Content uploaded by Robert T. Batey
Author content
All content in this area was uploaded by Robert T. Batey
Content may be subject to copyright.
STRUCTURAL INSIGHTS INTO THE SIGNAL
RECOGNITION PARTICLE
Jennifer A. Doudna
1
and Robert T. Batey
2
1
Department of Molecular and Cell Biology and Department of Chemistry, Howard
Hughes Medical Institute, University of California at Berkeley, Berkeley, California
94705; email: doudna@uclink.berkeley.edu
2
Department of Chemistry and Biochemistry, University of Colorado at Boulder,
Boulder, Colorado 80309; email: robert.batey@colorado.edu
Key Words SRP, signal sequence, SRP54, SRP19, SRP receptor, SRP RNA
fAbstract The signal recognition particle (SRP) directs integral membrane and
secretory proteins to the cellular protein translocation machinery during translation.
The SRP is an evolutionarily conserved RNA-protein complex whose activities are
regulated by GTP hydrolysis. Recent structural investigations of SRP functional
domains and interactions provide new insights into the mechanisms of SRP activity
in all cells, leading toward a comprehensive understanding of protein trafficking by
this elegant pathway.
CONTENTS
INTRODUCTION ........................................540
MECHANISM OF COTRANSLATIONAL PROTEIN TRAFFICKING .......541
Phylogenetic Conservation .................................541
Structural Features of the Conserved SRP Components ................542
Role of GTP Hydrolysis...................................544
SIGNAL RECOGNITION PARTICLE ASSEMBLY...................546
Structural and Functional Domains ............................546
Assembly of the Signal Sequence Binding Domain ..................547
THE SIGNAL RECOGNITION PARTICLE CYCLE ..................548
Signal Sequence Recognition................................548
Structure of the SRP54-RNA Complex..........................551
GTPase Stimulation in the SRP-SR Complex ......................552
Ribosome Interactions ....................................553
Nucleotide Exchange.....................................554
FUNCTIONS OF SIGNAL RECOGNITION PARTICLE RNA ............554
FUTURE DIRECTIONS ....................................555
Annu. Rev. Biochem. 2004. 73:539–57
doi: 10.1146/annurev.biochem.73.011303.074048
Copyright © 2004 by Annual Reviews. All rights reserved
First published online as a Review in Advance on March 9, 2004
5390066-4154/04/0707-0539$14.00
INTRODUCTION
In all cells, the signal recognition particle (SRP) targets proteins destined for
secretion or membrane insertion by binding to hydrophobic signal sequences at
the N terminus of polypeptides as they emerge from ribosomes. In a process
central to the ability of cells to communicate with other cells and the environ-
ment, the SRP recognizes ribosome-nascent chain complexes, docks with a
specific membrane-bound SRP receptor, releases the ribosome-associated
polypeptide into a translocon channel in the membrane, and dissociates from the
SRP receptor primed for another cycle of protein targeting (Figure 1). GTP
binding and hydrolysis by both the SRP and its receptor coordinate this process,
suggesting that induced conformational changes enable ordered binding and
Figure 1 The eukaryotic SRP-dependent protein targeting cycle. A nascent polypeptide
chain that is being actively translated by the ribosome (red) contains an amino-terminal
signal sequence (SP) that is bound by the SRP (yellow), which arrests translation. The
SRP-ribosome complex is targeted to the translocational complex (TC) embedded in the
endoplasmic reticulum membrane via an interaction with the membrane-bound receptor
complex SR
␣
/SR

. Following docking of the ribosome with the translocon, the signal
sequence is released from the SRP, and the SRP is released from the SRP receptor in a
GTP-dependent fashion.
540 DOUDNA yBATEY
release of the signal peptide, the ribosome, the SRP receptor, and the translocon.
Interestingly, all cytoplasmic SRPs are ribonucleoproteins that consist of one
RNA molecule and up to six proteins, a subset of which share sequence,
structural, and functional homology. Structural studies of SRP components and
complexes, together with genetic and biochemical experiments, have provided
significant insights into the mechanism of SRP-mediated protein trafficking as
well as the pathway of SRP assembly (reviewed in 1–5). Here we discuss recent
advances in understanding how both the bacterial and mammalian SRP particles
form functional complexes and how they bind specifically to SRP receptor
proteins and the ribosome. These data make possible future experiments to
understand the specificity of signal peptide recognition and the coupling of
nucleotide hydrolysis with structural dynamics that drive the protein transloca-
tion cycle.
MECHANISM OF COTRANSLATIONAL PROTEIN
TRAFFICKING
All cells localize secretory and integral membrane proteins, whose biosynthesis
begins in the cytoplasm, to specialized pores that enable cotranslational protein
export. The SRP is an RNA-protein complex that provides an evolutionarily
conserved mechanism for protein trafficking by recognizing the hydrophobic
signal sequence found on the N terminus of targeted proteins.
Phylogenetic Conservation
In mammalian cells, the SRP includes one RNA molecule (7S RNA) and six
proteins named according to their molecular weight: SRP72, SRP68, SRP54,
SRP19, SRP14, and SRP9. SRP54 and one stem-loop, helix 8 of the 7S RNA, are
conserved in archaea and eubacteria, and together with a second RNA helical
region (helix 6) constitute the signal peptide binding domain (S domain) of the
SRP. SRP54 also includes a GTPase domain and an N-terminal helical domain
that play roles in communicating the peptide-bound state of the SRP to the SRP
receptor, the ribosome, and the translocon pore. SRP19 induces a structural
change in the S domain of 7S RNA required for SRP54 binding. SRP72 and
SRP68 form a heterodimeric subcomplex that binds the middle segment of the 7S
RNA, whereas SRP14 and SRP9 bind cooperatively to the end opposite helix 8
in the SRP RNA to form the Alu domain responsible for transient translational
arrest during protein targeting.
In eubacteria, SRP comprises the smaller 4.5S RNA and an SRP54 homolog
called Ffh (Fifty-four homolog). All SRP RNAs include the highly conserved
binding site for the SRP54 protein, called domain IV in bacterial SRP RNA;
however, the less well conserved Alu domain is missing in gram-negative
bacteria. Both Ffh and 4.5S RNA are essential genes in Escherichia coli, and
541THE SIGNAL RECOGNITION PARTICLE
SRP54 and Ffh have been shown to be functional homologs. Furthermore, a
truncated form of 4.5S RNA, which includes just the Ffh binding site, supports
growth in a 4.5S RNA-depleted strain, showing that the critical function of the
RNA is contained within the peptide-recognition domain of the ribonucleopro-
tein complex.
The SRP receptor consists of a conserved protein, SR
␣
in mammals and FtsY
in bacteria, with GTPase activity, sequence, and structural homology to the
GTPase domain of SRP54/Ffh. In mammalian cells, SR
␣
binds a second subunit,
SR

