REVIEWS

Tail-anchored membrane protein

insertion into the endoplasmic
reticulum
Ramanujan S. Hegde* and Robert J. Keenan‡
Abstract | Membrane proteins are inserted into the endoplasmic reticulum (ER) by two
highly conserved parallel pathways. The well-studied co-translational pathway uses signal
recognition particle (SRP) and its receptor for targeting and the SEC61 translocon for
membrane integration. A recently discovered post-translational pathway uses an entirely
different set of factors involving transmembrane domain (TMD)-selective cytosolic
chaperones and an accompanying receptor at the ER. Elucidation of the structural and
mechanistic basis of this post-translational membrane protein insertion pathway highlights
general principles shared between the two pathways and key distinctions unique to each.

Chaperones All biological membranes contain a structurally diverse compartments of the secretory and endocytic pathways,
A large group of proteins that assortment of integral membrane proteins, which col­ and both nuclear envelope membranes9.
facilitate the folding, assembly, lectively constitute ~30% of the cellular proteome 1,2. Insertion into the ER membrane can occur either
transport and degradation of These proteins impart essential functionality to the lipid co-translationally or post-translationally, each of which
non-native polypeptides by
minimizing inappropriate
bilayer to allow a range of cellular activities, includin­g offers distinct advantages10,11. In the co-translational
interactions. transmembrane communication, transport and mem­ pathway, all of the steps from initial protein recognition
brane morphogenesis. The selective and asymmetric to final insertion into the membrane occur during pro­
insertion of membrane proteins is therefore an evo­­lu­ tein synthesis. By contrast, targeting and insertion via
tionarily ancient problem that was solved by the earlies­t post-translational pathways occur after complete syn­
life forms. thesis of the membrane protein substrate. Thus, the ribo­
The shared feature of all integral membrane pro­ some is a major functional component during all steps
teins is the highly hydrophobic transmembrane domain of co-translationa­l insertion, whereas its role in post-
(TMD), which in the final structure resides within the translationa­l pathways is limited to the very earliest steps.
lipid bilayer 3. Thus, a critical obstacle in membrane Although the co-translational pathway was discovered
protein insertion is the movement of these TMDs from over 30 years ago12–14 and has been extensively studied
the aqueous cytosol, where they are synthesized, into the in many systems15–19, the post-translational insertion
lipid bilayer, where they are energetically most stable4. pathway has only recently come into focus. Similarly to
*Medical Research Council This process necessitates selective TMD recognition, the post-translational translocation of soluble proteins
(MRC) Laboratory of shielding of the TMD from the aqueous cytosol, target­ into various organelles20–24, the basic paradigm of post-
Molecular Biology,
Hills Road, Cambridge,
ing to the membrane surface and integration of the translationa­l ER membrane protein insertion involves
CB2 0QH, UK. TMD into the lipid bilayer in the correct orientation. cytosolic chaperones (mediating recognition and shield­
‡
Department of Biochemistry All membrane protein insertion pathways must solve ing) that interact with a specific ER‑localized receptor
and Molecular Biology, these four problems, each of which typically involves (mediating targeting and insertion). Here, we review
The University of Chicago,
specialized and highly regulated factors in the cytosol this post-translational pathway and discuss how the
Gordon Center for Integrative
Science, Room W238, and target membrane. problems of recognition, shielding, targeting and inser­
Chicago, Illinois 60637, USA. In eukaryotes, membrane proteins synthesized tion are solved by its machinery. As we outline, there is
e-mails: on cytosolic ribosomes can be targeted to mitochon­ now sufficient information about each step to provide a
[email protected]; dria5, peroxisomes6, chloroplasts (in plants)7 and the plausible mechanistic framework for the whole pathway.
[email protected]
doi:10.1038/nrm3226
endoplasmic reticulum (ER)8. Among these, the ER These insights are starting to reveal common themes
Published online accommodates the largest number of proteins, encom­ that are likely to apply to all membrane protein insertio­n
16 November 2011 passing all membrane proteins of the plasma membrane, processes.

NATURE REVIEWS | MOLECULAR CELL BIOLOGY VOLUME 12 | DECEMBER 2011 | 787

© 2011 Macmillan Publishers Limited. All rights reserved

REVIEWS

C 542 After this initial recognition step, the SRP-bound

%
'4NWOGP TGEGRVQT 5'% ribosome–nascent chain complex is targeted to the
SRP receptor 32,33 at the ER membrane. The ribosome–
%[VQUQN )62 )62 +PVGITCVKQP 0 nascent chain complex is then released from SRP and
4GNGCUG transferred to the SEC61 complex, the central compo­
)62 )62 nent of a protein translocon in the ER16,34,35. The ribosome
subsequently completes the synthesis of the membrane
4GE[ENKPI 6CTIGVKPI protein while remaining bound to the SEC61 translocon.
)&2
The SEC61 complex is therefore positioned to recognize
0CUEGPVEJCKP 5KIPCNRGRVKFG
542 each TMD as it emerges from the ribosome and facilitate
their integration into the lipid bilayer 36–38.
A major advantage of the co-translational strategy is
4KDQUQOG 4GEQIPKVKQP
that the machinery for targeting and insertion is physi­
cally coupled to the ribosome near the polypeptide exit
D )GVs)GV tunnel. This spatial organization allows TMDs to be
'4NWOGP %
recognized, shielded and inserted with minimal expo­
sure to the bulk cytosol. The co-translational insertion
%[VQUQN #&2 #&2 +PVGITCVKQP
machiner­y therefore enjoys a considerable competitive
#&2 #&2 0 advantage in binding TMDs over many of the other
4GNGCUG potential binding partners in the cell.
More importantly, especially for multi-spanning
membrane proteins, the handling of TMDs as they
4GE[ENKPI 6CTIGVKPI
2TGVCTIGVKPI emerge from the ribosome substantially obviates a need
EQORNGZ
6/&
to maintain the solubility of highly hydrophobic, lengthy
and complicated proteins. Because the machinery
for TMD insertion is the same as that which mediates
#62 #62
#62 #62 the translocation of soluble proteins, membrane pro­
4GEQIPKVKQP )GV
4KDQUQOG 64% teins with large soluble domains that undergo trans­
Figure 1 | Membrane protein biosynthesis in eukaryotes. a | In the co-translational location do so by the co-translational pathway. Thus, a
pathway for the insertion of endoplasmic reticulum (ER) membrane proteins, signal near-universa­l theme is that membrane proteins with
0CVWTG4GXKGYU^/QNGEWNCT%GNN$KQNQI[
recognition particle (SRP) recognizes the hydrophobic signal peptide of the nascent multiple TMDs or large translocated domains use the co-
chain as it emerges from a translating ribosome. The ribosome–nascent chain– translationa­l mode of translocation. These proteins con­
SRP complex is targeted to the membrane by a GTP-dependent interaction with stitute the majority of membrane proteins in the cell, and
the SRP receptor, resulting in the release of the signal peptide and docking of the co-translational membrane protein insertion has been
ribosome–‌nascent chain complex to the SEC61 translocon. Translation then resumes, described for the ER, the topologically equivalent bacte­
and the nascent polypeptide is inserted into the membrane bilayer. After GTP hydrolysis, rial plasma membrane39, the mitochondrial inner mem­
SRP is recycled to the cytosol. b | In the post-translational pathway for the insertion of
brane40 and the thylakoids of chloroplasts41. The SEC61
tail-anchored (TA) ER membrane proteins, a soluble pre-targeting complex captures the
complex also mediates the post-translational transloca­
hydrophobic transmembrane domain (TMD) of the TA substrate after it emerges from
the ribosomal exit tunnel. After loading onto Get3 (TRC40 in mammals), the TA substrate tion of soluble proteins in bacteria and eukaryotes10,11; as
is targeted to the ER membrane by interaction with the Get1–Get2 receptor complex. this pathway is not known to mediate membrane protein
After ATP has been hydrolysed, the TA substrate is released for insertion into the bilayer. insertion, it is not considered in this Review.
ATP binding recycles Get3 (or TRC40) back to the cytosol. N, amino terminus.
Discovery of a new insertion pathway
As the SRP-dependent pathway was being elucidated,
The co-translational pathway it became clear that at least some membrane proteins
The extensively studied co-translational pathway pro­ might not be able to use this route for insertion. In par­
vides an important conceptual context for understand­ ticular, the apparently obligate recognition of membrane
ing post-translational membrane protein insertion. It is proteins by SRP during synthesis was noted to be incom­
therefore worth first summarizing the general features patible with proteins that have a single TMD near the
of co-translational membrane protein insertion8,11,25,26. carboxyl terminus42. This is because the TMD would be
This pathway begins when a hydrophobic segment of inside the ribosomal tunnel (which houses ~40 amino
the protein, typically the first TMD, emerges from the acids of the nascent polypeptide) when the termination
Ribosomal exit tunnel ribosome (FIG. 1). This hydrophobic domain is recog­ codon was reached. This means that TMD recognition
An internal channel in the
nized by signal recognition particle (SRP)27, a large would be required to occur after the termination of
large subunit of the ribosome
through which the nascent ribo­nucleoprotein complex that is composed of multiple translation (that is, post-translationally).
polypeptide travels before proteins and an RNA scaffold28. SRP has a high affin­ Among the first of these tail-anchored (TA) protein­s to
emerging into the cytosol. ity for ribosomes and binds these, through its 54 kDa be examined for its SRP-dependence was synapto­brevin
Various factors bound to the sub­unit (SRP54), near the ribosomal exit tunnel29. The — a SNARE protein that has key roles in intra­cellular
ribosome surface can affect
the folding and/or targeting
Met-rich domain (M‑domain) of SRP54, which directly vesicular trafficking. It was rigorously demonstrated
of the nascent polypeptide as binds to hydrophobic domains30,31, is therefore precisely that, as predicted, synaptobrevin uses an SRP- and
it emerges from the exit tunnel. poised to captur­e the nascent membrane protein. SEC61‑independent post-translational pathway for its

