Cell Division Defects in Escherichia coli Deficient in the Multidrug Efflux Transporter AcrEF-TolC

Strains, growth conditions, and drug susceptibility experiments.

The bacterial strains and plasmids used in this study are listed in Table 1. E. coli strains were grown at 37°C in Luria-Bertani (LB) medium (10 g of Bactotryptone, 10 g of yeast extract, and 5 g of NaCl per liter). Antibiotics were added when needed to the following final concentration: ampicillin (100 μg/ml), kanamycin (34 μg/ml), spectinomycin (50 μg/ml), tetracycline (25 μg/ml), and chloramphenicol (25 μg/ml).

The MICs of various antimicrobial agents were determined as described before (38).

Plasmid construction.

The pUC-AcrA plasmid was obtained by treatment of pUC151A with DraIII and NsiI restriction enzymes followed by incubations with T4 DNA polymerase and T4 DNA ligase. In pUC-AcrA, the C-terminal 716 amino acid residues of AcrB are deleted. Judging from complementation studies of the drug susceptible phenotype of AG100AX the truncated AcrB is nonfunctional (data not shown). To construct pUC-Cva-AcrA, a 163-bp PCR fragment encoding the transmembrane domain of CvaA was amplified from plasmid pHK11 using the forward primer 5′-GGCCGGCTCGAGGTTTACATATGAAGTGGCAGGGACGG-3′ and the reverse primer 5′-GGCCGGTGGCCACCTTGTTGGGCCTGGCTATAGGTACCAACAATAATG-3′. The PCR fragment was digested with XhoI and MscI restriction enzymes and inserted into plasmid pUC151A treated with the same restriction enzymes. This fragment replaced the coding sequence for the first 28 amino acids of AcrA.

To construct pUC-GFP-AcrA, the gfp gene was amplified from the pQB1T7 plasmid using primers with flanking XhoI restriction sites. The 747-bp PCR fragment was digested with XhoI and inserted into pUC-Cva-AcrA, treated with XhoI and calf intestinal alkaline phosphatase. To construct pACYC-AcrEF, the acrEF operon was amplified by PCR using E. coli K-12 chromosomal DNA as a template. The PCR fragment was treated with the AvaI and HindIII restriction enzymes and ligated into pACYC184 treated with the same enzymes. The 781-bp fragment of acrE was deleted from pACYC-AcrEF by digestion with AleI and ApaLI restriction enzymes followed by treatment with T4 DNA polymerase and T4 DNA ligase. The resulting plasmid, pACYC-AcrF, expressed AcrF, presumably under the native acrEF promoter.

Microscopy.

Freshly transformed bacterial strains were grown to an optical density at 600 nm (OD₆₀₀) of 0.6 to 0.8 in 5 ml of LB medium supplemented with 100 μg/ml of ampicillin.

Phase contrast microscopy.

We mixed 300 μl of cell culture with 15 μl of 25 μg/ml poly-l-lysine. The mixture was spread on coverslips and incubated for 5 min. Excess liquid was removed and the coverslips were rinsed six times in phosphate-buffered saline (PBS) solution (pH 7.3); 5 μl of PBS were spotted onto a slide and the coverslips were placed on the top. Slides were photographed with a Black and White Spot camera (Insight) mounted on an Olympus BX50 microscope through an UPlanFI ×100/1.3 oil-immersion objective.

Fluorescence microscopy.

To visualize chromosomes and membranes, cells were fixed by mixing 400 μl of cell culture with 6.6 ml of ice-chilled PBS/ethanol (75%). Cells were collected by centrifugation, washed once with PBS, and resuspended in 300 μl of PBS. Slides were prepared as described for the phase contrast microscopy. For the last step 5 μl of PBS supplemented with 1x Sypro Orange stain (a nonspecific hydrophilic fluorescent protein dye that does not stain nucleic acids) and 100 nM 4′,6′-diamidino-2-phenylindole (DAPI) (Molecular Probes) were used. For AcrA localization, live exponential-phase bacteria expressing green fluorescent protein (GFP)-CvaA-AcrA fusion protein were spotted onto glass slides and photographed as described above.

Transmission electron microscopy.