, containing a single transmembrane region. SR

is also a GTPase but has
only distant homology to the GTPases of SRP54 and SR
␣
. Interestingly,
mutation of the SR

GTPase domain, but not deletion of the transmembrane
domain, disrupts signal recognition (SR) function in vivo (6). FtsY weakly
associates with the bacterial inner membrane, perhaps through direct interaction
with membrane phospholipids (7–11).
Structural Features of the Conserved SRP Components
Crystal structures of the GTPase domains of Thermus aquaticus Ffh and E. coli
FtsY provided the first structural insights into GTPase function in SRP (12, 13).
Ffh and FtsY each contain three domains, two of which, the N and G domains,
are related at both the sequence and structural level and comprise the GTPase of
each protein (Figure 2a). The G domain adopts a classical GTPase fold in which
four conserved sequence motifs (I-IV) are organized around the nucleotide
binding site. Motif II is part of a sequence unique to the SRP GTPases called the
insertion-box domain (IBD) that extends by two strands the central

-sheet of the
domain. The amino-terminal N domain, a four-helix bundle, packs against the G
domain to form a contiguous unit referred to as the NG domain (Figure 2b). Side
chains from the C-terminal end of the G domain contribute to the hydrophobic
core of the N domain, an interface conserved in Ffh and FtsY that creates an axis
about which the relative orientations of the N and G domains vary. Nucleotide-
dependent changes in the relative N and G domain positions are proposed to
enable N domain detection of the GTP-bound state of the G domain (14–16).
Crystal structures of full-length T. aquaticus Ffh as well as the human SRP54
M domain revealed an all-helical domain featuring a prominent hydrophobic
cleft comprising helices
␣
M1,
␣
M2, and
␣
M4 and an extended flexible loop, the
“finger loop,”connecting
␣
M1 and
␣
M2 (Figure 3) (17, 18). Adjacent to this
cleft, helices
␣
M3 and
␣
M4 form a classical helix-turn-helix (HTH) motif that
contains a conserved sequence of serine, arginine, and glycine residues essential
for high-affinity binding to SRP RNA. The crystal structure of the E. coli Ffh M
domain bound to the phylogenetically conserved region of SRP RNA revealed
that the HTH motif binds the distorted minor groove of the RNA (Figure 4a)
(19). In the protein-RNA complex, nucleotides in the asymmetric loop of the
RNA wrap around the outside of the helix and make specific contacts to the M
domain. Comparison of the structure to that of the unbound RNA shows that a
542 DOUDNA yBATEY
Figure 2 (a) Cartoon of the alignment of the E. coli FtsY and Ffh proteins. The figure
emphasizes the conservation of the NG domain of each and includes the four conserved
sequence motifs (G1-G4) and an insertion element (IBD that is unique to the SRP-
associated ras-type GTPases. (b) Structure of the E. coli NG domain of the SRP receptor
protein FtsY (PDB ID: 1FTS). The N domain represents the four helix bundle on the top
and the ras-type GTPase domain (G domain) on the bottom. The four conserved sequence
motifs in the Ffh/FtsY family of proteins (G1–G4) are highlighted in blue and the
insertion-box domain (IBD) are highlighted in magenta.
543THE SIGNAL RECOGNITION PARTICLE
significant conformational change is induced upon M domain binding, coupled to
the ordered binding of metal ions and waters in the complex (19–23).
Role of GTP Hydrolysis
The protein targeting cycle is regulated by the coordinated action of GTPases,
SRP54 and SR
␣
/SR

in eukaryotes and Ffh/FtsY in bacteria, that control signal
peptide binding and release. In the GTP-bound state, SRP binds to a nascent
signal peptide and the SRP receptor, leading to localization of the associated
ribosome on the translocon. The GTPases in both SRP54/Ffh and SR
␣
/FtsY are
Figure 3 Crystal structure of the Thermus aquaticus Ffh protein (PDB ID: 2FFH). The
two principal domains, the NG domain (green) and the M domain (blue), are shown. Ffh
crystallized as a trimer of proteins in which a flexible linker between the two domains
(residues 308–318) (shown as a red dashed line) was disordered, and thus the orientation
of the two domains is ambiguous. Only one of the three possible pairs is shown.
544 DOUDNA yBATEY
Figure 4 Structures of the signal recognition domain of the SRP from the three kingdoms of life. (a) Structure of the
conserved domain of the E. coli 4.5S RNA (corresponding to domain IV of the eukaryotic SRP RNA) bound to Ffh
M domain (PDB ID: 1DUL). A 33-amino acid segment between helices one and two of the M domain (light blue) was
disordered in the structure (dashed red line). Nucleotides in the 4.5S RNA that are highly (green) and universally
(yellow) conserved are highlighted along with A39 of the asymmetric internal loop (AL). (b) Structure of the
Methanococcus jannaschii S domain RNA in complex with SRP19 (magenta) (PDB ID: 1LNG). Two adenosine
residues in domain III, highlighted in red, contact the asymmetric loop of domain IV by forming A-minor base triples
that potentially stabilize a conformation productive for M domain binding. (c) Structure of the ternary complex between
the human S domain RNA/SRP19/SRP54 M domain (PDB ID: 1MFQ). The adenosines from the asymmetric loop that
form A-minor triples with domain III are highlighted in red.
545THE SIGNAL RECOGNITION PARTICLE
mutually stimulated upon complex formation and have been proposed to act as
GAPs (GTPase activating proteins) for each other (24, 25). Following release of
the signal peptide from SRP54/Ffh into the translocon, hydrolysis of SRP- and
SR-bound GTP molecules causes dissociation of the SRP-SR complex and
resumption of ribosome-catalyzed polypeptide synthesis. In eukaryotes, GTP
hydrolysis by SR

