Quaternary Dynamics of the SecA Motor Drive Motor Drive Translocase Catalysis ... area), chains...

Molecular Cell, Volume 52 Supplemental Information

Quaternary Dynamics of the SecA

Motor Drive Translocase Catalysis Giorgos Gouridis, Spyridoula Karamanou, Marios Frantzeskos Sardis, Martin Alexander Schärer, Guido Capitani, and Anastassios Economou

Gouridis et al SecA quaternary dynamics-Supplemental Text and Figures

Supplemental Text and Figures

Fig. S1 Structural analysis of SecA dimers, related to Figures 1 and 2.

A. All the available crystal contacts from dimeric SecA structures with resolution

better than 3Å and Rfree ≤ 30% were re-examined using EPPIC (Duarte et al., 2010).

Runs were carried out in September 2012 using Uniprot 2012_07. Summary view of

EPPIC interface analysis

for entries 1NL3_1

(interface 1 in the lattice)

(M. tuberculosis)

(Sharma et al., 2003)

and 1M6N (B. subtilis)

(Hunt et al., 2002).

Columns from left to right

correspond to: PDB entry

code analyzed and

source organism thereof,

thumbnail of the

interface,

interface number in the

crystal lattice (sorting by

area), chains involved in

the interface,

crystallographic operator

generating the interface


(operator X,Y,Z indicates that the interface is contained in the asymmetric unit). The

four last columns on the right display the interface classification results (xtal, bio or

nopred) according to the three criteria used by EPPIC (geometry, core-rim entropy

and core surface entropy) and the final call obtained by consensus from those three

criteria.

In the 1NL3_1 structure, a 15-residue extension of the aminoterminus, that replaced

α0 during cloning, contributes to the interface. In the 1M6N structure (Ding et al.,

2003; Hunt et al., 2002) part of α0 and the preceding N-terminus of one protomer

participate in crystal contacts with the other protomer, in a fashion similar to that of

the 1NL3_1 interface. Superimposing protomers A of the two dimers, reveals that

residues Ile4 in 1M6N and Leu11 of the engineered 1NL3_1 extension occupy the

same spatial position and fulfill the same role in the dimerization interface. We

presume that the 1NL3_1 engineered extension partly mimics the interaction seen

around Ile4 in the 1M6N structure (corresponding to Leu5 in the ecSecA sequence

and Leu2 in the mtSecA sequence).

B. Testing the E.coli SecA intercalated dimer by engineered cysteines. Left: In the

dimeric ecSecA (Papanikolau et al., 2007) one protomer is colored (NBD=dark blue;

IRA2=light blue) and the other is depicted with a transparent grey color. Residues

mutated to cysteines are indicated with red spheres in the colored protomer. Right:

The colored protomer has undergone a ~90º rotation with respect to the y-axis. Mass

measurements of the oxidized proteins are presented in Table S2.

C. Comparison of the aminoterminal regions of SecA and other helicase motors. The

N-terminal RecA fold domain of the Escherichia coli SecA ATPase motor (PDB:

2FSF; residues 9-419) was structurally aligned with the corresponding domain from


the helicases Vasa (PDB: 2DB3) (Sengoku and Wagatsuma, 2006) and eF4A (PDB:

1FUU) (Caruthers et al., 2000). Structures are shown in the same orientation to

highlight differences. The three N-terminal SecA helices (ribbon representation;

yellow) differ significantly from the corresponding regions of other Superfamily 2

(SF2) helicases. The linker connecting to the second RecA fold domain (i.e. IRA2 in

SecA) is shown in light blue. Helix α0 of ecSecA, not present in the crystallized

protein, (Papanikolau et al., 2007) is shown schematically. Its location was identified

after structural modeling using PDB: 1NL3_1 [see Fig. S2; (Sharma et al., 2003)] and

a combination of secondary structure prediction and CLUSTALW alignment on the

PBIL server (http://npsa-pbil.ibcp.fr/cgi-

bin/npsa_automat.pl?page=/NPSA/npsa_clustalw.html).

D. Sequence alignment and secondary structure features of the SecA N-terminal

region. A schematic representation of the SecA amino-termini is shown as

determined from various crystal structures; E.coli (PDB: 2FSF) (Papanikolau et al.,

2007); T.thermophilus (PDB: 2IPC) (Vassylyev et al., 2006); M.tuberculosis, (PDB:

1NL3; SecA1) (Sharma et al., 2003); B.subtilis (SecA_BSUB_1; PDB: 1M6N) (Ding et

al., 2003), (SecA_BSUB_2; PDB: 1TF5) (Osborne et al., 2004) and (SecA_BSUB_3;

PDB: 2BIM) (Zimmer et al., 2006) and T.maritima (PDB: 3DIN) (Zimmer et al., 2008).

Dotted lines indicate unresolved parts of the structure.

Secondary structure prediction algorithms propose helix α0 to exist in most

SecA sequences, while it is absent from others as seen in the alignment below:

Symbols used: c= coil, h= α-helix, e= β-strand:


10 20 30 40 50 60

| | | | | |

A3ERJ8_9BACT --MLSGLFSSIFPSRNDRELKRISRIIEHINRLEEEIRDLEDESLTGKTREFRERLSKGETL-------

Sec.Cons. --cccceececccccchhhhhhhhhhhhhhhhhhhhhhhhhhhhhhcchhhhhhhhhccccc-------

SECA_NEIG1 MLTNIAKKIFGSRNDRLLKQYRKSVARINALEEQMQALSDADLQAKTAEFKQRLADGQ----TL-----

Sec.Cons. chhhhhhhhcccchhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhccc----cc-----

SECA_PROMA ----MLKLLLGDPNARKLKRYQPILTDINLFEDEIASLNDDELRGKTSDFRTRLDKSSDS--SIQE---

Sec.Cons. ----c?eeeccccchhhhhh?cchhhhhhhhhhhhhhhchhhhccccchhhehhcccccc--cc?c---

Q14PT9_SPICI ---------MAVSDRKIVKKHGKIADKIMALDKTMQALSDDALKTKTNEFKAKLAEGVSLNDILIEAFA

Sec.Cons. ---------ccchhhhhhhhh?hhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhccchhhhhhhhcc

SECA_STAAB --MGFLSKILD-GNNKEIKQLGKLADKVIALEEKTAILTDEEIRNKTKQFQTELADIDN---VKKQ---

Sec.Cons. --cc???hhhc-cchhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh---hccc---

SECA_CHLTR ---MMDFLKRFFGSSQERILKRFQKLVEEVNACDEKFSSLSDDELRKKTPQLKQRYQDG-ESLD-

Sec.Cons. ---chhhhhhh?ccchhhhhhhhhhhhhhhhhhhhhh??cchhhhhhh?hhhhhhh?cc-cccc-

SECA_FLAPJ MSFINNILKVFVGDKSQKDVKAIQPIIAKIRTLENSLSNLSHDELRAKTVYFKDIIKQA-R----

Sec.Cons. cchhhhheeeeecccchhhhhhhhhhhhhhhhhhh?h?h??hhhhhhhhhhhhhhhhhc-c----

SECA_BORBU --MLKAVLETTIGSKSKRDLKDYLPTLRNINKLERWALLLADEDFSKETEKLKDELKSG-NSL--

Sec.Cons. --c?hhhhhhhccccc?hhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhcc-ccc--

SECA_CHLTE ---MLKIIAKIFGSKHEKDIKKIQPIVDRINEIYGTLNALPDEAFRNKGVELRKKVRDK-LIPF-

Sec.Cons. ---chhhhhhhhcccchhhhhhh?hhhhhhhhhhhchc?cchhhhhhhhhhhhhhhhhh-?ccc-

SECA_DEIRA ---MFRVLNKVFDNNKRDVERIIQTVVKPVNALEEETMRVEN--LAEAFMDLRRRVQDGGESLDS

Sec.Cons. ---chhhhhhh?cccchhhhhhhhhheccchhhhhhhhhhhh--hhhhhhhhhhhhhcccccccc

A0GXK1_9CHLR --MLN-FFRRLLGDSNEKEIRRLQPIVEEINRLGPEFARLSDAELRAKTDEFRQRLADGETLD------

Sec.Cons. --chh-hhhhh?cccchhhhhhhhhhhhhhh?cc?hhhhhhhhhhhhhhhhhhhhhhcccccc------

SECA_AQUAE --MLGWIAKKIIGTKNEREVKRLRKFVNQINELEKELDALTNKELVELAQELHDKIRFDEEL-------

Sec.Cons. --chhhhhhhh?cccchhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhccc-------

B9XPF8_9BACT --MIGFIVKKFIGSRNDREVKKLRPTAVKINELELELQKLPDDALRQKTAEWKARFAKIEDK-------

Sec.Cons. --c?hh??eee?cccc?hhhhhhccchhhhhhhhhhhhhcchhhhhhhhhhhhhhhhhhhcc-------

SECA_CYTH3 --MLG-ILAKLFGTKSGRDIKKLQPLVERINEEFQKLHALDDNQLRAQTDKIKGIIDADLSGI------

Sec.Cons. --chh-hhhhhhcccccchhhhhhhhhhhhhhhhhhhhh??hhhhhhhhhhhhhhhhhccccc------

SECA_THEP1 ---------MILFDKNKRILKKYAKMVSKINQIESDLRSKKNSELIRLSMVLKEKVNSFEDADEHLFEA

Sec.Cons. ---------ccehhhhhhhhhhhhhhhhhhhhhhhhhhhhcchhhhhhhhhhhhhhhhhhhhhhhhhcc

A6CEP4_9PLAN MEFLD-KLGEWLTTVTAWLERFLTGLFGSSNERQIRKLGFVRDK-EGHDQIVPGSMLAEIDS-------

Sec.Cons. cchhh-hhhhhhhhhhhhhhhhhhccccccchhhhhhh?eehcc-ccccceccccceeehcc-------

Secondary structure predictions were carried out using CLUSTALW on the PBIL

server. Ten algorithms were used to derive the consensus sequence [DSC (King and

Sternberg, 1996), DPM (Deleage and Roux, 1987), GOR I (Garnier et al., 1978),

GOR III (Gibrat et al., 1987), GOR IV (Garnier et al., 1996), HNN (Guermeur, 1997),

