Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2018 Feb 1.
Published in final edited form as: Mol Microbiol. 2016 Dec 7;103(3):439–451. doi: 10.1111/mmi.13567

SecA functions in vivo as a discrete anti-parallel dimer to promote protein transport

Tithi Banerjee 1, Christine Lindenthal 1,†, Donald Oliver 1,*
PMCID: PMC5263173  NIHMSID: NIHMS830607  PMID: 27802584

Summary

SecA ATPase motor protein plays a central role in bacterial protein transport by binding substrate proteins and the SecY channel complex and utilizing its ATPase activity to drive protein translocation across the plasma membrane. SecA has been shown to exist in a dynamic monomer-dimer equilibrium modulated by translocation ligands, and multiple structural forms of the dimer have been crystallized. Since the structural form of the dimer remains a controversial and unresolved question, we addressed this matter by engineering ρ-benzoylphenylalanine along dimer interfaces corresponding to the five different SecA x-ray structures and assessing their in vivo photo-crosslinking pattern. A discrete anti-parallel 1M6N-like dimer was the dominant if not exclusive dimer found in vivo, whether SecA was cytosolic or in lipid or SecYEG-bound states. SecA bound to a stable translocation intermediate was crosslinked in vivo to a second SecA protomer at its 1M6N interface, suggesting that this specific dimer likely promotes active protein translocation. Taken together, our studies strengthen models that posit, at least in part, a SecA dimer-driven translocation mechanism.

Abbreviated Summary

SecA ATPase facilitates protein transport through the integral membrane SecY channel complex. The physiological form of the various SecA dimers that have been crystallized was explored by engineering a site-specific crosslinker into potential dimer interfaces and performing in vivo photo-crosslinking. The results indicate that a single discrete dimer species is present within the cell, and this species was also captured during arrested protein transport through the SecY channel utilizing a SecA-OmpA-GFP trimeric protein.

graphic file with name nihms830607u1.jpg

Introduction

A single universally conserved protein-conducting channel complex facilitates the transport of proteins into or across the plasma membrane of prokaryotes or the endoplasmic reticular membrane of eukaryotes (reviewed in (Park and Rapoport 2012)). The channel is largely formed from a single protein, SecY in Bacteria or Sec61α in Achaea and Eukarya, that forms an hourglass-shaped structure with at least three functionally important elements: a constriction ring at its center that helps to seal the channel and limit ion permeability, a lateral gate on one side that is important for signal/anchor sequence intercalation and channel expansion as well as lateral release of integral membrane proteins into the lipid bilayer, and a plug domain at the bottom that helps to regulate channel opening and activity. Both pre-secretory and integral membrane proteins are directed to this channel complex utilizing either co-translational or post-translational targeting routes depending on the nature of the substrate protein as well as the organism. In Bacteria, integral membrane proteins primarily utilize the signal recognition particle/signal recognition particle receptor system to target nascent chain-bearing polysomes to the SecY complex, while most pre-secretory proteins utilize SecA protein and its substrate-specific SecB chaperone for such targeting (reviewed in (Lycklama a Nijeholt and Driessen 2012)). SecA directly recognizes the signal peptides present on pre-secretory proteins as well as SecY protein, thus serving as a critical factor for assembly of the active translocation complex. Once bound to both substrate and the SecY channel complex, SecA ATPase activity promotes conformationally-driven domain movements that are responsible for step-wise protein translocation. Whether SecA functions in a piston-like fashion to actively push protein segments across the channel or serves as a Brownian ratchet to bias forward movement of the polypeptide chain, or both, remain active areas of investigation (for recent contrasting views see (Bauer, Shemesh et al. 2014, Allen, Corey et al. 2016)).

SecA protein is found in a salt and concentration-dependent monomer-dimer equilibrium, which would predict mainly a dimer state within cells (Woodbury, Hardy et al. 2002). However many of the critical ligands that regulate SecA activity, such as nucleotides, signal peptides, phospholipids, or SecYEG protein, have been shown to affect this dynamic equilibrium. Numerous biochemical or structural studies have been performed to determine the quaternary state of SecA in solution, bound to membranes or SecYEG protein in proteoliposomes, or during active translocation conditions. These studies have led to a conflicting literature regarding the affects of SecA ligands on its quaternary state and to claims that SecA functions either as a monomer or dimer or some combination of both (see references reviewed in (Sardis and Economou 2010)). Further complicating this picture is a similar controversy as to whether SecYEG protein functions as a monomer or dimer. Studies of SecA quaternary state under in vitro conditions suffer from a number methodological shortcomings: (i) the detergents or high salt conditions utilized in some studies would tend to artificially induce SecA monomerization, (ii) non-physiological levels of ligands utilized in other studies could produce biochemically correct but non-physiological results, (iii) the use of monomer or dimer-biased SecA mutants (either knowingly or unknowingly) or pre-activated SecY (PrlA4) protein may have lead to artificial findings, and (iv) the use of non-equilibrium methods to study this equilibrium process may have generated a biased picture. Clearly additional in vivo or equilibrium in vitro methods are required to help resolve this complex and critical matter.

While SecA has a highly conserved domain structure and protomer fold, no less than five different SecA dimer structures with different subunit interfaces have been reported, including two from the same bacterial species (Hunt, Weinkauf et al. 2002, Sharma, Arockiasamy et al. 2003, Vassylyev, Mori et al. 2006, Zimmer, Li et al. 2006, Papanikolau, Papadovasilaki et al. 2007). Indeed this multitude of quaternary states has lead to the proposal that the dimer may simply function as a storage form within the cell. In order to begin to clarify the physiological form of SecA dimer we recently examined its solution state structure using Förester resonance energy transfer and found that our data was most consistent with the antiparallel 1M6N dimer as a good working model (Auclair, Oliver et al. 2013). This approach and other equilibrium methods can now address SecA quaternary structure dynamics in the presence of SecA ligands, and most importantly, under active translocation conditions.

In the present study we have examined the quaternary state and structure of SecA protein in vivo utilizing site-specific photo-crosslinking. The low efficiency of this technique in comparison to disulfide crosslinking assures that it is minimally perturbing of any natural protein interaction equilibrium. This approach has been utilized previously to map both the SecA-SecY and SecA dimer interfaces (Mori and Ito 2006, Das and Oliver 2011, Yu, Wowor et al. 2013), although only cytosolic SecA was examined in the latter study with somewhat equivocal results.