788 | DECEMBER 2011 | VOLUME 12 www.nature.com/reviews/molcellbio

© 2011 Macmillan Publishers Limited. All rights reserved

 REVIEWS

Box 1 | Other routes into the ER membrane

in vitro observations55,56, each of which was eventually
explained by subsequent mechanistic studies.
In the absence of the Get–transmembrane domain-recognition complex (TRC) system, First, TRC40 associated with the TMDs of TA pro­
many tail-anchored (TA) proteins are still able to insert into a membrane in vitro and teins but not membrane proteins with internal TMDs55,56.
in vivo with at least some efficiency. This raises the possibility of additional pathways Second, it was an ATPase (which reconciled the ATP-
for TA protein insertion into the endoplasmic reticulum (ER), including an unassisted
dependence of TA protein membrane insertion).
pathway44,50,51, a chaperone-mediated pathway involving heat shock protein 70
(HSP70)52, and a pathway using signal recognition particle (SRP) in a post-translational
Third, it was highly conserved across all eukaryotes,
mode53. The mechanisms of these possible routes are not considered here in detail. was essentia­l in mammals59, and mutant phenotypes in
Nevertheless, it is worth considering whether these are bona fide insertion pathways other organisms were related to processes that involved
that are normally operational in vivo. TA proteins60–62. Fourth, an ATPase-deficient mutant
Several observations suggest that these may be ad hoc insertion mechanisms that acted as a dominant-negative, selectively binding to but
are only detectable under specialized conditions. First, deletion of Get components in not releasing TA proteins, thereby precluding their inser­
Saccharomyces cerevisiae produces TA proteins that are substantially aggregated and/‌or tion into the membrane55. Fifth, a fraction of TRC40 was
mislocalized65–68. Thus, although enough essential TA proteins do manage to insert into found on the ER membrane, and this could be released
the ER to maintain viability, target specificity and insertion efficiency are compromised by ATP55. And finally, it contained a proteinaceous bind­
for all substrates that have been examined. Second, essentially all of the evidence for
ing site (or sites) on ER microsomes55. Based on these
these pathways derives from in vitro analyses that use one membrane, so targeting
specificity is not assayed. Although these systems are powerful, the interpretation of
observations, TRC40 was proposed to be a targeting
in vitro studies merits some caution in the absence of further corroboration. Additionally, facto­r that selectively recognizes TA proteins in the
the use of crude translation lysates and microsomes poses a substantial problem for cytosol and delivers them to the ER for insertion in an
interpretation because it is now clear that they contain the Get–TRC-targeting ATP-dependent manner 55.
machinery55,66. Even if cytosol is replaced by purified factors52, the addition of microsomes The high level of conservation readily identified Get3
contributes substantial amounts of the Get–TRC machinery (including Get3; TRC40 in (guided entry of TA proteins 3; originally known as Arr4
mammals). Thus, any factors that can temporarily prevent TA proteins from aggregating (REF. 63), again owing to its similarity to ArsA) as the
may seem to be necessary simply by facilitating capture by the Get–TRC pathway. budding yeast homologue of TRC40. Synthetic genetic
And finally, in vitro crosslinking assays (the primary means of detecting potential and physical interaction studies had already defined
targeting factors) can lead to numerous minor ‘off-pathway’ interactions, given the
GET3 as part of a pathway involving at least two other
rather hydrophobic TA substrate and lengthy reaction times. Thus, although it is possible
that other specific pathways exist, the rigorous validation and demonstration
genes (termed GET1 and GET2), the loss of which led
of their physiological importance await additional studies. to phenotypes that were consistent with a role in Golgi–
ER trafficking (hence the original delineation with the
acrony­m Get)60,61,64. However, the physical and functional
insertion into the ER43. Although this and subsequent links of the mammalian homologue, TRC40, to TA pro­
studies showed that insertion was protein- and ATP- tein insertion into membranes, combined with the fact
dependent43–46, the molecular basis of this energy require­ that Golgi–ER trafficking depends on TA proteins, led to
ment and the factors involved in insertion remained a reassessment of the Get pathway. Parsimoniously, all of
obscure for over 10 years. the otherwise unconnected yeast phenotypes associated
During this intervening period, TA proteins were with the Get pathway were reconciled as secondary to
increasingly appreciated to be of broad physio­logical defects in TA insertion, meriting a change to the cur­
importance. Representing ~3–5% of all membrane pro­ rent ‘guided entry of TA proteins’ moniker for the GET
teins47–49, TA proteins are found in all cellular membranes genes65. Subsequent physical and genetic inter­action
and have functions that range from membrane biogen­ analysis of this Get pathway in yeast led to the identi­
esis to apoptosis, vesicular trafficking, protein degradation fication of three additional factors (termed Get4, Get5
and many others. With increased interest in this class of and small Glu-rich tetratricopeptide repeat-containing 2
Translocon proteins, greater attention was paid to their mechanism (Sgt2))66–68. This defined the major players in a single
A membrane channel that is of insertion. However, studies of different TA proteins by experimental system of budding yeast and placed them
associated with the transport
different methods led to diverse conclusions. Proposals into either early (cytosolic for Get3, Get4, Get5 and Sgt2)
of polypeptides into or across
cellular membranes. included one of an unassisted mechanism not requirin­g or late (membrane for Get1, Get2 and Get3) steps that
any insertion machinery (for cytochrome b5 studied made genetic sense. Subsequent insights into the mecha­
SNARE in vitro)44,50,51, one of a heat shock protein 70 (HSP70)- nistic roles of these factors have come from a combina­
(Soluble NSF mediated pathway 52, and one in which SRP and SEC61 are tion of structural and functional studies, primaril­y of the
(N‑ethylmaleimide-sensitive
factor) attachment protein
used in a post-translational mode53,54. The physio­logical yeast Get pathway.
(SNAP) receptor). A family relevance of these potential routes to the membrane
of tail-anchored coiled-coil remains largely unclear at prese­nt (BOX 1). Substrate recognition by Get3
proteins that regulate fusion Eventually, biochemical analysis of TA protein Critical to TA protein targeting is its selective and effi­
reactions and target specificity
insertion in cell-free translation extracts, combined cient recognition by Get3. The sensitivity of this inter­
in vesicle trafficking.
with protein crosslinking approaches, led to the identi­ action to detergent and its dependence on the presence
Heat shock protein 70 fication of a factor that associated with the TMDs of of a functional TMD strongly suggested a direct recogni­
(HSP70). A ubiquitous family TA proteins55,56 (FIG. 1). This factor, originally annotated tion of the TMD via hydrophobic interactions with Get3
of ~70 kDa heat-shock Asna1 (REF. 57) (for its similarity to ArsA, an arsenite- (REF. 55). Such an interaction could also shield and main­
proteins that serve as
molecular chaperones to
transportin­g ATPase in bacterial systems58), was renamed tain the solubility of the hydrophobic TMD as it transits
regulate polypeptide folding, TMD-recognition complex protein of 40 kDa (TRC40). through the cytosol. Insight into both TA substrate
translocation and degradation. Evidence for its role in TA insertion came from several recognition and shielding came from structural studies.

NATURE REVIEWS | MOLECULAR CELL BIOLOGY VOLUME 12 | DECEMBER 2011 | 789

© 2011 Macmillan Publishers Limited. All rights reserved

REVIEWS

2CTVKCNN[ENQUGF
1RGP
CRQ
/I s#&2
4GNGCUG
#&2

#&2 #&2

/GODTCPG
GXGPVU
2
4GE[ENKPI
6CTIGVKPI
#62

%[VQUQNKE
αJGNKECN GXGPVU

#62 #62 6/& #62 #62

4GEQIPKVKQP
#62CUG
%NQUGF

/I s#62

Figure 2 | Nucleotide-dependent conformational changes in the Get3 homodimer. Each Get3 monomer comprises
two distinct regions: an α‑helical subdomain and an ATPase subdomain. In the presence0CVWTG4GXKGYU^/QNGEWNCT%GNN$KQNQI[
of ATP, the Get3 helical subdomains
become intimately associated, forming an extended composite hydrophobic groove (FIG. 3) that recognizes and binds to
the transmembrane domain (TMD) of a tail-anchored (TA) substrate. ATP hydrolysis, which occurs at some stage before the
release of the TA substrate at the endoplasmic reticulum (ER) membrane, produces an ADP-bound Get3 homodimer that is
partially closed. Following the release of the TA substrate and ADP, Get3 shifts back to an open conformation. Subsequently,
ATP binding allows recycling of Get3 back to the cytosol in a closed conformation. The insets show crystal structures of the
fungal Get3 homodimer in the nucleotide-free state (Protein Databank (PDB) ID 2WOO), the Mg2+–ADP-bound state
(PDB ID 3IQX) and the Mg2+–ADP–AlF4–-bound state (which seems to mimic the ATP-bound state; PDB ID 2WOJ).

