverse nascent membrane and organellar proteins and membrane proteins are distinguished by the facilitate their targeted delivery to diverse different Hsp70-cochaperone pairs and thus intracellular membranes. engage the correct targeting pathway remains Once Hsp70 and/or their associated an outstanding question that lies at the heart Hsp40’s captures the preprotein substrates, of understanding the fidelity of protein how the substrates are guided to the correct localization.

10 8. Walter, P., and Johnson, A. E. (1994) Acknowledgment. Signal sequence recognition and protein I thank H. Cho for critical discussions and targeting to the endoplasmic reticulum comments on the manuscript. This work was membrane. Annu Rev Cell Biol 10, 87-119 supported by NIH grant GM107368, Gordon 9. Akopian, D., Dalal, K., Shen, K., Duong, F., and Betty Moore Foundation Grant and Shan, S. O. (2013) SecYEG activates GBMF2939, and a fellowship from the GTPases to drive the completion of Weston Havens foundation to S.-o.S. cotranslational protein targeting. J Cell Biol

Conflict of interest. 200, 397-405

The author declares no conflicts of interest 10. Voorhees, R. M., and Hegde, R. S. Downloaded from with the contents of this article. (2016) Toward a structural understanding of co-translational protein translocation. Curr References Opin Cell Biol 41, 91-99 11. Costa, E. A., Subramanian, K., Nunnari, 1. Becker, T., Song, J., and Pfanner, N. (2019) http://www.jbc.org/ Versatility of Preprotein Transfer from the J., and Weissman, J. S. (2018) Defining the Cytosol to Mitochondria. Trends Cell Biol 29, physiological role of SRP in protein-targeting 534-548 efficiency and specificity. Science 359, 689- 2. Chio, U. S., Cho, H., and Shan, S. O. (2017) 692 at CALIFORNIA INSTITUTE OF TECHNOLOGY on October 2, 2019 Mechanisms of Tail-Anchored Membrane 12. Jan, C. H., Williams, C. C., and Protein Targeting and Insertion. Annu Rev Weissman, J. S. (2014) Principles of ER Cell Dev Biol 33, 417-438 cotranslational translocation revealed by 3. Cichocki, B. A., Krumpe, K., Vitali, D. G., proximity-specific ribosome profiling. and Rapaport, D. (2018) Pex19 is involved in Science 346, 1257521 importing dually targeted tail-anchored 13. Aviram, N., Ast, T., Costa, E. A., Arakel, proteins to both mitochondria and E. C., Chuartzman, S. G., Jan, C. H., peroxisomes. Traffic 19, 770-785 Hassdenteufel, S., Dudek, J., Jung, M., Schorr, 4. Chen, Y., Pieuchot, L., Loh, R. A., Yang, J., S., Zimmermann, R., Schwappach, B., Kari, T. M., Wong, J. Y., and Jedd, G. (2014) Weissman, J. S., and Schuldiner, M. (2016) Hydrophobic handoff for direct delivery of The SND proteins constitute an alternative peroxisome tail-anchored proteins. Nat targeting route to the endoplasmic Commun 5, 5790 reticulum. Nature 540, 134-138 5. Yagita, Y., Hiromasa, T., and Fujiki, Y. 14. Hassdenteufel, S., Sicking, M., Schorr, (2013) Tail-anchored PEX26 targets S., Aviram, N., Fecher-Trost, C., Schuldiner, peroxisomes via a PEX19-dependent and M., Jung, M., Zimmermann, R., and Lang, S. TRC40-independent class I pathway. J Cell (2017) hSnd2 protein represents an Biol 200, 651-666 alternative targeting factor to the 6. Borgese, N., Coy-Vergara, J., Colombo, S. endoplasmic reticulum in human cells. FEBS F., and Schwappach, B. (2019) The Ways of Lett 591, 3211-3224 Tails: the GET Pathway and more. Protein J 15. Chitwood, P. J., Juszkiewicz, S., Guna, 38, 289-305 A., Shao, S., and Hegde, R. S. (2018) EMC Is 7. Chartron, J. W., Hunt, K. C., and Frydman, Required to Initiate Accurate Membrane J. (2016) Cotranslational signal-independent Protein Topogenesis. Cell 175, 1507-1519 SRP preloading during membrane targeting. e1516 Nature 536, 224-228