Cells were collected directly from LB plates supplemented with 100 μg/ml of ampicillin. Fixation was done as described in (5). Fixed cells enrobed in agar were postfixed with 1% OsO₄ in 0.05 M sodium cacodylate buffer for 1 h at room temperature. After OsO₄ fixation, cells were washed three times with Milli-Q H₂O. Dehydration was done at room temperature in 15-min sequential steps of 30%, 50%, 70%, 80%, and 95% ethanol followed by three 15-min wash in 100% ethanol. The dehydrated cells were embedded in Embed-812 resin and sectioned to a thickness of 50 to 70 nm. The sectioned blocks were stained 10 min with saturated uranyl acetate and 5 min with Sato's lead (13) and observed using JEOL 2000-FX electron microscope in the Samuel Roberts Noble Electron Microscopy Laboratory at the University of Oklahoma.

Sucrose density gradient fractionation.

Cells were grown in 100 ml of LB medium supplemented with ampicillin (100 μg/ml) to an OD₆₀₀ of 0.8 to 1.0 and harvested by centrifugation. The EDTA-lysozyme treatment was performed as described previously (34). The resulting spheroplasts were sonicated (Branson Sonifier 450) for 45 s on ice. After removal of unbroken cells by centrifugation, membrane vesicles were collected by ultracentrifugation in a Beckman 70-Ti rotor at 60,000 rpm for 1 h. The resulting membrane pellet was resuspended in solution of 20% sucrose and 5 mM EDTA. All subsequent steps for membrane preparation were done as described before (20, 39). Membranes separated in the sucrose gradient were collected in 0.1 to 0.2 ml fractions from the bottom of the centrifuge tube. Fractions were mixed with sodium dodecyl sulfate (SDS) sample buffer, boiled for 10 min, analyzed by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (PAGE) and visualized by silver nitrate staining.

Immunoblotting analysis.

Bacterial cells were grown to OD₆₀₀ of 0.6 to 0.8 in 5 ml of LB medium supplemented with 100 μg/ml of ampicillin and collected by centrifugation. Cells were disrupted by sonication in 0.5 ml of 10 mM Tris-HCl (pH 8.0) containing 1 mM EDTA and 1 mM phenylmethylsulfonyl fluoride. Then, total membranes were collected by ultracentrifugation in a Beckman TLA-55 rotor at 40,000 rpm for 40 min. Total membranes were resuspended directly in the SDS-sample buffer for 10% SDS-PAGE analysis. Alternatively, the total membrane pellet was resuspended in 100 μl of PBS solution containing 2% Triton and 10 mM MgCl₂ and left on ice for 30 min followed by another round of ultracentrifugation at 40,000 rpm for 40 min. The supernatant was collected as the inner membrane fraction. Protein concentrations were determined using the DC Protein Assay (Bio-Rad) with bovine serum albumin as a standard. The inner membrane fractions of equal protein amount (3 μg) were loaded onto standard 10% SDS-polyacrylamide gel and immunoblotting was performed by standard techniques. Polyclonal anti-AcrA and anti-AcrB antibodies were used for immunodetection (43).

Bocillin FL binding assay.

The bocillin binding assay was performed as described (45) with the following modifications. Cells were first grown in 5 ml LB medium supplemented with appropriate antibiotics overnight. Overnight cultures were inoculated into 200 ml of fresh medium and allowed to grow until an OD₆₀₀ of ∼1.0, and harvested at 5,000 × g for 20 min. Cell were washed once with 20 mM potassium phosphate buffer (pH 7.5) containing 140 mM NaCl and resuspended in 2 ml of the same buffer supplemented with 1 mM EDTA. Lysozyme was added to the mixture to a final concentration 100 μg/ml, cell were incubated on ice for 15 to 30 min and sonicated 3 times for 20 seconds on ice. Unbroken cells and cell debris were removed by centrifugation at 5,000 × g for 20 min. The supernatant was collected by centrifugation in Beckman TLA-55 rotor at 40,000 rpm for 40 min. The resulting pellet was washed once and resuspended in 0.5 ml of the 20 mM potassium phosphate buffer (pH 7.5) containing 140 mM NaCl. The membrane preparations were used for the bocillin FL binding assay as performed (45).

Reverse transcription-PCR.

