Biochimica et Biophysica Acta 1758 (2006) 1587 – 1595 www.elsevier.com/locate/bbamem
Verapamil, a Ca 2+ channel inhibitor acts as a local anesthetic and induces the sigma E dependent extra-cytoplasmic stress response in E. coli C.L. Andersen 1 , I.B. Holland, A. Jacq ⁎ Institut de Génétique et Microbiologie, UMR 8621, Université Paris-Sud (XI), Bâtiment 400, 91405 Orsay cedex, France Received 2 December 2005; received in revised form 26 April 2006; accepted 25 May 2006 Available online 7 June 2006
Abstract Verapamil is used clinically as a Ca2+ channel inhibitor for the treatment of various disorders such as angina, hypertension and cardiac arrhythmia. Here we study the effect of verapamil on the bacterium Escherichia coli. The drug was shown to inhibit cell division at growth sub inhibitory concentrations, independently of the SOS response. We show verapamil is a membrane active drug, with similar effects to dibucaine, a local anesthetic. Thus, both verapamil and dibucaine abolish the proton motive force and decrease the intracellular ATP concentration. This is accompanied by induction of degP expression, as a result of the activation of the RpoE (SigmaE) extra-cytoplasmic stress response, and activation of the psp operon. Such effects of verapamil, as a membrane active compound, could explain its general toxicity in eukaryotic cells. © 2006 Elsevier B.V. All rights reserved. Keywords: E. coli; Verapamil; Dibucaine; DegP; Membrane potential; PspA
1. Introduction Ca2+ channel inhibitors have been used extensively, both clinically and in research in eukaryote systems, in order to elucidate the role of Ca2+ channels and their mode of action. Verapamil (a phenyl alkylamine) is a specific inhibitor of eukaryotic voltagegated L-type calcium channels and is used in the range 10–20 μmolar in the treatment of various disorders such as angina, hypertension and cardiac arrhythmia [1]. It is generally believed that verapamil blocks the flow of Ca2+ ions through the channel by binding to a specific site on the internal face at the cytoplasmic end of the channel. The positively charged verapamil (at physiological pH) may interact with a putative Ca2+ binding site (negatively charged) and in this way block the flow of Ca2+ ions [2,3]. However, verapamil has also other effects in addition to the inhibition of Ca2+ channels. For example, at 50 μM, verapamil competitively inhibits the function of P-glycoprotein in mammalian cells [4]. P-glycoprotein (ABCB1), a plasma membrane protein over-ex-
⁎ Corresponding author. Tel.: +33 1 69 15 57 17; Fax: +33 1 69 15 66 78. E-mail address:
[email protected] (A. Jacq). 1 Present Address: Lilly Critical Care Europe, Eli Lilly Export S.A., 1214 Vernier, Geneva, Switzerland. 0005-2736/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.bbamem.2006.05.022
pressed in multidrug-resistant (MDR) cells, is responsible for ATP driven efflux of structurally unrelated cytotoxic drugs, and thus, gives rise to resistance to many chemotherapeutic drugs in cancer cells (see [5]). In this context, verapamil has been used as a chemosensitizing agent to improve the clinical efficacy of anti cancer therapy [6]. Similarly, verapamil can decrease the resistance to anti-helminths drugs of parasitic helminths overexpressing P-glycoprotein [7]. Verapamil was also found in vitro to have a synergistic effect with itraconazole against clinical isolates of Aspergillus fumigatus [8]. However, the use of verapamil is limited by its general toxicity and it is important to better understand the various pharmacological effects of this clinically important drug. Verapamil is an amphipath, i.e., it contains both hydrophilic and hydrophobic groups. In addition, it contains a tertiary amine which is protonated at physiological pH. The same chemical characteristics exist in some local anesthetics. These molecules can partition into the phospholipid bilayer of the membrane, with their hydrophobic portion intercalating between the fatty acyl chains, while the cationic groups may interact with the negatively charged head groups of particular acidic phospholipids [9]. These drugs appear to physically reorganize the lipid bilayer, such that their accumulation in the membrane increase its fluidity as well as affecting the ratio of bilayer to non-bilayer
1588
C.L. Andersen et al. / Biochimica et Biophysica Acta 1758 (2006) 1587–1595
structures [10–12]. These changes strongly affect the barrier properties of the membrane, increasing membrane permeability [13,14]. Importantly, the functioning of membrane proteins, including different sensors and the activity of embedded enzymes, is likely to be influenced by membrane changes induced by such drugs. Previously, we have reported that N43, an E. coli strain lacking the multidrug resistance AcrAB pump [15], is hypersensitive to Ca2+ channel blockers, including verapamil, suggesting that the drug might be a substrate of this pump [16]. Specific targets for verapamil action in E. coli have not, however, been identified and an objective of this study was to elucidate in more detail the effects of the drug. We found that verapamil acted in a similar way to dibucaine and induced both the RpoE mediated extracytoplasmic stress response and to some extent the phage-shock response (PSP). 2. Materials and methods 2.1. Bacterial strains, media and chemicals All bacterial strains in this study were derivatives of Escherichia coli K12. Strain N43 ( F −, acrA1, Δlac-85, ara-14, galK12, rpsL179, malA1, xyl-5, mtl-1, sup44 ) was described by Nakamura and Suganuma [17]. This strain has an IS2 element inserted in the acrA gene, which is polar on the downstream gene acrB [18]. A degP::kan allele (from KS474, a gift from Jon Beckwith) and a cpxR:: spc allele (from PND325, a gift from T. Silhavy) were introduced into N43 by P1 transduction according to Miller [19]. CAG16037 (araD, Δ(ara-leu)7697, Δ (codB-lacI), galK16, galE15, mcrA0, relA1, rpsL150, spoT1, mcrB9999, hsdR2, rpoHP3::lacZ) was from Carol Gross [20]. MC3 (F-araD139 ΔU169 rpsL150 relA1 flbB5301 deoC7 ptsF25 rbsR λΦ(pspA-lacZYA) is described in Bergler [21] and PND2000 (F-lacΔU169 araD139 rpsL150 thi-flbB5301 deoC7 ptsF25 relA1 λ RS88 [degP::lacZ ]) is described in Danese et al. [22]. Luria–Bertini (LB) broth was prepared as in Miller et al. [19] and M63 minimal medium as described in Laoudj et al. [23]. When necessary, the medium was supplemented with kanamycin (25 μg/ml), or spectinomycin (50 μg/ml). Cells were grown at 30°C with continuous shaking, and cell mass was monitored by measuring the absorbance at 600 nm (A600) using a Hitachi U-1100 spectrophotometer. Radioactivity and protein standards were purchased from Amersham International, UK Ltd. Different drugs were purchased from Sigma Chemicals Ltd., while most other organic or inorganic compounds were purchased from either Sigma Chemicals Ltd. or Merck Chemicals Ltd.