SIMPA96 (Levin et al., 1996), PHD (Rost et al., 1994), PREDATOR (Argos et al.,

1996), SOPM (Geourjon and Deleage, 1994)]. SecA sequences (provided with their


ExPASy codes) were obtained from the following organisms chosen so as to span all

the bacterial phylla: [Neisseria gonorrhoeae Q5F807; Spiroplasma Q14PT9;

Staphylococcus aureus Q2YSH6; Prochlorococcus marinus Q7V9M9; Chlamydia

trachomatis O84707 ; Flavobacterium psychrophilum A6GX63; Chlorobaculum

tepidum Q8KK18; Borrelia burgdorferi O07497; Deinococcus radiodurans Q9RWU0;

Thermotoga petrophila A5IM ; Planctomyces maris A6CEP4; Cytophaga hutchinsonii

Q11YU5; Aquifex aeolicus O67718; Leptospirillum ferriphilum A3ERJ8; Rickettsia

akari A8GP42; bacterium Ellin514 B9XPF8.

E. Schematic maps of SecA derivatives used in this study with either truncation of the

N-terminal decapentapeptide and/or 6 alanyl point substitutions in α1 (indicated with

“6A”.). Sequence alignment of SecAs reveals conservation at the level of the

physicochemical properties of the N-terminal residues (CLUSTALW alignment server,

PBIL France). Light blue color indicates >70% conservation, while grey color

indicates >50% conservation. Data were derived from a comparison of 50 SecA

sequences. SecA(GH) has a Gly-His N-terminal extension before the 1-901 SecA

sequence, as a result of cleaving the purified His6-linker-TEV site-SecA protein with

the TEV (Tobacco Etch Virus) protease. Red highlights the region that is found to

participate in dimerization contacts in the dimeric structures (Fig. 1C).


Fig. S2 SecA oligomerization is dependent on the N-terminal region, related to

Figures 2 and 3D.

A. Monomeric SecA(Δα0/α1-6A) was tested for dimerization at higher concentrations

than allowed by our analytical GPC-MALLS system (upper limit of ~8µM at the

chromatographic peak) by using preparative GPC (50mM Tris pH:8.0, 50mM KCl;

40C) on a Superdex Hi-Load 26/60 prepacked column (GE; in which ~10fold more

protein can be loaded). UV traces of these experiments were normalized to the main

monomer peak and are shown superimposed. The presence of a minor dimer


population that increases in amount as the total protein concentration rises is

apparent at ~150ml elution.

B. The monomeric and dimeric populations of SecA(Δα0/α1-6A) were quantified for

each protein concentration shown in Panel A, by measuring the areas under the

curve, using MATLAB (The Mathworks Inc.). The sum of monomer and dimer, for the

same protein concentration was considered 100%. The percentage of dimer (y axis)

was plotted as a function of protein concentration (x axis; measured at the main peak

of the chromatogram). Clearly, the concentrations tested are well below the

saturation concentration for dimer formation. However, the observed linear correlation

(R2= 0.98) allows for an approximate calculation of the equilibrium dissociation

constant (KD), by extrapolation to the concentration at which 50% of SecA(Δα0/α1-

6A) is expected to be dimeric (133µM; Table S2).

C. GPC-MALLS analysis of SecA derivatives (2.5µM at the chromatographic peak) at

various ionic strength regimes. Mass values (y axis), plotted as a function of salt

concentration (x axis), follow a sigmoidal decay curve. In the case of SecA(GH)

(green line) the lower plateau of the sigmoidal curve intersects the y axis at a value

significantly higher than ~105 kDa (which corresponds to the value of the SecA

monomer measured by GPC-MALLS), indicating that a population of molecules is

insensitive to salt (presumably because it represents the hydrophobic dimeric

conformer; SRD).

D. Preparative GPC of the indicated SecA derivatives, at high ionic strength (50mM

Tris-HCl pH:8.0, 1M KCl; 20µM at the peak) on a HiLoad 26/60 Superdex 200 (GE),

at 40C. Monomeric (M) and dimeric/hydrophobic (SRD) SecA populations, are

indicated. Similar experiments were performed at a range of protein concentrations


(5-35µM at the peak). The percentage of monomer and dimer SecA, for each protein

concentration, was estimated by measuring the areas under the curve (MATLAB).

This way, the oligomeric behavior of high protein concentration of SecA and

derivatives was evaluated at high ionic strength (1M KCl). Similar experiments led to

the quantification presented in Fig. 2C.

E. GPC of SecA(GH) in 50mM Tris pH:8.0, 1M KCl at the indicated protein

concentrations (next to the chromatographic peak; loading concentrations are ∼ 10x

higher). From these chromatograms, the percentage of the SRD (y axis), calculated

as a function of protein concentration at the peak (x axis), permitted determination of

the equilibrium dissociation constant of the SRD (KD = 10µM), presented in Table S2.


Fig. S3 Purification and functional analysis of a cross-linked SecA dimer

through oxidation of Cys98, related to Figure 3.

For the experiments below we used SecA(1-834) that carries the indigenous Cys98

(next to helix α1) as the only cysteinyl residue, following removal of the C-terminal

tail. This was to avoid interference from 3 cysteines present in the C-tail.

A. SecA(1-834) was incubated

in the absence or, presence of

1mM CuCl2 (as indicated), on

ice, for 10min. Proteins were

analyzed in a non-reducing

7.5% SDS-PAGE, in the

presence or, absence of 1mM

DTT (indicated) and visualized

with Coomassie blue staining.

Low amounts of pre-existing

dimer (lane 3) become more

than 60% of the total protein

following incubation with CuCl2

(lane 4). No other cross-linked

products are observed and no

cross-linking is seen if Cys98 is

substituted by alanine (lanes 5-

7), indicating that the reaction

is specific. Oxidation is fully


reversed upon DTT addition (lane 2).

B. Purification of oxidized SecA(1-834) by two consecutive GPC chromatographic

steps on a Superdex HR200 10/300 GL; flow rate 0.2ml/min. In the first

chromatography (indicated with dotted line), 50mM Tris-HCl pH=8.0; 1M NaCl; 8M

Urea was used in order to fully dissociate the non-covalent dimers into monomers

(single circle). Fractions were analyzed on non-reducing SDS-PAGE (7.5%) and

those containing the covalent dimer (lanes 3-5; double circle linked with red bar) were

pooled, concentrated and re-fractionated, in 50mM Tris-HCl pH=8.0, 50mM NaCl

(black line).

C. A SecA dimer, cross-linked via cysteine oxidation at its dimerization interface,

binds to SecYEG as evidenced by its float up on sucrose gradient experiments

together with the inverted membrane vesicles (see Supplemental Experimental

Procedures), either at oxidized or at reduced conditions (as indicated).

D. In vitro translocation of proPhoA by SecA(1-834) (lanes 2-5) or, its oxidized dimer

derivative (lanes 6-7) into the lumen of SecYEG-IMVs, in the absence or presence of

1mM DTT (as indicated), as described in Supplemental Experimental Procedures.

Lane 3 shows 5% the proPhoA input. When ATP is omitted from the reaction (lane 2)

or, IMVs are disrupted by 1% (v/v) Triton X-100 (lane 1), proPhoA is not protected

from proteolysis. DTT has no effect on translocation by SecA(1-834) (compare lane 4

to lane 5). On the contrary, translocation by the oxidized SecA(1-834) dimer does not

occur (lane 6) unless DTT is added (lane 7).


Fig. S4 Functional characterization of SecA mutants, related to Figure 4.

A. Secretion, translocation ATPase and triggering are affected by the SecA

dimerization properties

In vitro translocation (dotted line; left y axis), translocation ATPase (black line; left y

axis) and activation energy (grey boxes; right y axis) were determined (see

Supplemental Experimental Procedures) and plotted superimposed as a function of

temperature (4-39ºC; x axis), for various SecA derivatives (indicated within each


panel) bound to either wild type (left panels) or, PrlA4 (right panels) SecYEG-IMVs.

For normalization of translocation ATPase and in vitro translocation (left axis) the

values of SecA at 390C were considered 100% (top left panel). All other values (for all

derivatives) are expressed as a percentage of these values. Error bars of the in vitro

translocation values represent standard deviation values (n=2).

Activation Energy derived from the linear parts of Arrhenius plots, before and

after the transition points and is expressed in kJ/mole (right axis). In some cases, at

low temperature range, ATPase values are close to zero hence, activation energy

could not be determined (i.e.; 4-80C for SecA, top panel on the left).