Near-simultaneous reports of Get3 crystal struc­ recognition element in Ffh (the bacterial SRP54 homo­
tures from multiple fungal species and in multiple states logue) is constructed from an α‑helical protein scaf­
revealed that Get3 was a symmetric homodimer 69–73. fold that presents a large hydrophobic surface area for
Each Get3 monomer comprises a core ATPase domain substrate binding 74 (FIG. 3). Moreover, these scaffolds are
decorated with an α‑helical domain. The arrange­ highly dynamic. As two crystal structures of the SRP54–
ment of Get3 subunits depends on the nucleotide state, signal peptide complexes show, this flexibility can be
transitioning from a fully open state in the absence of leveraged to accommodate targeting signals of differ­
nucleotide to a fully closed conformation in the presence ent lengths and sequence75,76. A similar mechanism is
of Mg 2+–ADP–AlF4– (which seems to mimic the ATP- probably at play in the case of Get3, although this awaits
bound state) and a partially closed state in the presence furthe­r structural analysis of Get3–TMD complexes.
of Mg 2+–ADP (FIG. 2). In contrast to the relatively rigid In addition to flexibility in accommodating different
conformation of the ATPase domain, the conformation sequences, substrate recognition by Get3 must also be
of the α‑helical domain is sensitive to nucleotide bind­ selective in at least two ways. First, the C‑terminal TMDs
ing. In the fully closed, ATP-bound state, the helica­l of TA proteins must be distinguished from the internal
subdomains are in direct contact and define a large, TMDs of co-translational substrates. Second, Get3 must
hydrophobic groove that spans both Get3 monomers71 avoid the TMDs of TA proteins destined for other orga­
(FIGS 2,3). nelles (such as peroxisomes, mitochondria and chloro­
Three lines of evidence illustrated that the helical plasts). As there is little difference between the TMDs
domains mediate recognition. First, the size, shape, of these different substrates77, a key issue is how the Get
hydrophobicity, flexibility and the ATP-dependent form­ pathway selects only the correct substrates for targeting.
ation of the composite groove argued for this being the One clue comes from the observation that, in the
site of TMD binding 69–72. Second, hydrogen exchange absence of SRP (or other competing factors), Get3 can
mass spectrometry (HX-MS) studies demonstrated bind substrates containing internal TMDs. Conversely,
protection of the α‑helical subdomains upon binding SRP cannot recognize a TMD it normally binds when that
to a TA substrate69. Third, perturbing the composite same TMD is near the C terminus. This suggests that,
hydrophobic groove by introducing negative charges or under physiological conditions, SRP binding to inter­
disrupting dimerization reduced TA substrate binding nal TMDs is strongly favoured by its association with a
in vitro and led to growth defects in vivo71. translating ribosome. Hence, its very high local concen­
The process of substrate recognition in the co- and tration near the ribosomal exit tunnel ensures that SRP
post-translational pathways shows important func­ will out-compete any other available binding proteins,
tional and mechanistic similarities. As with Get3, the such as Get3.

790 | DECEMBER 2011 | VOLUME 12 www.nature.com/reviews/molcellbio

© 2011 Macmillan Publishers Limited. All rights reserved

 REVIEWS

the problem of how it could capture TA proteins after

C
their release from the ribosome in a sufficiently timely
manner to avoid inappropriate interactions and aggrega­
tion. Second, it seemed unlikely that the conformation
of Get3 that binds TA proteins, which exposes a large
hydrophobic surface, would be favoured or long-lived
in the aqueous cytosol. Third, it was unclear how Get3
could outcompete other chaperones in the cytosol that
also bind hydrophobic domains. Thus, a gap existed in
our knowledge between TA protein release from the
ribosome and subsequent recognition by Get3.
Genetic analysis of budding yeast had suggested that
D cytosolic cofactors Get4, Get5 and Sgt2 affected this pro­
cess, although it was unclear how 66,67. Insight into this
problem came from parallel biochemical studies in the
yeast and mammalian systems that converged in sup­
porting a conserved role for pre-targeting cofactors in
aiding substrate capture by Get3 or TRC40 (FIG.  4).
In yeast, Sgt2 was observed to bind directly to the TMDs
of TA proteins, and this interaction was critical for their
loading onto Get3 (REF. 79). Importantly, transfer of TA
proteins from Sgt2 to Get3 required the Get4–Get5
subcomplex. Interaction analysis further showed that
Get4–Get5 forms a scaffold that bridges Sgt2 (which
binds to Get5) and Get3 (which binds to Get4).
Importantly, Get4 may favour binding selectively to the
ATP-bound (and hence closed) conformation of Get3
(REF. 80). This means that Get3 is recruited to Get4–Get5
selectively in a conformation that exposes its hydrophobi­c
Figure 3 | Substrate recognition by the post- and co-translational targeting
0CVWTG4GXKGYU^/QNGEWNCT%GNN$KQNQI[
machinery. a | A large groove in Get3 is formed by the association of the two α‑helical TMD binding groove. Thus, an attractive model is that
subdomains (left panel, green and blue) in the ATP-bound, closed dimer conformation. substrates are transferred from Sgt2 to Get3 through the
This composite groove presents a large hydrophobic surface (yellow) for binding to the ability of Get4–Get5 to selectively recruit the correct con­
transmembrane domain (TMD) of a tail-anchored (TA) protein substrate. The TA protein formation of Get3 in proximity to substrate-bound Sgt2.
SEC61β (shown in red in the right panel; Protein Databank (PDB) ID 1RHZ) is modelled Precisely how this intricate handover occurs remains
inside the groove. b | Crystal structures of the Met-rich domain (M‑domain) of signal to be investigated.
recognition particle 54 kDa (SRP54) bound to a signal peptide (shown in red) (top panel, A similar process seems to operate in the mammalian
PDB ID 3KL4; bottom panel, PDB ID 3NBD). Note the different, but overlapping, peptide system78. In this case, an assay for substrate capture by
binding sites. For both the post-translational TA protein pathway (a) and the TRC40 was used to illustrate the need for other factors.
co-translational pathway (b), the dynamic properties of these helical, hydrophobic
Purification of a requisite factor revealed a three-protein
scaffolds probably allow them to accommodate different sequences during targeting.
complex composed of BAG6 (also known as BAT3 and
Scythe), TRC35 and UBL4A. TRC35 and UBL4A are
homologous to Get4 and Get5, respectively, and BAG6
Distinguishing ER‑destined TA proteins from other TA can interact with TA proteins, similarly to Sgt2 (although
proteins is a more difficult problem because they cannot no primary sequence homology is apparent). Depletion
be discriminated on the basis of topological constraints. of the BAG6 complex resulted in defective TA protein
In vitro crosslinking studies in mammalian translation capture by TRC40 (REF. 78) and reduced membrane
extracts show a clear dependence on hydrophobicity for protein insertion efficiency81. Thus, by homology and
TRC40 association, with even modes­t decreases abolish­ by functional analysis, pre-targeting cofactors facilitate
ing the interaction78. But it is difficult to envision how the substrate loading onto Get3 and TRC40 in the yeast and
flexible hydrophobic groove of Get3 or TRC40 could, mammalian systems, respectively.
by itself, provide tight discrimination between closely Although mammalian BAG6 and yeast Sgt2 are
related TA substrates. Instead, as described below, addi­ not related, it is noteworthy that BAG6 interacts with
tional cofactors acting in conjunction with Get3 or TRC40 SGTA 82, the mammalian homologue of Sgt2. This
probabl­y enhance the fidelity of substrate recognition. suggests a more parsimonious model, in which the
yeast and mammalian systems are even more similar
Cofactors for substrate loading than previously thought (FIG. 4). Get4–Get5 (in yeast)
Although the structure of Get3 explained how it could or TRC35–UBL4A–BAG6 (in mammals) is the scaf­
bind and shield the TMD of a TA protein, it was unclear fold that dynamically brings Sgt2 or SGTA into close
how the TA protein could get loaded onto Get3 in the proximity with Get3 or TRC40. This large and trans­
first place. At least three issues were especially problem­ ient assembly is loosely defined as the TRC, within
atic. First, Get3 did not appear to bind ribosomes, raising which a substrate would be sorted among these factors.