2.2. Protein labeling with 35S-methionine, preparation of heat stable protein fraction and protein analysis Cultures were grown in M63 minimal medium supplemented with 0.5% (w/v) glucose at 30 °C. When the cell density reached an A600 = 0.2, drug was added to the culture, which was incubated for 10 min before labeling with 10 μCi/ml of 35S-methionine for 10 min. The heat stable protein fraction was prepared as follows. Cells were harvested, resuspended in TEK buffer (100 mM KCl, 50 mM Tris–HCl pH 8, 1 mM EDTA), lysed by heating at 100 °C for 10 min and centrifuged at 13000 rpm for 15 min at 4 °C. The supernatant, containing the heat-stable proteins, was transferred to a new tube and proteins precipitated by 5% TCA on ice for at least 30 min; precipitates were collected by centrifugation at 13000 rpm for 10 min, washed once with acetone and dried. Then the protein residue was resuspended in SDS-sample buffer (0.125 mM Tris–HCl (pH 6.8), 10% (w/v) βmercaptoethanol, 4% (w/v) SDS, 20% (w/v) glycerol, 0.05% (w/v) Bromophenol blue) and boiled for 3 min prior to loading on gels for SDS-PAGE. SDS-PAGE was carried out according to the method of Laemmli [24]. The separating gels were 15% acrylamide with a stacking gel of 5%. Gels were fixed, dried and autoradiographed against an X-ray film until a suitable exposure was obtained.
2.3. ATP measurement N43 was grown in M63 at 30 °C. At an A600 = 0.16, the culture was treated with different concentrations of verapamil for 20 min. The cultures were kept on ice until ATP measurements. ATP measurements were carried out essentially according to the method described by Collura and Letellier [25]. 20 μl of culture was added to 80 μl dimethylsulfoxide (DMS0) to destroy the cell envelope. The lysed samples were diluted with 5 ml of cold sterile milliQ water to prevent any interference of DMSO in the luminescence assay. 100 ml of this suspension were withdrawn for ATP measurement with a luminometer (Lumac), using the luciferin/luciferase assay system.
2.4. Measurement of membrane potential (ΔΨ) The method was essentially according to Collura and Letellier [25]. A culture of N43 was grown as usual in M63 at 30 °C to an A600 = 0.2, when 200 ml of cells were harvested. The cells were permeabilised in 20 ml Tris–EDTA buffer (120 mM Tris–HCl pH 8, 1 mM EDTA) at 37 °C for 10 min with occasional agitation. After centrifugation, the cells were washed in 20 ml HEPES/NaCl buffer (10 mM HEPES, 150 mM NaCl, pH 7.2), recentrifuged and resuspended in 5.4 ml of the same buffer and kept on ice until assayed. 270 μl of permeabilised cells were transferred to Eppendorf tubes and energized by the addition of 0.5% glucose and 0.6 mM KCl to restore the intracellular concentration of K+. Different concentrations of the drugs to be tested were added and the samples incubated for 10 min at 30 °C. A mixture of non and radioactive 3 H-TPP (tetraphenylphosphonium ion, 10 mM, 25 μCi) was added and the samples were further incubated for 10 min at 30 °C. 100 μl (equivalent to 2 × 108 cells) of samples were filtered in triplicate on GF/C membranes and quickly rinsed twice with 5 ml HEPES/NaCl buffer. Filters were dried, transferred to scintillation vials, and scintillation liquid added. The radioactivity on the filters was counted wit a preset 3H/14C program. TPP+ uptake values were corrected for non-specific binding by subtracting the counts obtained for cells treated with 20 μM CCCP under the same conditions as above. The membrane potential was calculated using Nernst's equation: DΨ ¼ 2:3RT=F logðVout =Vin ½TPPin =½TPPout (For E. coli: Vout/Vin = 1000 for a suspension of 2 × 109 cells/ml. At 30 °C: 2.3 RT/F = 60 mV).
2.5. Western blotting Total cell proteins (equal A600 loadings) were separated by SDS-PAGE (15%) and transferred onto nitrocellulose membranes (Schleicher and Schull, 0.45 μM) by electroblotting, at 4 °C, for 45 min at 70 V in 25 mM Tris, 192 mM glycine, 20% (v/v) methanol, pH 8.3. Subsequent steps were carried out at room temperature. Proteins were fixed to the membrane by 0.025% glutaraldehyde for 40 min. Non-specific sites were blocked in 5% non-fat milk powder in PBS (prepared as described in [26]) for 3 h. Thereafter, the membrane was incubated overnight with primary antibody (1/1000 in PBS-buffer containing 0.5% non-fat milk) with gentle shaking. After 4 washings in PBS buffer, the membrane was incubated with the secondary antibody (anti-rabbit antibody conjugated to peroxidase) at a dilution of 1/10000 in PBS for 1 h. Cells were washed twice in PBS before detection by enhanced chemiluminescence (ECL), according to the manufacturer's (Amersham) instructions.