B. Introducing the W775A mutation in the SecA monomer leads to constitutive

elevated basal ATPase. W775A mutation was

previously shown to cause soluble SecA to

constitutibvely hydrolyze ATP (compare lane 2 to

lane 1; Vrontou et al., 2004). We introduced W775A

mutation in SecA(Δα0/α1-6A). Clearly, SecA(Δα0/α1-

6A)/W775A (lane 4) exhibits as high ATPase as

SecAW775A (lane 2). We conclude that the SecA

monomer does not suffer from an inherent inability to

perform high ATP turnovers. Error bars represent

standard deviation values (n=3).


Fig. S5 Constitutively triggered translocases, related to Figures 2E, 4 and 5.

A. PrlD and PrlA translocases exist in a constitutively triggered state.

In vivo translocation of the indicated proPhoA derivatives by wild type SecA or,

SecA/PrlD23 bound to either wild type SecYEG or, SecY/PrlA4-EG (as indicated). Prl

translocases (with either SecA or SecY Prl subunits; indicated in red) secrete more

efficiently than the wild type

translocase, substrates with

defective signal peptides or no

signal peptides (compare lanes

6-8 and 10-12 with lanes 2-4).

However, as previously

demonstrated (Gouridis et al.,

2009) the Prl translocases do not

restore secretion of defective

substrates back to wild type

efficiency (compare lanes 6-8

with lane 5 or, lanes 10-12 with

lane 9). Error bars represent

standard deviation values (n=3).

B. Activation Energy (Ea; kJ/mole; y axis) of soluble (lanes 1, 4 and 7), membrane-

bound (lanes 2, 5 and 8) and translocating (lanes 3, 6 and 9) SecA (A; as indicated in

cartoon form). Either wild type SecA or SecA/PrlD23(with the Y134S mutation; red

circle) were used. IMVs with wild type SecYEG (Y) or SecY/PrlA4-EG (red square)

were used. For wild type translocase (SecA-SecYEG) the lowering of Ea occurs only


in the presence of preprotein (lane 3). SecY/PrlA4 lowers the Ea at the membrane

bound state without any signal peptide or preprotein (white rectangle). In

SecA/PrlD23 the Ea is already low in solution in the absence of either SecY or signal

peptide (lane 7). This indicates that SecA/PrlD23 has a conformational state that

mimics the allosteric contribution coming to wild type SecA from both the channel and

the signal peptide.

Fig. S6 In vivo genetic complementation by SecA derivatives and growth

inhibition by over-expressed SecA, related to Figures 3C and 6.

A. In vitro translocation of proPhoA by wild type SecA into the lumen of wild type

SecYEG-IMVs was examined as a function of SecY:SecA molar ratio (as indicated).


B. and C. In vivo genetic complementation assay for SecA and mutant derivatives (as

indicated) using the BL21.19 secAts strain, at 42ºC (Mitchell and Oliver, 1993). Cells

transformed with an empty vector or its derivatives with cloned secA(pIMBB10) or

secA(α1-6A; pIMBB1278) or secA(Δα0; pIMBB571) or secA(Δα0/ α1-6A;

pIMBB1286) genes were grown in LB medium, overnight, at 30ºC. Cultures, freshly

diluted in LB (OD595nm=0.01) were: a) incubated at 300C until OD595=0.5; cells were

diluted and spotted on LB/ampicillin plates supplemented with IPTG (as indicated)

and then incubated at 42ºC for 14 hours (Panel B), b) supplemented with IPTG (as

indicated) and then growth was monitored by optical density, at 420C (Panel C). Error

bars in panel C represent standard deviation values (n=2). Cell viability and growth

rate are affected by the: a) SecA expression levels and b) dimerization properties of

the protein being expressed.


Table S1 Dimerization contact residues in SecA 1M6N and 1NL3_1, related to Figure 1.

1M6N

Chain A Chain B Interaction Interface Ile 3 Asp 682 α0-WD

Leu 5 Ile 659 α0-SD Leu 5 Leu 815 α0-IRA1 Leu 5 Leu 818 α0-IRA1 Leu 6 Phe 808 α0-IRA1 Leu 6 Phe 811 α0-IRA1 Leu 6 Leu 815 α0-IRA1 Lys 8 Glu 665 α0-SD Lys 8 Ser 661 α0-SD Lys 8 Gln 662 α0-SD Lys 8 Glu 665 α0-SD Arg 13 Asp 654 α0-SD

Asp 337 Lys 633 PBD-SD Glu 338 Arg 602 PBD-IRA2 His 339 Val 623 PDB-SD His 339 His 620 IRA2-SD Lys 421 Gln 695 IRA2-WD Pro 424 Pro 694 IRA2-WD Ala 599 Tyr 794 IRA2-IRA1 Arg 602 Glu 338 IRA2-PBD His 620 His 339 SD-PBD Ala 623 His 339 SD-PBD Asn 629 Arg 805 SD-IRA1 Arg 632 Arg 805 SD-IRA1 Lys 633 Lys 797 SD-IRA1 Lys 633 Asp 337 SD-PBD Lys 633 Glu 802 SD-IRA1 Ala 636 Gln 801 SD-IRA1 Arg 637 Gln 796 SD-IRA1 Asp 640 Lys 797 SD- IRA1 Lys 643 Glu 647 SD- SD Glu 647 Lys 643 SD- SD Asp 654 Arg 14 SD- α0 Ile 659 Leu 5 SD- α0 Gln 662 Lys 8 SD- α0 Glu 665 Lys 8 SD- α0 Glu 665 Lys 8 SD- α0 Asp 682 Ile 3 WD- α0 Pro 694 Pro 424 WD-IRA2 Gln 695 Lys 421 WD-IRA2 Tyr 794 Ala 599 IRA1-IRA2 Gln 796 Arg 637 IRA1-SD Lys 797 Lys 633 IRA1-SD Lys 797 Asp 640 IRA1- SD Gln 801 Ser 636 IRA1-SD Glu 802 Lys 633 IRA1-SD Arg 805 Asn 629 IRA1-SD Arg 805 Arg 632 IRA1-SD Phe 808 Leu 6 IRA1- α0 Phe 811 Leu 6 IRA1- α0 Leu 815 Leu 5 IRA1- α0 Leu 815 Leu 6 IRA1- α0 Leu 818 Leu 5 IRA1- α0

Hydrogen Bonds 23 Hydrophobic 16

Ionic 14

Type of interaction Average strength

(Kj/mole)/relative strengh Hydrophobic 7.5 (Y)

Hydrogen bonds 25 (3.3 Y) Ionic 52.5 (7 Y)

1NL3_1 Chain A Chain B Interaction Interface Leu 5 Ala 658 α0 – SD Leu 5 Phe 808 α0 – IRA1 Leu 5 Phe 811 α0 – IRA1 Lys 8 Asp 654 α0 – SD Val 9 Phe 808 α0 – IRA1

Arg 22 Asp 689 α1 – WD Lys 23 Asp 682 α1 – WD Lys 23 Glu 816 α1 – IRA1 Lys 23 Asp 682 α1 – WD Lys 23 Asp 816 α1 – IRA1 Asn 26 Lys 685 α1 – WD Ala 30 Glu 729 α1 – WD Lys 56 Glu 674 NBD-WD

Asp 654 Lys 8 SD-α0 Ala 658 Leu 5 SD-α0 Glu 674 Lys 56 WD-NBD Asp 682 Lys 23 SD-α1 Asp 682 Lys 23 SD-α1 Lys 685 Asn 26 WD-α1 Asp 689 Arg 22 WD-α1 Glu 729 Ala 30 WD- α1 Phe 808 Leu 5 IRA1- α0 Phe 808 Val 9 IRA1- α0 Phe 811 Leu 5 IRA1- α0 Glu 816 Lys 23 IRA1- α1 Glu 816 Lys 23 IRA1- α1

Hydrogen Bonds 8 Hydrophobic 8

Ionic 10


The two biologically relevant SecA dimerization interfaces (1M6N, 1NL3_1) predicted

with EPPIC (see Fig. S1A) were further inspected. To this end, the ecSecA sequence

was modeled into these two dimeric structures, and the contact interfaces were

analyzed in the PIC (Protein Interaction Calculator) server (Tina et al., 2007). The

upper two tables are showing the residues involved in the dimer formation in the

structures, the nature of the interactions (i.e. hydrophobic, ionic, hydrogen bonds)

and the SecA regions involved [α0, α1, NBD (Nucleotide binding domain), IRA1, or 2

(Intramolecular regulator of ATPase 1 or 2), SD (Scaffold domain), WD (Wing

domain)]. The bottom table summarizes average values of the strength (in Kj/mole) of

each interaction between a pair of residues (Dickinson, 1997; Eissa et al., 2006;

Knapp et al., 2005). Relative strength values between these interactions are also

indicated in the table at bottom right. The present analysis is only qualitative due to

the intrinsic accuracy limits of homology models, especially with respect to side chain

positioning.


Table S2 Dimensions, masses and nucleotide affinities of SecA and derivatives, related to Figures 2, 3 and 6.