NATURE REVIEWS | MOLECULAR CELL BIOLOGY VOLUME 12 | DECEMBER 2011 | 791

© 2011 Macmillan Publishers Limited. All rights reserved

REVIEWS

;GCUV /COOCNU including the recruitment of other chaperones, such

5WDUVTCVG as Hsp104 and Hsp70, both of which bind to Sgt2
%JCRGTQPGU (REF. 79). The precise nature of this key sorting step
)GV 5IV 64% 5)6# remains an important area of study, which will be
greatly aided by the determination of the structures
)GV of complexes combined with structure-based muta­
)GV 64%
genesis studies. The hydrophobic transfer process is
$#) 7$.# likely to be highly co­ordinated so that exposure of the
TMD to the cytosol is minimized.
The substrate recognition problem is, in many
ways, qualitatively different and considerably more
complex for post-translational substrates than for co-
translational ones. In the co-translational case, precise
positioning of the SRP54 M-domain at the ribosome
exit tunnel greatly simplifies the recognition problem
to one of linear scanning. This not only reduces com­
petition with other factors but also limits the degrees of
7PMPQYPFGUVKPCVKQP freedom for both the substrate and the targeting factor.
'4VCTIGVKPI &GITCFCVKQP
OKVQEJQPFTKC! By contrast, post-translational substrates are access­
Figure 4 | TA protein sorting by the TRC. The transmembrane domain (TMD)- ible to a large number of highly abundant chaperones
recognition complex (TRC) consists of a stable core complex (shown in grey) and several and co-chaperones, all of which primarily interact
dynamically associated components. In budding yeast, the TRC core consists of Get4 with proteins on the basis of their hydrophobicity.
0CVWTG4GXKGYU^/QNGEWNCT%GNN$KQNQI[
and Get5, with Get4 recruiting Get3 and Get5 recruiting small Glu-rich tetratricopeptide Nonetheless, effective sorting is achieved among these
repeat-containing 2 (Sgt2). Sgt2 can additionally recruit other chaperones. In mammals, different factors. TA protein sorting by the TRC may
TRC35 (which is homologous to Get4) and UBL4A (which is homologous to Get5) are
therefore provide a useful paradigm for understanding
in a complex with BAG6, which probably recruits SGTA (which is homologous to Sgt2).
Engagement of the TRC by a substrate (probably bound to Sgt2 or SGTA) results in protein triage among chaperone systems in general.
its sorting among any of several potential TMD-binding proteins (dashed arrows). This
sorting is presumably dictated by a combination of substrate features and availability of Protein capture at the ribosome
the binding partners. The substrate can therefore emerge from the TRC bound to any Although the existence of a pre-targeting factor helps
of multiple binding partners, each of which imparts a specific downstream fate. Get3 to explain how substrates might be sorted and loaded
or TRC40 association mediates endoplasmic reticulum (ER) targeting, whereas BAG6 onto Get3, the issue of how TA proteins are first cap­
binding can recruit an E3 ubiquitin ligase that mediates substrate degradation. The fates tured on release from the ribosome remains unclear.
of other complexes are not understood but could include targeting to other destinations, Some insight into this initial step comes from bio­
including the mitochondria. chemical analysis in the mammalian system, in which
the BAG6 complex was observed to interact with
ribosomes78. Its recruitment there might be mediated
A committed targeting complex would be generated by TRC35 or UBL4A, the yeast homologues of which
only if the substrate binds productively with Get3 or (Get4 and Get5) were found to be weakly associated
TRC40, a fate favoured by ER‑destined TA proteins. If with ribosomes in a proteomic analysis86.
the substrate is unsuitable for ER targeting, it would be Remarkably, recruitment of the BAG6 complex
transferred to other factors that impart alternative fates. to ribosomes was strongly favoured by the presence
These might include chaperones that are specific for of a TMD inside the ribosomal tunnel78. The impli­
other destinations, such as the mitochondria, or qualit­y cation of this observation is that the BAG6 complex
control factors that mediate degradation. Indeed, recent may be located favourably for initial substrate capture
work suggests that BAG6 is precisely such a quality con­ when the substrate is released from the ribosome. But,
trol factor that can bind a range of hydrophobic sub­ because TA protein release would occur very soon
strates, recruit an E3 ubiquitin ligase and route them after the TMD is synthesized, it was unclear how the
for proteasoma­l degradation83–85. BAG6 complex could be recruited to such ribosomes
Thus, the picture that is emerging is one of initia­l in time. This seems to be aided by a TMD-dependent
hydrophobic protein capture at the ribosome, followed delay in translation termination78.
by assembly into a highly dynamic sorting complex How the termination of translation might be con­
(the TRC) containing many potential binding partners. trolled in a substrate-specific manner remains com­
The substrate would then partition among the bind­ pletely obscure, although it has been observed in other
ing partners, with the final outcome dependin­g on the contexts87. Similarly, precisely how sequences inside
specific features of the substrate. TA proteins destined the ribosome could influence events at the surface to
for the ER would be transferred to Get3 or TRC40, promote recruitment of the BAG6 complex remains a
whereas other hydrophobic proteins would have mystery. It is possible that the presence of hydro­phobic
alternative fates. In this view, the evolutio­n of BAG6 sequences inside the ribosomal tunnel subtly alters
(which does not have an obvious yeast homologue) ribosome conformation in a manner that is exploited
could have occurred to allow enhanced quality control, by the BAG6 complex. A similar explanation has been
whereas yeast may use alternative mechanisms, put forward for how SRP might be recruited88,89 and how

792 | DECEMBER 2011 | VOLUME 12 www.nature.com/reviews/molcellbio

© 2011 Macmillan Publishers Limited. All rights reserved

 REVIEWS

sequences inside the ribosome can influence its inter­ binding to Get3, was defective for TA protein insertion
action with the SEC61 translocon90–92. Alternatively, into the membrane95. These observations suggested that
the BAG6 complex might be recruited to ribosomes Get2 functions to recruit the Get3–TA protein targeting
by signals that are present in the mRNA that encodes complex to the membrane.
the TA protein. Indeed, studies indicate that cis-actin­g The structure of nucleotide-free Get3 bound to
sequences in the TMD-coding region of bacterial the Get1 cytosolic fragment revealed two Get1 frag­
membrane protein mRNAs can direct these mRNAs ments bound to equivalent sites on opposite faces of
to the plasma membrane93,94. The initial capture step the open Get3 dimer 95,97 (FIG. 5). Strikingly, each Get1
by the BAG6 complex remains to be studied in mecha­ coiled-coil inserts itself between the two Get3 subunits
nistic detail, as does the apparent regulation of transla­ to completely disrupt the closed dimer interface. This
tional termination and the relationship between these observation immediately suggested that Get1 functions
two processes. to release substrate from Get3. Consistent with this,
functional analysis showed that Get1, but not Get2, pro­
Targeting and release at the ER motes substrate release95,96. Moreover, Get1 was unable
After a TA protein is successfully loaded onto Get3 by to promote substrate release from an ATPase-deficient
the action of the TRC, the resulting Get3–TA protein Get3 variant (in which Asp57 was replaced with Asn),
complex must next be targeted to the ER. In budding suggesting that it functions on a Get3–TA substrate
yeast, genetic studies have indicated that Get1 and Get2, complex in which the ATP has already been hydrolysed.
both of which are multi-spanning ER membrane pro­ The crystal structures also provided key insights into
teins, are needed for targeting 65. Furthermore, their how targeting and substrate release are coordinated by
abilit­y to form a complex with Get3 in the absence of the two receptor subunits. The Get1- and Get2‑binding
other factors61,66 suggested that these three proteins sites on Get3 are partially overlapping, and interaction
could be the minimal factors required for targeting, and analyses showed that the receptor subunits compete
possibly insertion into, the ER membrane. for binding to Get3 (REF. 95), which is consistent with a
Insight into the role of Get1 and Get2 came from sequential handover mechanism (FIG. 5). A complex of
recent reconstitution studies95,96. Genetic and bio­ Get3 bound simultaneously to portions of Get2 and Get1
chemical depletion and add-back experiments estab­ can be detected at high concentrations by NMR, and this
lished that Get1 and Get2 are each indispensable for may represent the transient intermediate during hand­
Get3‑dependent insertion of the TA substrate into the over 97. Taken together, these studies suggest a model in
membrane95,96. Remarkably, efficient targeting and inser­ which Get2 recruits the Get3–TA substrate targeting
tion could be achieved in proteoliposomes containing complex, with Get3 in a closed dimer conformation, and
only recombinant Get1 and Get2 at physiological con­ subsequently transfers it to Get1, which drives substrate
centrations95. Thus, Get1 and Get2 are both necessary release by disrupting the composite hydrophobic groove
and sufficient for the membrane-associated events of TA and stabilizing the open state of Get3.
protein targeting and insertion. The stoichiometry of the Get1–Get2 receptor com­
The availability of a simple reconstituted system plex remains to be established. The simplest possibility
using completely purified, recombinant components in view of the crystal structures and the symmetric Get3
permitted a detailed analysis of how these two mem­ dimer is that two Get1 and two Get2 subunits form a
brane proteins interact with, and regulate the func­ heterotetrameric assembly. The resulting high-avidity
tion of, Get3 (REF. 95). Interaction analysis illustrated interaction of two Get1 subunits with the Get3–TA
that Get1 and Get2 associate through their membrane substrate complex would facilitate substrate release at
domains and that they each interact with Get3 via their physio­logical concentrations. Nevertheless, a hetero­
prominent (non-homologous) cytosolic domains95,96. dimeric receptor is also plausible, with Get2 binding
The unusual feature of a receptor that interacts with to one side of the Get3 heterodimer and Get1 binding to
Get3 in two different ways suggested that these two the other. Evidence that this mechanism could work is
interactions serve distinct functions: targeting of the provided by the finding that artificially heterodimerized
Get3–TA protein complex to the ER, followed by sub­ Get1–Get2 cytosolic domains can mediate substrate
strate release at the ER membrane. This indeed proved release in vitro96.
to be the case, as revealed by a combination of structural This critical step of releasing a substrate from its
and functional studies. tightly bound targeting factor at the correct place and
Insight into both steps was provided by structures of time is a problem faced by all targeting pathways. In the
the complexes that form between Get3 and the cytosolic TA protein pathway, substrate release is first obliga­torily
Get1 and Get2 receptor fragments95,97 (FIG. 5). The struc­ ‘primed’ by nucleotide hydrolysis, whereas its actual
ture of the amino‑terminal end of the Get2 cytosolic release is promoted by the Get3–Get1 interaction95,96.
fragment in complex with Mg2+–ADP–AlF4–-bound Get3 This two-step mechanism is similar in concept, albeit
showed two Get2 fragments bound to equivalent sites on different in details, to the SRP-mediated co-translational
opposite faces of the closed Get3 dimer. Importantly, the insertion pathway. Here, a substrate bound to SRP54 is
Get3 hydrophobic groove was intact and accessible, with released in two successive steps, one involving nucleotide
the N‑terminal ends of Get2 tether­ed to the membrane binding and the other involving receptor interaction. The
by a long, flexible linker. A Get1–Get2 complex contain­ nucleotide-dependent step involves GTP binding to both
ing a structure-based mutation in Get2 that disrupts SRP and its receptor to allow targeting 98,99. The second

NATURE REVIEWS | MOLECULAR CELL BIOLOGY VOLUME 12 | DECEMBER 2011 | 793

© 2011 Macmillan Publishers Limited. All rights reserved

REVIEWS

)GVs)GV
%
'4

%[VQUQN
)GV
0
#&2 #&2
6#UWDUVTCVG
)GV
0
#&2 #&2

)GV

Figure 5 | Targeting and substrate release at the ER membrane. Nucleotide- (either ADP or ATP) and tail-anchored
(TA) substrate-bound Get3 is captured at the endoplasmic reticulum (ER) membrane0CVWTG4GXKGYU^/QNGEWNCT%GNN$KQNQI[
by the long, flexible amino termini
of Get2. The Get3–TA substrate complex, now in an ADP-bound, partially closed state, is transferred to Get1, which
wedges open the composite hydrophobic groove to promote TA substrate (and ADP) release. Finally, ATP re-binding
dissociates the stable Get1–Get3 ‘post-insertion’ complex to recycle Get3 back to the cytosol. Although depicted here
as a stable heterotetramer, the stoichiometry and subunit composition of the Get1–Get2 receptor complex is not known.
The insets show crystal structures of Mg2+–ADP–AlF4–-bound Get3 in complex with the cytosolic fragment of Get2
(left, Protein Databank (PDB) ID 3ZS9), and of nucleotide-free Get3 in complex with the cytosolic fragment of Get1
(right, PDB ID 3ZS8). C, carboxyl terminus.