leads to dissociation of its complex with SR
␣
. Key questions
about the mechanism of SRP center on the role of GTP hydrolysis in coordinat-
ing and controlling the timing of these events in the cell.
The role of SR

and why it is required in eukaryotic cells remains unclear, but
some clues were provided by the crystal structure of SR

bound to GTP and the
N-terminal interaction domain of SR
␣
(26). SR

has nanomolar affinity for GTP
(27), and in its GTP-bound state, it is catalytically inert when bound to SR
␣
,
requiring a GAP and a GEF (guanine-nucleotide exchange factor) to function as
a GTPase switch for release of the SR
␣
subunit (28). In the structure, an
extensive intermolecular interface includes both polar and hydrophobic contacts
with interdigitated side chains that produce a high-affinity complex (26). Recent
data from a fluorescence nucleotide exchange assay show that the

subunit of
the protein-conducting channel in the endoplasmic reticulum functions as the
GEF for SR

in yeast, and by analogy in other organisms as well (29). The nature
of the SR

GAP activity and how it correlates with ribosome and translocon
binding and release await further biochemical and structural studies.
SIGNAL RECOGNITION PARTICLE ASSEMBLY
SRP comprises discrete domains corresponding to the translational arrest and
peptide recognition functions of the particle. Results of structural and biochem-
ical experiments highlight the topology of SRP as well as the hierarchy of
interactions that result in a functional complex.
Structural and Functional Domains
Electron microscopy (EM) revealed that the eukaryotic SRP has an elongated
rod-like structure (⬃240x60Å) comprising 3 distinct regions (30, 31). The Alu
domain and the S domain, responsible for translation arrest and signal sequence
binding, respectively, lie at opposite ends of the complex. Connecting these
functional centers is a low-mass region thought to be a flexible RNA linker. This
global organization is consistent with simultaneous binding of the elongation
arrest domain at the ribosomal subunit interface and binding of the signal
recognition domain near the peptidyl exit site on the large ribosomal subunit. In
crystal structures of the murine Alu domain, which include the Alu region of the
SRP RNA bound to the SRP9/14 heterodimer, the RNA forms a U-turn that
connects two helical stacks (32, 33). In contrast to earlier models, however, this
structure does not resemble that of tRNA, implying that molecular mimicry is not
546 DOUDNA yBATEY
the mechanism of ribosome binding or elongation arrest. Part of the structural
organization of the S domain of the SRP has recently been revealed in several
crystal structures, which culminate with the human S domain RNA/SRP19/
SRP54 M domain complex (34–36). The cryo-EM reconstruction of the mam-
malian SRP bound to the 80S ribosome is eagerly awaited to reveal the
mechanism of simultaneous translational arrest and signal peptide binding by the
SRP (89).
Assembly of the Signal Sequence Binding Domain
The universally conserved ribonucleoprotein core of the cytosolic SRP comprises
the SRP54/Ffh protein and domain IV of the SRP RNA. In bacteria, these two
components interact through the methionine-rich M domain of SRP54/Ffh to
form a functional enzyme. The crystal structure of an E. coli SRP RNA/Ffh M
domain complex (19) revealed contacts between two internal loops in the RNA
and a series of strictly conserved amino acids in the M domain (Figure 4a).
Strikingly, the crystal structure of a ternary complex between the human
M-domain/SRP-19/S-domain RNA showed nearly identical architecture in the
protein and RNA at this site of contact as well as an identical set of contacts
between the two (36). Clearly, this interface remained constant over evolutionary
time.
The first internal loop, symmetric in all SRP RNAs, contains six universally
conserved nucleotides that form three noncanonical base pairs. Functional group
mutagenesis showed that disruption of either of two of these pairs, a sheared G-G
pair or a reverse Hoogsteen A-C pair, abolishes protein binding (22, 37);
similarly, a single point mutation in this loop is lethal in E. coli. In contrast to the
symmetric loop, the sequence of the asymmetric internal loop of the SRP RNA
is more variable (38, 39). In both the E. coli and human SRP complexes, the
asymmetric loop presents a 5⬘-side adenosine base to the M domain for extensive
recognition by three universally conserved amino acids (19, 36). Because all
phylogenetic variants of the SRP RNA have at least one adenosine on the 5⬘side
of the asymmetric loop, this set of contacts is probably universally conserved.
To overcome unfavorable electrostatic consequences of extruding the adeno-
sine, a series of metal ions interact specifically with the major groove of the
asymmetric internal loop (23). An extensive hydrogen bonding network between