2.6. β-galactosidase assays σE activity was assayed by monitoring β-galactosidase activity expressed from the chromosomal σE dependent lacZ reporter gene rpoHp3-lacZ [20]. Enzymatic activity was measured as described in Miller [19] and expressed in Miller Units.
3. Results 3.1. Effect of verapamil upon E. coli growth Previous results showed that strain N43, an acrA1 mutant of E. coli K12, which carries an IS2 insertion interrupting the acrAB operon (see also Materials and methods) is hypersensitive to
C.L. Andersen et al. / Biochimica et Biophysica Acta 1758 (2006) 1587–1595
several drugs including inhibitors of Ca2+ channels (verapamil and diltiazem) and to inhibitors of calmodulin (W7, 48/80, TFP), indicating that these drugs are probably substrates of the major multidrug resistance pump encoded by the acrAB genes [15,16]. The MIC of verapamil for the N43 strain was found to be 0.3 mM, a 10-fold reduction in the MIC for a wild-type E. coli K12 strain. We took advantage of this hypersensitivity of the acrA1 mutant to study the effect of verapamil on E. coli growth and physiology. Previous microscopic studies [27] indicated that a low concentration of verapamil inhibited cell division in N43. In this study we sought to confirm this effect using flow cytometry. Strain N43 was grown in minimal medium to early exponential phase, as described in Materials and methods, and treated with increasing concentrations of verapamil. At 300 μM verapamil, as shown in Fig. 1A, growth was inhibited after 2 to 3 generations (3 to 5 h). As shown in Fig. 1B, this inhibition could be reversed by the addition of Mg2+ or Ca+2. Since this effect was not specific for Ca2+, we hypothesized that the addition of these divalent cations contributes to a generally decreased permeability of the outer membrane to the drug. This result also indicated that the effect is reversible and hence that the drug is bacteriostatic rather than bacteriocidal. Determination of viable counts for verapamiltreated cultures confirmed that the drug is bacteriostatic, not bacteriocidal, at least up to 0.5 mM (data not shown). At a lower concentration (200 μM), verapamil had little if any effect on growth (Fig. 1A). However, microscope analysis indicated that this concentration of drug resulted in a heterogeneous cell size distribution with many cells 2–4 times normal length. This effect on cell division, giving a heterogeneous size distribution, was confirmed by flow cytometry (Fig. 1C). In order to determine whether the effect of verapamil on cell division was due to induction of the SOS response, the effect of the drug on DNA synthesis was measured. At concentrations (150–200 μM) affecting cell division, the drug had no detectable effects on DNA synthesis. Moreover, no increase in the level of the RecA protein (measured by Western blot, or using a sfiA-lacz fusion) could be detected in verapamil-treated cells (all data not shown). Thus, verapamil treatment results in division inhibition by a mechanism independent of the SOS response at concentrations which are below those required for growth inhibition. This mechanism might include effects on the cytoplasmic membrane that perturb the action of the division machinery in some way. Possible ways in which verapamil could affect the membrane are explored below. 3.2. Verapamil induces a specific set of heat-stable proteins We have previously shown [23] that treatment of E. coli N43 with the Ca2+ chelator EGTA, induces the synthesis of a specific family of heat-stable, low molecular weight acidic proteins, with apparent molecular weights of 12, 14.3, 17, 18, 21.5–23, 27, 34– 35 kDa. These include three potential Ca2+ binding proteins, P34, P18 and P12 (named from their apparent molecular weight in SDSPAGE), since they cross-reacted with antibodies to calmodulin and to the calmodulin-like protein calerythrin from Saccharopolyspora erythraea. We hypothesized that some or all of these proteins belong to a putative Ca2+ regulon and were induced in response to perturbation of intracellular Ca2+ homeostasis by the
1589
EGTA treatment. Verapamil is a Ca2+ channel blocker in eukaryotic cells, and it was interesting to test whether in bacteria, it could also disturb Ca2+ homeostasis and induce a similar set of proteins. The results, shown Fig. 2, indicate that verapamil, at a minimal concentration of 500 μM, specifically induced the synthesis of three major heat-stable polypeptides with apparent molecular weight corresponding to 14 kDa (P14), 17 kDa (P17) and 50 kDa (P50). Qualitatively, this profile, which was highly reproducible, is very different from that obtained after induction by EGTA [23] although the 14 kDa and 17 kDa had similar molecular weights to two proteins induced by EGTA. However, further analysis using 2-D gels showed that P14 and P17, induced by verapamil, migrated towards the basic end of the gel (data not shown) and were clearly different from the two heat-stable, acidic proteins identified previously [23]. 3.3. P50 is identical to DegP, the periplasmic serine protease Verapamil is an amphipathic drug also containing a tertiary amine group protonated at physiological pH and, as such, presents some characteristics similar to that of local anesthetics for example dibucaine, and can therefore be expected to disturb the membrane bilayer. It is now well established in bacteria that damage to the cell membrane or components of the cell envelope in general, triggers protective responses to such stress and we anticipated therefore that verapamil might generate such an extra-cytoplasmic stress [28,29]. In E. coli, there are two main pathways which control the response to such stress, the two-component cpxA/cpxR system and the extracytoplasmic stress response, that is mediated by the alternative sigma factor RpoE. DegP is a periplasmic protease of 48 kDa, whose synthesis is controlled by both pathways and therefore a possible candidate for P50. To test this, cells were treated with verapamil for 60 min and total cell proteins were separated by SDS-PAGE, followed by Western blotting using anti-DegP antibodies. The results shown in Fig. 3 clearly demonstrated that the amount of DegP is substantially increased in the presence of the drug. Moreover, when a N43degP::kan strain, which grows quite normally, was treated by verapamil, expression of P14, P17 but not P50 was induced, consistent with P50 being DegP (data not shown). 3.4. Induction of DegP is independent of the cpxA/R pathway Expression of degP is under the control of two pathways, the two-component cpx pathway [30] responding to surface attachment or alkaline pH [31] and the rseA/rpoE pathway [32], responding to accumulation of misfolded outer membrane proteins in the periplasm [20]. E. coli strains lacking CpxA (the sensor) or CpxR (the response regulator) fail to induce degP expression, following overproduction of NlpE, a known inducing signal of the pathway [30]. To test whether induction of DegP by verapamil was dependent upon the cpxA/R system, a cpxR mutant was analyzed for induction of [35S]-labeled DegP, P14 and P17 proteins by verapamil, as before. The results (Fig. 4) clearly showed that verapamil induction of P14, P17 and DegP, optimal at 0.5 mM verapamil, was not affected in the cpxR mutant. This indicated that the CpxA/R pathway is not induced by verapamil.
