This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental material
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow E-mail this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Arends, S. J. R.
Right arrow Articles by Weiss, D. S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Arends, S. J. R.
Right arrow Articles by Weiss, D. S.

 Previous Article  |  Next Article 

Journal of Bacteriology, June 2009, p. 3772-3784, Vol. 191, No. 12
0021-9193/09/$08.00+0     doi:10.1128/JB.00179-09
Copyright © 2009, American Society for Microbiology. All Rights Reserved.

ATP-Binding Site Lesions in FtsE Impair Cell Division {triangledown} ,{dagger}

S. J. Ryan Arends, Ryan J. Kustusch, and David S. Weiss*

Department of Microbiology, University of Iowa, Iowa City, Iowa 52242

Received 10 February 2009/ Accepted 8 April 2009


arrow
ABSTRACT
 
FtsE and FtsX of Escherichia coli constitute an apparent ABC transporter that localizes to the septal ring. In the absence of FtsEX, cells divide poorly and several membrane proteins essential for cell division are largely absent from the septal ring, including FtsK, FtsQ, FtsI, and FtsN. These observations, together with the fact that ftsE and ftsX are cotranscribed with ftsY, which helps to target some proteins for insertion into the cytoplasmic membrane, suggested that FtsEX might contribute to insertion of division proteins into the membrane. Here we show that this hypothesis is probably wrong, because cells depleted of FtsEX had normal amounts of FtsK, FtsQ, FtsI, and FtsN in the membrane fraction. We also show that FtsX localizes to septal rings in cells that lack FtsE, arguing that FtsX targets the FtsEX complex to the ring. Nevertheless, both proteins had to be present to recruit further Fts proteins to the ring. Mutant FtsE proteins with lesions in the ATP-binding site supported septal ring assembly (when produced together with FtsX), but these rings constricted poorly. This finding implies that FtsEX uses ATP to facilitate constriction rather than assembly of the septal ring. Finally, topology analysis revealed that FtsX has only four transmembrane segments, none of which contains a charged amino acid. This structure is not what one would expect of a substrate-specific transmembrane channel, leading us to suggest that FtsEX is not really a transporter even though it probably has to hydrolyze ATP to support cell division.


arrow
INTRODUCTION
 
Cell division in Escherichia coli is carried out by ~20 proteins that localize to the midcell, where they form a structure called the septal ring (also called the divisome or septalsome) (4, 27, 55, 57). One component of the septal ring is an apparent ABC transporter composed of the integral membrane protein FtsX and its associated cytoplasmic ATPase, FtsE (15, 47). FtsE and FtsX are widely conserved among gram-negative and gram-positive bacteria. ftsE and/or ftsX mutants exhibit division defects in E. coli, Neisseria gonorrhoeae, Aeromonas hydrophila, and Flavobacterium johnsoniae, indicating that FtsEX function in cell division is conserved in these organisms (33, 40, 43, 45). In contrast, FtsEX of Bacillus subtilis has no obvious role in cell division but instead regulates entry into sporulation (24).

One interesting property of E. coli ftsEX null mutants is that they can be rescued by a variety of osmotic protectants (44). For example, when grown in LB containing >0.5% NaCl, an E. coli ftsEX null mutant is viable and only mildly filamentous, but upon shift to LB lacking NaCl, the cells become filamentous and die (20, 47). A shift to low-osmolarity medium is also accompanied by a dramatic slowing of the overall rate of growth (mass increase) (47). We suspect that ftsEX contributes to both cell division and growth, but it has proven difficult to exclude the possibility that the growth defect is caused by attempts at cell division that go awry.

FtsEX contributes to cytokinesis by improving the assembly and/or stability of the septal ring. Septal ring assembly in an ftsEX mutant is fairly normal in LB that contains 1% NaCl but defective in LB that lacks NaCl (hereinafter referred to as LB0N) (47). More precisely, in LB0N, septal ring assemblies contain the "early" division proteins FtsZ, FtsA, and ZipA but lack the "late" proteins FtsK, FtsQ, FtsL, FtsI, and FtsN. The mechanism by which FtsEX contributes to septal ring assembly is still under investigation, but it probably involves protein-protein interactions, because FtsX has been shown to interact with FtsA and FtsQ in a bacterial two-hybrid system (31), while FtsE has been shown to interact with FtsZ in a coprecipitation assay (15). The FtsE-FtsZ interaction could be important for improving constriction rather than, or in addition to, septal ring assembly.

Remarkably, nothing is known about FtsEX's most obvious potential function—transporting a substrate involved in septum assembly. We are aware of only two studies that attempted to address this issue. The first concluded that FtsEX is needed for insertion of potassium transporters in the cytoplasmic membrane (54). However, in our view the data were not compelling and the connection to cell division, if any, is not obvious. The other study noted that ftsE and ftsX are cotranscribed with ftsY, which is a component of the signal recognition particle pathway for insertion of many proteins into the cytoplasmic membrane (20). That study therefore tested an ftsEX null mutant for defects in export of β-lactamase to the periplasm or insertion of leader peptidase into the cytoplasmic membrane. No such defects were found, so the authors concluded that FtsEX function is probably unrelated to FtsY. Besides these studies, at least one review article suggested that FtsEX might insert division proteins into the cytoplasmic membrane (8). Obviously, the finding that localization of several membrane proteins to the septal ring shows a leaky but pronounced dependence on FtsEX could be explained if the "missing" proteins were not getting into the membrane efficiently. Despite these speculations about potential FtsEX substrates, it is important to note that some members of the ABC "transporter" family do not move anything across cell membranes (reviewed in reference 19), so it cannot be taken for granted that FtsEX is a transporter at all.


arrow
MATERIALS AND METHODS
 
Media. LB consisted of 10 g of tryptone, 5 g of yeast extract, and 10 g NaCl per liter, with 15 g agar/liter for plates. For LB0N, the NaCl was omitted. Antibiotics were used at the following concentrations: chloramphenicol (Cam), 10 µg/ml; kanamycin (Kan), 40 µg/ml; ampicillin (Amp), 100 µg/ml for plasmids and 25 µg/ml for chromosomal alleles; and spectinomycin (Spc), 50 µg/ml. Isopropyl-β-D-thiogalactopyranoside (IPTG) was used at 0.1 mM, while L-arabinose (Ara) and D-glucose (Glu) were used at 0.2%. Cultures were grown at 30°C.

Bacterial strains. Strains and plasmids used in this study are listed in Table 1. Strains EC437, EC439, EC441, and EC449 were constructed from MC4100 using {lambda} InCh as described previously (10, 58), except that the respective fts genes were initially cloned into pDSW209 or pDSW210. EC1215, a markerless ftsEX deletion mutant derived from MG1655, was constructed in several steps. First, a complementing copy of ftsEX was introduced into EC251 (our laboratory's isolate of MG1655) by transformation with the CRIM helper plasmid pAH69 followed by integration of the CRIM plasmid pDSW525 (Spcr, pAH144::PBAD::ftsEX) into the chromosome at the HK022 att site to produce EC1068 (29). Second, the authentic copy of ftsEX in the 76-min region was replaced with a Camr marker using {lambda} Red methods (18). To do this, an 1,114-bp {Delta}ftsEX::Cam fragment was obtained by PCR on the Camr template plasmid pKD3 using P494 (AGAGGCACTTTTTGCCCGAGAGGATTAACAATGATTCGCTTTGAACATGTGTGTAGGCTGGAGCTGCTTC) and P495 (ATCAAACTTCGTTCCGAAAACCTGTGCCACTTCCGCAACCGCCGATGACACATATGAATATCCTCCTTAG) as primers. This fragment was used to replace ftsEX in EC1068/pKD46 (carries {lambda} Red recombinase). Recombinants were selected on LB Cam Ara; the arabinose serves to induce PBAD::ftsEX and was maintained in all subsequent steps. Recombinants were confirmed by colony PCR using the primers C1 and C2, described in reference 18, along with the gene-specific primers P505 (GCCTGCAGAACAAATCGCACC) and P506 (GACATAGAGATGAAATACCGG). An isolate with the desired deletion of ftsEX was saved as EC1111. Third, the FLP recombinase plasmid pCP20 (13) was used to excise the Cam marker, creating EC1117. Finally, the complementing PBAD::ftsEX was evicted using pAH83 (29) to create EC1215. The deletion in EC1215 extends from codon 6 of ftsE to codon 297 of ftsX and does not disrupt the major promoter for rpoH embedded in the 3' end of ftsX (22). To create EC1882, EC1215/pDSW610 was transduced to Ampr with P1 phage grown on EC436.