water, three cations, and the RNA, observed crystallographically (19), is essential
for stability; removal of all metal cations from the binding reaction reduced the
binding affinity of the complex by at least 10
6
-fold (23). Although metal ions
were observed throughout the major groove of the RNA, the metal ions in the
asymmetric loop appear to stabilize the protein-RNA complex. For example,
Mn
2⫹
preferentially stabilizes the complex compared to Mg
2⫹
by enabling
formation of an additional A-A pair that likely further stabilizes the bound
conformation (23). Additionally, Cs
⫹
enhances stability relative to other mono-
valent cations, in contrast to its behavior in other RNAs (40) and protein-RNA
complexes (41), because Cs
⫹
binds at the site where the backbone comes into
547THE SIGNAL RECOGNITION PARTICLE
close contact with itself and presumably alleviates unfavorable electrostatic
interactions (23). Thus, the two metals that most stabilize the protein-RNA
complex both affect the structure of the asymmetric internal loop.
Archaea and eukarya require the presence of a second protein, SRP19, for the
efficient binding of SRP54 to the SRP RNA. This protein binds to another
domain of the SRP RNA, domain III, and to the conserved GNRA tetraloop of
domain IV (35) (Figure 4b). A solution structure of SRP19 alone (42), as well as
crystal structures the S-domain RNA bound to SRP 19 (34, 35, 43), and
SRP19/SRP54-M domain (36), show that cooperative assembly occurs through
the stabilization of the bound form of the asymmetric internal loop motif via
A-minor base triples (37, 44) with a second RNA helix (helix 6) (Figure 4b,c)
(45–47). As in the E. coli structure, magnesium ions bind the asymmetric loop,
presumably further stabilizing this region of the RNA. Thus, cooperative assem-
bly of this RNP appears to be driven by the formation of new RNA-RNA contacts
rather than protein-protein contacts, similar to the assembly of the central domain
of the 30S ribosomal subunit (48, 49).
Although their interactions with SRP RNA are nearly identical, Ffh and
SRP54 in the bacterial and human SRP complexes, respectively, use different
strategies to stabilize the critical asymmetric loop. This begs the question as to
why the SRP19/domain III-mediated assembly arose. Currently no other role for
SRP19 in SRP function has been determined, suggesting that the primary role for
this protein in the SRP is facilitating its assembly and stabilizing the intact
particle. Thus, the domain III/SRP19 extension may be an evolutionary adapta-
tion of the SRP to enhance or control the kinetics of assembly that cannot be
achieved by metals alone.
THE SIGNAL RECOGNITION PARTICLE CYCLE
Nascent protein targeting involves a choreographed cycle of signal peptide
binding and release coupled to GTP hydrolysis and interactions with the ribo-
some and translocon pore. Exciting recent advances in understanding GTPase
activity and SRP-ribosome interactions have come from x-ray crystallography,
cryo-electron microscopy and mutagenesis studies.
Signal Sequence Recognition
Though a number of structures of various domains of the SRP have revealed a
wealth of information about protein-RNA, protein-protein, and protein-nucle-
otide interactions, few insights have emerged as to how the most critical ligand
of all—the signal sequence of the protein to be targeted—is specifically recog-
nized. A typical signal sequence comprises 9–12 large hydrophobic residues in
a row (50) that adopt an
␣
-helical conformation (51–53). In eukarya, these
signals are typically found at the N terminus of the protein; however, in E. coli
548 DOUDNA yBATEY
the SRP-dependent signal is often a transmembrane helix within inner membrane
proteins (54–57). The SRP appears to recognize any sequence that bears a critical
level of hydrophobicity, though flanking basic residues are also important (52,
58). Currently unanswered is the fascinating question of how the SRP recognizes
and productively binds almost any such hydrophobic
␣
-helix.
Early studies of target recognition by the SRP suggested that nascent polypep-
tide chains bearing signal sequences cross-linked to SRP54/Ffh through the M
domain and that the M domain was sufficient to mediate this interaction (59, 60).
From these data, the “methionine bristle”hypothesis was proposed in which a
flexible, methionine-rich pocket is used by the SRP to recognize almost any
given signal sequence (61). Consistent with this hypothesis, structures of the M
domain in the free and RNA-bound states show that the methionine-rich region
of the protein, the presumed signal binding site, is conformationally flexible
(Figure 5). In the structure of the T. aquaticus M domain (17), this region is
involved in extensive crystal packing contacts with the signal binding site of an
adjacent protein in the crystal lattice (Figure 5b). Consequently, it takes on a