Measured Diameter Measured Mass Dimerization Affinity (kD)

References SecA Protein

Buffer Loading Concent

ration (µM)

nm Method kDa Method Oligomeric state (%)

-ADP (µM)

+ADP (µM)

Y134C/A488C cross-linked

50mM Tris, pH:8, 50mM Kcl

10 N/M 204.7 (±0.89)

MALLS Dimer This study

Y134C/H484C cross-linked


10 N/M 203.6 (±0.78)

MALLS Dimer This study

WT 50mM Tris, pH:8, 50mM Kcl

10-75 10 (±0.05)

QELS 202.08 (±2.9)

MALLS Dimer <0.01 <0.01 This study

6-901 20mM AmAc 1-10 10.23 nES-IMS 196.0 (±2.3)

nES-IMS Dimer (Kapelios et al., 2011)

9-861 50mM sodium citrate, 6–9% (w/v) PEG35000, 6–10%

glycerol, 50mM ammonium

sulphate

N/A 15.8 X-ray, PDB code: 2FSF

N/A Dimer (Papanikolau et al., 2007)

Δα0/Δα1 50mM Tris, pH:8, 50mM Kcl

5-25 N/M 103.08 (±2.4)

MALLS Monomer >200 N.D. This study

α1-6A ED


10-75 10 (±0.05)

QELS 169.07 (±6.2)

MALLS ~68% Dimer 0.137 (±8)

0.204 (±11)

This study

α1-6A SRD


10-75 11 (±0.03)

QELS 175.9 (±1.88)

MALLS ~74% Dimer This study

Δα0


10-75 9.8 (±0.01)

QELS 185.3 (±1.3)

MALLS ~83% Dimer 0.014 (±4)

0.116 (±9)

This study

Δα0/α1-6A 50mM Tris, pH:8, 50mM Kcl

10-75 8.6 (±0.03)

QELS 107.1 (±0.2)

MALLS Monomer ~133 N.D. This study

ΔIRA1 50mM Tris, pH:8, 50mM Kcl

5-25 N/M 108.3 (±0.4)

MALLS Monomer >200 N.D. This study

Wt 50mM Tris, pH:8, 1M Kcl

10-75 8.6 (±0.01)

QELS 101.6 (±1.08)

MALLS Monomer This study

Δα0/α1-6A 50mM Tris, pH:8, 1M Kcl

10-75 8.6 (±0.03)

QELS 102.2 (±1.1)

MALLS Monomer This study

GH 50mM Tris, pH:8, 50mM Kcl

10-75 10.8 (±0.08)

QELS 206.4 (±3.07)

MALLS Dimer <0.01 <0.01 This study

GH SRD

50mM Tris, pH:8, 1M Kcl

10 N.D.

PrlD23 50mM Tris, pH:8, 50mM Kcl

10-75 11 (±0.08)

QELS - 185.1 (±1.5)

MALLS ~83% Dimer This study

1-834 (cross-linked

dimer)

20mM AmAc+1mMDTT

1-10 10.23 (±0.06)

nES-IMS 204.2 (±3.18)

nES-IMS Dimer This study

1-834 (cross-linked

dimer)

20mM AmAc 1-10 10.13 (±0.06)

nES-IMS 198.8 (±3.12)

nES-IMS Dimer This study

Values derived with QELS represent the hydrodynamic or Stokes diameter (DH), while values from

nES-IMS represent the electophoretic diameter (Kapelios et al., 2011). MALLS: multi-angle laser light

scattering; QELS: quasi-elastic light scattering; AmAc: ammonium acetate; nES-IMS: nano

electrospray ion mobility spectrometry. wt is SecA(1-901). Y134C/A488C and Y134C/H484C are

derivatives of SecA(6-901) (Fig. S1B) all other mutated residues are in SecA(1-901).

Dimer dissociation constants (KD) of SecA derivatives were calculated by Prism (GraphPad), from data

presented in Fig. S2.


Table S3 Dissociation constants (KD) and stoichiometries of SecA:SecY

complexes, related to Figures 3, 4 and 5. To WT SecYEG SecA

Dissociation Constant (KD, nM)

Stoichiometry SecYEG:SecA

Wt 50 (±14.9) 2:2.26 α1-6A 28.6(±11) - Δα0 48.8(±16.3) -

Δα0/α1-6A 19.2(±3.1) 2:0.98 GH 63.2(±7.1) 2:1.97

PrlD23 56.6(±9.7) 2:2.37 To proPhoA wt SecYEG:SecA

2:2 Dissociation Constant (KD, nM)

Stoichiometry SecYEG/SecA:proPhoA

Wt 237.4(±63.2) 2:1.17 α1-6A 170(±43.6) 2:1.29 Δα0 190(±30.1) 2:1.2

Δα0/α1-6A 105.9(±36.5) 2:1.03 GH 200.8(±41.6) 2:1.11

PrlD23 275(±30.6) 2:0.9 C98 oxidized dimer 230(±34.3) -

To PhoA wt SecYEG:SecA 2:2 Dissociation Constant

(KD, nM) Stoichiometry

SecYEG/SecA:PhoA Wt 514.6(±57.2) 2:1.37

α1-6A 340(±90.1) 2:1.09 Δα0 440(±99.2) 2:1.11

Δα0/α1-6A 235.6(±65.5) 2:1.16 GH 550(±40.8) 2:0.97

PrlD23 590(±30.3) 2:1.2

Dissociation constants and stoichiometries for the indicated SecA derivatives. (–) indicates the cases

that stoichiometries were not measured due to technical reasons. The C98 oxidized dimer is described

in Fig. S3. Values were determined as described in the Supplemental Experimental Procedures.


Supplemental movies

Supplemental movie S1, related to Figure 1B. Transition from the 1M6N bsSecA dimer (Hunt et al., 2002) to the 1NL3_1 mtSecA

dimer (Fig. S1A). In all cases ecSecA models of the structures were used (see legend

to Supplementary PDB file 1). Chain A (light green) of the 1M6N dimer was

superimposed (using the superimposition function of Pymol) to chain A of 1NL3_1

dimer. The interpolation between these two structures was accomplished with the use

of the proprietary clustering algorithm RIGIMOL that is incorporated into the

molecular graphics and manipulation program PYMOL (DeLano Scientific LLC, Palo

Alto, CA, USA). The movie was obtained using the emovie function in Pymol (Hodis

et al., 2007). The first part of the movie shows a cartoon representation of the

structures with the same color code and orientation as in Fig. 1B. In the second part

of the movie the α0/ α1 region is colored in red, while the two protomers are colored

with light green (fixed) or cyan (moving) color. Different orientations of the transition

between the two crystallographic snapshots are also presented to render obvious the

role of the α0/ α1 region, acting as a gudgeon. The importance of the α0/α1 region

for the transition between distinct dimeric states was analyzed in detail and confirmed

biochemically (see text for details). In the third part of the movie, surface

representation of the two protomers is shown and the color code is kept the same as

in the second part.


Supplemental PDB file S1, related to Figure 1.

Superimposition of the ecSecA 1M6N modeled structure onto the ecSecA

1NL3_1 modeled structure.

The two biologically relevant SecA dimers (1NL3_1, 1M6N) predicted by EPPIC (see

Fig. S1) were further inspected. To this end, the ecSecA sequence was modeled onto

these two dimeric structures by using SwissModel (Kiefer et al., 2009) assembling the

dimers in Coot (Emsley et al., 2010) and minimizing the resulting models with

Refmac5 (Murshudov et al., 2011). A PDB file is displaying the superimposition

based on chain A of the SecA dimer found in the crystal structure of PDB entry 1M6N

(B. subtilis) onto the 1NL3_1 dimer described above. Structures appear in cartoon

mode with 1NL3_1 in blue and 1M6N in magenta. See also corresponding

Supplemental Movie S1 displaying a model of the transition from between these

dimers.


Supplemental Experimental Procedures

Plasmids and strains: are described below or, previously (Baud et al., 2002;

Gouridis et al., 2009; Jarosik and Oliver, 1991; Karamanou et al., 1999; Keramisanou

et al., 2006; Papanikou et al., 2004; Sianidis et al., 2001; Vrontou et al., 2004)

His6SecA(Y134C/6-901), His6SecA(H484C/6-901) and His6SecA(A488C/6-901)

were constructed by “megaprimer” PCR using pIMBB7 [His6SecA(6-901) as template.

For His6SecA(Y134C/6-901), X284 (5’ CC GTC AAC GAC TGC CTG GCG CAA

CG 3’; mutagenic forward) and X278 (5’ G TTT ATA CAT TTC CGA GCT GTC TTC

3’; reverse) were used as primers. This PCR product was used in a second step with

X279 (5’ ACG ACT CAC TAT AGG GAG ACC ACA ACG 3’) as forward primer. The

BclI/BamHI fragment of the final PCR product replaced the corresponding one in

pIMBB7, resulting in pIMBB522.

For His6SecA(H484C/6-901), X317 (5’ CGC TTC GTT GGC GCA GAA TTT

GGC GTT 3’; mutagenic reverse) and X272 (5’ C TAC AAG CTG GAT ACC GTC

GTT GTT 3’; forward) were used as primers. This PCR product was used in a second

step with X273 (5’CTG ACG ACC AGA ACG ACC GCG CAA CTG 3’) as forward

primer. The BglII/KpnI fragment of the final PCR product replaced the corresponding

one in pIMBB7, resulting in pIMBB577.

For His6SecA(A488C/6-901), X320 (5' G AGC AAC AAT CGC GCA TTC GTT

GGC GTG GAA TTT 3'; mutagenic reverse) and X272 (5’ C TAC AAG CTG GAT

ACC GTC GTT GTT 3’; forward) were used as primers. This PCR product was used

in a second step with X273 (5’ CTG ACG ACC AGA ACG ACC GCG CAA CTG 3’) as

reverse primer. The BglII/KpnI fragment of the final PCR product replaced the

corresponding one in pIMBB7, resulting in pIMBB575.