11 16. Guna, A., Volkmar, N., Christianson, J. insertion of tail-anchored proteins. J. Cell Sci. C., and Hegde, R. S. (2018) The ER membrane 9, 1522-1530 protein complex is a transmembrane domain 25. Stefanovic, S., and Hegde, R.S. (2007) insertase. Science 359, 470-473 Identification of a targeting factor for post- 17. Shurtleff, M. J., Itzhak, D. N., translational membrane protein insertion Hussmann, J. A., Schirle Oakdale, N. T., Costa, into the ER. Cell 128, 1147-1159 E. A., Jonikas, M., Weibezahn, J., Popova, K. 26. Schuldiner, M., Collins, S.R., Thompson, D., Jan, C. H., Sinitcyn, P., Vembar, S. S., N.J., Denic, V., Bhamidipati, A., Punna, T., Hernandez, H., Cox, J., Burlingame, A. L., Ihmels, J., Andrews, B., Boone, C., Greenblatt,

Brodsky, J. L., Frost, A., Borner, G. H., and J.F., Weissman, J.S., and Krogna, N.J. (2005) Downloaded from Weissman, J. S. (2018) The ER membrane Exploration of the function and organization protein complex interacts cotranslationally of the yeast early secretory pathway through to enable biogenesis of multipass membrane an epistatic miniarray profile. Cell 123, 507- proteins. Elife 7 519 http://www.jbc.org/ 18. Linxweiler, M., Schick, B., and 27. Schuldiner, M., Metz, J., Schmid, V., Zimmermann, R. (2017) Let's talk about Secs: Denic, V., Rakwalska, M., Schmitt, H.D., Sec61, Sec62 and Sec63 in signal Schwappach, B., and Weissman, J.S. (2008) transduction, oncology and personalized The GET complex mediates insertion of tail- medicine. Signal Transduct Target Ther 2, anchored proteins into the ER membrane. at CALIFORNIA INSTITUTE OF TECHNOLOGY on October 2, 2019 17002 Cell 134, 634-645 19. Wu, X., Cabanos, C., and Rapoport, T. A. 28. Wang, F., Chan, C., Weir, N. R., and (2019) Structure of the post-translational Denic, V. (2014) The Get1/2 transmembrane protein translocation machinery of the ER complex is an endoplasmic-reticulum membrane. Nature 566, 136-139 membrane protein insertase. Nature 512, 20. Chartron, J. W., Clemons, W.M.Jr., and 441-444 Suloway, C. J.M. (2012) The complex process 29. Mariappan, M., Li, X., Stefanovic, S., of GETting tail-anchored membrane proteins Sharma, A., Mateja, A., Keenan, R., and to the ER. Curr. Opin. Struct. Biol. 22, 217- Hegde, R.S. (2010) A ribosome-associating 224 factor chaperones tail-anchored membrane 21. Hegde, R. S., and Keenan, R.J. (2011) proteins. Nature 466, 1120-1124 Tail-anchored membrane protein insertion 30. Wang, F., Brown, E.C., Mak, G., Zhuang, into the endoplasmic reticulum. Nature Rev. J., and Denic, V. (2010) A chaperone cascade Mol. Cell Biol. 12, 787-798 sorts proteins for posttranslational 22. Kutay, U., Harmann, E., and Rapoport, membrane insertion into the endoplasmic T.A. (1993) A class of membrane proteins reticulum. Molec. Cell 40, 1-13 with a C-terminal anchor. Trends Cell Biol. 3, 31. Mock, J. Y., Chartron, J. W., Zaslaver, 72-75 M., Xu, Y., Ye, Y., and Clemons, W. M., Jr. 23. Cho, H., and Shan, S. O. (2018) (2015) Bag6 complex contains a minimal tail- Substrate relay in an Hsp70-cochaperone anchor-targeting module and a mock BAG cascade safeguards tail-anchored domain. Proc Natl Acad Sci U S A 112, 106- membrane protein targeting. EMBO J 37 111 24. Favaloro, V., Vilardi, F., Schlecht, R., 32. Shao, S., Rodrigo-Brenni, M. C., Kivlen, Mayer, M.P., and Dobberstein, B. (2010) M. H., and Hegde, R. S. (2017) Mechanistic Asna1/TRC40-mediated membrane