-hairpin like structure, called the finger loop, that is partially inserted into a
neighboring hydrophobic groove. Because authentic signal sequences are prob-
ably
␣
-helical, the authors contend that the structure illustrates the inherent
flexibility rather than the signal sequence binding mechanism of the finger loop.
Underscoring its flexibility, this region of the M domain was entirely disordered
in the E. coli M domain-4.5S RNA complex structure (Figure 5a) (19, 22). This
may represent the true state of the signal recognition site in the SRP in solution,
consistent with the conformational flexibility required for productive binding of
heterogeneous targets.
Crystal structures of the human M domain present a different picture of how
the signal sequence potentially interacts with the SRP. In the absence (18) and
presence of SRP-RNA (36), the first helix of the M domain undergoes a “domain
swap”with an adjacent protein (Figure 5c). This has been observed in the crystal
structures of other proteins (62, 63) and suggests weak interactions between this
helix and the rest of the M domain. The result of the domain swap is that helix
one (h1⬘, Figure 5c) packs into a shallow, moderately hydrophobic groove of an
adjacent molecule. Thus, in an alternative model of how the M domain may
recognize signal sequences, helix 1⬘(h1⬘) occupies nearly the same position as
helix 1 in the E. coli and T. aquaticus structures, although helix 1 represents the
signal peptide (Figure 5d).
These models for peptide recognition leave unanswered the question concern-
ing the involvement of the NG domain and the RNA in target recognition by the
SRP. Even though the M domain can bind the signal on its own or in complex
with SRP RNA, the affinity of the interaction was improved by the NG domain,
and the NG domain can be cross-linked to signal peptides in solution (64, 65).
Whereas biochemical and crystallographic studies indicate that the NG and M
domains of Ffh are loosely associated in the absence of a bound signal sequence
and/or GTP (17, 66), a recent crystal structure of the intact SRP54 protein
549
THE SIGNAL RECOGNITION PARTICLE
550 DOUDNA yBATEY
provides clues to interdomain communication that may occur upon signal
binding (see below). It has also been proposed that the SRP RNA has a role in
signal recognition via electrostatic interactions between the backbone of the
RNA and positively charged residues adjacent to the hydrophobic sequence (19),
a hypothesis supported by mutagenesis and in vivo experiments (58, 67).
Although these data provide tantalizing clues to the mechanism of protein
targeting by the SRP, a structure of the SRP-signal complex will be an important
step toward an atomic-level understanding of SRP function.
Structure of the SRP54-RNA Complex
SRP54 is the only protein subunit conserved in all SRPs and controls commu-
nication with the SRP receptor, the ribosome, and the translocon. In the crystal
structure of T. aquaticus SRP54, the linker region between the G and M domains
was disordered and hence provided no information about the three-dimensional
domain arrangement of SRP54 or its organization in complex with SRP RNA.
Recent determination of structures of SRP54 from the archaeon Sulfolobus
solfataricus alone and complexed with helix 8 of SRP RNA reveal the architec-
ture of the complex and a hydrophobic contact between the M and N domains,
suggesting a possible mechanism for interdomain communication (67a). The
structures, solved by molecular replacement at a resolution of ⬃4Å, reveal an
L-shaped protein in which the NG domain represents the long arm and the M
domain represents the short arm. Helix 8 of SRP RNA lies parallel to the long
axis of the NG domain, giving the complex an overall U shape (Figure 6). Only
one region of interaction, which involves a short stretch of hydrophobic contacts
between the loop connecting helices
␣
N3 and
␣
N4 at the distal end of the N
domain, the N-terminal region of
␣
ML, and the C-terminal region of the short
␣
helix
␣
M1b adjacent to the finger loop, is observed between the N and M
domains (Figure 6). The high degree of evolutionary conservation of these
4™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™
Figure 5 Potential models of signal sequence recognition by the SRP54/Ffh M domain.
(a) The hydrophobic, methionine-rich pocket in the E. coli M domain (PDB ID: 1DUL) is
significantly disordered (dashed line), suggesting a conformationally dynamic binding site
that can recognize a variety of sequences. (b) The conformation of the finger loop between
helices one and two of the T. aquaticus M domain structure (PDB ID: 2FFH) is stabilized
through its interaction with a significantly hydrophobic groove of an adjacent M domain in
the crystal (note that the signal corresponds to the finger loop). (c) A model of signal
recognition proposed from the structure of the human SRP M domain (PDB ID: 1QB2).
Helix one (h1⬘) of an adjacent M domain in the crystal structure packs into a shallow,
moderately hydrophobic groove formed between helices one and two (h1, h2). (d)An
alternative model for signal sequence binding by the human M domain in which h1⬘is
considered part of the domain structure by virtue of its close superposition with helix 1 of
the E. coli and T. aquaticus structures and helix 1 represents the signal sequence.
551THE SIGNAL RECOGNITION PARTICLE
residues suggests a functional role of the contact, perhaps as a greasy hinge
enabling interdomain flexibility in SRP54. Although the relative orientations of
the NG and M domains are not affected by RNA binding, shape and charge
complementarity between the phosphate backbone of the RNA minor groove and
residues 121–126 of the G domain suggest a possible contact that might be
induced at some stage in the SRP cycle; this would enable communication
between the peptide binding and GTPase functionalities of the particle.
GTPase Stimulation in the SRP-SR Complex
Two central questions about the mechanism of SRP-mediated protein targeting
are how GTPase activities in the SRP54/Ffh and SR
␣
/FtsY proteins are enhanced
upon heterodimerization of the SRP with its receptor, and how GTP hydrolysis
is coupled to peptide binding and release. Crystal structures of the homologous
GTPase domains of these two proteins, determined in their apo- and nucleotide-
bound states, revealed a common two-domain NG fold comprising an
␣
-helical
Figure 6 Crystal structure of a complex between Sulfolobus solfataricus SRP domain IV
RNA and SRP54 (67a). The N domain of SRP54 (blue) contacts the N-terminal helix of the
M domain (red), while the M domain forms a complex with the SRP RNA (gray)ina
fashion similar to that of the bacterial and eukaryal variants. Figure courtesy of I. Sinning.
552 DOUDNA yBATEY
N domain packed against a G domain with a fold similar to those of other
members of the GTPase superfamily. Structural similarity between the G
domains of Ffh and FtsY and dimeric ATP-utilizing proteins led to a model for
the Ffh-FtsY complex in which the G domains dimerize in an antiparallel
orientation (68). Similar to other GTPases, four conserved motifs, I-IV, comprise
the SRP GTPases and include residues directly involved in GTP binding and
hydrolysis. In addition, the G domains of both Ffh and FtsY contain a subdomain,
termed the insertion box domain (IBD), that extends the core
␣
/