His6SecA-Cys-(Y134C/6-901) was constructed by “megaprimer” PCR using

pIMBB258 (His6SecA-Cys-) as template, X284 (5’ CC GTC AAC GAC TGC CTG

GCG CAA CG 3’; mutagenic forward) and X278 (5’ G TTT ATA CAT TTC CGA GCT

GTC TTC 3’; reverse) as primers. This product was used in a second PCR step with

X279 (5’ ACG ACT CAC TAT AGG GAG ACC ACA ACG 3’) as forward primer. The

BclI/BamHI fragment of the final PCR product replaced the corresponding one in

pIMBB258, resulting in pIMBB642.

His6SecA-Cys-(H484C/6-901) and His6SecA-Cys-(A488C/6-901) were

constructed by replacing the 0.413kbp BglII/KpnI restriction fragment of pIMBB258

with the corresponding one from pIMBB577 or pIMBB575, resulting in pIMBB584 and

pIMBB857 respectively.

His6SecA-Cys-(Y134C/H484C/6-901) was constructed by replacing the

0.88kbp BclI/BamHI restriction fragment of pIMBB584 with the corresponding one

from pIMBB642, resulting in pIMBB859.

His6SecA-Cys-(Y134C/A488C/6-901) was constructed by replacing the

0.886kbp BclI/BamHI restriction fragment of pIMBB857 with the corresponding one

from pIMBB642, resulting in pIMBB858.

His6SecA(Y134S/6-901) was constructed by “megaprimer” PCR using pIMBB7

as template, X283 (5’ CC GTC AAC GAC TCC CTG GCG CAA CG 3’; mutagenic

forward) and X278 (5’ G TTT AT A CAT TTC CGA GCT GTC TTC 3’; reverse) as

primers. The 324bp PCR product was used in a second step with X279 (5’ ACG ACT

CAC TAT AGG GAG ACC ACA ACG 3’) as forward primer. The BclI/BamHI

restriction fragment of the final PCR product replaced the corresponding one of

pIMBB7 resulting in pIMBB523.


His10SecA(R22A/K23A/1-901) was constructed by Quick-Change

mutagenesis, using pIMBB379 [pET16b/SecA(1-901)] as template, X659 (5’ CGC

ACC CTG CGC CGG ATG GCC GCA GTG GTC AAC ATC ATC AAT 3’;

mutagenic forward) and X660 (5’ ATT GAT GAT GTT GAC CAC TGC GGC CAT

CCG GCG CAG GGT GCG ATC GTT ACG 3’; mutagenic reverse) as primers,

resulting in pIMBB1274.

For His10SecA-4A [His10SecA(R19A/R20A/R22A/K23A)], pIMBB1274 was

the template, X699 (5’ CGT AAC GAT CGC ACC CTG GCC GCG ATG GCC

GCA GTG GTC AAC 3’; mutagenic forward) and X700 (5’ GTT GAC CAC TGC

GGC CAT CGC GGC CAG GGT GCG ATC GTT ACG 3’; mutagenic reverse) the

primers, resulting in pIMBB1275.

For His10SecA-6A [His10SecA(D15A/R16A/R19A/R20A/R22A/K23A)]

pIMBB1275 was the template, X816 (5’ GTT TTC GGT AGT CGT AAC GCT GCC

ACC CTG GCC GCG ATG GCC 3’; mutagenic forward) and X864 (5’ GGC CAT CGC

GGC CAG GGT GGC AGC GTT ACG ACT ACC GAA AAC 3’; mutagenic reverse)

the primers, resulting in pIMBB1276.

SecA(1-901) in pET3a was constructed by replacing the 2.1kbp NdeI/MfeI

restriction fragment of pIMBB379 (His-SecA1-901) with the corresponding one from

pIMBB571 [pET3a/SecA(15-901)], resulting in pIMBB1280.

SecA-6A(1-901) in pET3a was constructed by replacing the 2.1kbp NdeI/MfeI

restriction fragment of pIMBB1276 with the corresponding one from pIMBB571


SecA-6A(15-901) in pET3a was constructed by replacing the 2.1kbp NdeI/MfeI

restriction fragment of pIMBB571 with the corresponding one from the PCR product


that used pIMBB1276 as template and X1017 (GGGAATTCCAT ATG GCT GCC

ACC CTG GCC GCG) and X208 (CTC TTC ATG CAG TTC TGG TTC) as primers,


SecA(Y134S/1-901) was constructed by replacing the 2.6kbp NcoI restriction

fragment of pIMBB10 with the corresponding one from pIMBB523, resulting in

pIMBB1314.

SecA-6A(Y134S/15-901) was constructed by replacing the 2.6kbp NcoI

restriction fragment of pIMBB1286 with the corresponding one from pIMBB523,


For in vivo secretion experiments, SecA(1-901) was cloned under the control

of a tetracycline promoter (pTet; inducible with anhydrotetracycline) in pASK-

IBA7plus vector (IBA). For this purpose, the PCR product that derived using pIMBB10

[SecA(1-901)] as template and X174 (5’ATG GTA GGT CTC AGC GCC TAA TCA

AAT TGT TAA CTA AAG TTT TCG3’; inserts BsaI site) and X175 (ATG GTA GGT

CTC ATA TCA TTG CAG GCG GCC ATG GCA CTG3’; inserts BsaI site) as primers

was cloned in the BsaI site of pASK-IBA7plus vector, resulting in pIMBB276. Mutants

were transferred in this vector by subcloning.

For in vivo secretion experiments, preprotein genes were cloned under the

control of an arabinose promoter, using pBAD33.

proPhoA-his was constructed by PCR, using pIMBB882 as template, X629

(5’CGG GGT ACC GTT TAA CTT TAA GAA GGA GAT ATA CAT ATG3’; inserts a

Kpn I site) and X630 (5’AAA CCC AAG CTT TCA GTG GTG GTG GTG GTG GTG3’;

inserts a Hind III site) as primers. This fragment was ligated in the Kpn I-Hind III sites

of pBAD33, resulting in pIMBB932.


proPhoA(L8Q)-his was constructed by PCR using pIMBB883 as template,

X629 (5’CGG GGT ACC GTT TAA CTT TAA GAA GGA GAT ATA CAT ATG3’;

inserts a Kpn I site) and X630 (5’AAA CCC AAG CTT TCA GTG GTG GTG GTG

GTG GTG3’; inserts a Hind III site) as primers. This fragment was ligated in the Kpn I

- Hind III sites of pBAD33, resulting in pIMBB933.

PhoA(∆2-22)-his was constructed by PCR using pIMBB932 as template, X646

(5’GGG AAT TCC ATA TGA CCC CAG AAA TGC CTG TT3’; inserts a Nde I site) and

X561 (5’GAC CCG CTC GAG TTT CAG CCC CAG AGC GGC3’; inserts a Xho I site)

as primers. This fragment replaced the Nde I - Xho I fragment of pIMBB932, resulting

in pIMBB954.

Purification of SecA: SecA derivatives without an oligohistidine-tag, were purified as

follows: after lysis by French press (800 psi; 3-4 rounds), soluble material (50,000xg;

30min; 40C; Sorval) was loaded (50mM Tris–HCl, pH 8; 175mM KCl-10% glycerol;

2.5mM PMSF) on a column with an in-house prepared Cibacron-Blue SepharoseTM

CL-6B resin (Pharmacia). Bound proteins were washed (50mM Tris–HCl, pH 8.0; 50-

200mM KCl; 10% glycerol) and then eluted (50mM Tris–HCl, pH 8.0; 1M KCl-2%

glycerol). Fractions with SecA protein were pooled, concentrated (Vivacell-Sartorius;

Amicon-MIllipore) and loaded (50mM Tris-HCl pH:8.0; 1M KCl) on a Superdex 200

column (HiLoad 26/60; prep grade; GE Healthcare). Monomeric SecA (∼100kDa) was

collected, concentrated and loaded on a second Superdex 200 column (equilibrated

in 50mM Tris-HCl pH:8.0; 50mM KCl buffer). Dimeric SecA (∼200kDa) was collected,

concentrated, dialyzed (in 50mM Tris-HCl pH:8.0; 50mM KCl; 50% glycerol) and

stored at -80ºC. In the case of SecA(Δα0/α1-6A) only monomeric SecA was purified

in the last gel filtration step.


The "ED state" SecA was purified as a monomer in the first gel filtration step

(1M KCl) and a dimer in the second gel filtration step (50mM KCl). The "SRD state"

was purified as a dimer in both gel filtration steps. The "monomer state" was

purified as a monomer in both gel filtration steps. The "TD state" (SecA/PrlD23) was

purified as the "ED state" (see above).