12 basis for a molecular triage reaction. Science 42. Clerico, E. M., Tilitsky, J. M., Meng, W., 355, 298-302 and Gierasch, L. M. (2015) How hsp70 33. Rao, M., Okreglak, V., Chio, U. S., Cho, molecular machines interact with their H., Walter, P., and Shan, S. O. (2016) Multiple substrates to mediate diverse physiological selection filters ensure accurate tail- functions. J Mol Biol 427, 1575-1588 anchored membrane protein targeting. Elife 43. Costello, S. M., Plummer, A. M., 5 Fleming, P. J., and Fleming, K. G. (2016) 34. Krysztofinska, E. M., Evans, N. J., Dynamic periplasmic chaperone reservoir Thapaliya, A., Murray, J. W., Morgan, R. M. L., facilitates biogenesis of outer membrane

Martinez-Lumbreras, S., and Isaacson, R. L. proteins. Proc Natl Acad Sci U S A 113, Downloaded from (2017) Structure and Interactions of the TPR E4794-4800 Domain of Sgt2 with Yeast Chaperones and 44. Mateja, A., Szlachcic, A., Downing, M. Ybr137wp. Front Mol Biosci 4, 68 E., Dobosz, M., Mariappan, M., Hegde, R. S., 35. Chartron, J. W., Gonzalez, G.M., and and Keenan, R. J. (2009) The structural basis http://www.jbc.org/ Clemons, W.M. Jr. (2011) A structural model of tail-anchored membrane protein of the Sgt2 protein and its interactions with recognition by Get3. Nature 461, 361-366 chaperones and the Get4/Get5 complex. J. 45. D'Andrea, L. D., and Regan, L. (2003) Biol. Chem. 286, 34325-34334 TPR proteins: the versatile helix. Trends 36. Chio, U. S., Chung, S., Weiss, S., and Biochem Sci 28, 655-662 at CALIFORNIA INSTITUTE OF TECHNOLOGY on October 2, 2019 Shan, S. O. (2019) A Chaperone Lid Ensures 46. Meacham, G. C., Patterson, C., Zhang, Efficient and Privileged Client Transfer W., Younger, J. M., and Cyr, D. M. (2001) The during Tail-Anchored Protein Targeting. Cell Hsc70 co-chaperone CHIP targets immature Rep 26, 37-44 e37 CFTR for proteasomal degradation. Nat Cell 37. Kulak, N. A., Pichler, G., Paron, I., Biol 3, 100-105 Nagaraj, N., and Mann, M. (2014) Minimal, 47. Scheufler, C., Brinker, A., Bourenkov, encapsulated proteomic-sample processing G., Pegoraro, S., Moroder, L., Bartunik, H., applied to copy-number estimation in Hartl, F. U., and Moarefi, I. (2000) Structure eukaryotic cells. Nat Methods 11, 319-324 of TPR domain-peptide complexes: critical 38. Ghaemmaghami, S., Huh, W. K., Bower, elements in the assembly of the Hsp70- K., Howson, R. W., Belle, A., Dephoure, N., Hsp90 multichaperone machine. Cell 101, O'Shea, E. K., and Weissman, J. S. (2003) 199-210 Global analysis of protein expression in yeast. 48. Zhang, M., Windheim, M., Roe, S. M., Nature 425, 737-741 Peggie, M., Cohen, P., Prodromou, C., and 39. Schmid, D., Baici, A., Gehring, H., and Pearl, L. H. (2005) Chaperoned Christen, P. (1994) Kinetics of molecular ubiquitylation--crystal structures of the CHIP chaperone action. Science 263, 971-973 U box E3 ubiquitin ligase and a CHIP-Ubc13- 40. Mayer, M. P., Schroder, H., Rudiger, S., Uev1a complex. Mol Cell 20, 525-538 Paal, K., Laufen, T., and Bukau, B. (2000) 49. Chang, Y.-W., Chuang, Y.-C., Ho, Y.-C., Multistep mechanism of substrate binding Cheng, M.-Y., Sun, Y.-J., Hsiao, C.-D., and determines chaperone activity of Hsp70. Nat Wang, C. (2010) Crystal structure of Get4- Struct Biol 7, 586-593 Get5 complex and its interactions with Sgt2, 41. Powers, E. T., Powers, D. L., and Get3 and Ydj1. J. Biol. Chem. 285, 9962-9970 Gierasch, L. M. (2012) FoldEco: a model for 50. Chartron, J. W., VanderVeide, D.G., proteostasis in E. coli. Cell Rep 1, 265-276 Rao, M., and Clemons, W.M. Jr. (2012) Get5