Reverse transcription (RT)-PCR was performed by using QIAGEN OneStep RT-PCR, which allows reverse transcription and PCR to be carried out sequentially in the same reaction tube. Total RNA was isolated from exponentially growing cells at OD₆₀₀ of 0.8 to 1.0 and stationary phase cells using QIAGEN RNeasy kit. Purified total RNA (2 μg) was used as a template in RT-PCR. For AcrEF, forward primer GTAATGACGAAACATGCCAGGTTTTTCCTC and reverse primer GATTTATCCTTTAAAGCAACGGCGGATCACC were used to yield a product of 4,280 bp. For SulA, forward primer 5′-GAGTGAATTTTTAGCCCGGAAAGTTGTCTC-3′ and reverse primer 5′-GTACACTTCAGGCTATGCACATCGTTCTTC-3′ were used to yield a product of 482 bp.

Increased expression of AcrA leads to filamentation of E. coli cells deficient in AcrEF

. During studies of the AcrAB multidrug transporter we noticed that transformation of a high-copy number plasmid pUC151A expressing AcrA and AcrB under the native P_acrAB promoter into E. coli strains lacking AcrEF would often proceed with very low efficiency. Phase-contrast microscopy of cells deficient in either AcrEF (W4680E) or both AcrEF and AcrAB (AG100AX) showed that these strains grow as highly filamentous cells (the length of filaments is more than 50 μm) when the expression of AcrAB increased (Fig. 1A). In contrast, the cell morphology of wild-type E. coli cells (AG100) or cells lacking only AcrAB (AG100A) and carrying pUC151A was normal.

To investigate whether the activity of the multidrug transporter AcrB is toxic in acrEF mutants, cells were transformed with plasmids carrying either the periplasmic accessory protein AcrA (pUC-AcrA) or AcrB (pBP) alone. Surprisingly the increased expression of AcrA alone was sufficient to cause cell filamentation in AcrEF null cells (Fig. 1A). No detrimental effect of the AcrA overproduction was found in the wild-type or AcrAB null mutant. The overproduction of AcrB transporter was well tolerated and all cells carrying pBP were morphologically normal.

The filamentous AG100AX/pUC-AcrA cells grew more slowly than normal AG100AX/pUC18 cells with growth rates of 1.03 and 1.56 h⁻¹, respectively. Consistent with the filamentous phenotype, the CFU of AG100AX/pUC-AcrA were lower by two to three orders of magnitude than those of AG100AX/pUC18 (Fig. 1C).

The AcrA effect on cell morphology was dependent on the amount of AcrA. Low levels of AcrA expression induced only mild or no filamentation of AcrEF null cells (see below). The chromosomal levels of AcrA expression in AG100MB (ΔacrB ΔacrEF) (data not shown) or W4680E (Fig. 1A) did not cause cell division defects. Since cell filamentation depends on the amounts of AcrA, the difference in the AcrA toxicity in E. coli with different genetic backgrounds could also be due to variations in intracellular amounts of the overproduced AcrA. However, immunoblotting analysis showed that the level of AcrA overproduction does not depend on the genetic background of E. coli (Fig. 1B). In all tested strains the level of AcrA expression from pUC-AcrA was very similar and exceeded the chromosomal AcrA level by several folds. The pUC151A plasmid expressing both AcrA and AcrB in a single operon produced less AcrA in W4680E (ΔacrEF) and ECM2112 (ΔacrAB ΔtolC) strains. However, AcrA amounts in these cells still exceeded the chromosomal AcrA level. Thus, the cell division defects caused by the overproduction of AcrA are specific for the acrEF mutants.

AcrA-induced cell filamentation is due to the loss of AcrF function.

All characterized RND-type transporters of E. coli including AcrAB and AcrEF require for their activity the outer membrane protein TolC (30). To investigate whether a loss of the AcrEF activity or an aberration in the acrEF region of E. coli chromosome led to filamentation, we introduced plasmids pUC-AcrA and pUC151A into ECM2112 (ΔacrAB ΔtolC) (Fig. 1A). We found that similar to the AcrEF null mutants the increased expression of AcrA caused severe cell division defects in ECM2112. The growth rate of the filamentous ECM2112/pUC-AcrA cells was only 0.4 h⁻¹ compared to 1.39 h⁻¹ for the normal ECM2112/pUC18 cells (Fig. 1C). CFU for these two strains differed by 3 to 5 orders of magnitude. This result suggested that the function of AcrEF-TolC is required to maintain normal cell morphology under conditions of the increased production of AcrA.