Multi-angle laser light scattering (MALLS) and Quasi-elastic light scattering

(QELS) experiments: Laser light scattering measurements were carried out online

following gel permeation chromatography (GPC) using a Superdex HR200 10/300 GL

column and an HPLC system (LC10A-VP, Shimadzu) coupled to a quadruple

detector scheme connected in series as follows: a photodiode-array detector (SPD-

M10AVP; Shimadzu) for UV monitoring at 280 nm; an 18-angle MALLS detector

(DAWN-EOS, Wyatt) using a K5-type cell and a laser wavelength of 690 nm; a QELS

detector (WyattQELS; Wyatt) connected

through an optical fiber to the MALLS

instrument through laser detector 13;

and a refractive index detector (RID-

10A; Shimadzu). MALLS and QELS

data were collected, analyzed and

plotted using the Astra v.5.0 software (Wyatt). Proteins (2–100µM in 100µl injection

volumes) in 50mM Tris–HCl, pH 8.0, 50mM KCl (or at the indicated KCl

concentrations), were loaded onto the column and chromatographed at 22°C at a

flow rate of 0.8ml min-1. Concentration at the peak fraction was typically 1/10 of the

loaded concentration (data not shown). For QELS measurements of SecA


hydrodynamic dimensions a minimum of 30µM protein loaded (3µM at the peak) was

required. An example of an autocorrelation function graph from a QELS experiments

where the measured data (filled black circles), are fitted (red line) to derive the

hydrodynamic diameter of wild-type SecA (in 50mM Tris-HCl pH:8, 50mM KCl),

obtained for several consecutive fractions of the chromatographic peak during online

GPC-MALLS/QELS (see also Fig.2A).

nES-GEMMA: A nanoelectrospray instrument coupled to a gas-phase electrophoretic

mobility molecular analyzer (nES-GEMMA) with a condensation particle counter

(CPC) detector (TSI Inc., St. Paul, MN) was used. The instrument configuration

consisted of a nES source unit (Model 3480C) equipped with a neutralizing chamber

(210Po α-source; 5 mCi, model P-2042 Nucleospot local air ionizer; NRD, Grand

Island, NY), whereas the ion mobility spectrometer used for the GEMMA separation

was a differential mobility analyzer (TSI Inc., macroIMS, Series 3080C). Detection

was achieved using a butanol-based Ultrafine CPC (TSI Inc., Series 3025A). IMS

version 2.0.1.0 software (TSI Inc.) was used for data acquisition and data analysis.

Sample solutions in Eppendorf tubes were placed into the nES unit as described

(Kapellios et al., 2011) and were analyzed with a nES flow rate of ~70 nL min-1.

Calculation of activation energies (Ea): The activation energy (Ea) of SecA, derived

from Arrhenius plots using measured Kcat values (moles Pi/mol SecA protomer/min)

of basal or, membrane or, translocation ATPase activities of SecA as a function of

temperature. In an Arrhenius plot, the y axis represents the natural logarithm of the

Kcat values, expressed in sec and the x axis represents the inversed temperature

values (1/T), expressed in Kelvin. The Ea of the translocase, was calculated (in kJ


mole-1) using the slope of the linear part of the curve (always involving more than 30

experimental points with R2 ≥0.97), under different regimes.

Preparation of SecYEG-IMVs: was as described (Lill et al., 1990; Rhoads et al.,

1984; Yamada et al., 1989), from cells overexpressing a wild type (van der Does et

al., 1996) or, a PrlA4 secYEG operon. To obtain unilamellar vesicles the IMVs

suspension was manually extruded through a polycarbonate membrane (Avestin Inc,

LiposoFast).

[35S]-labeling of SecA or prophoA derivatives: was done using TNT Quick coupled

Transcription/Translation systems (Promega) and following its instructions. At the end

of labeling, a buffer exchange step, using G-50 resin equilibrated with either buffer B

(50mM Tris-HCl pH:8.0; 50mM KCl, 5mM MgCl2) for SecA proteins or, 6M-Urea-

50mM Tris pH:8, 50mM KCl for prophoA proteins was added. 1mM DTT was further

added to all proteins.

SecA-SecY equilibrium dissociation constants (KD) and maximum receptor

concentration (Bmax): SecA bound to SecY: SecA was serially diluted in buffer B

reactions (20µl; 50mM Tris-HCl pH:8.0; 50mM KCl, 5mM MgCl2; 1mg/ml BSA)

containing SecYEG-IMVs (80nM SecY), in order to achieve a 10-250nM SecA

concentration range. [35S]-SecA was added as tracer (2µl) to all reactions. ADP or


AMP-PNP was added (1mM; whenever/as indicated). Samples were incubated on

ice, for 20 min, overlaid with an equal volume of sucrose cushion (6.85% (w/v) in

buffer B) and ultracentrifuged (320,000xg; 30min; 40C). The membrane bound

material was resuspended in buffer B and immobilized on a nitrocellulose membrane

using a vacuum manifold (Bio-Rad). [35S]-signals were visualized on the

phosphorimager (Storm, GE), quantified using Image J and then extrapolated to the

amount of SecA bound to SecYEG taking into account that, each signal represented :

x concentration of non-labeled SecA (10-250nM)+2µl [35S]-SecA. Bound SecA to

SecYEG was plotted (y axis) against SecA concentration used in the reaction (x axis).

Fitting of the hyperbolic curves was done by nonlinear regression, for one binding

site; KD and Bmax were determined using Prism (Graph Pad).

With this assay, not only affinities, but also stoichiometries can be determined. By

knowing the concentration of the membrane bound receptor (SecY), the oligomeric

state of the soluble ligand (SecA) within the concentration range used in the

experiment, we can calculate the maximal concentration of SecA bound to SecY

(Bmax). From this we can estimate the stoichiometry of SecA and SecY molecules in

the complex. In the experiment presented above, we are using wild type SecY and

dimeric (WT) or monomeric (Δα0/α1-6A) SecA. According to the estimated KD

(presented in Table S2); within the concentration range (10-250nM) used in the

experiment, SecA (WT) is completely dimeric, while SecA (Δα0/α1-6A), is completely

monomeric. The maximal concentration of dimeric SecA bound to SecY is Bmax ∼ 90

nM, while the concentration of monomeric SecA is approximatelly half of it, Bmax ∼ 40

nM. In both cases the concentration of SecY used is 80 nM. From this, we conclude

that the stoiciometries are SecY: SecA (WT) = 80nM /90 nM ∼ 2:2, while SecY: SecA


(Δα0/α1-6A) = 80nM / 40nM = 2:1. Similarly, we can estimate the ratio of proPhoA or

PhoA bound to the SecY/A complex (see below). All results of these experiments are

presented in Fig. 3A and B.

proPhoA or PhoA to SecY-bound SecA: SecA (200nM) and SecYEG-IMVs (200nM

SecY), in 20µl buffer B (see above) were pre-incubated on ice for 10min. proPhoA or

phoA (preincubated with 2mM DTT, on ice, for 2hrs to maintain a translocation-

competent state) was diluted in those samples in order to achieve a 50-1400nM (for

proPhoA) or, a 100-2000nM (for PhoA) concentration range. [35S]-substrate was

added as tracer (2µl) to all reactions. ADP or AMP-PNP was added (1mM;

whenever/as indicated). Samples were incubated on ice, for 20 min, overlaid with an

equal volume of sucrose cushion (6.85% (w/v) in buffer B) and ultracentrifuged

(320,000xg; 30min; 40C). Pellet resuspension, immobilization of membrane bound

material on nitrocellulose membrane, visualization and quantitation of [35S]-signals,

calculation of the amount of non-labelled prophoA bound to SecA-SecYEG, data

analysis and determination of KD and

Bmax were done as described above.

Membrane insertion assay: Wild

type SecA (0.3µM) and wild type (or, as indicated) SecYEG-IMVs (0.6µM) were

incubated (5min; 370C) with the indicated nucleotide (2mM). Samples were incubated

with trypsin (0.3 mg/ml; 15min; 40C) and then Pefabloc was added (15mM; 10min;

40C). Proteins were analyzed by SDS-PAGE, immunostained with α-C-domain SecA

antibody and visualized by a CCD camera (LAS3000, Fujifilm). The amount of the

protease-protected SecA C-domain (Economou and Wickner, 1994) was quantified

using Image J (NIH). An example of an insertion experiment is shown (as indicated).


25 such experiments were quantified, values were averaged, expressed as a

percentage of the measured protease-resistant material in the presence of AMP-PNP

(considered as 100%) and results (including error bars) are presented in Fig. 5A.

In vivo translocation of proPhoA and its derivatives: secYEG (van der Does et

al., 1996) or prlA4[secY(I408N/F286Y]EG (Smith et al., 2005) operons were cloned in

pET610 (AmpR; pMB1/pBR322 ori) under a trc promoter (van der Does et al., 1996).

secA or secA/prlD23(Y134S) genes (Huie and Silhavy, 1995) were cloned in pASK-

IBA7plus (AmpR; ColE1 ori) under a tet promoter. proPhoA or derivatives were cloned


in pBAD33 (CmR; pACYC184/p15 ori), under an arabinose promoter (Li et al., 2007).

JM109 cells were co-transformed with either pET610 & pBAD33 or pASK-IBA7plus &

pBAD33 and grown at 37ºC (OD595nm=0.2). SecYEG or SecA synthesis was induced

by addition of 0.2mM IPTG (isopropylthiogalactoside) or 0.1µM tetracycline

respectively for 20min. Synthesis of proPhoA (or derivatives) was limited to the basal

read-through of the arabinose promoter (20min). Cells, harvested by centrifugation

(3,834xg; 4min), were resuspended in 1M Tris-HCl pH:8.0, 15mM p-nitrophenyl

phosphate and incubated at 37ºC until a strong yellow color was developed.

Reactions were stopped by adding 10% (v/v) of buffer X (1 volume 0.5M EDTA pH8,

4 volumes 2.5M K2HPO4) and (1% v/v) TX-100. Cell debris were removed by

centrifugation (17,000xg; 4min) and absorbance was measured at 420nm. Alkaline

phosphatase units were calculated (Derman et al., 1993) and then converted to mass

of secreted protein by using a standard curve of the phosphate released by known

amounts of purified PhoA. Secretion by chromosomal SecYEG or SecA under the

same conditions was measured (using empty pET610 or pASK plasmids) and

subtracted.