fold along the
face distal to the N domain and was thought to provide the site of interaction in
the targeting complex (12, 14, 69). Despite their similarities, the SRP GTPases
exhibit several distinct properties relative to other GTPases, which include low
affinity and rapid exchange of both GDP and GTP (70). GTP hydrolysis in an
Ffh/FtsY complex is stimulated about 10-fold above that observed in either
protein alone.
A significant advance in understanding GTPase activation has come from
recent structure determinations of a complex between the Ffh and FtsY NG
domains (70a, 70b). The two proteins associate longitudinally along the N and G
domains, bringing the two active sites into direct contact to form a contiguous
catalytic chamber that contains two bound GMPPCP molecules, two hydrated
magnesium ions, and several waters. The contact region between the domains
comprises most of the conserved sequence motifs of the SRP GTPases and buries
⬃1800 Å
2
on each protein surface. Mutation of residues across the interface
surface in FtsY disrupts GTPase activity in the complex, confirming the impor-
tance of the crystallographically observed interaction (70a). Formation of the
Ffh-FtsY interface requires several conformational changes relative to the struc-
tures of the domains alone, including a rigid body motion of the N and G domains
that translates the distal loops of the N domain by ⬃11–12 Å. This movement is
accompanied by a substantial shift in the position of the C-terminal helix, leading
to speculation that the N domain functions as a sensor of targeting complex
formation by triggering a change in the relative orientations of the M-domain and
A-domain that are found on the C-terminal ends of the full-length Ffh and FtsY
proteins, respectively. Conserved arginine sidechains in each protein that might
serve as arginine fingers, analogous to the Ras/GAP structure, are oriented
asymmetrically such that only the Arg in Ffh is positioned like the Arg supplied
by the RasGAP. Possibly, the arrangement of arginines alternates within the
chamber, perhaps contributing sequentially to hydrolysis of the two bound
nucleotides.
Ribosome Interactions
In the mammalian system, SRP54 cross-links to two ribosomal proteins, L23a
and L35, located near the polypeptide exit site on the ribosome (71). A similar
interaction is observed in eubacteria, where Ffh/4.5S RNA complexes can be
cross-linked to L23 (72, 73). Interestingly, protein L23 also cross-links to a
chaperone protein, trigger factor (TF), though L23-SRP and L23-TF interactions
553THE SIGNAL RECOGNITION PARTICLE
appear to be mutually exclusive. These data suggest that L23 may play a role in
directing nascent polypeptides into the translocation machinery, though this
hypothesis remains to be tested in vivo. The ribosome somehow induces a
structural change in SRP54 that leads to increased GTP binding affinity, as well
as increased affinity for the ribosome itself (74, 75). One possibility, suggested
by the SRP54/RNA structure, is that this structural rearrangement corresponds to
rotation of the NG domain with respect to the M domain that may occur upon
signal peptide binding, with consequent opening of the finger loop similar to the
conformation observed in the T. aquaticus Ffh structure (17). Signal peptide
binding to the M domain might therefore result in a similar structural rearrange-
ment in the NG domain interface as occurs upon GTP binding to the G domain,
effectively linking signal sequence binding to the M domain with GTP binding
to the G domain.
Nucleotide Exchange
The molecular mechanism of reciprocal GTPase activity in SRP and its receptor
remains poorly understood. Recent evidence indicates that structural changes
induced in the bacterial SRP receptor, FtsY, upon formation of the SRP-FtsY
complex enhance nucleotide binding specificity in FtsY (76). Mutagenesis
studies support a similar weak-binding affinity in the eukaryotic homolog SR
␣
(77, 78). Why might this occur? One idea is that loosely bound GTP in free FtsY
would prevent futile cycles of GTP hydrolysis in the substantial fraction of free
FtsY in the cytosol (79). In this way, nucleotide hydrolysis would be coupled to
binding of SRP and presumably to nascent signal peptides.
FUNCTIONS OF SIGNAL RECOGNITION
PARTICLE RNA
Why does cellular SRP include an essential RNA, and what does the RNA
contribute to SRP function as well as to other physiological activities in the cell?
Most of the evidence to date addressing these questions comes from studies in the
E. coli system. In vitro, the 4.5S RNA appears to stabilize the structure of the Ffh