The best-characterized activity of AcrEF is related to multidrug efflux. MIC measurements showed that AG100AX is reproducibly twofold more susceptible to erythromycin, puromycin, ethidium bromide, and nalidixic acid than strain AG100A (ΔacrAB) (Table 2). Furthermore, when AcrEF was produced from pACYC-AcrEF in AG100AX, AcrEF fully complemented the drug-susceptible phenotype of AG100AX and provided higher levels of resistance to the same set of antibiotics (Table 2).

To verify that AcrEF is expressed in different E. coli strains we used RT-PCR. Figure 2A shows that AG100AX/pACYC-AcrEF cells contained large amounts of the acrEF transcript in both exponential and stationary phases. As expected AcrEF null mutants did not contain acrEF transcripts. However, small quantities of the longer, mutated acrEF transcript could be detected in AG100AX/pUC-AcrA strain when mRNA was isolated from stationary phase cells. All other tested strains contained detectable quantities of acrEF mRNA in both exponential and stationary phases. We conclude that the acrEF operon is expressed under normal physiological conditions.

In AcrEF-TolC complex the inner membrane transporter AcrF is responsible for substrate recognition and efflux. To verify that the activity of AcrF is required to maintain normal cell morphology, pACYC-AcrF producing AcrF alone or pACYC-AcrEF carrying the complete acrEF operon were introduced into filamentous AG100AX/pUC-AcrA and AG100AX/pUC151A cells. Since AcrF is homologous to AcrB we used the anti-AcrB antibody to estimate the expression of AcrF. The level of AcrF expression from pACYC-AcrF was substantially lower than that from pACYC-AcrEF (Fig. 2B). In agreement with previous studies, the coexpression of AcrA and AcrF from the compatible pUC-AcrA and pACYC-AcrF plasmids led to the complementation of the drug susceptible phenotype of AG100AX, albeit only partially (Table 2). Probably this partial complementation is due to the low level of AcrF expression (Fig. 2B).

Unexpectedly we found that the production of AcrEF from pACYC-AcrEF does not restore normal morphology of the filamentous AG100AX/pUC-AcrA or AG100AX/pUC151A (Fig. 2C, top panel). We noticed however, that AG100AX/pUC18 cells carrying pACYC-AcrEF were slightly elongated, suggesting that AcrEF overexpression is damaging for these cells. AcrE shares 61% identity and 72% similarity with AcrA. We reasoned that the overproduction of AcrE could further aggravate the problem caused by AcrA overproduction. To test this possibility we took advantage of the fact that both proteins can function with the AcrF transporter to confer multidrug resistance (23). Indeed, we found that pACYC-AcrF producing AcrF alone completely restored the normal morphology of AG100AX/pUC151A. Thus, functional AcrF is required to maintain normal E. coli morphology under conditions of increased production of AcrAB. However, pACYC-AcrF failed to complement the cell division defect of AG100AX/pUC-AcrA (Fig. 2C). For unknown reasons AcrB is needed for the complementation of function by the overproduced AcrF. Perhaps AcrB sequesters a fraction of AcrA into functional complexes and by this means reduces the toxicity of AcrA overproduction.

Taken together, these results suggest that functional AcrF is produced in E. coli cells and is required to maintain the normal morphology of E. coli under conditions of increased production of AcrA.

AcrA localizes in a confined manner in the cytoplasmic membrane of normal and filamentous cells.

To investigate how increased expression of AcrA leads to filamentation of AcrEF-deficient cells, we compared AcrA localization in normal and filamentous cells using GFP-AcrA fusion protein. For this purpose, the cleavable N-terminal signal peptide of the AcrA precursor was replaced with a single transmembrane segment of CvaA, a structural homolog of AcrA (12). The constructed CvaA-AcrA chimera was expressed under the native PacrAB promoter in a single operon with AcrB transporter (pUC-Cva-AcrA). Judging by complementation of the drug-susceptible phenotype of AG100AX, the CvaA-AcrA chimeric protein was fully functional (Table 2). Next, the N terminus of the CvaA-AcrA chimera was fused with GFP. The resulting construct, pUC-GFP-AcrA, expressed a chimeric protein containing a cytoplasmic GFP domain linked to a periplasmic AcrA domain by a single CvaA transmembrane segment. Western blot analysis with anti-AcrA antibody revealed that the amounts of GFP-CvaA-AcrA produced from pUC-GFP-AcrA were roughly equal to levels of the chromosomally produced wild-type AcrA and about 20- to 30-fold lower than those of the wild-type AcrA produced from pUC151A (Fig. 3A). The expression of GFP-CvaA-AcrA complemented the drug-susceptible phenotype of AG100AX, suggesting that the GFP-CvaA-AcrA chimeric protein is functionally active.