1590
C.L. Andersen et al. / Biochimica et Biophysica Acta 1758 (2006) 1587–1595
Fig. 1. (A) Effect of verapamil on growth of E. coli. A culture of N43 was grown in M63 minimal medium at 30° and at A600 = 0.15 was treated with different concentrations of verapamil as follows: open diamonds: 0 mM; open squares: 0.2 mM; open circles: 0.3 mM; open triangles: 0.4 mM, closed squares: 0.5 mM; closed circles: 1 mM. Growth was followed by optical density measurement as described in Materials and methods. (B) Ca2+ or Mg2+ reverse the growth inhibition of verapamil. N43 was grown in M63 minimal medium at 30 °C and at A600 = 0.15, the cells were treated with 0, (squares) ; 0.3 (diamonds) or 0.5 mM (circles) verapamil. After 2 h of treatment, respectively 10 mM CaCl2 (black symbols) or MgCl2 (gray symbols) was added. Growth was followed by measurement of optical density. (C) The effect of verapamil on cell division of N43 analyzed by flow cytometry. N43 was grown in M63 minimal medium at 30 °C. At early exponential phase (A600 = 0.05), 0 (left panel) or 0.15 mM (right panel) verapamil, was added to the culture. After 4 h growth, aliquots of culture were taken and prepared for flow cytometry and analyzed as described in Materials and methods.
3.5. Verapamil induced the RpoE mediated extracytoplasmic stress response Since DegP induction by verapamil was independent of the cpxA/R two-component pathway, we surmised that its induction was due to activation of the RpoE pathway by verapamil. In the
absence of stress, RpoE is normally sequestered to the membrane by the inner membrane-localized anti-sigma factor RseA. Extracytoplasmic stress, resulting from accumulation of outer membrane proteins in the periplasm, leads to the activation of a proteolytic cascade. Thus, the essential periplasmic protease DegS, followed by the action of the cytoplasmic membrane protease
C.L. Andersen et al. / Biochimica et Biophysica Acta 1758 (2006) 1587–1595
Fig. 2. Effect of verapamil on the synthesis of heat stable proteins in mid exponential phase cultures. N43 was grown exponentially in M63 minimal medium at 30 °C. At an A600 = 0.2, verapamil was added to N43 at various concentration. After 10 min of treatment the cells were radio-labeled with 35Smethionine for 10 min. The samples were then processed as described in Materials and methods for preparation of the heat stable protein fraction before analysis by SDS-PAGE (15% acrylamide). Identical A600 equivalent of cells were loaded on each track. An autoradiogram of the fixed, dried gel is shown. The proteins P14, P17, and P50 most clearly induced by the drug are indicated by arrows.
YaeL/RseP, cleave RseA, releasing RpoE to the cytoplasm and triggering expression of the RpoE regulon [33]. Since an rpoE deletion mutation is lethal [34], in order to check whether verapamil activated the RpoE response, we used a lacZ transcriptional fusion, where lacZ was fused to the rpoHp3 promoter. This promoter is uniquely dependent for its transcription upon RpoE. As shown in Fig. 5, verapamil significantly (3fold at 3 mM) increased expression from this promoter, confirming that it could induce the extra-cytoplasmic stress response mediated by RpoE. Note that in this experiment, the range of verapamil concentrations employed was higher than in the case of N43, since the strain used is AcrAB+ and hence more resistant to verapamil than N43 (the MICs being 3 mM for an AcrAB+ strain versus 0.3 mM for an acrAB mutant).
1591
Fig. 4. The effect of a cpxR mutation on the induction of P14, P17 or DegP synthesis by verapamil. The radio-labeled heat stable proteins of cells from cultures of the parental strain N43 and its cpxR derivative treated with verapamil at the indicated concentration, as described in Fig. 2, were analyzed by SDS-PAGE.
thesis of heat-stable polypeptides analyzed by SDS-PAGE, was compared with the action of a number of different drugs (cerulenin, CCCP, nalidixic acid, novobiocin, rifampicin, chloramphenicol, tetracycline and the membrane active drug dibucaine). Only dibucaine treated cells displayed a similar pattern of induction to that shown by verapamil treated cells (Fig. 6 and data not shown). Dibucaine is a local anesthetic, a category of amphipathic drugs that affect the ratio of bilayer to non-bilayer structures in the membrane and increase membrane fluidity. Our results suggested therefore that verapamil in bacteria could act in a similar way to
3.6. Comparison of the action of verapamil and dibucaine In order to approach the question of how verapamil might provoke these changes in gene expression, its effect on the syn-
Fig. 3. Induction of DegP by verapamil treatment. A culture of N43 was grown in M63 at 30 °C, and at an A600 = 0.2, the culture was treated with different concentrations of verapamil for 60 min (indicated below) and samples of total proteins were prepared. Equal amounts of protein (A600 equivalent of cells) were loaded onto an acrylamide gel and after separation by SDS-PAGE electroblotted onto a nitrocellulose membrane. The Western blot was carried out as described in Materials and methods, using polyclonal anti-DegP antibody.