View this table:
[in this window]
[in a new window]

  		TABLE 1. Strains and plasmids

Molecular biological procedures and oligonucleotides. Standard procedures for cloning and analysis of DNA, PCR, electroporation, and transformation were used (5). Kits from Qiagen (Valencia, CA) were used to isolate DNA. Enzymes used to manipulate DNA were from New England Biolabs (Beverly, MA). Oligonucleotides were from Integrated DNA Technologies (Coralville, IA). Sequencing of DNA was performed by the DNA Core Facility of the University of Iowa using dye terminator cycle-sequencing chemistry. All constructs made by PCR were sequenced to verify their integrity.

Site-directed mutagenesis and plasmids. Megapriming (46) was used to introduce lesions into the ATP-binding site of ftsE in the plasmid pDSW610, a pBAD33 derivative that carries wild-type ftsEX (28, 47). The ftsE(K41Q) mutation was introduced by PCR with the primers P629 (GGCGGATCCATTCGCTTTGAACATGTC) and P657 (CAGCTTCAGGAGGGTACTTTGCCCTGCGCCGGAATGACCG), using pDSW610 as a template. The resulting 147-bp PCR product was purified by agarose gel electrophoresis and used as a megaprimer together with P630 (GCGAAGCTTTTATTCATGGCCCACGCC) to PCR-amplify ftsE from pDSW610. This produced a 684-bp fragment that was digested with PstI and SphI; these sites are internal to the PCR fragment rather than encoded by the primers. Finally, the restriction fragment was ligated into PstI/SphI-digested pDSW610 to create pDSW703 (PBAD::ftsE(K41Q)X). Plasmid pDSW700 (PBAD::ftsE(D162A)X) was constructed similarly except that the primers used to make the megaprimer were P652 (CGGGATCCTCATCGTGAACCTCGTAC) and P630. The resulting megaprimer was used together with P629 to create an ftsE(D162A) PCR fragment that was cloned into PstI/SphI-digested pDSW610. Plasmid pDSW712 (PBAD::ftsE(E163A)X) was constructed using P629 and P720 (CCAGGTTACCAGTCGGTGCGTCCGCCAGCAG) to make the megaprimer, which was then used together with P632 (GCGAAGCTTTTATTCAGGCGTAAAGTGG) to make a longer PCR product that was used to replace the 1,663-bp PstI-HindIII fragment of pDSW610.

Plasmid pDSW609 is derived from pTH18-kr (copy number, ~5) and carries Plac::ftsE-3xHA (30, 47). To construct derivatives that fused a 3x hemagglutinin (HA) tag (3xHA) to ftsE alleles with ATP-binding site mutations, the 565-bp PstI-SphI fragment of pDSW609 was replaced with the corresponding fragment from pDSW700, pDSW703, and pDSW712. A derivative that lacks the 3xHA tag was constructed by PCR amplification of ftsE with its stop codon from pDSW610 using P629 and P630 as primers. The 684-bp product was cut with PstI and HindIII and ligated into the same sites of pDSW609 to create pDSW904 (Plac::ftsE). It should be noted that the 3x HA tag was removed by the PstI-HindIII digestion. A derivative that expresses ftsE and ftsX with their authentic coupling (the first 7 bp of ftsX overlap the 3' end of ftsE) was created by subcloning the 1,663-bp PstI-HindIII fragment carrying ftsE and ftsX from pDSW610 into the same sites of pDSW609 to create pDSW906 (Plac::ftsEX). Derivatives of ftsE that encode the K41Q, D162A, and E163A lesions were created similarly from pDSW703, pDSW700, and pDSW712, respectively. These plasmids are pDSW907 [Plac::ftsE(K41Q)X], pDSW988 [Plac::ftsE(D162A)X], and pDSW989 [Plac::ftsE(E163A)X]. A derivative that expresses only ftsX was created by PCR amplification of ftsX from pDSW610 using the primers P911 (GGCGAATTCAAATAAGCGCGATGCAATCAATC) and P632. The 1,073-bp product was cut with EcoRI and HindIII and then ligated into the same sites of pDSW609 to create pDSW905 (Plac::ftsX).

The phoA fusions used to determine the topology of FtsX were constructed by PCR methods. Briefly, a 'phoA fragment that encodes most of phoA was cloned into pBAD33 to create pDSW438. PhoA fusions were then constructed by PCR amplification of fragments of ftsX of various lengths, using a common 5' primer (P686) and any of several 3' primers (see Table S1 in the supplemental material). PCR products were digested with SacI (cuts in the 5' primer) and either StuI or ScaI (cut in the 3' primers), both of which generate blunt ends. The ftsX' fragments were ligated into pDSW438 that had been digested with SacI and StuI. The 3' primers and resulting fusion sites in FtsX are as follows: P688 (R47), P689 (S69), P690 (S108), P691 (R145), P692 (S187), P693 (R213), P694 (R245), P695 (R270), P696 (R297), P697 (S300), and P698 (E352). A positive-control ftsI'-'phoA fusion was constructed by PCR amplification of the first 81 codons of ftsI; this product was digested with SacI and StuI and then ligated into the same sites of pDSW438 to create pDSW439.

General microscopy methods. Our microscope, camera, and software have been described previously (39, 59). Where indicated, nucleoids were stained with 4',6-diamidino-2-phenylindole from Vector Laboratories (Burlingame, CA) and membranes were stained with FM4-64 [N-(3-triethylammoniumpropyl)-4-(6-(4-(diethyl-amino)phenyl)hexatrienyl) pyridinium dibromide] from Molecular Probes (Eugene, OR). Digital images were imported to the Adobe Photoshop software program for cropping and minor adjustments to brightness and contrast. Final figures were assembled in Canvas. Depending on the experiment, cells were either fixed with formaldehyde-glutaraldehyde (42) or methanol (49) or live cells were immobilized on an agarose pad and photographed without fixation. For microscopy of live cells, slides were prepared by pipetting 40 µl of molten 1% agarose (in water) onto a prewarmed glass slide with two spacers made of laboratory tape. A second glass slide was placed on top to create a thin agarose pad. Once cooled, the two slides were separated and 5 µl of live culture in growth medium was applied to the agarose pad and covered with a no. 1 coverslip.

Western blotting. Western blotting was done essentially as described previously (58). Primary antibodies against FtsE-3xHA and PhoA fusions were monoclonal HA.11 (BabCo, Berkeley, CA) and polyclonal anti-PhoA (5 Prime->3 Prime, Boulder, CO), respectively. Anti-FtsK, anti-FtsQ, anti-FtsI, and anti-FtsN were all polyclonal antibodies raised in rabbits and have been described previously (12, 56, 58, 59). Secondary antibodies were from Pierce (Rockford, IL) and were conjugated to horseradish peroxidase. Blots were developed with chemiluminescent substrates and visualized with an LAS1000 imager from Fuji (Stamford, CT).

Testing effect of FtsEX depletion on levels of Fts membrane proteins. Strain EC1335 was grown in LB0N containing either Glu or Ara as described previously (47), except that the culture volume was increased to 500 ml. After about 5 h, cells in the glucose culture were ~20 µm long while those in the arabinose culture were ~5 µm long. Cells were harvested by centrifugation, resuspended in 2.5 ml of Tris-NaCl buffer (50 mM Tris, 50 mM NaCl, pH 8.0), and stored frozen at –80°C. For lysis, cells were thawed and EDTA and lysozyme were added to final concentrations of 5 mM and 0.1 mg/ml, respectively. Then, Benzonase (Novagen, Madison, WI) and protease inhibitor cocktail set V (Calbiochem, La Jolla, CA) were added, and samples were incubated on ice for 30 min. Lysis was completed by sonication. The lysate was centrifuged for 10 min at 8,000 x g to remove debris. The resulting supernatant was termed the crude extract, and a sample was saved for analysis by Western blotting. The remainder was centrifuged at 100,000 x g for 75 min. The supernatant was saved as the soluble fraction, while the pellet was resuspended in 0.4 ml of Tris-NaCl buffer and considered the membrane fraction. A Bradford assay was used to determine the protein levels in all three fractions, with bovine serum albumin as a standard. Ten µg of protein from each fraction were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis and Western blotting to determine levels of FtsK, FtsQ, FtsI, and FtsN.

Complementation assays. EC1215 transformants harboring pTH18kr-based plasmids (including the positive and negative controls, pDSW906 and TH18-kr) were grown overnight in LB with Kan (LB Kan) containing glucose to reduce leaky expression. To test the efficiency of colony formation, overnight cultures were adjusted to 109 cells/ml and serial dilutions were spotted onto LB and LB0N plates containing Kan and IPTG (LB Kan IPTG and LBDN Kan IPTG). Plates were incubated at 30°C for 18 h before photography. To test the ability of the mutant proteins to support growth and cell division upon a shift to LB0N, log-phase cultures (optical density at 600 nm [OD600], ~0.8) growing in LB Kan were diluted 10-fold into LB Kan IPTG or LB0N Kan IPTG. Growth was monitored by measuring the OD600, and cell morphology was monitored by phase-contrast microscopy. The mean number of cell division events during the first 120 min after the shift to LB0N was calculated under the assumption that growth (mass increase) was exponential using the following relationship: N = log2 {(e[ln(2) x 120 min/G])(Ltime0/Ltime120)}. Here, N is the mean number of division events, ln is the natural logarithm, G is the mass doubling time in minutes, and L is the average cell length.