The localization of GFP-CvaA-AcrA was characterized in living cells taken from freshly grown colonies or from mid- to late-logarithmic liquid cultures. Since GFP-CvaA-AcrA is produced from pUC-GFP-AcrA in substantially lower amounts than AcrA produced from pUC151A, the AcrEF- and TolC-deficient cells with the chromosomal copy of AcrA and pUC-GFP-AcrA were mildly filamentous (Fig. 3B). The morphology of the AG100AX/pUC-GFP-AcrA strain, which does not contain a chromosomal copy of AcrA, was normal (data not shown). The localization of the GFP-CvaA-AcrA protein was restricted to the membrane region in both normal and filamentous AcrEF null cells. Although the degradation product of GFP-CvaA-AcrA fusion can be seen on Western blots blots (Fig. 3A), the membrane localization was specific for GFP-CvaA-AcrA fusion.

We found that GFP-CvaA-AcrA fusion protein was highly unstable in TolC null cells. In these cells, GFP fluorescence was evenly distributed throughout the cytoplasm (data not shown). In wild-type and filamentous AcrEF null cells, the membrane bound GFP-CvaA-AcrA was found to be concentrated into foci at the poles and at intermediate positions of growing cells suggesting that the lateral diffusion of AcrA in the inner membrane and periplasm is restricted. The GFP-CvaA-AcrA foci appeared as pairs of dots usually not perpendicular to the long axis of the cells. The focal localization of AcrA was similar in normal and filamentous cells. We conclude that the cell division defects of acrEF mutants are not caused by mislocalization of the overproduced AcrA.

Increased expression of AcrA does not cause membrane aberrations.

Overproduction of membrane proteins could lead to the morphological and compositional changes of the cytoplasmic and/or the outer membrane (24). Figure 4 shows transmission electron microscopic images of normal AG100AX/pUC18 and filamentous AG100AX/pUC-AcrA cells. In both preparations an evenly distributed cytoplasm packed with ribosomes and a nucleoid can be seen (Fig. 4A and 4C). The cell envelope was clearly seen around majority of cells and consisted of the inner membrane, peptidoglycan layer, and outer membrane (Fig. 4B and 4D). We did not find accumulation of membrane-rich structures previously documented in prokaryotic cells overproducing certain membrane proteins (2, 24). Furthermore, overall membrane morphology was very similar between normal and filamentous cells. However, some differences can also be seen. We did not find complete or partial septa in filaments. In contrast, septum formation, even in the very early stages of cell division, was clearly detected in normal cells. This result suggested that filamentation could be caused by defects in the septum assembly.

In some of the cross-sections shown in Fig. 4, a contraction in the thickness of the peripheral membrane, peptidoglycan, and outer membrane can be seen in filamentous but not in normal cells. In the other regions of each cell, these structures are similar between two morphological species. Since in our experiments we used a traditional fixation followed by the negative staining approach, this contraction could be a result of sample preparation and processing. Thus, we conclude that the overproduction of AcrA does not perturb the structure of the E. coli cell envelope. The lack of septa in filamentous AcrEF null cells with increased expression of AcrA suggested that filamentation could be due to defects in septum assembly.

Previous studies showed that defects or inhibition of penicillin-binding proteins (PBPs) lead to aberrations in cell morphology and septum formation. Specifically, thermosensitive mutants of PBP3, also known as FtsI, produced filaments at nonpermissive temperature (35). Located in the periplasm PBP3 is a membrane transpeptidase required for peptidoglycan synthesis at the septum generated by cell division (36). If the expression of PBP3 is defective in acrEF cells, then large amounts of AcrA in the periplasm could further compromise the PBP3 function leading to formation of filaments. To investigate if the expression of PBPs is altered in acrEF mutants, crude membrane fractions were purified and treated with fluorescent β-lactam antibiotic bocillin. Figure 5A shows the composition of PBPs in E. coli cells with different genetic background. The composition and amounts of PBPs in all tested E. coli strains were identical.