13 carboxyl-terminal domain is a novel Hsp90 Machinery. Cold Spring Harb Perspect dimerization motif that tethers an extended Biol Get4/Get5 complex. J. Biol. Chem. 287, 59. Alvira, S., Cuellar, J., Rohl, A., 8310-8317 Yamamoto, S., Itoh, H., Alfonso, C., Rivas, G., 51. Simon, A. C., Simpson, P. J., Goldstone, Buchner, J., and Valpuesta, J. M. (2014) R. M., Krysztofinska, E. M., Murray, J. W., Structural characterization of the substrate High, S., and Isaacson, R. L. (2013) Structure transfer mechanism in Hsp70/Hsp90 folding of the Sgt2/Get5 complex provides insights machinery mediated by Hop. Nat Commun 5, into GET-mediated targeting of tail- 5484

anchored membrane proteins. Proc Natl 60. Rosenzweig, R., Moradi, S., Zarrine- Downloaded from Acad Sci U S A 110, 1327-1332 Afsar, A., Glover, J. R., and Kay, L. E. (2013) 52. Gristick, H. B., Rome, M. E., Chartron, J. Unraveling the mechanism of protein W., Rao, M., Hess, S., Shan, S. O., and disaggregation through a ClpB-DnaK Clemons, W. M., Jr. (2015) Mechanism of interaction. Science 339, 1080-1083 http://www.jbc.org/ Assembly of a Substrate Transfer Complex 61. Seyffer, F., Kummer, E., Oguchi, Y., during Tail-anchored Protein Targeting. J Biol Winkler, J., Kumar, M., Zahn, R., Sourjik, V., Chem 290, 30006-30017 Bukau, B., and Mogk, A. (2012) Hsp70 53. Gristick, H. B., Rao, M., Chartron, J. W., proteins bind Hsp100 regulatory M domains Rome, M. E., Shan, S. O., and Clemons, W. M., to activate AAA+ disaggregase at aggregate at CALIFORNIA INSTITUTE OF TECHNOLOGY on October 2, 2019 Jr. (2014) Crystal structure of ATP-bound surfaces. Nat Struct Mol Biol 19, 1347-1355 Get3-Get4-Get5 complex reveals regulation 62. Winkler, J., Tyedmers, J., Bukau, B., of Get3 by Get4. Nat Struct Mol Biol 21, 437- and Mogk, A. (2012) Hsp70 targets Hsp100 442 chaperones to substrates for protein 54. Chio, U. S., Chung, S., Weiss, S., and disaggregation and prion fragmentation. J Shan, S. O. (2017) A protean clamp guides Cell Biol 198, 387-404 membrane targeting of tail-anchored 63. Lee, J., Kim, J. H., Biter, A. B., Sielaff, B., proteins. Proc Natl Acad Sci U S A 114, Lee, S., and Tsai, F. T. (2013) Heat shock E8585-E8594 protein (Hsp) 70 is an activator of the Hsp104 55. Albanese, V., Yam, A. Y., Baughman, J., motor. Proc Natl Acad Sci U S A 110, 8513- Parnot, C., and Frydman, J. (2006) Systems 8518 analyses reveal two chaperone networks 64. Shan, S. O. (2016) ATPase and GTPase with distinct functions in eukaryotic cells. Tangos Drive Intracellular Protein Transport. Cell 124, 75-88 Trends Biochem Sci 41, 1050-1060 56. Hessa, T., Sharma, A., Mariappan, M., 65. Gasper, R., Meyer, S., Gotthardt, K., Eshleman, H.D., Gutierrez, E., and Hegde, R.S. Sirajuddin, M., and Wittinghofer, A. (2009) It (2011) Protein targeting and degradation are takes two to tango: regulation of G proteins coupled for elimination of mislocalized by dimerization. Nat Rev Mol Cell Biol 10, proteins. Nature 475, 394-397 423-429 57. Wegele, H., Wandinger, S. K., Schmid, 66. Bange, G., and Sinning, I. (2013) SIMIBI A. B., Reinstein, J., and Buchner, J. (2006) twins in protein targeting and localization. Substrate transfer from the chaperone Nat Struct Mol Biol 20, 776-780 Hsp70 to Hsp90. J Mol Biol 356, 802-811 67. Leipe, D. D., Koonin, E. V., and Aravind, 58. Biebl, M. M., and Buchner, J. (2019) L. (2003) Evolution and classification of P- Structure, Function, and Regulation of the