Localization of FtsE proteins with ATP-binding site lesions. Transformants that produced the FtsE-3xHA derivatives from low-copy-number plasmids were grown in LB or LB0N containing Kan and IPTG to an OD600 of ~0.3. A sample was taken to determine protein levels by Western blotting, and a second sample was fixed with methanol and processed for immunofluorescence microscopy using anti-HA antibody as described previously (47).

Effect of FtsE(D162A) on localization of Fts proteins in wild-type background. Cultures in exponential growth (OD600 of ~0.5) in LB Kan were induced by diluting them 1:50 into LB Kan IPTG. After 3 h of further growth, cells were fixed with formaldehyde-glutaraldehyde and the various green fluorescent protein (GFP)-Fts protein fusions were visualized by fluorescence microscopy.

Ability of mutant FtsE proteins to recruit FtsI and FtsN to septal rings. (i) For GFP-FtsI, these experiments used an FtsEX depletion strain designated EC1882 [{Delta}ftsEX P204::gfp-ftsI/pDSW610(PBAD::ftsEX)] carrying various pTH18-kr derivatives that express alleles of ftsE along with wild-type ftsX under the control of Plac. Cultures were grown overnight in LB Kan Cam Ara. The next morning, cultures were diluted about 500-fold (to achieve a starting OD600 of 0.005) in LB0N Kan Cam with either Glu or Ara, which allows for depletion or maintenance of FtsEX production, respectively. After 2 h of growth, IPTG was added to induce chromosomal gfp-ftsI and plasmid-borne ftsEX constructs. Growth was continued for another 4 to 5 h, by which time the OD600 was ~0.4 and cells depleted of FtsEX were ~30 µm long. Live cells were then immobilized on an agarose pad and photographed. (ii) For FtsN and FtsZ, the ftsEX null strain EC1215 carrying pTH18-kr derivatives was grown overnight in LB Kan with glucose to limit leaky expression from the plasmids. The next morning, cultures were diluted 1:50 into LB Kan IPTG (except that in the case of ftsE(D162A), the IPTG was omitted until the next step). These cultures were grown to midlog phase and then diluted 10-fold into LB Kan IPTG or LB0N Kan IPTG. After 2 to 3 h of further growth, the OD600 was ~0.3, and cells were fixed and processed for immunolocalization of FtsN or FtsZ as described previously (42, 59).

Ability of FtsE and FtsX to localize independently of each other. To localize FtsX-GFP in the absence of FtsE, the ftsEX null strain EC1215 harboring pDSW513 (P204::ftsX-gfp) was grown in LB Kan IPTG to an OD600 of ~0.8, diluted 10-fold into the same medium with or without NaCl, and grown for 2 to 3 h. Fluorescence microscopy was used to visualize GFP in both live and formaldehyde-glutaraldehyde-fixed cells. Localization of FtsE was tested similarly using a plasmid that expresses 3xHA-tagged FtsE. Cells were fixed with methanol, and FtsE was detected with anti-HA antibody (47).

Computer-predicted topology of FtsX. The following programs were used with default parameters to predict the membrane topology of FtsX: TopPred (http://bioweb.pasteur.fr/seqanal/interfaces/toppred.html), TMPred (http://www.ch.embnet.org/software/TMPRED_form.html), HMMTOP (http://www.enzim.hu/hmmtop/html/submit.html), and TMHMM (http://www.cbs.dtu.dk/services/TMHMM-2.0/) (14, 36, 53).

Alkaline phosphatase assay. Strain KS272 is phoA minus and was used as the host strain to determine the topology of FtsX. Overnight cultures grown in LB Cam were subcultured 1:100 into 5 ml of the same medium and grown for 3 h before 70 µl of 20% arabinose solution was added to induce expression of the phoA fusions. Samples were taken 1 h after induction, and alkaline phosphatase activity was determined as described previously (38). All constructs were assayed in three independent experiments in duplicate. Samples were also taken for Western blotting analysis to determine the steady-state level of the PhoA fusion protein.


arrow
RESULTS
 
FtsEX does not transport Fts proteins into the cytoplasmic membrane. We demonstrated previously that several proteins needed for cell division in E. coli localize poorly to the septal ring when cells growing in LB0N medium are depleted of FtsEX (47). Among these proteins are FtsK, FtsQ, FtsI, and FtsN—all of which reside in the cytoplasmic membrane (9, 11, 16, 21). To determine whether insertion of these division proteins into the membrane depends on FtsEX, we used a previously described depletion strain named EC1335 that expresses ftsEX from an arabinose-inducible pBAD plasmid (47). Cells were grown in LB0N containing arabinose or glucose until the latter were filamentous; average cell lengths at time of harvest were ~5 µm and ~20 µm for the arabinose and glucose cultures, respectively. Then, cells were fractionated and examined by Western blotting with polyclonal antibodies against FtsK, FtsQ, FtsI, and FtsN. Depletion of FtsEX did not have any obvious effect on the abundance of any of the four proteins examined (Fig. 1), despite the fact that division had essentially ceased in the depleted cells. This finding argues that FtsEX does not insert any of the proteins we studied into the cytoplasmic membrane.


Figure 1
View larger version (73K):
[in this window]
[in a new window]

  		FIG. 1. Steady-state levels of Fts membrane proteins in cells depleted of FtsEX. The FtsEX depletion strain EC1335 was grown in LB0N Cam and either Ara or Glu for about 5 h, at which time the Glu culture was filamentous. Cells were fractionated, and FtsK, FtsQ, FtsN, or FtsI was detected by Western blotting. Only the relevant portion of each blot is shown. The position of a molecular mass marker and the protein detected are indicated to the left and right, respectively, of each blot. crude, crude extract; sol., soluble fraction; mem., membrane fraction.

Mutant FtsE proteins with lesions in the ATP-binding motifs. Site-directed mutagenesis was used to introduce amino acid substitutions into the Walker A (K41Q) or Walker B (D162A, E163A) ATP-binding motifs (Fig. 2A). Similar lesions have been made in numerous other ABC transporters. Based on this literature, the K41Q and D162A mutant proteins are not expected to bind ATP, while the E163A protein should bind ATP but not hydrolyze (see reference 19 and references therein). However, the literature on ABC transporters is enormous, and not all published reports are consistent with these predictions or each other. We attempted to characterize the ATP-binding properties of our mutant proteins but encountered severe problems with solubility of FtsE, and this line of investigation was abandoned.


Figure 2
View larger version (56K):
[in this window]
[in a new window]

  		FIG. 2. ATP-binding site mutants of ftsE. (A) Cartoon of FtsE. Vertical black bars indicate invariant residues in an alignment of FtsE proteins from E. coli, Salmonella enterica serovar Typhimurium, Yersinia pestis, Vibrio cholerae, Haemophilus influenzae, Pseudomonas aeruginosa, and Pasteurella multocida. Sequences of the Walker A, Walker B, and ABC signature motifs are shown below, with residues targeted for mutagenesis indicated by arrows. (B) Steady-state level of mutant FtsE proteins. Strain EC251 (wild type) transformed with plasmids that direct expression of ftsE-3xHA alleles was induced with 0.1 mM IPTG, and whole cells were analyzed by Western blotting with anti-HA antibody. The location of molecular mass markers is shown at the left. FtsE-3xHA is predicted to be 28.5 kDa. The plasmids used were pDSW609, pDSW890, pDSW891, pDSW892, and pTH18-kr. (C) Complementation test. The {Delta}ftsEX null mutant EC1215 was transformed with low-copy-number Plac::ftsEX plasmids (or controls) that direct production of the indicated proteins. Transformants were grown overnight in LB Kan, adjusted to 109 CFU/ml in LB0N, and 10-fold serial dilutions were spotted onto LB0N Kan with 0.1 mM IPTG. Plates were incubated overnight at 37°C and then photographed. The plating efficiency was ~100% for all strains when plates contained LB rather than LB0N (not shown). The plasmids used were pTH18-kr, pDSW904, pDSW905, pDSW906, pDSW907, pDSW988, and pDSW989.

The mutant ftsE alleles were fused to three tandem copies of the hemagglutinin epitope tag (3xHA) and expressed from a low-copy-number plasmid with an IPTG-inducible Plac promoter. We demonstrated previously that wild-type FtsE-3xHA localizes to the septal ring and supports cell division (47). Western blotting of whole-cell extracts with anti-HA antibody revealed that all three mutant FtsE proteins were produced at levels comparable to that of the wild type (Fig. 2B).