step is a GTP-dependent interaction between SRP and a hydrophobic TMD to the membrane surface is insuf­
the SRP receptor that results in structural re­arrangements ficient for insertion, and that the SEC61 complex has a
that expose the M‑domain–signal sequence module to crucial role. On the basis of structural and functional
facilitate release to the translocon100. analysis, this crucial function is twofold102–104. First, SEC61
seems to directly recognize substrate TMDs via a spe­
The elusive insertion step cific bindin­g site within its membrane domain. Second,
Virtually nothing is known about how TA substrates are this binding site also serves as a ‘lateral gate’, which pro­
inserted into the ER membrane. Following its release vides TMDs direct access to the lipid bilayer. Thus, the
from Get3, the substrate must avoid improper inter­ TMD takes a route through the centre of SEC61 to bypass
actions (in particular, aggregation) and insert into the phospholipid headgroups that otherwise preclude
the membrane bilayer in the correct orientation. On the facil­e access to the hydrophobic core of the membrane.
basis of the rigid interaction between Get1 and Get3, Applying these principles to the TA protein pathway,
the TMD is presumably released parallel to and abut­ it is attractive to speculate that Get1 and Get2, both of
ting the bilayer surface. From this position, insertion which are multi-spanning membrane proteins, interact
requires the hydrophobic TMD to cross the polar head directly with the TA protein’s TMD to chaperone it into
groups of the phospholipids and reach the hydrophobic the bilayer. Such a recognition event could provide an
membrane core. This could either occur ‘spontaneously’ additional layer of proofreading, as has been ascribed
(that is, without direct assistance from any factors) or the to the SEC61 complex 105–107. More importantly, it would
Get1–Get2 complex could chaperone the TA protein into explain how the TMD efficiently accesses the hydro­
the membrane (FIG. 6). Although in vitro studies show that phobic core of the bilayer with minimal possibilit­y of
less-hydrophobic TMDs can insert spontaneously into ‘off-pathway’ events such as aggregation or inappro­
liposomes44,51, most TA proteins cannot. The mechanistic priate interactions. One speculative model for how
basis of this final step awaits additional studies, but some chaperoning could be achieved via coupled conforma­
insight can be gleaned from experiments done in the tional changes in Get1–Get2–Get3 is depicted in FIG. 6.
co-translational system. Alternatively, Get1–Get2 could distort the lipid bilayer in
A type  I membrane protein containing a single its vicinity to facilitate TMD insertion. The biochemica­l
N‑terminal TMD, a short luminal domain and an exten­ strategies used to study SEC61‑mediated insertion in
sive cytosolic C‑terminal domain can be efficiently tar­ the co-translational pathway 101,104,108–110 will clearly be
geted to the membrane surface via the SRP pathway, useful in determining the insertion mechanism of the
but subsequent insertion fails in the absence of the TA protei­n pathway. However, distinguishing between
SEC61 complex 101. This indicates that simply targeting what can happen in a simplified system and what does

794 | DECEMBER 2011 | VOLUME 12 www.nature.com/reviews/molcellbio

© 2011 Macmillan Publishers Limited. All rights reserved

 REVIEWS

C7PCUUKUVGF D%JCRGTQPGCUUKUVGF
)GVs)GV n5VTCKPGFo n4GNCZGFo
% %
'4

%[VQUQN

6#UWDUVTCVG
0 0
#&2 #&2 #&2 #&2

)GV
Figure 6 | Alternative models for the insertion of TA proteins into the ER membrane. a | After release from Get3,
0CVWTG4GXKGYU^/QNGEWNCT%GNN$KQNQI[
the tail-anchored (TA) substrate might transiently associate with the membrane surface before inserting ‘spontaneously’
into the lipid bilayer. b | Alternatively, the transmembrane domains (TMDs) of Get1 and Get2 may interact directly with the
TA substrate to chaperone it into the bilayer. In the model shown here, binding of two Get1 subunits to the partially closed
targeting complex results in a ‘strained’ configuration of the Get1–Get2 receptor complex. This arrangement might
provide a hydrophobic gate through which the TMD could diffuse into the bilayer. ‘Wedging open’ the Get3 dimer for
substrate release would then allow the Get1–Get2 receptor complex to ‘relax’ back into a low-energy configuration that
is sealed-off from the cytosol. Thus, conformational changes in Get3 might be coupled to those of the receptor TMDs to
promote TA substrate integration. C, carboxyl terminus; ER, endoplasmic reticulum; N, amino terminus.

happen in a physiological context may be challenging, that distinct Get1–Get2–Get3 and Get3–Get4–Get5
as the role of Get1–Get2 in insertion might simply be to complexes can be isolated from fractionated yeast 66,
accelerate an already favourable reaction, and any con­ and provides an elegant mechanism for spatially
sequences may only be apparent in a highly crowded regulating Get3 activity in the cytosol and at the ER
context that reflects the in vivo situation. membrane.

Factor recycling The ATPase cycle

After the TA substrate has been released, Get3–Get1 Biochemical, genetic and structural studies of the various
must dissociate so that Get3 can be recycled back to steps in the TA protein pathway have begun to clarify the
the cytosol and vacate Get1 for the next substrate. The role of ATP binding and hydrolysis by Get3 in the inser­
first clue for how this might be accomplished came tion cycle. As is the case for SRP in the co-translational
from early studies that showed ATP-dependent release pathway, the overarching theme is that the nucleotide
of TRC40 from the ER55. Structural studies provided state directly influences the conformation of Get3, which
the second clue. The crystal structure of nucleotide-free in turn regulates its interactions with the TA substrate,
Get3 bound to the cytosolic fragment of Get1 revealed targeting cofactors and receptors at the ER membrane.
the stable, high-affinity ‘post-insertion’ complex 95,97. Some of these interactions subsequently change the
In this open dimer conformation, the conserved nucleotide state. This allows Get3 to proceed undirec­
hairpin loop of Get1 inserts into the Get3 active site. tionally through its cycle of conformational changes
This interaction is both sterically and electrostatically and thereby selectively bind substrate in the cytosol and
incompatible with ATP binding, suggesting that the release it at the membrane.
high intracellular concentration of free ATP could be Under physiological conditions, Get3 in the cytosol
used to displace Get3 from Get1 after substrate release. is probably in an ATP-bound closed state95. The closed
Consistent with this, interaction studies showed that conformation preferentially incorporates into the TRC
ATP binding could quantitatively disrupt the Get3– owing to the nucleotide-dependence of the Get3–Get4
Get1 interaction95–97. This recycling mechanism con­ interaction80,96. Thus, substrate loading onto Get3 via
trasts with the co-translational pathway, in which, after the TRC requires ATP binding. Once loaded, the TA
substrate release, the SRP–SRP receptor complex is substrate bridges the helical domains across the hydro­
disso­ciated by GTP hydrolysis111. phobic groove of Get3. This locks Get3 into a closed
After release from Get1, what prevents Get3–ATP conformation, and traps the nucleotide in the dimer
from re-binding to Get2 in a ‘dead-end’ complex that interface. It is possible that substrate binding stim­ulates
cannot recruit a new TA substrate? Previous studies ATP hydrolysis by Get3, which would weaken its inter­
identified Get3 mutations that disrupt ATP-dependent action with Get4 and promote disengagement from
binding to Get4 (REF. 80). Remarkably, these mutations TRC. This would generate a committed ADP-bound
map to the overlapping Get1- and Get2‑binding sites Get3–TA substrate complex that is stabilized in a closed
on Get3. Thus, by competing for the same site as the conformation by the tightly bound substrate.
Get1 and Get2 receptor complex, Get4 binding could This targeting complex would be captured at the mem­
sequester the recycled Get3–ATP complex in the cyto­ brane by Get2 and then transferred to Get1. Although
sol and promote another round of substrate loading by targeting seems ‘agnostic’ to the nucleotide state of Get3,
the TRC. This model is consistent with the observation the subsequent transfer reaction absolutely requires ATP