Results

A 1M6N-like dimer is the dominant SecA dimer within the cytosol in vivo

We located structure-specific “signature” residues at the various interfaces of the five known SecA dimer x-ray structures for incorporation of ρ-benzoylphenylalanine (pBpA) (Fig. 1). Selection criteria included (i) a chosen residue should be close (generally <5 Å) to the nearest residue of the adjacent protomer for the structure of interest but distant (generally >15 Å) to the nearest residue of the adjacent protomer for the other four structures, (ii) the selected residue should preferably be in a well structured region, and (iii) highly conserved residues that could potentially perturb biological structure or function should preferably be avoided. Given the smaller subunit interfaces for certain structures as well as a degree of overlap in the residues that constituted their interfaces, we also selected five multi-structure signature residues that were diagnostic for two or three structures. The selected structure-specific signature residues are given in Supporting Information Table S1.

Fig. 1. SecA dimer x-ray structures with structure-specific signature residues.

Fig. 1

Open in a new tab

A–E. Residues unique to a given structural interface are depicted as red spheres, while interfacial residues shared by two or three structures are depicted as spheres in other colors with the homologous E. coli residue number given at the right of the color key.

A. B. subtilis SecA 1M6N (Hunt, Weinkauf et al. 2002).

B. B subtilis SecA 2IBM (Zimmer, Li et al. 2006).

C. T. thermophilus SecA 2IPC (Vassylyev, Mori et al. 2006).

D. E. coli SecA 2FSF (Papanikolau, Papadovasilaki et al. 2007).

E. M. tuberculosis SecA 1NL3 (Sharma, Arockiasamy et al. 2003).

F. B. subtilis SecA 1M6N with signature residues numbered according to E. coli coordinates utilizing sequence alignment homology.

To conduct in vivo photo-crosslinking we employed a genetic system that utilizes an orthogonal amber suppressor tRNA and tRNA synthetase that incorporates pBpA at engineered amber codons within SecA as described previously (Ryu and Schultz 2006, Das and Oliver 2011). In addition, we utilized a plasmid-borne SecA over-expression system given the low efficiency of pBpa-induced photo-crosslinking (estimated on average at ~10–15% based on quantification of our western blots shown below). Incorporation of pBpA into a typical SecA mutant, SecA653, was verified (Supporting Information Fig. S1). The functionality of our collection of SecA signature mutants was also verified in vivo by their ability to complement the SecA temperature sensitive mutant BL21.1 at the restrictive temperature but solely in the presence of pBpA (Supporting Information Fig. S2). We next screened our collection of SecA signature mutants by photo-crosslinking followed by western blotting of the cytosolic fraction of cells. Nine out of ten 1M6N signature mutants displayed slower mobility, UV-dependent species that were identified as SecA dimers based on their co-migration with dimers generated from the identical SecA mutant that was first purified and subsequently irradiated (Fig. 2A). Given that the purified SecA mutant proteins were predicted to be nearly all dimers at the salt and protein concentrations employed here (>90%) (Woodbury, Hardy et al. 2002, Das, Stivison et al. 2008), and their level of crosslinking was often less than or similar to that observed in vivo, we conclude that cytosolic SecA is mainly a dimer in vivo. Dimers of two different electrophoretic mobilities were observed: one migrating at ~140–150 kDa (SecA339, SecA653, SecA815), and another migrating at ~180 kDa (SecA658, SecA678, SecA682, SecA695, SecA811, SecA815, SecA819). Whether a larger species migrating at ~235 kDa (SecA678, SecA682, SecA695) is a dimer or trimer was not investigated. The existence of at least two classes of dimers and their migration as a family of adjacent species is not surprising given the broad chemical reactivity of pBpA and its ability to create dimers of varying axial radii depending on the positions of the photo-chemically-reactive donor and acceptor residues on the primary amino acid sequences of the conjoined pair. A similar observation has been made in a SecA-SecY photo-crosslinking study (Mori and Ito 2006). Furthermore, certain SecA mutants produced additional major or minor crosslinked species that were not present or detectable in the purified SecA mutant sample (SecA658, SecA678, SecA682, SecA695, SecA811, SecA819; see bands labeled with asterisks). Whether these represent additional SecA dimer species induced only under in vivo conditions or SecA interactions with other proteins was not investigated. We did probe our blots with SecB antisera, but the results were negative. Since within the cell SecA dimer is conformationally modulated by interactions with both small (e.g. adenosine nucleotides) and larger ligands (e.g. ribosomes, SecB, signal peptides and mature regions of protein substrates, SecYEG, and anionic phospholipids), the observed differences between in vivo and in vitro pBpa-induced crosslinking patterns would be expected.

Fig. 2. Analysis of SecA signature mutants by in vivo photo-crosslinking.

Fig. 2

Open in a new tab

A, B. Western blots probed with SecA antisera of either the cytosolic fraction of the indicated SecA signature mutant subjected to in vivo photo-crosslinking for the indicated times or the identical purified SecA mutant protein subjected to in vitro photo-crosslinking for 20 min. SecA dimer bands (A-A) or higher molecular weight dimer or trimer bands (A-A?) are indicated. SecA crosslinked species of unknown identity are labeled by asterisks. The blots shown are representative of three independent experiments.

A. 1M6N signature mutants.

B. Signature mutants for the remaining structures as indicated.

The one 1M6N signature mutant that was negative for crosslinking, SecA665, was analyzed in silico. While this residue is very close (2–3 Å) to a potential acceptor in the neighboring protomer in the 1M6N dimer, we noted that it is within a confined and rigid region of the structure that might hamper the crosslinking reaction, which has been found to require a 108.9° angle for optimal reaction of the UV-induced pBpA di-radical (Dorman and Prestwich 1994). Yu et al. also obtained a similar negative result for this residue in their parallel study (Yu, Wowor et al. 2013).

The mutant analysis for the four other SecA dimer structures was definitive: none of the four (for 2IPC) or five (for 2IBM, 2FSF or 1NL3) signature mutants tested for each structure displayed any SecA dimer (Fig. 2B). These results strongly suggest that these latter four dimers are only relevant in the context of SecA crystallography or that they constitute only a minor fraction of SecA quaternary state in the cytosol in vivo. Given our rate of negative results in selecting 1M6N signature mutants (10%), the probability of achieving uniformly negative results for all four or five mutants for each of the other structures if a legitimate dimer interface existed would be 10−4 or 10−5, respectively. Clearly only a single dominant SecA dimer form (1M6N-like) exists physiologically within the cytoplasm.