M domain, as indicated by circular dichroism and proteolysis experiments (66,
80), and studies on the kinetics of Ffh-FtsY complex formation show that 4.5S
RNA enhances both association and dissociation of the complex (81, 82).
Intriguingly, several lines of evidence support an additional role for 4.5S RNA in
translation on the ribosome, because of the observation that the deleterious
effects of 4.5S RNA depletion can be suppressed by mutations in translation
factor EF-G or in the 16S or 23S ribosomal RNAs (83–86). Recent evidence
implies that although the SRP RNA interaction with EF-G homologs is con-
served in archaea, the essential activity of SRP RNA is in fact as part of the
signal-sequence binding particle rather than on the ribosome (90). It is intriguing
554 DOUDNA yBATEY
to note that in chloroplasts the SRP RNA has apparently been replaced by a
protein, indicating that study of this SRP may provide clues to the role of the
RNA (87, 88).
FUTURE DIRECTIONS
With many structures of individual components of the SRP pathway now known,
attention is focused on understanding how these molecules interact to enable
efficient protein targeting. How signal peptide binding is achieved, how peptide
binding and release is controlled, and how the SRP coordinates interactions with
its receptor, the ribosome, and the translocon remain fascinating unanswered
questions. A combination of genetic, biochemical, and structural approaches will
be required to address these issues and fully illuminate the function of one of the
most ancient of the cellular ribonucleoproteins.
ACKNOWLEDGMENTS
We thank Irmgard Sinning for sharing unpublished data and for preparation of
Figure 6. This work was supported in part by the NIH (GM 22778 to J.A.D.).
The Annual Review of Biochemistry is online at http://biochem.annualreviews.org
LITERATURE CITED
1. Keenan RJ, Freymann DM, Stroud RM,
Walter P. 2001. Annu. Rev. Biochem. 70:
755–75
2. Driessen AJ, Manting EH, van der
Does C. 2001. Nat. Struct. Biol. 8:
492–98
3. Eichler J, Moll R. 2001. Trends Micro-
biol. 9:130–36
4. Sauer-Eriksson AE, Hainzl T. 2003.
Curr. Opin. Struct. Biol. 13:64–70
5. Nagai K, Oubridge C, Kuglstatter A,
Menichelli E, Isel C, Jovine L. 2003.
EMBO J. 22:3479–85
6. Fulga TA, Sinning I, Dobberstein B,
Pool MR. 2001. EMBO J. 20:2338–47
7. de Leeuw E, Poland D, Mol O, Sinning I,
ten Hagen-Jongman CM, et al. 1997.
FEBS Lett. 416:225–29
8. de Leeuw E, te Kaat K, Moser C,
Menestrina G, Demel R, et al. 2000.
EMBO J. 19:531–41
9. Herskovits AA, Bibi E. 2000. Proc. Natl.
Acad. Sci. USA 97:4621–26
10. Herskovits AA, Seluanov A, Rajsbaum
R, ten Hagen-Jongman CM, Henrichs T,
et al. 2001. EMBO Rep. 2:1040–46
11. Moll RG. 2003. Biochem. J. 374:247–54
12. Freymann DM, Keenan RJ, Stroud RM,
Walter P. 1997. Nature 385:361–64
13. Montoya G, Svensson C, Luirink J, Sin-
ning I. 1997. Nature 385:365–68
14. Freymann DM, Keenan RJ, Stroud RM,
Walter P. 1999. Nat. Struct. Biol. 6:
793–801
15. Ramirez UD, Minasov G, Focia PJ,
Stroud RM, Walter P, et al. 2002. J. Mol.
Biol. 320:783–99
16. Shepotinovskaya IV, Freymann DM.
2002. Biochim. Biophys. Acta 1597:
107–14
17. Keenan RJ, Freymann DM, Walter P,
Stroud RM. 1998. Cell 94:181–91
555THE SIGNAL RECOGNITION PARTICLE
18. Clemons WM Jr, Gowda K, Black SD,
Zwieb C, Ramakrishnan V. 1999. J. Mol.
Biol. 292:697–705
19. Batey RT, Rambo RP, Lucast L, Rha B,
Doudna JA. 2000. Science 287:1232–39
20. Schmitz U, James TL, Lukavsky P,
Walter P. 1999. Nat. Struct. Biol. 6:
634–38
21. Jovine L, Hainzl T, Oubridge C, Scott
WG, Li J, et al. 2000. Struct. Fold. Des.
8:527–40
22. Batey RT, Sagar MB, Doudna JA. 2001.
J. Mol. Biol. 307:229–46
23. Batey RT, Doudna JA. 2002. Biochem-
istry 41:11703–10
24. Miller JD, Wilhelm H, Gierasch L, Gil-
more R, Walter P. 1993. Nature 366:
351–54
25. Powers T, Walter P. 1995. Science 269:
1422–24
26. Schwartz T, Blobel G. 2003. Cell 112:
793–803
27. Bacher G, Pool M, Dobberstein B. 1999.
J. Cell Biol. 146:723–30
28. Legate KR, Falcone D, Andrews DW.
2000. J. Biol. Chem. 275:27439–46
29. Helmers J, Schmidt D, Glavy JS, Blobel
G, Schwartz T. 2003. J. Biol. Chem. 278:
23686–90
30. Andrews DW, Walter P, Ottensmeyer
FP. 1985. Proc. Natl. Acad. Sci. USA 82:
785–89
31. Andrews DW, Walter P, Ottensmeyer
FP. 1987. EMBO J. 6:3471–77
32. Weichenrieder O, Wild K, Strub K,
Cusack S. 2000. Nature 408:167–73
33. Weichenrieder O, Stehlin C, Kapp U,
Birse DE, Timmins PA, et al. 2001. RNA
7:731–40
34. Wild K, Sinning I, Cusack S. 2001. Sci-
ence 294:598–601
35. Hainzl T, Huang S, Sauer-Eriksson AE.
2002. Nature 417:767–71
36. Kuglstatter A, Oubridge C, Nagai K.