Fig. 5. Induction of RpoE activity by verapamil. CAG16037 (rpoHp3-lacZ acrA+) was grown in LB until A600 = 0.3. Verapamil was then added at increasing concentrations (0, 0.5, 1, 2, 3 mM). After 30 min, samples were withdrawn and β-galactosidase activity was assayed as described in Materials and methods.
1592
C.L. Andersen et al. / Biochimica et Biophysica Acta 1758 (2006) 1587–1595
Fig. 6. Verapamil and dibucaine have similar effects upon the synthesis of heatstable proteins. An exponentially growing culture of N43 at 30 °C (A600 = 0.2) was mixed with different concentrations of verapamil or dibucaine and after 10 min of treatment radio labeled with 35S-methionine for a further 10 min. The samples were processed as described in Materials and methods for preparation of the heat-stable protein fraction before analysis by 15% SDS-PAGE. An autoradiogram of the dried gel is shown. The positions of P14, P17 and DegP are indicated by arrows on the right.
that of local anesthetics, rather than as a specific Ca2+ channel blocker. 3.7. Verapamil and dibucaine affect the membrane electrical potential and induce the synthesis of PspA Previous studies [25] showed that dibucaine treatment of E. coli, led to an increase in ATP levels at low drug concentrations, while higher concentrations (>500 μM) led to a substantial reduction in internal ATP levels. They also showed that dibucaine had a marked effect on the membrane electrical potential (ΔΨ ), that is, enhancing at low concentrations, inhibiting at high concentrations. As shown in Fig. 7, verapamil also behaves in the same way with respect to ATP levels and the membrane electrical potential. PspA (phage shock induced protein) is an inner membrane of 25.3 kDa, which is induced in response to envelope stress and seems to play a role in the maintenance of the proton motive force [35]. Accordingly, we investigated whether verapamil and dibucaine could also induce the synthesis of this protein, as detected by western blotting, using anti-PspA antibodies. As shown in Fig. 8, when exponentially growing cells were treated with low concentrations of verapamil or dibucaine for 30 min in M63-medium at 30 °C, moderately enhanced synthesis of PspA was observed. Note that unlike DegP, PspA is not heat-resistant and total cell proteins, rather than the heat-resistant fraction, were
Fig. 7. Effect of verapamil on the internal ATP concentration (A) and on the transmembrane electrical potential (ΔΨ ) of N34 cells (B). The level of ATP was measured after 20 min treatment with different concentrations of verapamil as described in Materials and methods. ΔΨ was measured in EDTA-permeabilised cells using 3H-TPP+ as a radioactive probe, also as described in Materials and methods The drug was added 10 min prior to the addition of TPP+.
analyzed in Fig. 8. PspA induction was detected at sub-inhibitory concentrations of verapamil, as low as 50–100 μM, although higher levels of PspA were induced at 1.0 mM (data not shown). Since verapamil induced the synthesis of PspA we considered the possibility that the products of pspB (9 kDa) and pspC
Fig. 8. Both verapamil and dibucaine induce the synthesis of the phage shock protein PspA. Samples of total proteins of N43 cells treated with either verapamil or dibucaine for 60 min, were analyzed by Western blotting using PspA polyclonal antibodies.
C.L. Andersen et al. / Biochimica et Biophysica Acta 1758 (2006) 1587–1595
Fig. 9. Verapamil induces the synthesis of DegP and PspA in a wild type Acr+ strain. Strains MC3 ( pspA-lacZ ) and PND2000 (degP-lacZ ) were grown in LB until A600 = 0.4. Verapamil was then added at increasing concentrations (0, 0.5, 1, 2, 3 mM). After 30 min, samples were withdrawn and β-galactosidase activity was assayed as described in Materials and methods.
(14 kDa), which are co-regulated with pspA [36], might correspond to P14 and P17 induced by verapamil. However, both these proteins, together with DegP, were still induced to the normal high levels, when the pspB and pspC genes respectively were disrupted (data not shown). PspC and PspB are normally required for high-level synthesis of PspA. PspC is thought to be a membrane sensor which may detect changes in the proton motive force [35]. Since in the absence of PspC or PspB, P14 and P17 are still induced to high levels, these proteins are not in any way coregulated with the products of the psp operon. 3.8. Verapamil induces the synthesis of DegP and PspA independently of the presence of the acrA1 mutation In this study we mostly used the N43 strain, which carries an acrA1 mutation, in order to bypass the resistance to verapamil due to the AcrAB pump, the main multidrug resistance pump in E. coli. However, the fact that verapamil was able to induce the RpoE mediated response in a wild-type background, albeit at a 10-fold higher concentration (Fig. 5) suggested that the effect of verapamil was not dependent upon the specific N43 background. To confirm this fact, we looked whether verapamil could induce DegP and PspA in a wild-type, acrA+ strain, using degP-lacZ and a pspA-lacZ reporter fusion strains. As shown in Fig. 9, treatment of wild-type E. coli by increasing concentration of verapamil, in the range used in the case of the RpoE response (0.5 to 3 mM) led to the induction of DegP at concentrations of 1 mM and above. PspA synthesis was also found to be induced, although to a less extent than DegP and at 2 mM. Hence, those results confirm in a wild-type background the results obtained in the N43 background and show that the verapamil effect we describe in this study was not due to an indirect effect on the cells of the acrA1 mutation. 4. Discussion The utilization of strain N43 and its derivatives, all lacking a functional AcrAB multidrug resistant pump [15,18], has enabled