Only a 1M6N-like dimer is found in lipid and SecYEG-bound states in vivo

Cellular fractionation studies have shown that under normal physiological conditions approximately two-thirds of SecA is present in the cytosol, while the remainder is divided approximately equally between lipid and SecYEG-bound states (Cabelli, Dolan et al. 1991). These latter two states can be readily distinguished by treatment of membranes with chaotropic agents like urea that remove lipid-bound SecA while preserving SecYEG-bound SecA (Eichler, Rinard et al. 1998). SecA associated with the channel complex is particularly important, since it represents, at least in part, SecA actively engaged in promoting membrane translocation steps. SecYEG-bound SecA has been variously reported to be a monomer or dimer in vitro depending on the method of analysis (see references reviewed in (Sardis and Economou 2010)). In order to address which SecA quaternary states are present within these two membrane pools, we performed cell fractionation studies with our SecA signature mutants after photo-crosslinking. The absence of the integral membrane protein OmpA from the cytosolic and peripheral membrane fractions after treatment of total membranes with 6 M urea was utilized to ensure the integrity of our fractionation methodology (Supporting Information Fig. S3). In general, we found a similar pattern of SecA dimer species in all three fractions for a given mutant for the five 1M6N signature mutants studied (Fig. 3A). However the amount of dimer in the peripheral membrane fraction was generally reduced (less so for the strongly crosslinking mutants SecA653 and SecA815), consistent with the ability of phospholipids to induce SecA monomerization in vitro (Or, Navon et al. 2002, Benach, Chou et al. 2003). In certain cases we noted additional crosslinked species present only within the cytosolic or integral membrane fractions (SecA658, SecA678; see bands labeled with asterisks). Whether these represent novel SecA dimer species or compartment-specific interaction of SecA with other proteins was not investigated further.

Fig. 3. Analysis of cytosolic, peripheral and integral membrane fractions of SecA signature mutants subjected to in vivo photo-crosslinking.

Fig. 3

Open in a new tab

A–B. Western blots probed with SecA antisera of cytosolic, total membrane, peripheral or integral membrane fractions of the indicated SecA signature mutant subjected to in vivo photo-crosslinking for the indicated times or the identical purified SecA mutant protein subjected to in vitro photo-crosslinking for 20 min. The same blot for all fractions of a given mutant is shown to facilitate comparison. SecA dimer bands (A-A) or higher molecular weight dimer or trimer bands (A-A?) are indicated. SecA crosslinked species of unknown identity are labeled by asterisks. The blots shown are representative of three independent experiments.

A. 1M6N signature mutants.

B. Signature mutants for the remaining structures as indicated.

A very different result was obtained in our fractionation analysis of signature mutants for the four other SecA dimer structures. We analyzed 11 mutants that tested each structure at four positions given the use of multi-structure-specific signature mutants. Consistent with our results above (Fig. 2B), there were no detectable SecA dimers present in any of the fractions analyzed (Fig. 3B). Three mutants gave fraction-specific SecA crosslinked species (SecA137, SecA366, SecA611), but their rapid electrophoretic mobility was inconsistent with a dimer. The possibility of SecA crosslinked to either the small SecG or SecE proteins of the channel complex was not investigated, although SecA-SecG interaction has been detected previously utilizing this approach (Das and Oliver 2011). We conclude that only a single dominant dimer form (1M6N-like) exists in the three SecA sub-cellular pools examined (albeit substantially reduced in its lipid-bound state).

SecA-dependent protein translocation requires a 1M6N-like dimer interaction

The foregoing analysis begs the question of whether or not a 1M6N-like SecA dimer is required to promote active protein transport within the cell. In order to directly address this question we utilized a SecA-OmpA-GFP trimera that has functional SecA tethered to an artificial substrate that jams protein transport and creates a stable protein translocation intermediate within the SecYEG channel (Fig. 4A). The correct membrane topology of this intermediate was assessed previously by disulfide crosslinking of the C-terminal end of the OmpA signal peptide (Cys21) with the SecY plug domain (Cys68) on the trans side of the membrane (Zheng, Blum et al. 2016). We asked whether this tethered SecA protomer containing a 1M6N signature residue (SecA653) interacts with a free SecA protomer in this arrested state utilizing our in vivo photo-crosslinking strategy. The use of the tightly regulated, arabinose-inducible, OmpA-GFP chimera to effectively jam protein transport and create a stable translocation intermediate to mimic and study the substrate-engaged, active translocation state has been described previously (Park and Rapoport 2012).

Fig. 4. Detection of 1M6N-like SecA dimer at an arrested translocation intermediate by in vivo crosslinking.

Fig. 4

Open in a new tab

A. Scheme depicting the experimental strategy where the OmpA signal peptide (black bar) portion of the SecA-OmpA-GFP trimera traverses the channel which can be verified by cysteine disulfide crosslinking with the SecY plug domain, rapidly-folding GFP jams the channel to create the arrested translocation intermediate with the accumulation of cytosolic pre-MBP (diamonds), and tethered SecA with the 653 signature residue (star) is utilized to detect association with a second SecA protomer at the 1M6N interface by in vivo photo-crosslinking (joining line).

B. Western blot probed with MBP antisera assessing the extent of translocon jamming by accumulation of pre-MBP within the cytosolic/membrane fraction. A time course of jamming is shown where cells were fractionated into total (T), cytosolic/membrane (C), and periplasmic (P) fractions at the indicated times (in minutes). The positions of pre-MBP and MBP are indicated.

C–E. Western blots probed with anti-SecA (C), anti-GFP (D), or anti-c-Myc (E) antibodies showing the association of the SecA-OmpA-GFP trimera with SecY by in vivo disulfide crosslinking 45 min post-induction. Cells treated with copper phenanthroline (CuPh3) or DTT are indicated along with the positions of the trimera crosslinked to SecY (Trimera-Y), the trimera, a trimera degradation product, SecA, SecY, and a presumed SecY dimer (Y-Y).