In vitro translocation of proPhoA: Translocation was performed in 100µl reactions

(50mM Tris-HCl pH:8.0, 50mM KCl, 5mM MgCl2, 0.5mg ml-1 BSA, 1mM DTT, ATP

(2.5-10mM; as indicated), SecA or derivatives (0.4 or as indicated), SecYEG IMVs or

derivatives (1µM SecY; or as indicated) and proPhoA (8µM; or as indicated).

Reactions were incubated (37ºC; 12min; or as indicated) and translocation of proteins

into the lumen of the IMVs was terminated by addition of proteinase K (1mg ml-1, 20-

60min; or as indicated, 4ºC). Proteins were precipitated with trichloroacetic acid

(TCA; 15% w/v), analyzed by SDS–PAGE (13% acrylamide) together with known


amounts of purified proPhoA-his and immunostained with α-proPhoA antisera.

Chemiluminescent signals were visualized with a CCD-imager (LAS-3000; Fujifilm)

and measured using Image J software (Wayne Rasband; NIH). Translocated

immunostained material was transformed to protein mass using standard curves of

purified proPhoA-his immunostained on the same blots. The standard curves were

also used to ensure that in all the experiments the translocated material was in the

linear range (R2 ≥ 0.95).

Trapping of the polypeptide chain in the translocase holoenzyme: Procedure

accomplished as previously described (Gouridis et al., 2009). Briefly, the translocase

holoenzyme, assembled on SecYEG IMVs (1.0µM SecY) by addition of 0.4µM SecA

in buffer B, was incubated on ice for 10 min with freshly prepared [35S]-PhoA (∼600

fmoles), overlaid on an equal volume of BSA/sucrose cushion (prepared as

previously described) and ultracentrifuged (320,000g; 30 min; 4 °C). The SecYEG

bound [35S]-PhoA present in the pellet was isolated and resuspended in buffer B and

then incubated for 2 min at 37°C with ATP (1 mM) and synthetic proPhoA signal

peptide (50µM). At this point reactions were chased with excess of non-radiolabelled

PhoA (1.5µM). Translocation into the lumen of the IMVs was accomplished at 37°C

for 10 min and then terminated by proteinase K addition (1mg ml-1, 20 min, 4ºC).

Proteins were precipitated with trichloroacetic acid (TCA; 15% w/v), analyzed by

SDS–PAGE (13% acrylamide). The gel was incubated with sodium salicilate (1 h, 1

M) and then visualized by phosphor-imaging.

In vitro translocation of proPhoA as a function of SecA concentration:

Translocation reactions were set up as described (Supplemental Experimental

Procedures). SecA was used in the range of 0.4-30µM. Translocated immunostained


material was transformed to protein mass using standard curves of purified proPhoA-

his immunostained on the same blots. The use of standard curves ensured that the

translocated material was in the linear range in all the experiments (R2 ≥0.95). The

maximum in vitro translocation efficiency of SecA (0.2pmoles proPhoA /pmole

SecA/min) was achieved using 0.2µM SecA and a ratio of SecA1:SecY2.

In vivo translocation of proPhoA as a function of SecA concentration:

Translocation of proPhoA as a function of SecA concentration was examined in vivo.

SecA wt and the indicated derivatives were examined. The BL21.19 secAts strain

was transformed with vectors carrying secA genes. Cultures, grown at 300C

overnight, were diluted in fresh LB medium (OD595nm=0.01) supplemented with IPTG

(0, 1, 2, 3, 4, 5, 10, 15, 30 µM) and incubated at 42ºC, for 3 hours. Cells were

harvested by centrifugation; 1) secretion of proPhoA to the periplasm (see

Supplemental Experimental Procedures) and 2) intracellular SecA concentration were

measured (see Supplemental Experimental Procedures). In addition, the in vivo SecA

concentrations were adjusted accordingly to the in vitro ones using the normalized in

vitro and in vivo secretion values of SecA wt. The in vitro SecA secretion values are

linearly correlated to the SecA concentration (y=-0.66x+99.66, R2=0.77). The in vivo

SecA secretion values are also linearly correlated to the calculated intracellular

concentration, up to 60µM, at which the remaining secretion activity is approximately

15% (y=-1.83x + 107.65, R2=0.98). From these two equations, the concentration at

which the remaining activity would be 50%; is calculated to be 75µM (for the in vitro

measurements) or 31µM (for the in vivo measurements). The fact that the in vitro and

in vivo results were co-linear within a concentration range, allowed us to adjust the in

vivo calculated concentrations with the in vitro used concentrations. This was made


possible by multiplying the in vivo concentration values by a “correction factor” of 2.4

(=75/31). We interpret this difference as suggesting that the effective in vivo

concentration is ~2.4 times than what is nominally calculated. We attribute this

difference to poorly understood in vivo parameters such as macromolecular

crowding, biased sub-cellular protein topologies and lower actual available water

volume available for macromolecular diffusion (Minton, 2006; Zimmerman and Trach,

1991). Both data-sets permitted observation of proPhoA secretion as a function of

SecA concentration at the full concentration and secretion range (Fig. 5E).

Intracellular concentration of SecA or SecY: E.

coli cells with plasmids expressing SecA or SecY,

empty plasmids or no plasmid were grown to mid-

log phase and then harvested (17,000g; 10min).

The volume of liquid culture needed to yield

250ng SecA or 1000ng SecY was determined for

each strain separately (i.e. 1ml of E. coli K12 cells

O.D.595nm=0.5). Cells were resuspended in

3XLaemli buffer (250µl) supplemented with

protease inhibitor cocktail (EDTA Free

SigmaFASTTM). For SecA, cells were lyzed by incubating at 95ºC (10min; vigorous

vortexing). For SecY, cell lysis was achieved by a 30min incubation at 25oC, followed

by 6 cycles of 30 sec sonication (Branson 1510) and an additional 10 min incubation

at 370C. Various amounts of these lysates were analyzed, together with known

amounts of purified SecA or SecY proteins, by SDS-PAGE (10% or 13%) and

immunostained with α-SecA or α-SecY antisera. Proteins were quantitified as before


(see SupplementalExperimental Procedures In vitro translocation of proPhoA). An

example of SecY quantification is presented here. The SDS-PAGE gel (13%) shows

various amounts of lysates analyzed (lanes 1-4) together with purified, known

amounts of SecY (lanes 5-8). The curves from the densitometry analysis are also

presented.

Values were then transformed to intracellular concentrations by1 taking into

account the number of cells used for loading and the volume of a single cell. In a

liquid culture of E.coli, an O.D.595nm=1 corresponds to ~8x108 cells/ml (Moran et al.,

2010; Volkmer and Heinemann, 2011) The average estimated volume of an E. coli

cell (at midlog phase; in LB medium) is 2.5x10-15 Lt (Moran et al., 2010; Volkmer and

Heinemann, 2011).

Ultracentrifugation sedimentation experiments: were carried out in a bench-top

ultracentrifuge (TLX120 Optima; TLA120.2 rotor; Beckman), using polypropylene

tubes (0.2ml) (Gelis et al., 2007). Volumes and densities of the sucrose layers were

modified in order to achieve optimal flotation, as follows: reactions (10μl in buffer B

(see above) containing oxidized or reduced (10mM DTT; 12h) His6SecA6-834(C98)

(0.6μM) in the presence or absence of IMVs (1.5μM) were adjusted to 2M final

sucrose concentration and deposited over 15μl of 2.4M sucrose . Samples were

overlaid with one layer (30μl) of 1.7M sucrose and two consecutive layers, (75μl) of

1.4M and (20μl) of 1M sucrose in buffer B following centrifugation (4˚C; 436.000xg;

180min). Nine fractions of 22μl were removed, analyzed by SDS-PAGE, and

visualized by immunostaining of SecA with α-SecA antibodies to label bound SecA to

SecYEG containing IMVs.


Miscellaneous: Protein concentration was determined using the Bradford reagent

with BSA as standard. Nucleotides were from Roche; DNA enzymes from Minotech;

oligonucleotides from Metabion (Germany); dNTPs from Promega(Madison, USA);

chromatography materials and molecular weight markers from GE (U.K.).