1593
us to study the effects of this amphipathic drug in E. coli, in the μM range. Although we have no direct evidence that verapamil is a substrate for the AcrAB pump, in its absence the MIC for verapamil in N43 is reduced 10-fold [16]. In the acrA strain, inhibition of cell division is now detectable at sub-growth inhibitory concentrations of drug and the synthesis of three proteins, P14, P17 and P50 (DegP), is strongly induced, whilst this is not observed in the wild type strains. This indicates that verapamil is normally removed by the AcrAB pump reducing its accumulation in the cytoplasmic membrane. In this paper, we show that low concentrations of verapamil (100–150 μM) reduced the frequency of cell division, producing a heterogeneous population with cell sizes, 2–4 times that of untreated cells. At these concentrations of drug, the synthesis of DNA, RNA and protein was not detectably affected. The synthesis of RecA protein was also not induced by verapamil treatment (all data not shown) and therefore induction of the SDSresponse was not responsible for the inhibition of cell division. However, in the light of the evidence discussed below, that verapamil likely perturbs the integrity of the membrane bilayer, disruption of the assembly of the FtsZ complex at mid cell by verapamil, is a possibility. One reason to undertake this study was to explore the potential role of Ca2+ in cell cycle regulation in bacteria, anticipating that verapamil, as in eukaryotic cells, might inhibit potential Ca2+ channels. However, the effect of verapamil and EGTA [23] as indicated by changes in gene expression, were quite different. Moreover, studies directly measuring free Ca2+ levels in E. coli, provided further evidence that verapamil does not affect Ca2+ levels. Thus, while EGTA treatment does reduce levels of free intracellular Ca2+, as we have reported previously [37], verapamil does not [38]. Although so far, we have no data on the identity of the two heat-stable proteins, P14 and P17 induced by verapamil treatment, we were able to demonstrate that P50 is the periplasmic protease DegP. degP expression has been shown to be regulated by two different pathways, the two-component cpxA/R pathway [22], that responds to a variety of envelope stresses and the RpoE mediated response to extracytoplasmic stress. Verapamil induction of DegP synthesis (and P14 and P17 synthesis) was clearly shown to be independent of cpxR. A direct test of the involvement of RpoE in the synthesis of verapamil induced proteins, was not possible since we found that strains lacking rpoE were apparently non viable (see also [34]). However, we were able to demonstrate that verapamil could induce the expression of the RpoEdependent promoter rpoHp3-lacZ, clearly indicating that the expression of degP also proceeds through the RpoE pathway. The RpoE regulon comprises next to one hundred genes, with a great number of those genes involved in the synthesis, assembly and homeostasis of lipopolysaccharide and outer membrane proteins in addition to a protein degrading enzyme like DegP [39]. We note that the degP mutant is in fact no more sensitive to verapamil than its parental strain (data not shown), therefore other genes in the regulon are normally important for verapamil resistance. Amongst a wide variety of drugs tested in the course of this work only the amphipaths, W7, diltiazem (data not shown) and dibucaine, in addition to verapamil, induced the synthesis of P14, P17 and DegP. All these compounds share the property of
1594
C.L. Andersen et al. / Biochimica et Biophysica Acta 1758 (2006) 1587–1595
affecting the ratio of bilayer to non-bilayer structures in the membrane and would be expected to increase membrane fluidity. Irrespective of the precise pathway that facilitates verapamil activation of RpoE, the drug-induced changes in fluidity are likely to affect the activity of one or more membrane proteins. Considering the effects of verapamil and local anesthetics on membrane structure, we can envision two alternative mechanisms to activate RpoE. The current model of activation of the RpoE response is an accumulation of C-terminal domains of unfolded proteins that triggers the activation of the periplasmic protease DegS, the subsequent cleavage of RseA by DegS, and then by YaeL/RseP, finally leading to activation of RpoE. Thus, a mechanism for the action of verapamil could be an accumulation of unfolded outer membrane proteins and/or an LPS defect, as a consequence of perturbed transport through the inner membrane or as a direct effect of verapamil on the outer membrane. Alternatively, we cannot exclude that structural changes in the inner membrane caused by verapamil (as well as by dibucaine) might be sensed directly, leading to activation of RpoE by mechanisms dependent or independent of RseA (see also [40]). The effect of verapamil (this study) and dibucaine [25] reducing both the electrical membrane potential and intracellular ATP levels, could stem from inhibition of the ATP-synthase complex, through bilayer changes. Indeed, recent studies with the Lactococcus lactis OpuA (Glycine-betaine transporter) protein reconstituted into membrane vesicles, have demonstrated that both verapamil and a local anesthetic, tetracaine, dramatically affect transport activity, apparently by accumulation of the drugs in the inner leaflet of the bilayer (B. Poolman, personal communication; see also [41]). Alternatively, the