F–G. Western blots probed with anti-SecA (F), anti-GFP (G), or anti-c-Myc (H) antibodies showing the association of the SecY-crosslinked SecA-OmpA-GFP trimera with a second SecA protomer by a combination of in vivo disulfide and photo-crosslinking 45 min post-induction. Cells treated with UV irradiation, copper phenanthroline (CuPh3) or DTT are indicated along with the positions of the putative doubly crosslinked SecY-trimera-SecA species (Y-Trimera-A), the trimera singly crosslinked to either SecA or SecY that co-migrate (Trimera-A/Y), a presumed SecY dimer (Y-Y), as well as the non-crosslinked species (Trimera, a Trimera degradation product, SecA, and SecY).

In conducting this experiment we monitored the course of translocon jamming after arabinose induction by looking at the accumulation of the normally periplasmic maltose-binding protein within the cytosolic and membrane fraction (Fig. 4B). Cytosolic and membrane pre-MBP could be detected as early as 10 min post induction, and its level began to plateau by 20 min. We also verified the insertion of the OmpA signal peptide of the trimera into the SecY channel by disulfide crosslinking. An appropriately sized, copper phenanthroline-dependent and DTT-reversible trimera-c-Myc-tagged SecY species was readily detected with antibodies to SecA, GFP, or c-Myc (Fig. 4C-E). When cells were subjected to solely photo-crosslinking, a UV-dependent species containing the trimera was detected with a similar mobility, but it clearly did not contain SecY protein and was mostly likely the trimera crosslinked to SecA protein (Fig 4F–H). When both photo and chemical crosslinking were employed, a doubly-crosslinked species was detected consisting of three proteins, the trimera, SecY and most likely SecA protein. In order to verify this latter assignment, we designed two different 1M6N signature cysteine pairs (Cys653, Cys402 or Cys661, Cys12) with one cysteine of the pair located on the SecA portion of the trimera and the other cysteine present on a plasmid copy of SecA (no cysteines were included on either OmpA or SecY in this particular experiment) (Fig. 5A). When cells were subjected to disulfide crosslinking both strains clearly showed the presence of the trimera-SecA crosslinked species (Fig. 5B). Strains that had only one cysteine of the pair gave no crosslinked species as expected. We conclude that SecA covalently bound to an arrested translocation intermediate interacts with an adjoining SecA protomer utilizing the 1M6N interface.

Fig. 5. Verification of 1M6N-like dimer at an arrested translocation intermediate by in vivo disulfide crosslinking.

Fig. 5

Open in a new tab

A. 1M6N dimer structure depicting the two cysteine pairs utilized for in vivo disulfide crosslinking. The Cys12 (red) and Cys661 (blue) pair are shown, along with the Cys402 (orange) and Cys653 (green) pair.

B. Western blots probed with either Anti-GFP or Anti-SecA antibodies as labeled demonstrating the association of the SecA-OmpA-GFP trimera with a second SecA protomer by in vivo disulfide crosslinking 45 min post-induction. Cells treated with copper phenanthroline (CuPh3) or DTT are indicated along with the positions of the trimera crosslinked to another SecA protomer (Trimera-A), as well as the non-crosslinked species (Trimera, a trimera degradation product, and SecA).

In order to strengthen this conclusion we made use of the well-described MalE-LacZ chimera that was originally shown to jam protein transport as well as a free copy of SecA containing a 1M6N signature residue (SecA653) for in vivo photo-crosslinking (Supporting Information Fig. S4A) (Ito, Bassford et al. 1981). Analysis of this alternative system, under maltose regulon control, showed that translocon jamming took longer than its arabinose-regulated counterpart, but it began to plateau by 75 min post-induction as evidenced by the cytosolic and membrane accumulation of pre-MBP (Supporting Information Fig. S4B). Photo-crosslinking analysis revealed that SecA dimer level continued to increase in both the cytosolic and membrane fractions during the course of jamming (Supporting Information Fig. S4C–D). These results are consistent with those presented above, indicating that a 1M6N-like dimer persists during jamming which is presumably associated with the various translocation intermediates that accumulate under these conditions (Ito, Bassford et al. 1981).

Discussion

Numerous studies have contributed to the complex literature attempting to define the active form of SecA, and these have been complicated by the dynamic nature of the SecA monomer-dimer equilibrium, its modulation by translocation ligands, and uncertainty regarding the oligomeric status of active SecYEG protein. Two recent studies have weighed into this matter with important findings. One study utilized dual-color fluorescence-burst analysis under equilibrium conditions to observe SecA interactions with SecYEG-containing liposomes at the single molecule level (Kusters, van den Bogaart et al. 2011). It was found that SecA associated with SecYEG as a dimer, where one SecA protomer bound solely to its SecA counterpart in a salt-dependent manner, while the SecYEG-bound SecA protomer remained salt-resistant. The authors concluded that the observed salt sensitivity of in vitro protein transport indicated that SecA functions at least in part as a dimer in their system. A second study identified two different SecA dimer forms: a major salt-sensitive form equivalent to 1M6N, and a minor salt-resistant form (~5% of total) that appears to utilize an overlapping interface related to 1M6N by a 60° rotation of the two protomers (Gouridis, Karamanou et al. 2013). Through analysis of SecA monomer or dimer-specific mutants, the authors concluded that SecA alternates between these two dimers as well as a monomeric state during the translocation cycle, with the two dimers catalyzing the initial steps in protein transport (formation of the SecA-SecYEG binary and substrate-SecA-SecYEG ternary complexes as well as the initial insertion of the substrate protein into the channel), while the monomer catalyzes the subsequent steps in the process (step-wise membrane traversal of the mature regions of the substrate).

Our study adds to the growing literature supporting a role for SecA dimer in promoting active protein transport. In particular, it clarifies the major (and perhaps only) type of dimer present in vivo, which is maintained during SecA association with SecYEG protein and to a lesser extent with lipid. In that regard the increased SecA intracellular concentrations utilized here would help to populate dimer forms with smaller subunit interfaces (and higher dissociation constants), and yet our data only supported the existence of a 1M6N-like dimer, which has one of the largest subunit interfaces (along with 2IBM). However given the limited number of residues examined in our study, we cannot rule out the existence of a minor alternative dimer state, particularly the one described by Gouridis et al. given its low abundance and the fact that it shares most of the 1M6N interface (Gouridis, Karamanou et al. 2013). Clearly more sensitive methods of structure-specific detection will be required to address the existence of minor dimer forms of SecA.