Phenotype of ftsE ATP-binding site mutants. We tested the ability of our mutant FtsE proteins to rescue the growth, division, and viability of an ftsEX null mutant in LB0N. To this end, we constructed a set of low-copy-number Plac plasmids that express both ftsE and ftsX, preserving the normal 7-bp overlap of these genes but omitting the 3xHA tag from ftsE. We also constructed a new ftsEX null mutant, named EC1215, that has a markerless deletion extending from codon 6 of ftsE to codon 297 of ftsX. This deletion leaves intact the major promoter for {sigma}32 embedded in the 3' end of ftsX (22). We constructed EC1215 because the mutant used in our previous study, RG60, has a nonexcisable Kanr determinant in ftsE, making it incompatible with our Plac::ftsEX plasmids, which confer Kanr (20, 47). Phenotypically EC1215 is indistinguishable from RG60. By way of clarification, it should be noted that RG60 was originally reported to be rescued by electrolytes but not osmolytes (20). However, Reddy subsequently demonstrated that both electrolytes and osmolytes rescue an independently constructed ftsEX null mutant (44). We reinvestigated this issue and found that RG60 and EC1215 are both rescued by osmolytes (data not shown), so osmotic rescue of E. coli ftsEX null mutants is a general phenomenon.

To test the functionality of the FtsE proteins with ATP-binding site lesions, EC1215 transformants were grown in LB and then transferred to LB0N. The results are summarized in Fig. 3 and Table 2.


Figure 3
View larger version (15K):
[in this window]
[in a new window]

  		FIG. 3. Effect of ATP-binding site lesions in FtsE on growth in LB0N. Cultures of the {Delta}ftsEX null mutant EC1215 carrying various ftsEX plasmids (or controls) were grown to mid-log phase in LB Kan with 0.1 mM IPTG. At time zero, these cultures were diluted into LB0N Kan with 0.1 mM IPTG. The plasmids used were pTH18-kr (black circles), pDSW904 (medium-gray circles), pDSW905 (open circles), pDSW906 (black squares), pDSW907 (dark-gray squares), pDSW988 (light-gray squares), and pDSW989 (open squares).


View this table:
[in this window]
[in a new window]

  		TABLE 2. Growth and division of ftsE mutants in LB and LB0Na

Transferring the null mutant (EC1215/Plac vector) to LB0N caused a sudden and dramatic slowing of growth. Prior to the shift, the mass doubling time was 40 min and the average cell length was ~12 µm. During the first 120 min postshift, the mass doubling time was ~70 min and the average cell length increased to ~17 µm. The modest increase in cell length reflects both the low rate of growth and the leakiness of the division defect, since each cell divided ~1 time during the first 2 h in LB0N. For comparison, the wild type (EC1215/Plac::ftsEWTX) grew with a doubling time of ~40 min in LB0N and divided three times during 2 h. While the K41Q and E163A mutant proteins rescued growth in the sense of mass increase (doubling times, ~50 min), the D162A protein did not (doubling time, ~80 min), and none of the three mutant proteins rescued cell division (~1 division event in 2 h, comparable to results for the empty vector control). As a result, all three transformants (EC1215/Plac::ftsEmutX) became ~16 µm long during the first 2 h after the shift to LB0N, only slightly shorter than the empty vector control (~17 µm). By 5 h postshift, growth had essentially stopped and about half of the cells appeared to be dead as judged by phase-contrast microscopy and 4',6-diamidino-2-phenylindole staining, which revealed many ghost cells and gross defects in nucleoid structure. Consistent with this, the plating efficiency of EC1215/Plac::ftsEmutX on LB0N was ~0.001% (Fig. 2C), and even this value is probably too high, because restreaking revealed that most of the colonies obtained had acquired suppressor mutations. Efforts to find conditions of salt and temperature under which the ATP-binding site mutant proteins rescued colony formation compared to that for an empty vector control were unsuccessful.

FtsE(D162A) is a mild dominant negative that impairs constriction of the septal ring in wild-type E. coli. Although none of the mutant FtsE proteins studied here were strong dominant negatives, the D162A mutant protein impaired cell division modestly when produced in a wild-type background. We therefore transformed the low-copy-number Plac::ftsE(D162A)X plasmid into a set of wild-type strains that harbored chromosomal gfp fusions to various division genes. Production of FtsE(D162A) increased the proportion of cells in the population with septal rings containing the late division proteins FtsL or FtsI from about ~50% to about ~95%, and we readily observed filamentous cells with multiple septal rings (Fig. 4; see Table S2 in the supplemental material). The fraction of cells exhibiting septal GFP-FtsN also increased, in this case from ~40% to ~60%, but long cells with multiple sites of GFP-FtsN localization were not observed because GFP-FtsN suppressed the dominant-negative effect of FtsE(D162A) (compare length of GFP-FtsN to that of the GFP control in Table S2 in the supplemental material). Similar overproduction of wild-type FtsE was not deleterious to division and had little or no effect on Fts protein localization (see Table S2 in the supplemental material). Taken together, these findings imply FtsE(D162A) inhibits division at a step after septal ring assembly.


Figure 4
View larger version (89K):
[in this window]
[in a new window]

  		FIG. 4. Effect of FtsE(D162A) on cell division and septal ring assembly. Wild-type cells expressing both the indicated gfp construct and ftsE(D162A) (together with ftsX) were fixed, and GFP was visualized by fluorescence microscopy. Arrows indicate sites of Fts protein localization. The strains shown are EC437 (gfp-ftsI), EC439 (gfp-ftsL), EC441 (gfp-ftsN), EC449 (ftsZ-gfp), EC450 (zipA-gfp), and EC452 (gfp). All carry plasmid pDSW988 (Plac::ftsE(D162A)X).

FtsE proteins with ATP-binding site lesions localize to septal rings. Our observations concerning the dominant-negative effect of FtsE(D162A) suggested that FtsE uses ATP to promote septal ring constriction rather than septal ring assembly. They also suggested that the mutant protein assembles into the septal ring. If it had blocked division by mislocalizing, it should have sequestered other Fts proteins away from the septal ring, which was not the case.

To examine localization of the mutant FtsE proteins directly, we fused them to a C-terminal 3xHA tag and produced them from a low-copy-number Plac plasmid in both a wild-type and {Delta}ftsEX null mutant background. Because FtsE localization depends upon FtsX (see below), when working with the {Delta}ftsEX strain, we supplied FtsX from a compatible PBAD plasmid. Transformants were grown with inducers in LB and LB0N and then fixed and processed for immunofluorescence microscopy with anti-HA antibody. All three mutant forms of FtsE localized to septal rings in both the wild type (not shown) and a {Delta}ftsEX null mutant (Fig. 5; Table 3) grown with or without NaCl.


Figure 5
View larger version (26K):
[in this window]
[in a new window]

  		FIG. 5. ATP-binding site mutants of FtsE localize to potential division sites. EC1215 ({Delta}ftsEX) cotransformed with plasmids that express ftsX (pDSW621) and the indicated ftsE-3xHA allele (pDSW609, pTH18-kr [(–)], pDSW890, pDSW891, or pDSW892) were grown in LB0N containing antibiotics, Ara (to induce ftsX), and 0.1 mM IPTG (to induce ftsE). Cells were fixed, and FtsE was visualized by immunofluorescence microscopy, with a secondary antibody conjugated to Alexa 488 used to detect FtsE-3xHA. Membranes were stained with FM4-64. Arrows indicate sites of FtsE localization.


View this table:
[in this window]
[in a new window]

  		TABLE 3. Localization of FtsE proteins with a mutant ATP-binding sitea

FtsE ATP-binding site mutants support recruitment of FtsI and FtsN to septal rings. The finding that mutant FtsE proteins localize to the septal ring led us to ask whether they support recruitment of downstream division proteins. To test this hypothesis, we examined recruitment of the last two essential proteins in the recruitment hierarchy, FtsI and FtsN (3, 58), since their presence would imply septal rings are complete.

We lack good enough antibodies against FtsI to localize it by immunofluorescence, so we used a chromosomal gfp-ftsI fusion instead. This, however, was not without its own set of challenges, because in our hands, ftsEX null strains do not tolerate expression of gfp-ftsI, which is surprising given that the gfp-ftsI fusion is functional as assessed by complementation (58). For these reasons, we used an FtsEX depletion background that expresses wild-type ftsEX from an arabinose-inducible PBAD plasmid and gfp-ftsI from an IPTG-inducible chromosomal gene. The ftsE allele to be tested was introduced (together with ftsX) on an IPTG-inducible Plac::ftsEX plasmid. Strains were grown in LB0N containing glucose and IPTG until the Plac empty vector control became filamentous and live cells were visualized by fluorescence microscopy (Fig. 6 and Table 4).