Supplemental References

Baud, C., Karamanou, S., Sianidis, G., Vrontou, E., Politou, A.S., and Economou, A. (2002). Allosteric communication between signal peptides and the SecA protein DEAD motor ATPase domain. J Biol Chem 277, 13724-13731. Caruthers, J.M., Johnson, E.R., and McKay, D.B. (2000). Crystal structure of yeast initiation factor 4A, a DEAD-box RNA helicase. Proc Natl Acad Sci U S A 97, 13080-13085. Deleage, G., and Roux, B. (1987). An algorithm for protein secondary structure prediction based on class prediction. Protein Eng 1, 289-294. Derman, A.I., Puziss, J.W., Bassford, P.J., Jr., and Beckwith, J. (1993). A signal sequence is not required for protein export in prlA mutants of Escherichia coli. Embo J 12, 879-888. Dickinson, R.B. (1997). A Dynamic Model for the Attachment of a Brownian Particle Mediated by Discrete Macromolecular Bonds. J Colloid Interface Sci 190, 142-151. Ding, H., Hunt, J.F., Mukerji, I., and Oliver, D. (2003). Bacillus subtilis SecA ATPase exists as an antiparallel dimer in solution. Biochemistry 42, 8729-8738. Duarte, J.M., Srebniak, A., Scharer, M.A., and Capitani, G. (2012). Protein interface classification by evolutionary analysis. BMC bioinformatics 13, 334. Eissa, A.S., Puhl, C., Kadla, J.F., and Khan, S.A. (2006). Enzymatic cross-linking of beta-lactoglobulin: conformational properties using FTIR spectroscopy. Biomacromolecules 7, 1707-1713. Emsley, P., Lohkamp, B., Scott, W.G., and Cowtan, K. (2010). Features and development of Coot. Acta Crystallogr D Biol Crystallogr 66, 486-501. Garnier, J., Gibrat, J.F., and Robson, B. (1996). GOR method for predicting protein secondary structure from amino acid sequence. Methods Enzymol 266, 540-553. Garnier, J., Osguthorpe, D.J., and Robson, B. (1978). Analysis of the accuracy and implications of simple methods for predicting the secondary structure of globular proteins. J Mol Biol 120, 97-120. Geourjon, C., and Deleage, G. (1994). SOPM: a self-optimized method for protein secondary structure prediction. Protein Eng 7, 157-164. Gibrat, J.F., Garnier, J., and Robson, B. (1987). Further developments of protein secondary structure prediction using information theory. New parameters and consideration of residue pairs. J Mol Biol 198, 425-443. Gouridis, G., Karamanou, S., Gelis, I., Kalodimos, C.G., and Economou, A. (2009). Signal peptides are allosteric activators of the protein translocase. Nature 462, 363-367.


Hodis, E., Schreiber, G., Rother, K., and Sussman, J.L. (2007). eMovie: a storyboard-based tool for making molecular movies. Trends Biochem Sci 32, 199-204. Huie, J.L., and Silhavy, T.J. (1995). Suppression of signal sequence defects and azide resistance in Escherichia coli commonly result from the same mutations in secA. J Bacteriol 177, 3518-3526. Hunt, J.F., Weinkauf, S., Henry, L., Fak, J.J., McNicholas, P., Oliver, D.B., and Deisenhofer, J. (2002). Nucleotide control of interdomain interactions in the conformational reaction cycle of SecA. Science 297, 2018-2026. Jarosik, G.P., and Oliver, D.B. (1991). Isolation and analysis of dominant secA mutations in Escherichia coli. J Bacteriol 173, 860-868. Kapellios, E.A., Karamanou, S., Sardis, M.F., Aivaliotis, M., Economou, A., and Pergantis, S.A. (2011). Using nanoelectrospray ion mobility spectrometry (GEMMA) to determine the size and relative molecular mass of proteins and protein assemblies: a comparison with MALLS and QELS. Anal Bioanal Chem 399, 2421-2433. Karamanou, S., Vrontou, E., Sianidis, G., Baud, C., Roos, T., Kuhn, A., Politou, A.S., and Economou, A. (1999). A molecular switch in SecA protein couples ATP hydrolysis to protein translocation. Mol Microbiol 34, 1133-1145. Keramisanou, D., Biris, N., Gelis, I., Sianidis, G., Karamanou, S., Economou, A., and Kalodimos, C.G. (2006). Disorder-order folding transitions underlie catalysis in the helicase motor of SecA. Nat Struct Mol Biol 13, 594-602. Kiefer, F., Arnold, K., Kunzli, M., Bordoli, L., and Schwede, T. (2009). The SWISS-MODEL Repository and associated resources. Nucleic Acids Res 37, D387-392. King, R.D., and Sternberg, M.J. (1996). Identification and application of the concepts important for accurate and reliable protein secondary structure prediction. Protein Sci 5, 2298-2310. Knapp, J.E., Bonham, M.A., Gibson, Q.H., Nichols, J.C., and Royer, W.E., Jr. (2005). Residue F4 plays a key role in modulating oxygen affinity and cooperativity in Scapharca dimeric hemoglobin. Biochemistry 44, 14419-14430. Li, W., Schulman, S., Boyd, D., Erlandson, K., Beckwith, J., and Rapoport, T.A. (2007). The plug domain of the SecY protein stabilizes the closed state of the translocation channel and maintains a membrane seal. Mol Cell 26, 511-521. Lill, R., Dowhan, W., and Wickner, W. (1990). The ATPase activity of SecA is regulated by acidic phospholipids, SecY, and the leader and mature domains of precursor proteins. Cell 60, 271-280. Minton, A.P. (2006). How can biochemical reactions within cells differ from those in test tubes? Journal of cell science 119, 2863-2869. Moran, U., Phillips, R., and Milo, R. (2010). SnapShot: key numbers in biology. Cell 141, 1262-1262 e1261. Murshudov, G.N., Skubak, P., Lebedev, A.A., Pannu, N.S., Steiner, R.A., Nicholls, R.A., Winn, M.D., Long, F., and Vagin, A.A. (2011). REFMAC5 for the refinement of macromolecular crystal structures. Acta Crystallogr D Biol Crystallogr 67, 355-367. Osborne, A.R., Clemons, W.M., Jr., and Rapoport, T.A. (2004). A large conformational change of the translocation ATPase SecA. Proc Natl Acad Sci U S A 101, 10937-10942. Papanikolau, Y., Papadovasilaki, M., Ravelli, R.B., McCarthy, A.A., Cusack, S., Economou, A., and Petratos, K. (2007). Structure of dimeric SecA, the Escherichia coli preprotein translocase motor. Journal of molecular biology 366, 1545-1557.


Papanikou, E., Karamanou, S., Baud, C., Sianidis, G., Frank, M., and Economou, A. (2004). Helicase Motif III in SecA is essential for coupling preprotein binding to translocation ATPase. EMBO Rep 5, 807-811. Rhoads, D.B., Tai, P.C., and Davis, B.D. (1984). Energy-requiring translocation of the OmpA protein and alkaline phosphatase of Escherichia coli into inner membrane vesicles. J Bacteriol 159, 63-70. Rost, B., Sander, C., and Schneider, R. (1994). PHD--an automatic mail server for protein secondary structure prediction. Comput Appl Biosci 10, 53-60. Scharer, M.A., Grutter, M.G., and Capitani, G. (2010). CRK: an evolutionary approach for distinguishing biologically relevant interfaces from crystal contacts. Proteins 78, 2707-2713. Sengoku, S., and Wagatsuma, K. (2006). Comparative studies of spectrochemical characteristics between axial and radial observations in inductively coupled plasma optical emission spectrometry. Anal Sci 22, 245-248. Sharma, V., Arockiasamy, A., Ronning, D.R., Savva, C.G., Holzenburg, A., Braunstein, M., Jacobs, W.R., Jr., and Sacchettini, J.C. (2003). Crystal structure of Mycobacterium tuberculosis SecA, a preprotein translocating ATPase. Proc Natl Acad Sci U S A 100, 2243-2248. Sianidis, G., Karamanou, S., Vrontou, E., Boulias, K., Repanas, K., Kyrpides, N., Politou, A.S., and Economou, A. (2001). Cross-talk between catalytic and regulatory elements in a DEAD motor domain is essential for SecA function. Embo J 20, 961-970. Smith, M.A., Clemons, W.M., Jr., DeMars, C.J., and Flower, A.M. (2005). Modeling the effects of prl mutations on the Escherichia coli SecY complex. J Bacteriol 187, 6454-6465. Tina, K.G., Bhadra, R., and Srinivasan, N. (2007). PIC: Protein Interactions Calculator. Nucleic Acids Res 35, W473-476. van der Does, C., den Blaauwen, T., de Wit, J.G., Manting, E.H., Groot, N.A., Fekkes, P., and Driessen, A.J. (1996). SecA is an intrinsic subunit of the Escherichia coli preprotein translocase and exposes its carboxyl terminus to the periplasm. Mol Microbiol 22, 619-629. Vassylyev, D.G., Mori, H., Vassylyeva, M.N., Tsukazaki, T., Kimura, Y., Tahirov, T.H., and Ito, K. (2006). Crystal structure of the translocation ATPase SecA from Thermus thermophilus reveals a parallel, head-to-head dimer. Journal of molecular biology 364, 248-258. Volkmer, B., and Heinemann, M. (2011). Condition-dependent cell volume and concentration of Escherichia coli to facilitate data conversion for systems biology modeling. PLoS One 6, e23126. Vrontou, E., Karamanou, S., Baud, C., Sianidis, G., and Economou, A. (2004). Global co-ordination of protein translocation by the SecA IRA1 switch. J Biol Chem 279, 22490-22497. Yamada, H., Tokuda, H., and Mizushima, S. (1989). Proton motive force-dependent and -independent protein translocation revealed by an efficient in vitro assay system of Escherichia coli. J Biol Chem 264, 1723-1728. Zimmer, J., Li, W., and Rapoport, T.A. (2006). A novel dimer interface and conformational changes revealed by an X-ray structure of B. subtilis SecA. Journal of molecular biology 364, 259-265.


Zimmer, J., Nam, Y., and Rapoport, T.A. (2008). Structure of a complex of the ATPase SecA and the protein-translocation channel. Nature 455, 936-943. Zimmerman, S.B., and Trach, S.O. (1991). Estimation of macromolecule concentrations and excluded volume effects for the cytoplasm of Escherichia coli. Journal of molecular biology 222, 599-620.

Quaternary Dynamics of the SecA Motor Drive Motor Drive Translocase Catalysis ... area), chains...

Documents

Transcript of Quaternary Dynamics of the SecA Motor Drive Motor Drive Translocase Catalysis ... area), chains...