NATURE REVIEWS | MOLECULAR CELL BIOLOGY VOLUME 12 | DECEMBER 2011 | 795

© 2011 Macmillan Publishers Limited. All rights reserved

REVIEWS

to have been hydrolysed95,96. This is probably because BAG6, Sgt2, Get3, HSP70 and possibly others)? This crit­
the binding site for Get1 is partially buried in the fully ical checkpoint determines the fate of the hydrophobi­c
closed ATP-bound state of Get3 (REF. 95). Get1 binding substrate: insertion into the ER membrane (via Get3),
causes Get3 to fully open, thereby favouring substrate insertion into the mitochondrial outer membrane
release. Once Get3 opens, the ADP is no longer trapped (by an unknown mechanism), or ubiquitylation (in a
and probably dissociates rapidly. The re-binding of ATP process involving BAG6) and proteasomal degradation for
to Get3 dissociates it from Get1. This effectively means misfolded membrane proteins. Defining the mechanism
that Get1 acts as a nucleotide exchange factor that dis­ of this process is therefore an important future goal.
places ADP (via opening of the Get3 dimer) and allows The biochemical and structural framework estab­
its replacemen­t with ATP. lished over the past few years also sets the stage for
Whereas most parts of the ATPase cycle have strong detailed mechanistic studies of events at the ER mem­
experimental support, the timing of ATP hydrolysis brane. Determining the structure and organization of the
remains unclear. It must occur at a point after the sub­ Get1–Get2 receptor complex will be critical for under­
strate binds to Get3 and before the interaction of this standing insertion. Is the receptor a stable hetero­dimeric,
protein with Get1. Although it is attractive to posit that heterotetrameric or higher order assembly? Or are there
hydrolysis stimulated by either substrate or Get2 inter­ temporal variations in its stoichiometry and subunit
action provides an additional checkpoint, this may not be composition? Similarly, the order and timing of Get3–
necessary. In fact, a slow rate of intrinsic hydrolysis might substrate inter­actions with the receptor have not been
provide a mechanism for kinetic proofreading. In this directly established. Most significantly, almost nothing is
model, loosely bound substrates would dissociate before known about how the TA substrate inserts into the bilayer
hydrolysis, thereby preventing their membrane-localized after its release from Get3. Does it insert spontaneously?
release by Get1. This might reduce inappropriate target­ Or is it actively chaperoned by the conserved TMDs of
ing of mitochondrial TA proteins (which generally have Get1 and Get2? Answers to these questions will require
TMDs of lower hydrophobicity than ER‑directed TA high-resolution structural analysis of Get1–Get2–Get3
proteins)47–49,77 to improve the overall fidelity of sorting. complexes trapped at different stages of this cycle. This
Similar kinetic proofreading mechanisms have been promises to be technically challenging, but the goal is in
uncovered in the SRP pathway and are thought to maxi­ reach now that robust expression systems for functional
mize the sorting efficiency of an otherwise promiscuous Get1 and Get2 proteins are in hand.
interaction between SRP and various proteins112. Finally, there may be much to learn from studying TA
protein biogenesis in other organisms. The recent discov­
Future challenges and perspectives ery of a conserved Get3 orthologue in archaea indicates
Since the discovery of the TA protein insertion pathway in that the post-translational TA pathway is more broadly
2007, rapid advances have been made through the appli­ conserved than previously appreciated116,117. No obviou­s
cation of genetic, biochemical and structural approaches. sequence homologues have been identified for other
Over the past 5 years, all of the core components have been cytosolic components (for example, BAG6, Sgt2, Get4 or
identified, and a general mechanistic understanding of the Get5), but functional data are consistent with the pres­
pathway from ribosome to the ER is now in hand. But the ence of an orthologous integral membrane receptor in
details of many key steps that take place in the cytosol archaea117. If such a receptor exists in archaea, it shares
and at the membrane remain a mystery. For example, how only limited sequence homology with the yeast recep­
does the TRC selectively interact with cytosolic ribosomes tor subunits, Get1 and Get2. There are also intriguing
that contain a TMD in the exit channel? Structural studies evolutionary differences between the yeast and mam­
are needed to identify the ribo­somal binding site of the malian pathways. The absence of a BAG6 homologue in
TRC, and to determine how the TRC senses the presence yeast suggests increasing complexity in the pre-targeting
of a TMD in the exit channel. After this initial capture machinery of higher eukaryotes. And, whereas the human
step, what is the mechanism for TA substrate transfer Trp-rich basic (WRB) protein shows sequence and func­
from the TRC to Get3? Although crystal structures exist tional homology to yeast Get1 (REF. 118), no obvious
for Get4 and portions of Get5 and Sgt2 (REFS 80,113–115), sequence homologue to Get2 has been found in higher
mechanistic insight awaits structural analysis of the intact eukaryotes. Thus, the identification and characterization
TRC. More broadly, how does the TRC sort different of components in the archaeal and higher eukaryotic
hydrophobic substrates between its multiple hydropho­ pathways promises new insight into the mechanism of
bic region-binding factors (including the chaperones TA protein targeting and insertion.

1. Wallin, E. & von Heijne, G. Genome-wide analysis 4. White, S. H. & von Heijne, G. How translocons select 7. Inaba, T. & Schnell, D. J. Protein trafficking to plastids:
of integral membrane proteins from eubacterial, transmembrane helices. Annu. Rev. Biophys. 37, one theme, many variations. Biochem. J. 413, 15–28
archaean, and eukaryotic organisms. Protein Sci. 7, 23–42 (2008). (2008).
1029–1038 (1998). 5. Chacinska, A., Koehler, C. M., Milenkovic, D., 8. Shao, S. & Hegde, R. S. Membrane protein insertion
2. Fagerberg, L., Jonasson, K., von Heijne, G., Lithgow, T. & Pfanner, N. Importing mitochondrial at the endoplasmic reticulum. Annu. Rev. Cell Dev.
Uhlen, M. & Berglund, L. Prediction of the human proteins: machineries and mechanisms. Cell 138, Biol. 27, 25–56 (2011).
membrane proteome. Proteomics 10, 1141–1149 628–644 (2009). 9. Palade, G. Intracellular aspects of the process of
(2010). 6. Heiland, I. & Erdmann, R. Biogenesis of protein synthesis. Science 189, 347–358 (1975).
3. White, S. H. & von Heijne, G. Transmembrane helices peroxisomes. Topogenesis of the peroxisomal 10. Wickner, W. & Schekman, R. Protein translocation
before, during, and after insertion. Curr. Opin. Struct. membrane and matrix proteins. FEBS J. 272, across biological membranes. Science 310,
Biol. 15, 378–386 (2005). 2362–2372 (2005). 1452–1456 (2005).