Figure 6
View larger version (52K):
[in this window]
[in a new window]

  		FIG. 6. ATP-binding site mutants of FtsE support recruitment of FtsI and FtsN to septal rings. Left panel set: localization of GFP-FtsI. Cells were fixed and photographed under phase-contrast and fluorescence microscopy. Arrows indicate sites of GFP-FtsI localization. The cells shown are derivatives of EC1882 carrying plasmids pTH18-kr (empty vector), pDSW906 (ftsEX), pDSW904 (ftsE alone), pDSW905 (ftsX alone), pDSW907 [ftsE(K41Q) and ftsX], pDSW988 [ftsE(D162A) and ftsX], and pDSW989 [ftsE(E163A) and ftsX] in rows A to G, respectively. Right panel set: localization of FtsN. Cells were fixed, processed for immunofluorescence microscopy to visualize FtsN, and stained with FM4-64 to visualize membranes. Arrows indicate sites of FtsN localization. The cells shown are derivatives of EC1215 carrying the same plasmids used to study GFP-FtsI recruitment in the left panel set.


View this table:
[in this window]
[in a new window]

  		TABLE 4. Effect of FtsE lesions on septal localization of FtsI, FtsN, and FtsZa

Because the leaky dependence of FtsI on FtsEX makes these results somewhat complex, we will describe them in detail, starting with the positive and negative controls. When the depletion strain was grown in the presence of glucose but a complementing Plac::ftsEX plasmid was induced, cells averaged ~9 µm in length, ~75% exhibited at least one septal ring containing GFP-FtsI, and the frequency of GFP-FtsI rings when normalized to cell length (a proxy for mass) was ~1 per 11 µm. In the negative control, which carried a Plac empty vector, depletion of FtsEX increased the cell length to ~30 µm but only ~45% of these filaments had a septal ring containing GFP-FtsI, and the frequency of GFP-FtsI rings was ~1 per 40 µm. These results are similar to those we observed previously using different strains (47).

When cells depleted of FtsEX contained either the Plac::ftsE or Plac::ftsX plasmid, the results were essentially identical to those observed with an empty Plac vector, namely, filamentous cells with ~1 GFP-FtsI band per 45 µm. This implies that an FtsEX complex is needed for recruitment of downstream proteins to the septal ring. Notably, producing an FtsE ATP-binding site mutant protein together with FtsX largely restored GFP-FtsI localization, implying that FtsE probably does not have to bind and hydrolyze ATP to fulfill its function in septal ring assembly. For example, comparing Plac::ftsX to any of the three Plac::ftsEmutX plasmids showed that an FtsE protein with an ATP-binding site lesion increased the fraction of filaments with at least one GFP-FtsI ring from ~40% to ~80% and increased the frequency of rings from ~1 per 45 µm to ~1 per 15 µm. This two- to threefold improvement in GFP-FtsI localization was not associated with a comparable improvement in division, since the filaments averaged ~30 µm and ~26 µm with Plac::ftsX and Plac::ftsEmutX, respectively. Although the ring spacing with Plac::ftsEmutX was not as good as that observed with Plac::ftsEX (15 µm versus 11 µm), we think this modest defect is an indirect effect of filamentation (see below).

To study FtsN localization, we transformed our set of Plac plasmids into an ftsEX null strain and visualized the endogenous FtsN protein by immunofluorescence microscopy. Filaments obtained from the null mutant in this experiment were shorter than those obtained from the depletion strain used for GFP-FtsI because the null mutant became sick and grew poorly after the shift to LB0N (Fig. 2) (47); sickness was especially an issue when the null mutant expressed ftsE(D162A), which had to be induced for a shorter period of time than the other mutant ftsE alleles. Nevertheless, the results for FtsN were analogous to the results for GFP-FtsI. FtsN localization exhibited a leaky but marked dependence on FtsEX, with ~1 band per 9 µm in the positive control (Plac::ftsEX) compared to ~1 per 35 µm in the negative control (Plac vector). Neither FtsE alone nor FtsX alone rescued localization of FtsN, but any of the three ATP-binding site mutants of FtsE could substitute for the wild type when coproduced with FtsX. Thus, when only FtsX was produced from the Plac plasmid, ~40% of the filaments exhibited at least one site of FtsN localization and the frequency of FtsN bands normalized to cell length was ~1 per 35 µm. In contrast, when a mutant FtsE protein was produced together with FtsX, FtsN localization was observed in ~70% of the filaments and the spacing of FtsN bands was ~1 per 17 µm. By either metric, this is roughly a twofold improvement in FtsN localization. Nevertheless, cell length was about the same, 18 µm for FtsX versus ~15 µm for FtsEmutX.

As a further control during the FtsN experiments, a portion of the filaments was used for immunolocalization of FtsZ, which does not depend on FtsEX (47). Although FtsZ rings were readily detected regardless of the FtsE or FtsX content of the cells, there were only about half as many FtsZ rings per unit cell length in filaments as in shorter cells from a population that was dividing (Table 4). For example, when cells carried the Plac::ftsEWTX plasmid, they averaged 4.5 µm in length and the ring spacing was ~1 per 5 µm. In all other cases, the cells were about 15 µm long and ring spacing increased to ~1 per 10 µm. The twofold reduction in the frequency of FtsZ rings in filaments was expected because we observed something similar for FtsA and ZipA in our previous study of FtsEX (47) and because many reports have documented that depletion or inactivation of late division proteins has a modest deleterious effect on FtsZ rings (2, 12, 26, 39, 58). This observation may be relevant to our finding that GFP-FtsI and FtsN rings were spaced farther apart in filaments than in dividing cells. While this finding could be interpreted to mean that mutant FtsE proteins are not fully competent for recruitment of late division proteins, we suspect it is a secondary consequence of filamentation, as evidenced by the fact that we observe a similar change for FtsZ, which is not FtsEX dependent.

FtsX localizes without FtsE. To determine whether FtsE and FtsX could localize to the septal ring independently of each other, we transformed the ftsEX null mutant EC1215 with plasmids that expressed either ftsE-3xHA or ftsX-gfp. FtsX-GFP localized even in cells grown in LB0N (Fig. 7; also see Table S3 in the supplemental material). Localization was better in live cells than in fixed cells. In contrast, we saw few, if any, convincing instances of localization of FtsE-3xHA in cells that lacked FtsX (see Table S3 in the supplemental material), in agreement with a recent report that GFP-FtsE and FLAG-FtsE require FtsX for localization (15).


Figure 7
View larger version (22K):
[in this window]
[in a new window]

  		FIG. 7. FtsX-GFP localizes in cells that lack FtsE. Fluorescence microscopy was used to image FtsX-GFP in live and fixed cells of EC251 (wild type) or EC1215 ({Delta}ftsEX) harboring pDSW513 (P204::ftsX-gfp). Arrows indicate sites of FtsX-GFP localization.

FtsX has only four TM segments. If FtsEX serves as a transporter, then the transmembrane (TM) segments in FtsX would be expected to form a substrate channel. We predicted the topology of FtsX with four publicly available computer algorithms using default parameters. All four programs predicted FtsX to contain four TM segments, with the N terminus and C terminus in the cytoplasm and a large periplasmic loop (~160 amino acids) between TM1 and TM2 (Fig. 8A). To determine the number and location of all TM segments experimentally, we constructed a set of eleven ftsX'-'phoA fusions, using the predicted structure as a guide. The alkaline phosphatase activity of each fusion was measured and compared to that of a positive control ftsI'-'phoA fusion known to place the PhoA moiety in the periplasm (9). Three independent experiments showed the reporter strain harboring the ftsI'-'phoA fusion had an alkaline phosphatase activity of approximately 125 Miller units. Six of the eleven ftsX'-'phoA fusions (listed as 3, 4, 5, 6, 9, and 10 in Fig. 8A) displayed alkaline phosphatase activity higher than that of the positive control (166 to 195 Miller units). In contrast, the other five ftsX'-'phoA fusions and the empty vector control exhibited <12 Miller units of activity. The results were fully consistent with the predicted topology.


Figure 8
View larger version (65K):
[in this window]
[in a new window]

  		FIG. 8. Topology of FtsX. (A) Model of FtsX. Arrowheads and black residues indicate junction points of FtsX'-'PhoA fusions. Arrowheads are colored based on their activity as either high (>125 units; black) or low (<12 units; white). Arrows with circled "+" or "–" symbols indicate charged residues predicted by some computer topology programs to be within the membrane. Shaded residues are highly conserved (≥80% identity) in an alignment of FtsX proteins from E. coli, Photorhabdus luminescens, Xylella fastidiosia, V. cholerae, H. influenzae, P. aeruginosa, Shewanella oneidensis, Photobacterium profundum, P. multocida, Xanthomonas axonopodis, and Neisseria meningitidis. (B) Western blot of FtsX'-'PhoA fusions. Lane numbers correspond to the fusions in panel A. "(–)" is a strain with the 'phoA fusion vector, which lacks a translational start site. "(+)" is an ftsI'-'phoA fusion. Molecular mass markers are indicated to the left of the blot.