14 loop kinases and related proteins. J Mol Biol and Sinning, I. (2011) Structural basis for tail- 333, 781-815 anchored membrane protein biogenesis by 68. Bozkurt, G., Stjepanovic, G., Vilardi, F., the Get3-receptor complex. Science 333, Amlacher, S., Wild, K., Gert, B., Favaloro, V., 758-762 Rippe, K., Hurt, E., Dobberstein, B., and 76. Kubota, K., Yamagata, A., Sato, Y., Sinning, I. (2009) Structural insights into tail- Goto-Ito, S., and Fukai, S. (2012) Get1 anchored protein binding and membrane stabilizes an open dimer conformation of protein insertion by Get3. Proc. Natl. Acad. Get3 ATPase by binding two distinct Sci. U. S. A. 106, 21131-21136 interfaces. J. Mol. Biol. 422, 366-375

69. Hu, J., Li, J., Qian, X., Denic, V., and Sha, 77. Wang, F., Whynot, A., Tung, M., and Downloaded from B. (2009) The crystal structures of yeast Get3 Denic, V. (2011) The mechanism of Tail- suggest a mechanism for tail-anchored anchored protein insertion into the ER protein membrane insertion. Plos One 4, membrane. Molec. Cell 43, 1-13 e8061 78. Burmann, B. M., Wang, C., and Hiller, S. http://www.jbc.org/ 70. Suloway, C. J. M., Chartron, J.W., (2013) Conformation and dynamics of the Zaslaver, M., and Clemons, W.M.Jr. (2009) periplasmic membrane-protein-chaperone Model for eukaryotic tail-anchored protein complexes OmpX-Skp and tOmpA-Skp. Nat binding based on the structure of Get3. Proc. Struct Mol Biol 20, 1265-1272 Natl. Acad. Sci. U. S. A. 106, 14849-14854 79. Hiller, S. (2019) Chaperone-Bound at CALIFORNIA INSTITUTE OF TECHNOLOGY on October 2, 2019 71. Mateja, A., Paduch, M., Chang, H. Y., Clients: The Importance of Being Dynamic. Szydlowska, A., Kossiakoff, A. A., Hegde, R. S., Trends Biochem Sci 44, 517-527 and Keenan, R. J. (2015) Protein targeting. 80. Holdbrook, D. A., Burmann, B. M., Structure of the Get3 targeting factor in Huber, R. G., Petoukhov, M. V., Svergun, D. I., complex with its membrane protein cargo. Hiller, S., and Bond, P. J. (2017) A Spring- Science 347, 1152-1155 Loaded Mechanism Governs the Clamp-like 72. Rome, M., Rao, M., Clemons, W.M., Dynamics of the Skp Chaperone. Structure 25, and Shan, S. (2013) Precise timing of ATPase 1079-1088 e1073 activation drives targeting of tail-anchored 81. Marcinowski, M., Holler, M., Feige, M. proteins. Proc. Natl. Acad. Sci. U. S. A. 110, J., Baerend, D., Lamb, D. C., and Buchner, J. 7666-7671 (2011) Substrate discrimination of the 73. Rome, M. E., Chio, U. S., Rao, M., chaperone BiP by autonomous and Gristick, H., and Shan, S. O. (2014) cochaperone-regulated conformational Differential gradients of interaction affinities transitions. Nat Struct Mol Biol 18, 150-158 drive efficient targeting and recycling in the 82. Mapa, K., Sikor, M., Kudryavtsev, V., GET pathway. Proc Natl Acad Sci U S A 111, Waegemann, K., Kalinin, S., Seidel, C. A., E4929-4935 Neupert, W., Lamb, D. C., and Mokranjac, D. 74. Mariappan, M., Mateja, A., Dobosz, M., (2010) The conformational dynamics of the Bove, E., Hegde, R.S., and Keenan, R.J. (2011) mitochondrial Hsp70 chaperone. Mol Cell 38, The mechanism of membrane associated 89-100 steps in tail-anchored protein insertion. 83. Schlecht, R., Erbse, A. H., Bukau, B., Nature 477, 61-66 and Mayer, M. P. (2011) Mechanics of Hsp70 75. Stefer, S., Reitz, S., Wang, F., Wild, K., chaperones enables differential interaction Pang, U.Y., Schwarz, D., Jorg, B., Hein, C., with client proteins. Nat Struct Mol Biol 18, Lohr, F., Bernhard, F., Denic, V., Dotsch, V., 345-351