The preservation of a 1M6N-like dimer in the SecYEG bound state is an important finding of our study and is consistent with the two studies cited above demonstrating that only one of the two SecA protomers within the dimer binds directly to SecYEG. We found that an additional SecA protomer is not accommodated into the T. maritima SecA-SecYEG 3DIN structure utilizing the entire 1M6N interface (Zimmer, Y. et al. 2008). However we were able to dock another T. maritima SecA protomer onto the SecA-SecYEG complex utilizing a significant portion of the 1M6N interface that included suitable adjacencies (<9 Å) for eight out of our nine positive 1M6N signature residues. The only residue that was not accommodated by this simulation was that homologous to E. coli SecA339 that was 18 Å away from its nearest neighbor, but this residue is within a weakly structured loop region whose position is likely to be uncertain under physiological conditions. The resultant dimer in the SecA2-SecYEG ternary complex is similar to the alternative dimer proposed by Gouridis et al. which should be at least marginally stable based on the surface area of its interface (2093 Å2) (Gouridis, Karamanou et al. 2013). In fact weakening of the 1M6N dimer interface in this case may be a deliberate strategy to promote better subunit exchange, which has recently been invoked as one of the sources of SecA-driven protein translocation (see “reciprocating piston model” below). Accordingly, we feel it is unnecessary to posit a completely novel SecYEG-bound SecA state in order to incorporate a 1M6N-like bound dimer into this complex. The SecA-SecY interaction surface in the 3DIN structure was validated previously by data from two in vivo crosslinking studies (Mori and Ito 2006, Das and Oliver 2011), and the only inconsistency noted was the absence of a second SecYEG channel complex interacting with the SecA nucleotide-binding domains that was lost in the original purification of the complex (Zimmer, Y. et al. 2008).

Beyond its structural importance, our study also points to a functional role of the 1M6N-like dimer in protein translocation. While in vitro approaches are now needed to rigorously address which biochemical steps are executed by the dimer and clarify whether a monomer participates in transport as well, our experiment with the SecA-OmpA-GFP trimera suggests that a dimer is involved in transport at a point when the signal peptide has already inserted into the channel. The earliest steps in protein transport appear to utilize SecA dimer, since efficient SecB-dependent delivery of substrates to SecA and subsequent activation steps have been shown to require the dimer. Extensive mapping of the interaction surfaces of this ternary complex and the mode of substrate transfer to SecA are most consistent with a 1M6N-like dimer (see references in (Suo, Simon et al. 2015)). By contrast, it is more difficult to accommodate a SecYEG-bound SecA dimer during ongoing substrate translocation, since these steps must allow for a variety of dynamic SecA domain movements which include: (i) movements of the two nucleotide-binding motor domains of SecA that drive its overall conformational cycle and protein transport, (ii) the mobility of the two-helix finger region of SecA that appears to push substrate proteins into the channel in a piston-like fashion (for an alternative view, see (Whitehouse, Gold et al. 2012)), (iii) the dynamic action of the SecA substrate-binding clamp that opens and closes to position successive segments of the substrate protein above the channel, and (iv) some sort of SecA membrane insertion that is required for channel activation and protein transport (see references in (Kusters and Driessen 2011)). A SecYEG-bound SecA monomer would better afford the necessary domain flexibility required to execute the later stages of protein transport.

Taken as a whole our study supports more recent thinking in the field that suggests that SecA alternates between a dimer and monomer as part of the protein translocation cycle (referred to as the reciprocating piston model in (Kusters and Driessen 2011)). In essence a SecYEG-bound SecA dimer is involved in the initial steps of substrate protein binding and channel insertion of the signal peptide and early mature region of the substrate. ATP hydrolysis then promotes release of the peripheral SecA subunit at this stage of the process. Re-binding of another SecA protomer to the substrate-containing SecYEG-bound SecA monomer promotes translocation of an additional polypeptide segment similar to the “active rolling” model proposed for ATP-dependent helicases that are homologous to the nucleotide-binding domains of SecA (Lohman, Thorn et al. 1998). In this and related models, both SecB and SecY interactions with SecA exploit the 1M6N dimer interface to transfer the substrate to SecA or induce its monomerization, respectively. Clearly our study now sets the stage to directly test these and related ideas utilizing more direct biochemical and biophysical approaches at both the ensemble as well as single molecule level.

Experimental procedures

Media, chemicals, strains and plasmids

LB (Miller) broth, IPTG and arabinose were purchased from Fisher Scientific. pBpA and maltose were purchased from Bachem and Difco, respectively. Quik-Change™ and Wizard Plus SV Miniprep DNA Purification System kits were obtained from Stratagene and Promega, respectively. The WesternBright™ Sirius enhanced chemi-luminescence kit was obtained from Advansta. Protease Inhibitor Cocktail was obtained from Sigma Aldrich. Most other common laboratory chemicals were obtained from the latter supplier or Fisher Scientific and were laboratory grade or better. Mouse anti-c-Myc monoclonal antibody and HRP-conjugated goat anti-mouse antibody were obtained from Genescript, while chicken anti-GFP antibody and HRP-conjugated goat anti-rabbit antibody were obtained from Abcam. HRP-conjugated goat anti-chicken antibody was obtained from Jackson ImmunoResearch Laboratories. Rabbit antisera to purified E. coli SecA protein, OmpA or MBP was prepared by Cocalico Biologicals (Reamstown, PA). Peptide affinity-purified SecY antisera was prepared by hyper-immunizing rabbits to a peptide identical to the carboxyl terminus of SecY (CYESALKKANLKGYGR) conjugated to keyhole limpet hemocyanin by maleimide chemistry utilizing Tana Laboratories, LC (Houstin, TX) for peptide synthesis, protein carrier conjugation, immunization, and peptide affinity purification of the antisera. E. coli BLR(λDE3) [F−ompT hsdS (rB−mB−) gal dcm Δ(srl-recA)306::Tn10 (TetR)] was obtained from Stratagene. BL21.1 is a secA51(Ts) leu::Tn10 derivative of BL21(λDE3). MM18.7 is a recA derivative of MM18 containing the malE-lacZ72-47(Hyb) fusion that has been described previously (Ito, Bassford et al. 1981). MC4100.2(λDE3) is a recA1 srl::Tn10 derivative of MC4100 that was lysogenized with λDE3, which has been previously described (Studier, Rosenberg et al. 1990). The pSup-pBpARS-6TRN plasmid encoding the Methanococcus jannaschii amber suppressor tRNA-tRNA synthetase pair that efficiently incorporates pBpA in place of an amber codon has been described previously (Wang, Xie et al. 2006). The pT7secA-his plasmid containing full-length secA with a carboxyl-terminal hexa-histidine tag under control of the T7 promoter has been reported (Jilaveanu, Zito et al. 2005), along with pBE2, which contains the secM secA operon cloned between the EcoRV and BamHI sites of pACYC184. The pCDFT7secYEG plasmid with secYEG under control of the T7 promoter with an amino-terminal c-Myc tag on secY has been described (Das and Oliver 2011). The pBAD-secA-OmpA-GFP plasmid containing full-length SecA fused at its carboxyl-terminus to the amino-terminus of OmpA-GFP has been reported previously (Zheng, Blum et al. 2016). All substitutions to amber or cysteine codons within plasmid-borne secA, secY or ompA genes were made by QuikChange™ mutagenesis and verified by DNA sequence analysis utilizing the University of Pennsylvania DNA Sequencing Facility.