796 | DECEMBER 2011 | VOLUME 12 www.nature.com/reviews/molcellbio

© 2011 Macmillan Publishers Limited. All rights reserved

 REVIEWS

11. Rapoport, T. A. Protein translocation across the 35. Deshaies, R. J. & Schekman, R. A yeast mutant 56. Favaloro, V., Spasic, M., Schwappach, B. &
eukaryotic endoplasmic reticulum and bacterial defective at an early stage in import of secretory Dobberstein, B. Distinct targeting pathways for the
plasma membranes. Nature 450, 663–669 protein precursors into the endoplasmic reticulum. membrane insertion of tail-anchored (TA) proteins.
(2007). J. Cell Biol. 105, 633–645 (1987). J. Cell Sci. 121, 1832–1840 (2008).
12. Blobel, G. & Dobberstein, B. Transfer of proteins 36. Becker, T. et al. Structure of monomeric yeast and 57. Kurdi-Haidar, B. et al. Isolation of the ATP-binding
across membranes. I. Presence of proteolytically mammalian Sec61 complexes interacting with the human homolog of the ArsA component of the
processed and unprocessed nascent immunoglobulin translating ribosome. Science 326, 1369–1373 bacterial arsenite transporter. Genomics 36,
light chains on membrane-bound ribosomes of murine (2009). 486–491 (1996).
myeloma. J. Cell Biol. 67, 835–851 (1975). 37. Menetret, J. F. et al. The structure of ribosome- 58. Chen, C. M., Misra, T. K., Silver, S. & Rosen, B. P.
13. Blobel, G. & Dobberstein, B. Transfer of proteins channel complexes engaged in protein translocation. Nucleotide sequence of the structural genes for
across membranes. II. Reconstitution of functional Mol. Cell 6, 1219–1232 (2000). an anion pump. The plasmid-encoded arsenical
rough microsomes from heterologous components. 38. Beckmann, R. et al. Architecture of the protein- resistance operon. J. Biol. Chem. 261,
J. Cell Biol. 67, 852–862 (1975). conducting channel associated with the translating 15030–15038 (1986).
14. Walter, P., Ibrahimi, I. & Blobel, G. Translocation of 80S ribosome. Cell 107, 361–372 (2001). 59. Mukhopadhyay, R., Ho, Y. S., Swiatek, P. J.,
proteins across the endoplasmic reticulum. I. Signal 39. Ulbrandt, N. D., Newitt, J. A. & Bernstein, H. D. The Rosen, B. P. & Bhattacharjee, H. Targeted disruption
recognition protein (SRP) binds to in‑vitro‑assembled E. coli signal recognition particle is required for the of the mouse Asna1 gene results in embryonic
polysomes synthesizing secretory protein. J. Cell Biol. insertion of a subset of inner membrane proteins. lethality. FEBS Lett. 580, 3889–3894 (2006).
91, 545–550 (1981). Cell 88, 187–196 (1997). 60. Auld, K. L. et al. The conserved ATPase Get3/Arr4
15. Hann, B. C. & Walter, P. The signal recognition particle 40. Ott, M. et al. Mba1, a membrane-associated modulates the activity of membrane-associated
in S. cerevisiae. Cell 67, 131–144 (1991). ribosome receptor in mitochondria. EMBO J. 25, proteins in Saccharomyces cerevisiae. Genetics 174,
16. Gorlich, D., Prehn, S., Hartmann, E., Kalies, K. U. & 1603–1610 (2006). 215–227 (2006).
Rapoport, T. A. A mammalian homolog of SEC61p and 41. Nilsson, R. & van Wijk, K. J. Transient interaction 61. Schuldiner, M. et al. Exploration of the function and
SECYp is associated with ribosomes and nascent of cpSRP54 with elongating nascent chains of the organization of the yeast early secretory pathway
polypeptides during translocation. Cell 71, 489–503 chloroplast-encoded D1 protein; ‘cpSRP54 caught through an epistatic miniarray profile. Cell 123,
(1992). in the act’. FEBS Lett. 524, 127–133 (2002). 507–519 (2005).
17. Hartmann, E. et al. Evolutionary conservation of 42. Kutay, U., Hartmann, E. & Rapoport, T. A. A class of 62. Kao, G. et al. ASNA‑1 positively regulates insulin
components of the protein translocation complex. membrane proteins with a C‑terminal anchor. secretion in C. elegans and mammalian cells.
Nature 367, 654–657 (1994). Trends Cell Biol. 3, 72–75 (1993). Cell 128, 577–587 (2007).
18. Jungnickel, B., Rapoport, T. A. & Hartmann, E. Protein 43. Kutay, U., Ahnert-Hilger, G., Hartmann, E., 63. Shen, J., Hsu, C. M., Kang, B. K., Rosen, B. P. &
translocation: common themes from bacteria to man. Wiedenmann, B. & Rapoport, T. A. Transport route for Bhattacharjee, H. The Saccharomyces cerevisiae
FEBS Lett. 346, 73–77 (1994). synaptobrevin via a novel pathway of insertion into the Arr4p is involved in metal and heat tolerance.
19. Poritz, M. A. et al. An E. coli ribonucleoprotein endoplasmic reticulum membrane. EMBO J. 14, Biometals 16, 369–378 (2003).
containing 4.5S RNA resembles mammalian signal 217–223 (1995). 64. Ho, Y. et al. Systematic identification of protein
recognition particle. Science 250, 1111–1117 Demonstrates that a model TA protein is inserted complexes in Saccharomyces cerevisiae by mass
(1990). into the ER via an SRP- and SEC61‑independent spectrometry. Nature 415, 180–183 (2002).
20. Stuart, R. A., Cyr, D. M., Craig, E. A. & Neupert, W. post-translational process that requires ATP 65. Schuldiner, M. et al. The GET complex mediates
Mitochondrial molecular chaperones: their role in hydrolysis and a protease-sensitive microsomal insertion of tail-anchored proteins into the ER
protein translocation. Trends Biochem. Sci. 19, 87–92 factor (or factors). membrane. Cell 134, 634–645 (2008).
(1994). 44. Brambillasca, S. et al. Transmembrane topogenesis Shows that Get proteins identified in references 60
21. Soll, J. & Tien, R. Protein translocation into and across of a tail-anchored protein is modulated by membrane and 61 are part of the TA pathway in yeast. Get3 is
the chloroplastic envelope membranes. Plant Mol. lipid composition. EMBO J. 24, 2533–2542 (2005). shown to be the TRC40 orthologue, and Get1 and
Biol. 38, 191–207 (1998). 45. Yabal, M. et al. Translocation of the C terminus of Get2 are shown to form an ER‑localized Get
22. Rapoport, T. A., Matlack, K. E., Plath, K., Misselwitz, B. a tail-anchored protein across the endoplasmic receptor complex.
& Staeck, O. Posttranslational protein translocation reticulum membrane in yeast mutants defective in 66. Jonikas, M. C. et al. Comprehensive characterization
across the membrane of the endoplasmic reticulum. signal peptide-driven translocation. J. Biol. Chem. of genes required for protein folding in the
Biol. Chem. 380, 1143–1150 (1999). 278, 3489–3496 (2003). endoplasmic reticulum. Science 323, 1693–1697
23. Mori, H. & Cline, K. Post-translational protein 46. Steel, G. J., Brownsword, J. & Stirling, C. J. (2009).
translocation into thylakoids by the Sec and ΔpH- Tail-anchore­d protein insertion into yeast ER requires 67. Costanzo, M. et al. The genetic landscape of a cell.
dependent pathways. Biochim. Biophys. Acta 1541, a novel posttranslational mechanism which is Science 327, 425–431 (2010).
80–90 (2001). independent of the SEC machinery. Biochemistry 41, 68. Copic, A. et al. Genomewide analysis reveals novel
24. Ma, C., Agrawal, G. & Subramani, S. Peroxisome 11914–11920 (2002). pathways affecting endoplasmic reticulum
assembly: matrix and membrane protein biogenesis. 47. Beilharz, T., Egan, B., Silver, P. A., Hofmann, K. & homeostasis, protein modification and quality control.
J. Cell Biol. 193, 7–16 (2011). Lithgow, T. Bipartite signals mediate subcellular Genetics 182, 757–769 (2009).
25. Keenan, R. J., Freymann, D. M., Stroud, R. M. & targeting of tail-anchored membrane proteins in 69. Bozkurt, G. et al. Structural insights into tail-anchored
Walter, P. The signal recognition particle. Annu. Rev. Saccharomyces cerevisiae. J. Biol. Chem. 278, protein binding and membrane insertion by Get3.
Biochem. 70, 755–775 (2001). 8219–8223 (2003). Proc. Natl Acad. Sci. USA 106, 21131–21136
26. Grudnik, P., Bange, G. & Sinning, I. Protein targeting 48. Kalbfleisch, T., Cambon, A. & Wattenberg, B. W. A (2009).
by the signal recognition particle. Biol. Chem. 390, bioinformatics approach to identifying tail-anchored 70. Hu, J., Li, J., Qian, X., Denic, V. & Sha, B. The crystal
775–782 (2009). proteins in the human genome. Traffic 8, 1687–1694 structures of yeast Get3 suggest a mechanism for tail-
27. Walter, P. & Blobel, G. Purification of a membrane- (2007). anchored protein membrane insertion. PLoS ONE 4,
associated protein complex required for protein 49. Kriechbaumer, V. et al. Subcellular distribution of e8061 (2009).
translocation across the endoplasmic reticulum. tail-anchored proteins in Arabidopsis. Traffic 10, 71. Mateja, A. et al. The structural basis of tail-anchored
Proc. Natl Acad. Sci. USA 77, 7112–7116 (1980). 1753–1764 (2009). membrane protein recognition by Get3. Nature 461,
28. Walter, P. & Blobel, G. Signal recognition particle 50. Colombo, S. F., Longhi, R. & Borgese, N. The role of 361–366 (2009).
contains a 7S RNA essential for protein translocation cytosolic proteins in the insertion of tail-anchored 72. Suloway, C. J., Chartron, J. W., Zaslaver, M. &
across the endoplasmic reticulum. Nature 299, proteins into phospholipid bilayers. J. Cell Sci. 122, Clemons, W. M. Jr. Model for eukaryotic tail-anchored
691–698 (1982). 2383–2392 (2009). protein binding based on the structure of Get3.
29. Halic, M. et al. Structure of the signal recognition 51. Brambillasca, S., Yabal, M., Makarow, M. & Proc. Natl Acad. Sci. USA 106, 14849–14854
particle interacting with the elongation-arrested Borgese, N. Unassisted translocation of large (2009).
ribosome. Nature 427, 808–814 (2004). polypeptide domains across phospholipid bilayers. 73. Yamagata, A. et al. Structural insight into the
30. Zopf, D., Bernstein, H. D., Johnson, A. E. & Walter, P. J. Cell Biol. 175, 767–777 (2006). membrane insertion of tail-anchored proteins by
The methionine-rich domain of the 54 kd protein 52. Abell, B. M., Rabu, C., Leznicki, P., Young, J. C. & Get3. Genes Cells 15, 29–41 (2010).
subunit of the signal recognition particle contains an High, S. Post-translational integration of tail-anchored References 69–73 describe the crystal structure
RNA binding site and can be crosslinked to a signal proteins is facilitated by defined molecular of fungal Get3 in different nucleotide-bound states.
sequence. EMBO J. 9, 4511–4517 (1990). chaperones. J. Cell Sci. 120, 1743–1751 (2007). Reference 71 reveals a large, hydrophobic groove in
31. High, S. & Dobberstein, B. The signal sequence 53. Abell, B. M., Pool, M. R., Schlenker, O., Sinning, I. & the Mg2+–ADP–AlF4–-bound complex. References
interacts with the methionine-rich domain of the High, S. Signal recognition particle mediates post- 69 and 71 implicate this groove as the TMD-
54‑kD protein of signal recognition particle. J. Cell translational targeting in eukaryotes. EMBO J. 23, binding site by using mutational and biophysical
Biol. 113, 229–233 (1991). 2755–2764 (2004). analyses.
32. Gilmore, R., Walter, P. & Blobel, G. Protein 54. Abell, B. M. et al. Tail-anchored and signal-anchored 74. Keenan, R. J., Freymann, D. M., Walter, P. &
translocation across the endoplasmic reticulum. II. proteins utilize overlapping pathways during Stroud, R. M. Crystal structure of the signal sequence
Isolation and characterization of the signal membrane insertion. J. Biol. Chem. 278, 5669–5678 binding subunit of the signal recognition particle.
recognition particle receptor. J. Cell Biol. 95, (2003). Cell 94, 181–191 (1998).
470–477 (1982). 55. Stefanovic, S. & Hegde, R. S. Identification of a 75. Janda, C. Y. et al. Recognition of a signal peptide by
33. Meyer, D. I., Krause, E. & Dobberstein, B. Secretory targeting factor for posttranslational membrane the signal recognition particle. Nature 465, 507–510
protein translocation across membranes—the role protein insertion into the ER. Cell 128, 1147–1159 (2010).
of the ‘docking protein’. Nature 297, 647–650 (2007). 76. Hainzl, T., Huang, S., Merilainen, G., Brannstrom, K.
(1982). Reports the discovery of the mammalian & Sauer-Eriksson, A. E. Structural basis of signal-
34. Simon, S. M. & Blobel, G. A protein-conducting TA protein pathway and the identification of sequence recognition by the signal recognition
channel in the endoplasmic reticulum. Cell 65, the TRC40 ATPase as the central cytosolic particle. Nature Struct. Mol. Biol. 18, 389–391
371–380 (1991). component. (2011).