2002. Nat. Struct. Biol. 9:740–44
37. Doherty EA, Batey RT, Masquida B,
Doudna JA. 2001. Nat. Struct. Biol. 8:
339–43
38. Regalia M, Rosenblad MA, Samuelsson
T. 2002. Nucleic Acids Res. 30:3368–77
39. Rosenblad MA, Gorodkin J, Knudsen B,
Zwieb C, Samuelsson T. 2003. Nucleic
Acids Res. 31:363–64
40. Basu S, Rambo RP, Strauss-Soukup J,
Cate JH, Ferre-D’Amare AR, et al. 1998.
Nat. Struct. Biol. 5:986–92
41. Conn GL, Gittis AG, Lattman EE, Misra
VK, Draper DE. 2002. J. Mol. Biol. 318:
963–73
42. Pakhomova ON, Deep S, Huang Q,
Zwieb C, Hinck AP. 2002. J. Mol. Biol.
317:145–58
43. Oubridge C, Kuglstatter A, Jovine L,
Nagai K. 2002. Mol. Cell 9:1251–61
44. Nissen P, Ippolito JA, Ban N, Moore PB,
Steitz TA. 2001. Proc. Natl. Acad. Sci.
USA 98:4899–903
45. Diener JL, Wilson C. 2000. Biochemistry
39:12862–74
46. Rose MA, Weeks KM. 2001. Nat. Struct.
Biol. 8:515–20
47. Yin J, Yang CH, Zwieb C. 2001. RNA
7:1389–96
48. Agalarov SC, Williamson JR. 2000. RNA
6:402–8
49. Agalarov SC, Prasad GS, Funke PM,
Stout CD, Williamson JR. 2000. Science
288:107–13
50. Valent QA, Kendall DA, High S, Kusters
R, Oudega B, Luirink J. 1995. EMBO J.
14:5494–505
51. McKnight CJ, Rafalski M, Gierasch LM.
1991. Biochemistry 30:6241–46
52. Lee HC, Bernstein HD. 2001. Proc. Natl.
Acad. Sci. USA 98:3471–76
53. Adams H, Scotti PA, De Cock H,
Luirink J, Tommassen J. 2002. Eur.
J. Biochem. 269:5564–71
54. Ulbrandt ND, Newitt JA, Bernstein HD.
1997. Cell 88:187–96
55. Beck K, Wu LF, Brunner J, Muller M.
2000. EMBO J. 19:134–43
56. Park SK, Jiang F, Dalbey RE, Phillips
GJ. 2002. J. Bacteriol. 184:2642–53
57. Sijbrandi R, Urbanus ML, ten Hagen-
556 DOUDNA yBATEY
Jongman CM, Bernstein HD, Oudega B,
et al. 2003. J. Biol. Chem. 278:4654–59
58. Peterson JH, Woolhead CA, Bernstein
HD. 2003. J. Biol. Chem. 278:46155–62
59. Romisch K, Webb J, Lingelbach K,
Gausepohl H, Dobberstein B. 1990.
J. Cell Biol. 111:1793–802
60. Zopf D, Bernstein HD, Johnson AE,
Walter P. 1990. EMBO J. 9:4511–17
61. Bernstein HD, Poritz MA, Strub K,
Hoben PJ, Brenner S, Walter P. 1989.
Nature 340:482–86
62. Clemons WM Jr, Davies C, White SW,
Ramakrishnan V. 1998. Structure 6:
429–38
63. Liu Y, Eisenberg D. 2002. Protein Sci.
11:1285–99
64. Newitt JA, Bernstein HD. 1997. Eur.
J. Biochem. 245:720–29
65. Cleverley RM, Gierasch LM. 2002.
J. Biol. Chem. 277:46763–68
66. Zheng N, Gierasch LM. 1997. Mol. Cell
1:79–87
67. Huang Q, Abdulrahman S, Yin J, Zwieb
C. 2002. Biochemistry 41:11362–71
67a. Rosendal, KR, Wild, K, Montoya, G,
Sinning, I. 2003. Proc. Natl. Acad. Sci.
USA 100:14701–6
68. Montoya G, Kaat K, Moll R, Schafer G,
Sinning I. 2000. Struct. Fold. Des. 8:
515–25
69. Padmanabhan S, Freymann DM. 2001.
Structure 9:859–67
70. Lu Y, Qi HY, Hyndman JB, Ulbrandt
ND, Teplyakov A, et al. 2001. EMBO J.
20:6724–34
70a. Egea, PF, Shan, S, Napetschnig, J, Sav-
age, DF, Walter, P, Stroud, RM. 2004.
Nature 427:215–21
70b. Focia, PJ, Shepotinovskaya, IV, Seidler,
JA, Freymann, DM. 2004. Science 303:
373–7
71. Pool MR, Stumm J, Fulga TA, Sinning I,
Dobberstein B. 2002. Science 297:
1345–48
72. Gu SQ, Peske F, Wieden HJ, Rodnina
MV, Wintermeyer W. 2003. RNA 9:
566–73
73. Ullers RS, Houben EN, Raine A, ten
Hagen-Jongman CM, Ehrenberg M, et
al. 2003. J. Cell Biol. 161:679–84
74. Bacher G, Lutcke H, Jungnickel B, Rap-
oport TA, Dobberstein B. 1996. Nature
381:248–51
75. Flanagan JJ, Chen JC, Miao Y, Shao Y,
Lin J, et al. 2003. J. Biol. Chem. 278:
18628–37
76. Shan SO, Walter P. 2003. Proc. Natl.
Acad. Sci. USA 100:4480–85
77. Kusters R, Lentzen G, Eppens E, van
Geel A, van der Weijden CC, et al. 1995.
FEBS Lett. 372:253–58
78. Rapiejko PJ, Gilmore R. 1997. Cell 89:
703–13
79. Luirink J, ten Hagen-Jongman CM, van
der Weijden CC, Oudega B, High S, et
al. 1994. EMBO J. 13:2289–96
80. Cleverley RM, Zheng N, Gierasch LM.
2001. J. Biol. Chem. 276:19327–31
81. Peluso P, Herschlag D, Nock S, Frey-
mann DM, Johnson AE, Walter P. 2000.
Science 288:1640–43
82. Peluso P, Shan SO, Nock S, Herschlag
D, Walter P. 2001. Biochemistry 40:
15224–33
83. Brown S. 1987. Cell 49:825–33
84. Brown S, Thon G, Tolentino E. 1989. J.
Bacteriol. 171:6517–20
85. Brunelli CA, O’Connor M, Dahlberg
AE. 2002. FEBS Lett. 514:44–48
86. Rinke-Appel J, Osswald M, von Kno-
blauch K, Mueller F, Brimacombe R, et
al. 2002. RNA 8:612–25
87. Eichacker LA, Henry R. 2001. Biochim.
Biophys. Acta 1541:120–34
88. Groves MR, Mant A, Kuhn A, Koch J,
Dubel S, et al. 2001. J. Biol. Chem. 276:
27778–86
89. Halic M, Becker T, Pool MR, Spahn
CM, Grassucci RA, Frank J, Beckmann
R. 2004. Nature 427:808–14
90. Sagar MB, Lucast L, Doudna JA. 2004.
RNA. In press
557THE SIGNAL RECOGNITION PARTICLE