15 84. Rosenzweig, R., Sekhar, A., Nagesh, J., Buchner, J., and Rapaport, D. (2018) and Kay, L. E. (2017) Promiscuous binding by Cytosolic Hsp70 and Hsp40 chaperones Hsp70 results in conformational enable the biogenesis of mitochondrial beta- heterogeneity and fuzzy chaperone- barrel proteins. J Cell Biol 217, 3091-3108 substrate ensembles. Elife 6 91. Papic, D., Elbaz-Alon, Y., Koerdt, S. N., 85. Deshaies, R. J., Koch, B. D., Werner- Leopold, K., Worm, D., Jung, M., Schuldiner, Washburne, M., Craig, E. A., and Schekman, M., and Rapaport, D. (2013) The role of Djp1 R. (1988) A subfamily of stress proteins in import of the mitochondrial protein Mim1 facilitates translocation of secretory and demonstrates specificity between a

mitochondrial precursor polypeptides. cochaperone and its substrate protein. Mol Downloaded from Nature 332, 800-805 Cell Biol 33, 4083-4094 86. Becker, J., Walter, W., Yan, W., and 92. Opalinski, L., Song, J., Priesnitz, C., Craig, E. A. (1996) Functional interaction of Wenz, L. S., Oeljeklaus, S., Warscheid, B., cytosolic hsp70 and a DnaJ-related protein, Pfanner, N., and Becker, T. (2018) http://www.jbc.org/ Ydj1p, in protein translocation in vivo. Mol Recruitment of Cytosolic J-Proteins by TOM Cell Biol 16, 4378-4386 Receptors Promotes Mitochondrial Protein 87. Mihara, K., and Omura, T. (1996) Biogenesis. Cell Rep 25, 2036-2043 e2035 Cytoplasmic chaperones in precursor 93. Harper, C. C., Berg, J. M., and Gould, S. targeting to mitochondria: the role of MSF J. (2003) PEX5 binds the PTS1 independently at CALIFORNIA INSTITUTE OF TECHNOLOGY on October 2, 2019 and hsp 70. Trends Cell Biol 6, 104-108 of Hsp70 and the peroxin PEX12. J Biol Chem 88. Young, J. C., Hoogenraad, N. J., and 278, 7897-7901 Hartl, F. U. (2003) Molecular chaperones 94. Harano, T., Nose, S., Uezu, R., Shimizu, Hsp90 and Hsp70 deliver preproteins to the N., and Fujiki, Y. (2001) Hsp70 regulates the mitochondrial import receptor Tom70. Cell interaction between the peroxisome 112, 41-50 targeting signal type 1 (PTS1)-receptor 89. Caplan, A. J., Cyr, D. M., and Douglas, Pex5p and PTS1. Biochem J 357, 157-165 M. G. (1992) YDJ1p facilitates polypeptide 95. Tripathi, A., Mandon, E. C., Gilmore, R., translocation across different intracellular and Rapoport, T. A. (2017) Two alternative membranes by a conserved mechanism. Cell binding mechanisms connect the protein 71, 1143-1155 translocation Sec71-Sec72 complex with 90. Jores, T., Lawatscheck, J., Beke, V., heat shock proteins. J Biol Chem 292, 8007- Franz-Wachtel, M., Yunoki, K., Fitzgerald, J. 8018 C., Macek, B., Endo, T., Kalbacher, H.,

16 Downloaded from http://www.jbc.org/ at CALIFORNIA INSTITUTE OF TECHNOLOGY on October 2, 2019

Figure 1 Multiple pathways form a robust network to deliver membrane proteins with diverse TMD topologies to the ER. Proteins with early TMDs are co-translationally targeted by the SRP and the SRP receptor to the Sec61p translocase and, in some cases, to the EMC complex (left pathway). Proteins with internal TMDs can be delivered by the SRP (dashed arrow a) or the less characterized SND components (middle pathway). Tail-anchored proteins harboring a late TMD are post-translationally targeted by the GET pathway to the Get1/2 receptors (right pathway), with the SND components serving as a backup route (dashed arrow b). Less hydrophobic TAs suboptimal for the GET pathway can be targeted to the ER via alternative pathways (dashed arrow c).