Western blotting with anti-PhoA antibody confirmed that most of the fusions were expressed at comparable levels (Fig. 8B). The only exceptions were the two earliest fusions, both of which were poorly expressed and had very little PhoA activity. Thus, the low PhoA activities of these two fusions cannot be used to infer topology. Nevertheless, the N terminus of FtsX can be reliably localized to the cytoplasm, because we demonstrated previously that a GFP-FtsX fusion supports division and is fluorescent (47). GFP exported to the periplasm by the Sec system does not fold properly and fails to fluoresce (23).

To our knowledge, the TMs of most transporters contain charged amino acids that interact with charges on the substrate(s). Our model for FtsX does not show any charged residues in the membrane, but four charged residues (R223, E292, D321, and E322) are shown adjacent to the lipid bilayer on the periplasmic side, and some predictions placed some of these residues in the membrane. None of these residues is invariant among FtsX proteins from different bacterial species (Fig. 8A), making it unlikely that they play a critical role in substrate transport. To verify this inference, each charged residue was changed to alanine by site-directed mutagenesis. All four alanine substitution derivatives complemented as well as the wild type (data not shown). It is unlikely that any of these residues is in the membrane, but even if they are, none is important for FtsEX function.


arrow
DISCUSSION
 
Our previous work demonstrated that FtsEX is a component of the septal ring and improves septal ring assembly, especially in media of low osmotic strength, such as LB0N (47). In particular, we showed that when cells are depleted of FtsEX, the "early" proteins in the dependency hierarchy for septal ring assembly, such as FtsZ, FtsA, and ZipA, still localize well. In contrast, the "late" proteins localize very poorly. That study left open a number of important questions, including whether FtsEX is a transporter, the mechanism(s) by which FtsEX improves septal ring assembly, and the specific contributions of FtsE and FtsX to this process. Several of the findings reported here have implications for those questions.

Evidence that FtsE uses ATP to promote septal ring constriction. We constructed and characterized three mutant FtsE proteins with lesions in the predicted ATP-binding site. These lesions had little or no effect on the ability of FtsE to localize to septal rings and recruit FtsI and FtsN. Because these are the last two (known) essential proteins in the recruitment hierarchy (3, 58), we infer that sepal rings assembled with FtsE ATP-binding site mutant proteins contain a full complement of essential division proteins. Nevertheless, the mutant ftsE alleles did not effectively rescue division, and filamentous cells with multiple septal rings were readily observed. Taken together, these findings imply FtsE has to handle ATP to support cell division. More intriguing, the ATP requirement appears to be for septal ring constriction rather than septal ring assembly. At the very least, ATP appears to be more important for constriction than assembly. Consistent with this interpretation, we also found one of the mutant FtsE proteins, FtsE(D162A), is a mild dominant negative that impaired division by interfering with septal ring constriction rather than septal ring assembly.

In a previous article, we suggested that FtsX might serve as a membrane anchor while FtsE uses ATP hydrolysis to improve constriction of the septal ring (47). In the meantime, Corbin and coworkers have shown that FtsE interacts directly with FtsZ (15), a tubulin-like protein that uses energy from GTP hydrolysis to drive cytokinesis (41). While evidence that FtsEX facilitates septal ring constriction dovetails nicely with the FtsE-FtsZ interaction, it is important to note that neither we (this study) nor Margolin's group have been able to show that FtsE localizes to division sites in the absence of FtsX, despite some effort. This is puzzling, because the FtsE-FtsZ interaction was strong (it stood up to coprecipitation) and independent of FtsX (15).

Two caveats concerning our inference that ATP is needed for constriction rather than assembly of the septal ring deserve mention. First, we have not investigated whether the mutant FtsE proteins support recruitment of several nonessential proteins to the septal ring (6, 7, 25, 32, 50). Although failure to recruit any one of these proteins would not account for the division defects reported here, failure to recruit several of them might. Second, we were unable to assay the ability of FtsE and its mutant derivatives to bind or hydrolyze ATP. Contrary to a previous report (20), in our hands His6-tagged FtsE proteins were insoluble, and even the wild type failed to bind ATP reproducibly in the presence (or absence) of urea and other solubilization reagents. Thus, the mutant proteins studied here might retain some residual ability to bind and hydrolyze ATP.

Implications for the role of FtsEX in septal ring assembly. We found that depletion of FtsEX to the point that cells could no longer divide had no noticeable effect on the amounts of the late proteins FtsK, FtsQ, FtsI, or FtsN in the membrane fraction. This result all but rules out the hypothesis that FtsEX serves as a specialized ABC transporter that helps to insert some of the late Fts proteins into the cytoplasmic membrane. We favor the hypothesis that FtsEX improves septal ring assembly by engaging in protein-protein interactions. Three interactions involving FtsEX have been reported: FtsX-FtsA, FtsX-FtsQ, and FtsE-FtsZ (15, 31). Here we showed that FtsX localized reasonably well in the absence of FtsE, but localization of FtsE was not observed unless FtsX was also present. Together, these findings imply that FtsX plays a more important role than FtsE in targeting FtsEX to the septal ring, and they underscore the potential significance of the FtsX-FtsA interaction (31). We also showed that both FtsE and FtsX had to be present to facilitate recruitment of FtsI or FtsN to septal rings. This implies an FtsEX complex is needed for recruitment of late proteins despite the fact that FtsX localizes in the absence of FtsE and may bind to FtsQ (31). The FtsE requirement could mean that FtsE interacts with late proteins directly, but using bacterial two-hybrid systems, we have yet to find any interactions involving FtsE other than those expected of an ABC transporter, namely, FtsE-FtsE and FtsE-FtsX (S. J. R. Arends and D. S. Weiss, unpublished). Alternatively, FtsE might improve the ability of FtsX to interact with late proteins by improving dimerization of FtsX or modulating the conformation of FtsX's periplasmic domain. Finally, we note that the relatively long time lag that separates FtsZ ring formation from recruitment of the late proteins (1) suggests that recruitment of the late proteins is triggered by something the FtsZ ring does, in which case FtsEX could improve septal ring assembly by stimulating FtsZ ring function rather than interacting with late division proteins directly.

Membrane topology of FtsX. Any substrate that FtsEX moves across the cytoplasmic membrane would presumably pass through a channel formed by a dimer of FtsX. Here we showed that FtsX has only four TM segments, which would provide for a total of eight TM segments in an FtsE2X2 complex. Other noteworthy features of the topology are the absence of charged amino acids in the TM segments and the presence of a large periplasmic loop between TM1 and TM2. An independent study arrived at a similar topology model for FtsX based on a combination of computer predictions, one GFP fusion and one PhoA fusion (17). The structure of FtsX seems most compatible with a substrate that is small and uncharged. But in our view, the structure of FtsX is not very suggestive of a transporter at all, and we wonder whether FtsE uses ATP hydrolysis to drive an event in the periplasm, where FtsX has a conspicuously large domain. If so, the TM segments of FtsX would transmit a conformation change rather than forming a substrate channel.

Support for this hypothesis comes from consideration of the two ABC transporters most closely related to FtsEX. These are MacB and LolCDE, both of which have only four TM segments per integral membrane domain and assemble into ABC transporters with a total of eight TM segments (8, 17, 34). LolC and LolD work together with the cytoplasmic ATPase LolE to transfer lipoproteins from the outer surface of the cytoplasmic membrane to the periplasmic chaperone LolA (52, 60). This activity does not involve moving a substrate across a membrane, and the TM segments of LolC and LolD serve to transmit ATP-driven conformation changes rather than forming a substrate channel. The role of the TM segments in MacB is less clear. On one hand, MacB helps to secrete heat-stable enterotoxin II from the periplasm across the outer membrane (61). Enterotoxin II crosses the outer membrane via TolC rather than MacB, so the TM segments of MacB do not function as a substrate channel in this process. On the other hand, MacB also affords weak protection against macrolide antibiotics. These drugs might cross the cytoplasmic membrane via a channel in MacB, but there are unresolved discrepancies among the various studies, and the actual role of MacB in this process is obscure (35, 37, 51).


arrow
ACKNOWLEDGMENTS
 
We thank Joe Lutkenhaus for anti-FtsK, Andrew Hawkins for a phoA vector, L. Paolozzi, G. Karimova, and D. Ladant for bacterial two-hybrid plasmids and strains, Bruce Ayati for the formula used to calculate division events for mutants in LB0N, and Kari Schmidt and Jennifer Wendt for constructing some of the strains and plasmids used in this study. We are grateful to members of the Weiss laboratory for helpful discussions.