In vivo photo and chemical crosslinking and cell fractionation

In general a freshly struck out single colony of the strain utilized for photo or chemical crosslinking was inoculated into LB media supplemented with appropriate antibiotics (25 μg/ml chloramphenicol, 50 μg/ml streptomycin, and 100 μg/ml ampicillin as needed) and grown overnight at 37°C with shaking at 250 rpm. The overnight culture was diluted 1:50 into LB media supplemented with appropriate antibiotics and 1 mM pBpA (when photo-crosslinking was needed), and SecA, SecYEG, the SecA-OmpA-GFP trimera, or the MalE-LacZ chimera were induced as follows: (i) for analysis of the SecA signature mutants, BLR(λDE3) containing pSup-pBpARS-6TRN and pT7SecA-his plasmids with the indicated signature residue was used, and induction with 0.5 mM IPTG started at an OD600 of 0.4 to 0.6 and continued for an addition 1 h, (ii) for the translocon jamming experiments involving the SecA-OmpA-GFP trimera, MC4100.2(λDE3) containing pSup-pBpARS-6TRN, pCDFT7secYEG, and pBAD-secA-OmpA-GFP plasmids (Fig 4) or pBE2 and pBAD-secA-OmpA-GFP plasmids (Fig 5) with the indicated SecA signature or SecA, SecY or OmpA cysteine residue(s) was used, 30 μM IPTG and 0.2% maltose were included in the grown media, when necessary, and the trimera was induced with 0.2% arabinose at OD600 of 0.15 for 45 min or for the times indicated, and (iii) for the translocon jamming experiments involving the MalE-LacZ chimera, MM18.7 containing the pSup-pBpARS-6TRN and pT7secA-his plasmids with the indicated SecA signature residue was used, 30 μm IPTG was included in the growth media, and the MalE-LacZ chimera was induced with 0.2% maltose for the times indicated. All subsequent steps were done at 4°C or on ice. To adjust for somewhat different cell densities, 6 OD600 cell equivalent of each culture was harvested by sedimentation in a microfuge at 14,000 rpm for 5 min, washed with 5 ml PBS (10 mM sodium phosphate, pH 7.5, 140 mM NaCl), and resuspended in 6 ml PBS buffer. 2 ml samples were UV irradiated on ice at 365 nm for 10 or 20 min using a Rayonet 2000 UV crosslinker (Southern New England Ultraviolet Company), while 2 ml non-irradiated samples served as negative controls. The pattern of in vivo photo-crosslinking of each SecA signature mutant was compared to that of the same purified SecA mutant protein irradiated at 1 μM concentration in 25 mM Tris-HCl, pH 7.5, 25 mM KCl, 0.5 mM EDTA, 0.5 mM PMSF. SecA protein was purified utilizing a His-Bind resin column (Novagen) according to the manufacturer’s protocol. After irradiation, each cell sample was sedimented, resuspended in 1 ml breakage buffer (10 mM Tris-HCl, pH 7.5, 1 mM EDTA, 0.25 mM PMSF, 1 mM DTT, 100 μg/ml RNase, 100 μg/ml DNase, 60 μg/ml lysozyme, 1X Protease inhibitor cocktail), and cells were placed in a polycarbonate tube and disrupted with 30 sec bursts of a cup horn sonicator (Heat Systems) until near clarity. Unbroken cells were removed by sedimentation in a microfuge at 14,000 rpm for 5 min. The clarified supernatant (total cell fraction) was isolated and saved as needed. The clarified supernatant was further processed where necessary by sedimentation in a Sorvall S120 AT2 rotor at 82,000 rpm for 30 min at 4°C, and this supernatant (cytosolic fraction) was saved, while the membrane pellet (total membrane fraction) was solubilized in 60 μl ABB buffer (5% SDS, 10 mM Tris-Cl, pH 8, 1 mM EDTA) with constant stirring for 1 h at 37°C, when 20 μl 4X sample buffer (8% SDS, 500 mM Tris-HCl, pH 6.8, 20% 2-mercaptoethanol, 60% glycerol, 0.02% bromophenol blue) was added, and stirring continued for an additional 10 min. When additional fractionation was needed, each membrane pellet was resuspended in 150 microliters of 6 M urea in 10 mM Tris-HCl, pH 7.5 on ice for 30 min, and then re-sedimented as above. The supernatant (peripheral membrane fraction) was saved, while the membrane pellet (integral membrane fraction) was solubilized similarly to the total membrane fraction. Generally 15 μl samples were loaded onto a 10% SDS-PAGE gel, which was run at 100 V at 4°C until the dye front reached the bottom. Western transfers were performed at 100 V for 1 h, and nitrocellulose membranes were blocked overnight with 10 ml TBS buffer (20 mM Tris-HCl, pH 7.5, 140 mM NaCl, 0.25% Tween 20) supplemented with 10% non-fat dry milk. The membrane was probed with appropriate primary and secondary antibodies as indicated in the figure legends. All primary and secondary antibodies were used at a 1:5,000 dilution with the exception of SecA and GFP antibodies that were used at a 1:10,000 dilution. Proteins were visualized using a WesternBright™ Sirius kit as described by the manufacturer.