17 Downloaded from http://www.jbc.org/ at CALIFORNIA INSTITUTE OF TECHNOLOGY on October 2, 2019

Figure 2 Current model of TA targeting by the GET pathway. TAs released from the ribosome (step 1) are captured by Ssa1 (step 2), which effectively competes with the aggregation of TAs in the cytosol (dashed arrow a). Ssa1 transfers the bound TA to Sgt2 (steps 3-4) and then to Get3 (steps 5-6), which delivers TAs to the Get1/2 receptors for insertion into the ER membrane (step 7). Both of the TA transfers likely occur via concerted mechanisms enabled by interaction motifs or scaffold protein (Get4/5) that bridge the upstream and downstream chaperones (transfer complexes in brackets). Sgt2 provides a selection filter at which suboptimal TAs can be rejected (dashed arrow b) and re-sorted to alternative targeting pathways (dashed arrows c-e).

18 Downloaded from http://www.jbc.org/ at CALIFORNIA INSTITUTE OF TECHNOLOGY on October 2, 2019

Figure 3 Current model of the Sgt2•Get4/5•Get3 substrate transfer complex. Sgt2 is in blue, Get4 is in orange, Get5 is in red, Get3 is in yellow/tan, and the TA substrate is in green. The left inset shows the crystal structure of the Sgt2 TPR domain (cyan; PDB 3SZ7), superimposed onto the structure of the HOP TPR1 domain (grey) bound to the Hsc70 C-terminal peptide (magenta; PDB 2BYI). The lower inset shows the structure of the Sgt2 N-terminal domain (N) bound to the Get5 UBL domain (PDB 2LXC). The right inset shows the co-crystal structure of Get4/5N bound to ATP-loaded Get3 (PDB 4PWX), with the bound nucleotides in spacefill.

19 Downloaded from http://www.jbc.org/ at CALIFORNIA INSTITUTE OF TECHNOLOGY on October 2, 2019

Figure 4 Nucleotide, substrate, and interaction partners induced conformational changes in Get3 drive the targeted delivery of TAs to the ER. (A) Current model of the Get3 ATPase cycle. In the cytosol, ATP drives Get3 into the closed conformation (step 1). Get4/5 selectively recognizes closed Get3 and further inhibits its ATPase activity (step 2), which primes Get3 to capture the TA substrate from Sgt2 (steps 3-4). TA loading induces conformational breathing in Get3 to sample open states, which facilitates its release from Get4/5 (step 5). Without Get4/5 bound, the TA substrate activates ATP hydrolysis on Get3 (step 6). ADP-bound Get3•TA more extensively samples open conformations and is primed for capture and remodeling by the Get1/2 receptors at the ER (step 7). Lighter shaded molecules depict alternative conformations sampled by Get3. (B) Crystal structures of Get3 in the closed (left; PDB 2WOJ) and wide-open (right; PDB 3SJB) conformations. The ATPase and helical domains of Get3 are highlighted. Get3 is bound to - ADP•AlF4 in the closed state (left, spacefill) and to the cytosolic domain of Get1 (Get1CD) in the wide-open state (right, dark red). Green and red stars depict the approximate location of the FRET donor and acceptor dyes, respectively, used in the single molecule study in (54). (C) Sequential opening of Get3 upon substrate loading drive the ER targeting of TA. The FRET histograms of doubly labeled Get3 are compared at different stages of the GET pathway. Adapted with permission from (54).

20 Guiding Tail-anchored Membrane Proteins to the ER In a Chaperone Cascade Shu-ou Shan J. Biol. Chem. published online October 1, 2019

Access the most updated version of this article at doi: 10.1074/jbc.REV119.006197

Alerts: • When this article is cited Downloaded from • When a correction for this article is posted

Click here to choose from all of JBC's e-mail alerts http://www.jbc.org/ at CALIFORNIA INSTITUTE OF TECHNOLOGY on October 2, 2019