This work was supported in part by a grant from the National Institutes of Health (GM59893) to D.S.W. and by funds from the Department of Microbiology. S.J.R.A. was supported by an NIH Training Grant in Biotechnology (T32 GM08365-13). The DNA Facility is supported through the Holden Comprehensive Cancer Center's Cancer Center Support Grant 2 P30 CA086862 from the National Cancer Institute/NIH and the Carver College of Medicine, University of Iowa.


arrow
FOOTNOTES
 
* Corresponding author. Mailing address: Department of Microbiology, University of Iowa, 3-403 BSB, Iowa City, IA 52242. Phone: (319) 335-7785. Fax: (319) 335-9006. E-mail: david-weiss{at}uiowa.edu Back

{triangledown} Published ahead of print on 17 April 2009. Back

{dagger} Supplemental material for this article may be found at http://jb.asm.org/. Back


arrow
REFERENCES
 
    1
  1. Aarsman, M. E., A. Piette, C. Fraipont, T. M. Vinkenvleugel, M. Nguyen-Disteche, and T. den Blaauwen. 2005. Maturation of the Escherichia coli divisome occurs in two steps. Mol. Microbiol. 55:1631-1645.[CrossRef][Medline]
    2. 2
  2. Addinall, S. G., E. Bi, and J. Lutkenhaus. 1996. FtsZ ring formation in fts mutants. J. Bacteriol. 178:3877-3884.[Abstract/Free Full Text]
    3. 3
  3. Addinall, S. G., C. Cao, and J. Lutkenhaus. 1997. FtsN, a late recruit to the septum in Escherichia coli. Mol. Microbiol. 25:303-309.[CrossRef][Medline]
    4. 4
  4. Arends, S. J. R., K. B. Williams, R. J. Kustusch, and D. S. Weiss. 2007. Cell division, p. 173-197. In M. Ehrmann (ed.), The periplasm. ASM Press, Washington, DC.
    5. 5
  5. Ausubel, F. A., R. Brent, R. E. Kingston, D. D. Moore, J. G. Seidman, J. A. Smith, and K. Struhl. 1998. Current protocols in molecular biology. John Wiley & Sons, New York, NY.
    6. 6
  6. Bernhardt, T. G., and P. A. de Boer. 2003. The Escherichia coli amidase AmiC is a periplasmic septal ring component exported via the twin-arginine transport pathway. Mol. Microbiol. 48:1171-1182.[CrossRef][Medline]
    7. 7
  7. Bernhardt, T. G., and P. A. de Boer. 2004. Screening for synthetic lethal mutants in Escherichia coli and identification of EnvC (YibP) as a periplasmic septal ring factor with murein hydrolase activity. Mol. Microbiol. 52:1255-1269.[CrossRef][Medline]
    8. 8
  8. Bouige, P., D. Laurent, L. Piloyan, and E. Dassa. 2002. Phylogenetic and functional classification of ATP-binding cassette (ABC) systems. Curr. Protein Pept. Sci. 3:541-559.[CrossRef][Medline]
    9. 9
  9. Bowler, L. D., and B. G. Spratt. 1989. Membrane topology of penicillin-binding protein 3 of Escherichia coli. Mol. Microbiol. 3:1277-1286.[CrossRef][Medline]
    10. 10
  10. Boyd, D., D. S. Weiss, J. C. Chen, and J. Beckwith. 2000. Towards single-copy gene expression systems making gene cloning physiologically relevant: lambda InCh, a simple Escherichia coli plasmid-chromosome shuttle system. J. Bacteriol. 182:842-847.[Abstract/Free Full Text]
    11. 11
  11. Carson, M. J., J. Barondess, and J. Beckwith. 1991. The FtsQ protein of Escherichia coli: membrane topology, abundance, and cell division phenotypes due to overproduction and insertion mutations. J. Bacteriol. 173:2187-2195.[Abstract/Free Full Text]
    12. 12
  12. Chen, J. C., D. S. Weiss, J. M. Ghigo, and J. Beckwith. 1999. Septal localization of FtsQ, an essential cell division protein in Escherichia coli. J. Bacteriol. 181:521-530.[Abstract/Free Full Text]
    13. 13
  13. Cherepanov, P. P., and W. Wackernagel. 1995. Gene disruption in Escherichia coli: TcR and KmR cassettes with the option of Flp-catalyzed excision of the antibiotic-resistance determinant. Gene 158:9-14.[CrossRef][Medline]
    14. 14
  14. Claros, M. G., and G. von Heijne. 1994. TopPred II: an improved software for membrane protein structure predictions. Comput. Appl. Biosci. 10:685-686.[Free Full Text]
    15. 15
  15. Corbin, B. D., Y. Wang, T. K. Beuria, and W. Margolin. 2007. Interaction between cell division proteins FtsE and FtsZ. J. Bacteriol. 189:3026-3035.[Abstract/Free Full Text]
    16. 16
  16. Dai, K., Y. Xu, and J. Lutkenhaus. 1996. Topological characterization of the essential Escherichia coli cell division protein FtsN. J. Bacteriol. 178:1328-1334.[Abstract/Free Full Text]
    17. 17
  17. Daley, D. O., M. Rapp, E. Granseth, K. Melen, D. Drew, and G. von Heijne. 2005. Global topology analysis of the Escherichia coli inner membrane proteome. Science 308:1321-1323.[Abstract/Free Full Text]
    18. 18
  18. Datsenko, K. A., and B. L. Wanner. 2000. One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc. Natl. Acad. Sci. USA 97:6640-6645.[Abstract/Free Full Text]
    19. 19
  19. Davidson, A. L., E. Dassa, C. Orelle, and J. Chen. 2008. Structure, function, and evolution of bacterial ATP-binding cassette systems. Microbiol. Mol. Biol. Rev. 72:317-364.[Abstract/Free Full Text]
    20. 20
  20. de Leeuw, E., B. Graham, G. J. Phillips, C. M. ten Hagen-Jongman, B. Oudega, and J. Luirink. 1999. Molecular characterization of Escherichia coli FtsE and FtsX. Mol. Microbiol. 31:983-993.[CrossRef][Medline]
    21. 21
  21. Dorazi, R., and S. J. Dewar. 2000. Membrane topology of the N-terminus of the Escherichia coli FtsK division protein. FEBS Lett. 478:13-18.[CrossRef][Medline]
    22. 22
  22. Erickson, J. W., V. Vaughn, W. A. Walter, F. C. Neidhardt, and C. A. Gross. 1987. Regulation of the promoters and transcripts of rpoH, the Escherichia coli heat shock regulatory gene. Genes Dev. 1:419-432.[Abstract/Free Full Text]
    23. 23
  23. Feilmeier, B. J., G. Iseminger, D. Schroeder, H. Webber, and G. J. Phillips. 2000. Green fluorescent protein functions as a reporter for protein localization in Escherichia coli. J. Bacteriol. 182:4068-4076.[Abstract/Free Full Text]
    24. 24
  24. Garti-Levi, S., R. Hazan, J. Kain, M. Fujita, and S. Ben-Yehuda. 2008. The FtsEX ABC transporter directs cellular differentiation in Bacillus subtilis. Mol. Microbiol. 69:1018-1028.[CrossRef][Medline]
    25. 25
  25. Gerding, M. A., Y. Ogata, N. D. Pecora, H. Niki, and P. A. de Boer. 2007. The trans-envelope Tol-Pal complex is part of the cell division machinery and required for proper outer-membrane invagination during cell constriction in E. coli. Mol. Microbiol. 63:1008-1025.[CrossRef][Medline]
    26. 26
  26. Ghigo, J. M., D. S. Weiss, J. C. Chen, J. C. Yarrow, and J. Beckwith. 1999. Localization of FtsL to the Escherichia coli septal ring. Mol. Microbiol. 31:725-737.[CrossRef][Medline]
    27. 27
  27. Goehring, N. W., and J. Beckwith. 2005. Diverse paths to midcell: assembly of the bacterial cell division machinery. Curr. Biol. 15:R514-R526.[CrossRef][Medline]
    28. 28
  28. Guzman, L. M., D. Belin, M. J. Carson, and J. Beckwith. 1995. Tight regulation, modulation, and high-level expression by vectors containing the arabinose PBAD promoter. J. Bacteriol. 177:4121-4130.[Abstract/Free Full Text]
    29. 29
  29. Haldimann, A., and B. L. Wanner. 2001. Conditional-replication, integration, excision, and retrieval plasmid-host systems for gene structure-function studies of bacteria. J. Bacteriol. 183:6384-6393.[Abstract/Free Full Text]
    30. 30
  30. Hashimoto-Gotoh, T., M. Yamaguchi, K. Yasojima, A. Tsujimura, Y. Wakabayashi, and Y. Watanabe. 2000. A set of temperature sensitive-replication/-segregation and temperature resistant plasmid vectors with different copy numbers and in an isogenic background. Gene 241:185-191.[CrossRef][Medline]
    31. 31
  31. Karimova, G., N. Dautin, and D. Ladant. 2005. Interaction network among Escherichia coli membrane proteins involved in cell division as revealed by bacterial two-hybrid analysis. J. Bacteriol. 187:2233-2243.[Abstract/Free Full Text]
    32. 32