For chemical crosslinking, strains were grown identically as described above. 3 OD600 cell equivalent was harvested, resuspended in 2 ml protoplast buffer (100 mM sodium phosphate, pH 7.5, 5 mM EDTA, 10 mM phenanthroline), and treated with copper phenanthroline (180 mM phenanthroline, 60 mM CuSO4, 50 mM NaH2PO4) to a final concentration of 300 μM for 10 min at 30°C as indicated (Bunn and Ordal 2003). The reaction was quenched by addition of n-ethylmaleimide and EDTA to final concentrations of 1 mg/ml and 5 μM, respectively, for 10 min. For treatment with reducing agent, DTT was added to a final concentration of 60 mM for 10 min. Either the total cell or membrane fraction was isolated and subsequently analyzed as described above except that the final sample was solubilized in 40 μl ABB buffer and an equal volume of 2X non-reducing sample buffer (40 mM Tris– HCl, pH 7.8, 16 mM NaH2PO4, 2% (w/v) SDS, 50 μg/ml bromophenol blue, 20 mM EDTA, 2 mg/ml N-Ethylmaleimide, and 0.10 g/ml sucrose).

Assessment of translocon jamming

The extent of translocon jamming was assessed by MBP fractionation. For each time point analyzed, 0.6 OD600 cell equivalent was harvested by sedimentation in the microfuge at 14,000 rpm for 5 min at 4°C, and resuspended in 200 μl ice cold 20% sucrose-0.03 M Tris-HCl, pH 8. Half the sample was saved as the total cell control, while the other half was spheroplasted by addition of 10 μl of 1 mg/ml lysozyme in 0.1M EDTA, pH 8 for 15 min on ice. Treated cells were sedimented in the microfuge at 5,500 rpm for 5 min at 4°C, and the supernatant (periplasmic fraction) was removed, while the cell pellet (cytoplasm/membrane fraction) was resuspended in 100 μl of buffer A (10 mM Tris-HCl, pH 7.5, 150 mM NaCl, 5 mM MgSO4). All samples were boiled for 10 min in the presence of sample buffer, and 5 μl aliquots were analyzed by SDS-PAGE and western blotting as described above.

Supplementary Material

Supp data

Acknowledgments

The authors thank Stephanie Day for assistance with the early portion of the project, and Tom Rapoport (Harvard Medical School) for suggesting the use of the SecA-OmpA-GFP trimera. This work was supported by a grant from the National Institutes of Health (GM110552 to D.O.)

Footnotes

Conflict of interest

The authors declare that they have no conflict of interest.

Supporting information

Additional supporting information may be found in the online version of this article at the publisher’s web-site.