We next studied the protein composition of cell envelopes isolated from normal and filamentous cells. For this purpose, E. coli cells were converted into spheroplasts by EDTA-lysozyme treatment (34). Crude membranes were first purified by sedimentation in a two-step sucrose gradient and then subjected to sedimentation through a 30%-60% sucrose gradient following the technique of Ishidate et al. (20). Previously, using this approach we found that the majority of the wild type AcrA remains associated with the inner membrane during sucrose density centrifugation (39). However, about 30% of AcrA reproducibly comigrated with the outer membrane fractions.

Figure 5B shows the purified cell envelopes from the normal and filamentous AG100AX strain carrying pUC18 and pUC-AcrA, respectively. The inner and outer membranes from both cell types migrated to the expected positions on a sucrose gradient, 1.16 to 1.18 g/ml and 1.22 to 1.26 g/ml for the inner membrane and outer membrane, respectively (data not shown). Thus, the overall lipid and protein compositions of cell envelopes from normal and filamentous cells are very similar.

This result was confirmed by staining the membrane fractions separated by SDS-PAGE with silver nitrate. Composition of major inner membrane and outer membrane proteins of normal and filamentous cells appears to be very similar (Fig. 5B). However, two major bands in the intermediate fraction and a few minor bands in the inner membrane fraction showed different expression levels in normal and filamentous cells. The major protein migrated in the intermediate density fractions of the cell envelopes isolated from normal cells was identified by the N-terminal protein sequencing as flagellin encoded by fliC gene. The amounts of minor bands in inner membrane fractions correlated with the amount of FliC and were lower in the filamentous cells. Thus, the differences in protein composition seen on SDS-PAGE gels could be largely attributed to either the differential expression of flagellum genes or posttranslational effects on flagellum in normal and filamentous cells. Judging by immunoblotting the second major protein of the intermediate fraction is AcrA (data not shown).

Taken together these results suggest that the filamentation of AcrEF null cells is not caused by an adverse effect of the increased levels of AcrA on the membrane morphology or composition. Although acrEF cells readily produce filaments when AcrA expression increases, these cells contain normal amounts of PBP3.

Increased levels of AcrA in acrEF null cells interfere with chromosome segregation.

In E. coli, assembly of the septum at the cell division site is negatively regulated by the nucleoid in a phenomenon called nucleoid occlusion (8). The cell division arrest and cell filamentation are often induced by DNA damage or defects in DNA segregation. To analyze chromosome structure and positioning in the normal and filamentous acrEF null cells, log-phase cells grown at 37°C in LB medium were fixed, stained with DAPI and Sypro Orange, and subjected to fluorescence microscopy to visualize DNA, septa, and/or the cell envelope. This analysis revealed that about half of the filamentous AG100AX/pUC-AcrA cells contain abnormal nucleoids. The abnormal nucleoid-staining patterns observed include nucleoids that are asymmetrically positioned in the filament and large aggregates of nucleoids or decondensed nucleoids occupying an extensive part of the filament (Fig. 6B). Some filaments however contained normal nucleoids, evenly positioned throughout the filament. The abnormal nucleoid-staining patterns were also absent in AcrEF null cells containing pUC18 alone analyzed in parallel (Fig. 6A), indicating that these abnormal staining patterns were not a fixation artifact.

Occasionally the long filaments contained a few normally developed septa suggesting that the cell division apparatus is present and functional in the filaments. However, formation of septa did not correlate with chromosome segregation as evidenced by the presence of large fragments of filaments devoid of both the chromosome and the septum (Fig. 6B). These results therefore show that the acrEF mutants with increased expression of AcrA are defective in chromosome segregation and condensation.

E. coli reacts to defects in chromosome segregation by activating the SOS response, which ceases cell division (15). The SOS response induces the expression of > 30 genes (10). These genes are involved in a variety of processes, but the majority is involved in DNA repair. One of the SOS genes is sulA. The protein is synthesized in large amounts during SOS response, reaching >0.3% of total protein synthesis. SulA is sufficient to halt cell division by binding to the tubulin-like GTPase FtsZ and preventing septum formation (4, 18). To investigate whether cell division arrest of acrEF cells is caused by activation of the SOS response we analyzed the amounts of sulA transcript in normal and filamentous acrEF cells by RT-PCR. We found that the amounts of sulA transcript were similar in normal and filamentous cells. We conclude that the chromosome segregation defects in acrEF null cells with increased production of AcrA are not related to the activation of SOS response.