decrease of ΔΨ itself could be due to an increased passive flux of protons, following an increase in membrane permeability, resulting from transient pore formation, which has been proposed to accompany the action of such membrane active drugs [13,14,42]. Whatever the precise mechanism, loss of membrane potential following verapamil treatment, would be expected to induce the phage-shock-protein operon (psp) involved in protection against dissipation of the proton-motive force [35,36,43]. PspA synthesis is also induced when other outer membrane proteins are overexpressed, or in response to stress induced by uncouplers, cerulenin, high temperature or by ethanol, some of which also induced the RpoE dependent response [36]. A previous study [35] also showed that the 26 kDa PspA protein, which is peripherally bound to the cytoplasmic membrane, is implicated in maintaining the proton motive force (PMF). Moreover, mutations affecting ATP synthase were found to induce the psp operon both in Yersinia enterocolitica and in E. coli [44]. Here we show that both verapamil, and dibucaine, cause dissipation of the PMF, and particularly dibucaine, induce expression of PspA, supporting the hypothesis of a role for the PMF in induction of the psp operon. However, the verapamil induced proteins P14 and P17 are apparently not identical to PspC and PspB, and since the psp regulon was determined recently to be limited to the pspABCDE operon and the unlinked gene, pspG/yjbO, encoding a 9 kDa protein, P14 and P17 may therefore belong to the large RpoE regulon. In summary, in this report we showed that treatment by verapamil induced two extra-cytoplasmic stress responses, the
RpoE and the Psp responses. In contrast, verapamil did not appear to activate the cpxA/R mediated response. The action of verapamil strongly resembles that of the local anesthetic, dibucaine. Altogether, our results show that verapamil is a membrane active drug, which could explain its toxicity in vivo. In addition, verapamil, as well as anesthetics such as benzyl alcohol and ether, have been found to abolish P-glycoprotein (Pgp) ATPase activity, thus modulating efflux from multidrug-resistant cells [12]. Verapamil is a competitive inhibitor of P-glycoprotein. However, the results by Regev et al. [12] suggest an alternative, additional mechanism for modulation of Pgp activity, that of an indirect inhibition of Pgp activity through fluidization of the lipid phase of the membrane. Our findings are in favor of such a potential mode of action and are clearly relevant to the clinical use of verapamil. Acknowledgements A special thanks to L. Letellier for help in ATP measurements and for helpful discussions. We would also like to acknowledge J. Tommassen for the gift of anti-PspA antibody, R. Kolter for the anti-DegP antibody and P. Bouloc for strain CAG16037. In addition, we acknowledge the support from ARC, Ministère pour la Recherche et l'Enseignement Supérieur (contract no. 92C0313), CNRS, Université de Paris-XI, and the Human Frontier Program (contract no. RG0386-95 M). C.L. Andersen also wishes to acknowledge the support of “Le Secours des Amis des Sciences” and le Ministère pour la Recherche et l'Enseignement Supérieur. References [1] A. Fleckenstein, Historical overview. The calcium channel of the heart, Ann. N. Y. Acad. Sci. 522 (1988) 1–15. [2] W.A. Catterall, J. Striessnig, Receptor sites for Ca2+ channel antagonists, Trends Pharmacol. Sci. 13 (1992) 256–262. [3] G. Varadi, Y. Mori, G. Mikala, A. Schwartz, Molecular determinants of Ca2+ channel function and drug action, Trends Pharmacol. Sci. 16 (1995) 43–49. [4] J.A. Endicott, V. Ling, The biochemistry of P-glycoprotein-mediated multidrug resistance, Annu. Rev. Biochem. 58 (1989) 137–171. [5] D. Leveque, F. Jehl, P-glycoprotein and pharmacokinetics, Anticancer Res. 15 (1995) 331–336. [6] V. Ling, Multidrug resistance: molecular mechanisms and clinical relevance, Cancer Chemother. Pharmacol. 40 (1997) S3–S8 (Suppl.). [7] D. Kerboeuf, W. Blackhall, R. Kaminsky, G. von Samson-Himmelstjerna, P-glycoprotein in helminths: function and perspectives for anthelmintic treatment and reversal of resistance, Int. J. Antimicrob. Agents 22 (2003) 332–346. [8] J. Afeltra, R.G. Vitale, J.W. Mouton, P.E. Verweij, Potent synergistic in vitro interaction between nonantimicrobial membrane-active compounds and itraconazole against clinical isolates of Aspergillus fumigatus resistant to itraconazole, Antimicrob. Agents Chemother. 48 (2004) 1335–1343. [9] B. Shi, H.T. Tien, Action of calcium channel and beta-adrenergic blocking agents in bilayer lipid membranes, Biochim. Biophys. Acta 859 (1986) 125–134. [10] J.M. Seddon, Structure of the inverted hexagonal (HII) phase, and nonlamellar phase transitions of lipids, Biochim. Biophys. Acta 1031 (1990) 1–69. [11] S. Maruyama, T. Hata, H. Matsuki, S. Kaneshina, Effects of pressure and local anesthetic tetracaine on dipalmitoylphosphatidylcholine bilayers, Biochim. Biophys. Acta 1325 (1997) 272–280. [12] R. Regev, Y.G. Assaraf, G.D. Eytan, Membrane fluidization by ether, other anesthetics, and certain agents abolishes P-glycoprotein ATPase activity
C.L. Andersen et al. / Biochimica et Biophysica Acta 1758 (2006) 1587–1595
[13] [14]
[15]
[16]
[17] [18]
[19] [20]
[21]
[22]
[23]
[24] [25]
[26]
[27]
[28]