NATURE REVIEWS | MOLECULAR CELL BIOLOGY VOLUME 12 | DECEMBER 2011 | 797

© 2011 Macmillan Publishers Limited. All rights reserved

REVIEWS

77. Borgese, N., Brambillasca, S. & Colombo, S. How tails during membrane integration. J. Cell Biol. 195, 108. Mothes, W., Prehn, S. & Rapoport, T. A. Systematic
guide tail-anchored proteins to their destinations. 41–54 (2011). probing of the environment of a translocating
Curr. Opin. Cell Biol. 19, 368–375 (2007). 93. Prilusky, J. & Bibi, E. Studying membrane proteins secretory protein during translocation through
78. Mariappan, M. et al. A ribosome-associating factor through the eyes of the genetic code revealed a strong the ER membrane. EMBO J. 13, 3973–3982
chaperones tail-anchored membrane proteins. uracil bias in their coding mRNAs. Proc. Natl Acad. (1994).
Nature 466, 1120–1124 (2010). Sci. USA 106, 6662–6666 (2009). 109. Martoglio, B., Hofmann, M. W., Brunner, J. &
79. Wang, F., Brown, E. C., Mak, G., Zhuang, J. & 94. Nevo-Dinur, K., Nussbaum-Shochat, A., Dobberstein, B. The protein-conducting channel in the
Denic, V. A chaperone cascade sorts proteins for Ben-Yehuda, S. & Amster-Choder, O. Translation- membrane of the endoplasmic reticulum is open
posttranslational membrane insertion into the independent localization of mRNA in E. coli. laterally toward the lipid bilayer. Cell 81, 207–214
endoplasmic reticulum. Mol. Cell 40, 159–171 Science 331, 1081–1084 (2011). (1995).
(2010). 95. Mariappan, M. et al. The mechanism of membrane- 110. Do, H., Falcone, D., Lin, J., Andrews, D. W. &
References 78 and 79 describe a cytosolic protein associated steps in tail-anchored protein insertion. Johnson, A. E. The cotranslational integration of
complex that captures TA proteins after synthesis Nature 477, 61–66 (2011). membrane proteins into the phospholipid bilayer
and transfers them to Get3 or TRC40 for Defines the minimal set of factors required is a multistep process. Cell 85, 369–378 (1996).
subsequent targeting to the ER membrane. for TA protein insertion by biochemical 111. Connolly, T., Rapiejko, P. J. & Gilmore, R. Requirement
80. Chartron, J. W., Suloway, C. J. M., Zaslaver, M. a. & reconstitution using purified, recombinant Get1, of GTP hydrolysis for dissociation of the signal
Clemons, W. M. Structural characterization of the Get2 and Get3. recognition particle from its receptor. Science 252,
Get4/Get5 complex and its interaction with Get3. 96. Wang, F., Whynot, A., Tung, M. & Denic, V. 1171–1173 (1991).
Proc. Natl Acad. Sci. USA 107, 12127–12132 (2010). The mechanism of tail-anchored protein insertion 112. Zhang, X., Rashid, R., Wang, K. & Shan, S. O.
81. Leznicki, P., Clancy, A., Schwappach, B. & High, S. into the ER membrane. Mol. Cell 43, 738–750 Sequential checkpoints govern substrate selection
Bat3 promotes the membrane integration of (2011). during cotranslational protein targeting. Science 328,
tail-anchored proteins. J. Cell Sci. 123, 2170–2178 References 95 and 96 describe a detailed 757–760 (2010).
(2010). biochemical analysis of targeting, substrate release 113. Chang, Y.‑W. et al. Crystal structure of Get4–Get5
82. Winnefeld, M. et al. Human SGT interacts with and Get3 recycling. complex and its interactions with Sgt2, Get3, and
Bag-6/Bat-3/Scythe and cells with reduced levels of 97. Stefer, S. et al. Structural basis for tail-anchored Ydj1. J. Biol. Chem. 285, 9962–9970 (2010).
either protein display persistence of few misaligned membrane protein biogenesis by the Get3‑receptor 114. Bozkurt, G. et al. The structure of Get4 reveals an
chromosomes and mitotic arrest. Exp. Cell Res. 312, complex. Science 333, 758–762 (2011). α-solenoid fold adapted for multiple interactions in
2500–2514 (2006). References 95 and 97 provide structural snapshots tail-anchored protein biogenesis. FEBS Lett. 584,
83. Hessa, T. et al. Protein targeting and degradation are of Get3 bound to cytosolic fragments of Get1 and 1509–1514 (2010).
coupled for elimination of mislocalized proteins. Get2. 115. Chartron, J. W., Gonzalez, G. M. & Clemons, W. M. Jr.
Nature 475, 394–397 (2011). 98. Egea, P. F. et al. Substrate twinning activates the A structural model of SGT2 and its interactions with
84. Wang, Q. et al. A ubiquitin ligase-associated signal recognition particle and its receptor. chaperones and GET4/GET5. J. Biol. Chem. 286,
chaperone holdase maintains polypeptides in soluble Nature 427, 215–221 (2004). 34325–34334 (2011).
states for proteasome degradation. Mol. Cell 42, 99. Focia, P. J., Shepotinovskaya, I. V., Seidler, J. A. & 116. Borgese, N. & Righi, M. Remote origins of tail-
758–770 (2011). Freymann, D. M. Heterodimeric GTPase core of the anchored proteins. Traffic 11, 877–885 (2010).
85. Minami, R. et al. BAG‑6 is essential for selective SRP targeting complex. Science 303, 373–377 117. Sherrill, J., Mariappan, M., Dominik, P., Hegde, R. S.
elimination of defective proteasomal substrates. (2004). & Keenan, R. J. A conserved archaeal pathway for
J. Cell Biol. 190, 637–650 (2010). 100. Ataide, S. F. et al. The crystal structure of the signal tail-anchored membrane protein insertion. Traffic 12,
86. Fleischer, T. C., Weaver, C. M., McAfee, K. J., recognition particle in complex with its receptor. 1119–1123 (2011).
Jennings, J. L. & Link, A. J. Systematic identification Science 331, 881–886 (2011). 118. Vilardi, F., Lorenz, H. & Dobberstein, B. WRB is the
and functional screens of uncharacterized proteins 101. Heinrich, S. U., Mothes, W., Brunner, J. & receptor for TRC40/Asna1‑mediated insertion of
associated with eukaryotic ribosomal complexes. Rapoport, T. A. The Sec61p complex mediates the tail-anchored proteins into the ER membrane.
Genes Dev. 20, 1294–1307 (2006). integration of a membrane protein by allowing lipid J. Cell Sci. 124, 1301–1307 (2011).
87. Cao, J. & Geballe, A. P. Coding sequence-dependent partitioning of the transmembrane domain.
ribosomal arrest at termination of translation. Mol. Cell 102, 233–244 (2000). Acknowledgements
Cell. Biol. 16, 603–608 (1996). 102. Van den Berg, B. et al. X‑ray structure of a Research in the Hegde laboratory is supported by the
88. Berndt, U., Oellerer, S., Zhang, Y., Johnson, A. E. & protein-conducting channel. Nature 427, 36–44 Medical Research Council, UK. Research in the Keenan
Rospert, S. A signal-anchor sequence stimulates signal (2004). laborator­y is supported by grants from the US National
recognition particle binding to ribosomes from inside 103. Tsukazaki, T. et al. Conformational transition of Sec Institutes of Health (R01 GM086487), the Camille and Henry
the exit tunnel. Proc. Natl Acad. Sci. USA 106, machinery inferred from bacterial SecYE structures. Dreyfus Foundation and the Edward Mallinckrodt, Jr
1398–1403 (2009). Nature 455, 988–991 (2008). Foundation. The authors thank members of their laboratories
89. Bornemann, T., Jockel, J., Rodnina, M. V. & 104. Cannon, K. S., Or, E., Clemons, W. M. Jr, Shibata, Y. and other colleague­s for helpful discussions.
Wintermeyer, W. Signal sequence-independent & Rapoport, T. A. Disulfide bridge formation between
membrane targeting of ribosomes containing short SecY and a translocating polypeptide localizes the Competing interests statement
nascent peptides within the exit tunnel. Nature Struct. translocation pore to the center of SecY. J. Cell Biol. The authors declare no competing financial interests.
Mol. Biol. 15, 494–499 (2008). 169, 219–225 (2005).
90. Liao, S., Lin, J., Do, H. & Johnson, A. E. Both lumenal 105. Flower, A. M., Doebele, R. C. & Silhavy, T. J. PrlA and
and cytosolic gating of the aqueous ER translocon PrlG suppressors reduce the requirement for signal DATABASES
pore are regulated from inside the ribosome during sequence recognition. J. Bacteriol. 176, 5607–5614 Protein Databank: http://www.pdb.org/pdb/home/home.do
membrane protein integration. Cell 90, 31–41 (1994). 1RHZ | 2WOJ | 2WOO | 3IQX | 3KL4 | 3NBD | 3ZS8 | 3ZS9
(1997). 106. Osborne, R. S. & Silhavy, T. J. PrlA suppressor
91. Lin, P. J., Jongsma, C. G., Pool, M. R. & Johnson, A. E. mutations cluster in regions corresponding to three FURTHER INFORMATION
Polytopic membrane protein folding at L17 in the distinct topological domains. EMBO J. 12, Ramanujan S. Hegde’s homepage: http://www2.mrc-lmb.
ribosome tunnel initiates cyclical changes at the 3391–3398 (1993). cam.ac.uk/group-leaders/h-to-m/r-hegde
translocon. J. Cell Biol. 195, 55–70 (2011). 107. Jungnickel, B. & Rapoport, T. A. A posttargeting Robert J. Keenan’s homepage:
92. Lin, P. J., Jongsma, C. G., Liao, S. & Johnson, A. E. signal sequence recognition event in the http://keenanlab.bsd.uchicago.edu
Transmembrane segments of nascent polytopic endoplasmic reticulum membrane. Cell 82, ALL LINKS ARE ACTIVE IN THE ONLINE PDF
membrane proteins control cytosol/ER targeting 261–270 (1995).

798 | DECEMBER 2011 | VOLUME 12 www.nature.com/reviews/molcellbio

© 2011 Macmillan Publishers Limited. All rights reserved