  32. Karimova, G., C. Robichon, and D. Ladant. 2009. Characterization of YmgF, a 72-residue inner membrane protein that associates with the Escherichia coli cell division machinery. J. Bacteriol. 191:333-346.[Abstract/Free Full Text]
    33. 33
  33. Kempf, M. J., and M. J. McBride. 2000. Transposon insertions in the Flavobacterium johnsoniae ftsX gene disrupt gliding motility and cell division. J. Bacteriol. 182:1671-1679.[Abstract/Free Full Text]
    34. 34
  34. Kobayashi, N., K. Nishino, T. Hirata, and A. Yamaguchi. 2003. Membrane topology of ABC-type macrolide antibiotic exporter MacB in Escherichia coli. FEBS Lett. 546:241-246.[CrossRef][Medline]
    35. 35
  35. Kobayashi, N., K. Nishino, and A. Yamaguchi. 2001. Novel macrolide-specific ABC-type efflux transporter in Escherichia coli. J. Bacteriol. 183:5639-5644.[Abstract/Free Full Text]
    36. 36
  36. Krogh, A., B. Larsson, G. von Heijne, and E. L. Sonnhammer. 2001. Predicting transmembrane protein topology with a hidden Markov model: application to complete genomes. J. Mol. Biol. 305:567-580.[CrossRef][Medline]
    37. 37
  37. Lin, H. T., V. N. Bavro, N. P. Barrera, H. M. Frankish, S. Velamakanni, H. W. van Veen, C. V. Robinson, M. I. Borges-Walmsley, and A. R. Walmsley. 2009. MacB ABC transporter is a dimer whose ATPase activity and macrolide-binding capacity are regulated by the membrane fusion protein MacA. J. Biol. Chem. 284:1145-1154.[Abstract/Free Full Text]
    38. 38
  38. Manoil, C. 1991. Analysis of membrane protein topology using alkaline phosphatase and beta-galactosidase gene fusions. Methods Cell Biol. 34:61-75.[Medline]
    39. 39
  39. Mercer, K. L., and D. S. Weiss. 2002. The Escherichia coli cell division protein FtsW is required to recruit its cognate transpeptidase, FtsI (PBP3), to the division site. J. Bacteriol. 184:904-912.[Abstract/Free Full Text]
    40. 40
  40. Merino, S., M. Altarriba, R. Gavin, L. Izquierdo, and J. M. Tomas. 2001. The cell division genes (ftsE and X) of Aeromonas hydrophila and their relationship with opsonophagocytosis. FEMS Microbiol. Lett. 198:183-188.[CrossRef][Medline]
    41. 41
  41. Osawa, M., D. E. Anderson, and H. P. Erickson. 2008. Reconstitution of contractile FtsZ rings in liposomes. Science 320:792-794.[Abstract/Free Full Text]
    42. 42
  42. Pogliano, J., K. Pogliano, D. S. Weiss, R. Losick, and J. Beckwith. 1997. Inactivation of FtsI inhibits constriction of the FtsZ cytokinetic ring and delays the assembly of FtsZ rings at potential division sites. Proc. Natl. Acad. Sci. USA 94:559-564.[Abstract/Free Full Text]
    43. 43
  43. Ramirez-Arcos, S., H. Salimnia, I. Bergevin, M. Paradis, and J. A. Dillon. 2001. Expression of Neisseria gonorrhoeae cell division genes ftsZ, ftsE and minD is influenced by environmental conditions. Res. Microbiol. 152:781-791.[Medline]
    44. 44
  44. Reddy, M. 2007. Role of FtsEX in cell division of Escherichia coli: viability of ftsEX mutants is dependent on functional SufI or high osmotic strength. J. Bacteriol. 189:98-108.[Abstract/Free Full Text]
    45. 45
  45. Salmond, G. P., and S. Plakidou. 1984. Genetic analysis of essential genes in the ftsE region of the Escherichia coli genetic map and identification of a new cell division gene, ftsS. Mol. Gen. Genet. 197:304-308.[CrossRef][Medline]
    46. 46
  46. Sarkar, G., and S. S. Sommer. 1990. The "megaprimer" method of site-directed mutagenesis. BioTechniques 8:404-407.[Medline]
    47. 47
  47. Schmidt, K. L., N. D. Peterson, R. J. Kustusch, M. C. Wissel, B. Graham, G. J. Phillips, and D. S. Weiss. 2004. A predicted ABC transporter, FtsEX, is needed for cell division in Escherichia coli. J. Bacteriol. 186:785-793.[Abstract/Free Full Text]
    48. 48
  48. Strauch, K. L., and J. Beckwith. 1988. An Escherichia coli mutation preventing degradation of abnormal periplasmic proteins. Proc. Natl. Acad. Sci. USA 85:1576-1580.[Abstract/Free Full Text]
    49. 49
  49. Sun, Q., X. C. Yu, and W. Margolin. 1998. Assembly of the FtsZ ring at the central division site in the absence of the chromosome. Mol. Microbiol. 29:491-503.[CrossRef][Medline]
    50. 50
  50. Tarry, M., S. J. R. Arends, P. Roversi, E. Piette, F. Sargent, B. C. Berks, D. S. Weiss, and S. M. Lea. 2009. The Escherichia coli cell division protein and model Tat substrate SufI (FtsP) localizes to the septal ring and has a multicopper oxidase-like structure. J. Mol. Biol. 386:504-519.[CrossRef][Medline]
    51. 51
  51. Tikhonova, E. B., V. K. Devroy, S. Y. Lau, and H. I. Zgurskaya. 2007. Reconstitution of the Escherichia coli macrolide transporter: the periplasmic membrane fusion protein MacA stimulates the ATPase activity of MacB. Mol. Microbiol. 63:895-910.[Medline]
    52. 52
  52. Tokuda, H., and S. Matsuyama. 2004. Sorting of lipoproteins to the outer membrane in E. coli. Biochim. Biophys. Acta 1693:5-13.[Medline]
    53. 53
  53. Tusnady, G. E., and I. Simon. 2001. The HMMTOP transmembrane topology prediction server. Bioinformatics 17:849-850.[Abstract/Free Full Text]
    54. 54
  54. Ukai, H., H. Matsuzawa, K. Ito, M. Yamada, and A. Nishimura. 1998. ftsE(Ts) affects translocation of K+-pump proteins into the cytoplasmic membrane of Escherichia coli. J. Bacteriol. 180:3663-3670.[Abstract/Free Full Text]
    55. 55
  55. Vicente, M., A. I. Rico, R. Martinez-Arteaga, and J. Mingorance. 2006. Septum enlightenment: assembly of bacterial division proteins. J. Bacteriol. 188:19-27.[Free Full Text]
    56. 56
  56. Wang, L., and J. Lutkenhaus. 1998. FtsK is an essential cell division protein that is localized to the septum and induced as part of the SOS response. Mol. Microbiol. 29:731-740.[CrossRef][Medline]
    57. 57
  57. Weiss, D. S. 2004. Bacterial cell division and the septal ring. Mol. Microbiol. 54:588-597.[CrossRef][Medline]
    58. 58
  58. Weiss, D. S., J. C. Chen, J. M. Ghigo, D. Boyd, and J. Beckwith. 1999. Localization of FtsI (PBP3) to the septal ring requires its membrane anchor, the Z ring, FtsA, FtsQ, and FtsL. J. Bacteriol. 181:508-520.[Abstract/Free Full Text]
    59. 59
  59. Wissel, M. C., and D. S. Weiss. 2004. Genetic analysis of the cell division protein FtsI (PBP3): amino acid substitutions that impair septal localization of FtsI and recruitment of FtsN. J. Bacteriol. 186:490-502.[Abstract/Free Full Text]
    60. 60
  60. Yakushi, T., K. Masuda, S. Narita, S. Matsuyama, and H. Tokuda. 2000. A new ABC transporter mediating the detachment of lipid-modified proteins from membranes. Nat. Cell Biol. 2:212-218.[CrossRef][Medline]
    61. 61
  61. Yamanaka, H., H. Kobayashi, E. Takahashi, and K. Okamoto. 2008. MacAB is involved in the secretion of Escherichia coli heat-stable enterotoxin II. J. Bacteriol. 190:7693-7698.[Abstract/Free Full Text]

Journal of Bacteriology, June 2009, p. 3772-3784, Vol. 191, No. 12
0021-9193/09/$08.00+0     doi:10.1128/JB.00179-09
Copyright © 2009, American Society for Microbiology. All Rights Reserved.



This article has been cited by other articles:

  • Arends, S. J. R., Williams, K., Scott, R. J., Rolong, S., Popham, D. L., Weiss, D. S. (2010). Discovery and Characterization of Three New Escherichia coli Septal Ring Proteins That Contain a SPOR Domain: DamX, DedD, and RlpA. J. Bacteriol. 192: 242-255 [Abstract] [Full Text]  

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental material
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow E-mail this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Arends, S. J. R.
Right arrow Articles by Weiss, D. S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Arends, S. J. R.
Right arrow Articles by Weiss, D. S.

Copyright © 2010 by the American Society for Microbiology.   For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»