and modulates efflux from multidrug-resistant cells, Eur. J. Biochem. 259 (1999) 18–24. J. Sikkema, J.A. de Bont, B. Poolman, Mechanisms of membrane toxicity of hydrocarbons, Microbiol. Rev. 59 (1995) 201–222. F.J. Weber, J.A. de Bont, Adaptation mechanisms of microorganisms to the toxic effects of organic solvents on membranes, Biochim. Biophys. Acta 1286 (1996) 225–245. D. Ma, D.N. Cook, M. Alberti, N.G. Pon, H. Nikaido, J.E. Hearst, Genes acrA and acrB encode a stress-induced efflux system of Escherichia coli, Mol. Microbiol. 16 (1995) 45–55. S. Casaregola, M. Chen, N. Bouquin, V. Norris, A. Jacq, M. Goldberg, S. Margarson, M. Tempete, S. McKenna, H. Sweetman, et al., Analysis of a myosin-like protein and the role of calcium in the E. coli cell cycle, Res. Microbiol. 142 (1991) 201–207. H. Nakamura, A. Suganuma, Membrane mutation associated with sensitivity to acriflavine in Escherichia coli, J. Bacteriol. 110 (1972) 329–335. D. Ma, D.N. Cook, M. Alberti, N.G. Pon, H. Nikaido, J.E. Hearst, Molecular cloning and characterization of acrA and acrE genes of Escherichia coli, J. Bacteriol. 175 (1993) 6299–6313. J.H. Miller, A Short Course in Bacterial Genetics, Laboratory Press, Cold spring harbor, 1992. J. Mecsas, P.E. Rouviere, J.W. Erickson, T.J. Donohue, C.A. Gross, The activity of Sigma E, an Escherichia coli heat-inducible sigma-factor, is modulated by expression of outer membrane proteins, Genes Dev. 7 (1993) 2618–2628. H. Bergler, D. Abraham, H. Aschauer, F. Turnowsky, Inhibition of lipid biosynthesis induces the expression of the pspA gene, Microbiology 140 (Pt. 8) (1994) 1937–1944. P.N. Danese, W.B. Snyder, C.L. Cosma, L.J. Davis, T.J. Silhavy, The Cpx two-component signal transduction pathway of Escherichia coli regulates transcription of the gene specifying the stress-inducible periplasmic protease, DegP, Genes Dev. 9 (1995) 387–398. D. Laoudj, C.L. Andersen, A. Bras, M. Goldberg, A. Jacq, I.B. Holland, EGTA induces the synthesis in Escherichia coli of three proteins that crossreact with calmodulin antibodies, Mol. Microbiol. 13 (1994) 445–457. U.K. Laemmli, Cleavage of structural proteins during the assembly of the head of bacteriophage T4, Nature 227 (1970) 680–685. V. Collura, L. Letellier, Mechanism of penetration and of action of local anesthetics in Escherichia coli cells, Biochim. Biophys. Acta 1027 (1990) 238–244. J. Sambrook, E.F. Fritsch, T. Maniatis, Molecular Cloning—A Laboratory Manual, Second Edition ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, 1989. M. Goldberg, The isolation and characterisation of an E. coli mutant resistant to the voltage operated calcium channel inhibitor verapamil, Genetics, Leicester, Leicester, 1995. P.N. Danese, T.J. Silhavy, The sigma(E) and the Cpx signal transduction systems control the synthesis of periplasmic protein-folding enzymes in Escherichia coli, Genes Dev. 11 (1997) 1183–1193.
1595
[29] L. Connolly, A. De Las Penas, B.M. Alba, C.A. Gross, The response to extracytoplasmic stress in Escherichia coli is controlled by partially overlapping pathways, Genes Dev. 11 (1997) 2012–2021. [30] W.B. Snyder, L.J. Davis, P.N. Danese, C.L. Cosma, T.J. Silhavy, Overproduction of NlpE, a new outer membrane lipoprotein, suppresses the toxicity of periplasmic LacZ by activation of the Cpx signal transduction pathway, J. Bacteriol. 177 (1995) 4216–4223. [31] P.A. DiGiuseppe, T.J. Silhavy, Signal detection and target gene induction by the CpxRA two-component system, J. Bacteriol. 185 (2003) 2432–2440. [32] K. Hiratsu, M. Amemura, H. Nashimoto, H. Shinagawa, K. Makino, The rpoE gene of Escherichia coli, which encodes Sigma E, is essential for bacterial growth at high temperature, J. Bacteriol. 177 (1995) 2918–2922. [33] B.M. Alba, J.A. Leeds, C. Onufryk, C.Z. Lu, C.A. Gross, DegS and YaeL participate sequentially in the cleavage of RseA to activate the sigma(E)dependent extracytoplasmic stress response, Genes Dev. 16 (2002) 2156–2168. [34] A. De Las Penas, L. Connolly, C.A. Gross, SigmaE is an essential sigma factor in Escherichia coli, J. Bacteriol. 179 (1997) 6862–6864. [35] M. Kleerebezem, W. Crielaard, J. Tommassen, Involvement of stress protein PspA (phage shock protein A) of Escherichia coli in maintenance of the protonmotive force under stress conditions, EMBO J. 15 (1996) 162–171. [36] P. Model, G. Jovanovic, J. Dworkin, The Escherichia coli phage-shockprotein (psp) operon, Mol. Microbiol. 24 (1997) 255–261. [37] I.B. Holland, H.E. Jones, A.K. Campbell, A. Jacq, An assessment of the role of intracellular free Ca2+ in E. coli, Biochimie 81 (1999) 901–907. [38] H.E. Jones, I.B. Holland, H.L. Baker, A.K. Campbell, Slow changes in cytosolic free Ca2+ in Escherichia coli highlight two putative influx mechanisms in response to changes in extracellular calcium, Cell Calcium 25 (1999) 265–274. [39] V.A. Rhodius, W.C. Suh, G. Nonaka, J. West, C.A. Gross, Conserved and variable functions of the sigmaE stress response in related genomes, PLoS Biol. 4 (2006) e2. [40] B.M. Alba, C.A. Gross, Regulation of the Escherichia coli sigmadependent envelope stress response, Mol. Microbiol. 52 (2004) 613–619. [41] B. Poolman, J.J. Spitzer, J.M. Wood, Bacterial osmosensing: roles of membrane structure and electrostatics in lipid–protein and protein–protein interactions, Biochim. Biophys. Acta 1666 (2004) 88–104. [42] A. Shibata, K. Maeda, H. Ikema, S. Ueno, Y. Suezaki, S. Liu, Y. Baba, I. Ueda, Local anesthetics facilitate ion transport across lipid planar bilayer membranes under an electric field: dependence on type of lipid bilayer, Colloids Surf B Biointerfaces 42 (2005) 197–203. [43] A.J. Darwin, The phage-shock-protein response, Mol. Microbiol. 57 (2005) 621–628. [44] M.E. Maxson, A.J. Darwin, Identification of inducers of the Yersinia enterocolitica phage shock protein system and comparison to the regulation of the RpoE and Cpx extracytoplasmic stress responses, J. Bacteriol. 186 (2004) 4199–4208.