References

  1. Allen W, Corey R, Oatley P, Sessions R, Baldwin S, Redford S, Tuma R, Collinson I. Two-way communication between SecY and SecA suggest a Brownian ratchet mechanism for protein translocation. Elife. 2016;5:15598. doi: 10.7554/eLife.15598. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Auclair S, Oliver D, Mukerji I. Defining the solution sate dimer structure of Escherichia coli SecA using Forster resonance energy transfer. Biochemistry. 2013;52:2388–2401. doi: 10.1021/bi301217t. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bauer BW, Shemesh T, Chen Y, Rapoport TA. A “Push and Slide” mechanism allow sequence-insensitive translocation of secretory proteins by the SecA ATPase. Cell. 2014;157:1416–1429. doi: 10.1016/j.cell.2014.03.063. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Benach J, Chou YT, Fak JJ, Itkin A, Nicolae DD, Smith PC, Wittrock G, Floyd DL, Golsaz CM, Gierasch LM, Hunt JF. Phospholipid-induced monomerization and signal-peptide-induced oligomerization of SecA. J Biol Chem. 2003;278:3628–3638. doi: 10.1074/jbc.M205992200. [DOI] [PubMed] [Google Scholar]
  5. Bunn M, Ordal G. Transmembrane organization of the Bacillus subtilis chemorecptor McpB deduced by cysteine disulfide crosslinking. J Mol Biol. 2003;331:941–949. doi: 10.1016/s0022-2836(03)00834-9. [DOI] [PubMed] [Google Scholar]
  6. Cabelli RJ, Dolan KM, Qian L, Oliver DB. Characterization of membrane-associated and soluble states of SecA protein from wild-type and secA51(Ts) mutant strains of Escherichia coli. J Biol Chem. 1991;266:24420–24427. [PubMed] [Google Scholar]
  7. Das S, Oliver D. Mapping of the SecA-SecY and SecA-SecG interfaces by site-directed in vivo photocrosslinking. J Biol Chem. 2011;286:12371–12380. doi: 10.1074/jbc.M110.182931. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Das S, Stivison E, Folta-Stogniew E, Oliver D. Reexamination of the role of the amino terminus of SecA in promoting its dimerization and functional state. J Bacteriol. 2008;190:7302–7307. doi: 10.1128/JB.00593-08. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Dorman G, Prestwich GD. Benzophenone photophores in biochemistry. Biochemistry. 1994;33:5661–5673. doi: 10.1021/bi00185a001. [DOI] [PubMed] [Google Scholar]
  10. Eichler J, Rinard K, Wickner W. Endogenous SecA catalyzes preprotein translocation at SecYEG. J Biol Chem. 1998;273:21675–21681. doi: 10.1074/jbc.273.34.21675. [DOI] [PubMed] [Google Scholar]
  11. Gouridis G, Karamanou S, Sardis M, SM, Capitani G, Economou A. Quaternary dynamics of the SecA motor drive translocase catalysis. Mol Cell. 2013;52:655–666. doi: 10.1016/j.molcel.2013.10.036. [DOI] [PubMed] [Google Scholar]
  12. Hunt JF, Weinkauf S, Henry L, Fak JJ, McNicholas P, Oliver DB, Deisenhofer J. Nucleotide control of interdomain interactions in the conformational reaction cycle of SecA. Science. 2002;297:2018–2026. doi: 10.1126/science.1074424. [DOI] [PubMed] [Google Scholar]
  13. Ito K, Bassford P, Beckwith J. Protein localization in E. coli: is there a common step in the secretion of periplasmic and outer-membrane proteins? Cell. 1981;24:707–717. doi: 10.1016/0092-8674(81)90097-0. [DOI] [PubMed] [Google Scholar]
  14. Jilaveanu LB, Zito CR, Oliver D. Dimeric SecA is essential for protein translocation. Proc Natl Acad Sci U S A. 2005;102:7511–7516. doi: 10.1073/pnas.0502774102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Kusters I, Driessen A. SecA: a remarkable nanomachine. Cell Mol Life Sci. 2011;68:2053–2066. doi: 10.1007/s00018-011-0681-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Kusters I, van den Bogaart G, Kedrov A, Krasnikov V, Fulyani F, Poolman B, Driessen A. Quaternary structure of SecA in solution and bound to SecYEG probed at the single molecule level. Structure. 2011;19:430–439. doi: 10.1016/j.str.2010.12.016. [DOI] [PubMed] [Google Scholar]
  17. Lohman T, Thorn K, Vale R. Staying on track: common features of DNA helicases and microtubule motors. Cell. 1998;93:9–12. doi: 10.1016/s0092-8674(00)81139-3. [DOI] [PubMed] [Google Scholar]
  18. Lycklama a Nijeholt J, Driessen A. The bacterial Sec-translocase: structure and mechanism. Philos Trans R Soc Lond B Biol Sci. 2012;367:1016–1028. doi: 10.1098/rstb.2011.0201. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Mori H, Ito K. Different modes of SecY-SecA interactions revealed by site-directed in vivo photo-crosslinking. Proc Natl Acad Sci U S A. 2006;103:16159–16164. doi: 10.1073/pnas.0606390103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Or E, Navon A, Rapoport T. Dissociation of the dimeric SecA ATPase during protein translocation across the bacterial membrane. EMBO J. 2002;21:4470–4479. doi: 10.1093/emboj/cdf471. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Papanikolau Y, Papadovasilaki M, Ravelli R, McCarthy A, Cusack S, Economou A, Petratos K. Structure of dimeric SecA, the Escherichia coli preprotein translocase motor. J Mol Biol. 2007;366:1545–1557. doi: 10.1016/j.jmb.2006.12.049. [DOI] [PubMed] [Google Scholar]
  22. Park E, Rapoport TA. Bacterial protein translocation requires only one copy of SecY complex in vivo. J Cell Biol. 2012;198:881–893. doi: 10.1083/jcb.201205140. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Park E, Rapoport TA. Mechanisms of Sec61/SecY-mediated protein translocation across membranes. Annu Rev Biophys. 2012;41:21–40. doi: 10.1146/annurev-biophys-050511-102312. [DOI] [PubMed] [Google Scholar]
  24. Ryu Y, Schultz P. Efficient incorporation of unnatural amino acids into proteins in Esherichia coli. Nature Methods. 2006;3:263–265. doi: 10.1038/nmeth864. [DOI] [PubMed] [Google Scholar]
  25. Sardis M, Economou A. SecA: a tale of two protomers. Mol Microbiol. 2010;76:1070–1081. doi: 10.1111/j.1365-2958.2010.07176.x. [DOI] [PubMed] [Google Scholar]
  26. Sharma V, Arockiasamy A, Ronning DR, Savva CG, Holzenburg A, Braunstein M, Jacobs WR, Sacchettini JC. Crystal structure of Mycobacterium tuberculosis SecA, a preprotein tranlsocating ATPase. Proc Natl Acad Sci USA. 2003;100:2243–2248. doi: 10.1073/pnas.0538077100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Studier WF, Rosenberg AH, Dunn JJ, Dubendorff JW. Use of T7 RNA polymerase to direct expression of cloned genes. Methods Enzymol. 1990;185:60–89. doi: 10.1016/0076-6879(90)85008-c. [DOI] [PubMed] [Google Scholar]
  28. Suo Y, Simon J, Randall L. The basis of asymmetry in the SecA:SecB complex. J Mol Biol. 2015;427:887–900. doi: 10.1016/j.jmb.2014.12.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Vassylyev D, Mori H, Nassylyeva M, Tsukazaki T, Kimura Y, Tahirov T, Ito K. Crystal structure of the translocation ATPase SecA from Thermus thermophilus reveals a parallel, head-to-head dimer. J Mol Biol. 2006;364:248–258. doi: 10.1016/j.jmb.2006.09.061. [DOI] [PubMed] [Google Scholar]
  30. Wang L, Xie J, Schultz P. Expanding the genetic code. Annu Rev Biophys Biomol Struct. 2006;35:225–249. doi: 10.1146/annurev.biophys.35.101105.121507. [DOI] [PubMed] [Google Scholar]
  31. Whitehouse S, Gold V, Robson A, Allen W, Sessions R, Collinson I. Mobility of the SecA 2-helix-finger is not essential for polypeptide translocation via the SecYEG complex. J Cell Biol. 2012;199:919–926. doi: 10.1083/jcb.201205191. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Woodbury RL, Hardy S, Randall L. Complex behavior in solution of homodimeric SecA. Protein Science. 2002;11:875–882. doi: 10.1110/ps.4090102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Yu D, Wowor AJ, Cole JL, Kendall D. Defining the Escherichia coli SecA dimer interface residues through in vivo site-specific photo-cross-linking. J Bacteriol. 2013;195:2817–2825. doi: 10.1128/JB.02269-12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Zheng Z, Blum A, Banerjee T, Wang Q, Dantis V, Oliver D. Determination of the oligomeric state of SecYEG protein secretion channel complex using in vivo photo- and disulfide cross-linking. J Biol Chem. 2016;291:5997–6010. doi: 10.1074/jbc.M115.694844. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Zimmer J, Li W, Rapoport TA. A novel dimer interface and conformational changes revealed by an X-ray structure of B. subtilis SecA. J Mol Biol. 2006;364:259–265. doi: 10.1016/j.jmb.2006.08.044. [DOI] [PubMed] [Google Scholar]
  36. Zimmer J, NY, Rapoport TA. Structure of a complex of the ATPase SecA and the protein-translocation channel. Nature. 2008;455:936–943. doi: 10.1038/nature07335. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supp data
NIHMS830607-supplement-Supp_data.pdf (4.2MB, pdf)

RESOURCES