Active Efflux of Organic Solvents by Pseudomonas putida S12 Is Induced by Solvents

Abstract

Induction of the membrane-associated organic solvent efflux system SrpABC of Pseudomonas putida S12 was examined by cloning a 312-bp DNA fragment, containing the srp promoter, in the broad-host-range reporter vector pKRZ-1. Compounds that are capable of inducing expression of the srpABC genes include aromatic and aliphatic solvents and alcohols. General stress conditions such as pH, temperature, NaCl, or the presence of organic acids did not induce srp transcription. Although the solvent efflux pump in P. putida S12 is a member of the resistance-nodulation-cell division family of transporters, the srpABC genes were not induced by antibiotics or heavy metals.

Several Pseudomonas putida strains possess an intrinsic resistance to a wide variety of structurally unrelated hydrophobic solvents (1, 8, 20, 29) that are lethal for most other gram-negative bacteria. The susceptibility of bacteria to hydrophobic solvents is due to the accumulation of these compounds in the membrane (25, 26), causing an adverse effect on its physicochemical properties. Solvent-resistant bacteria are able to counterbalance these effects through a variety of mechanisms mostly affecting the lipid content of the cell membrane: isomerizing cis-unsaturated fatty acids to the more rigid trans-unsaturated fatty acids (4, 5, 7), changing the head group composition of membrane fatty acids (28), increasing the phospholipid content (20), or increasing the basal rate of phospholipid synthesis (19). These adaptations of the membrane are static, acting as a physical but still permeable barrier, and cannot explain the exceptional resistance of some P. putida strains to organic solvents. Therefore, it was anticipated that a dynamic system for exporting a broad range of structurally unrelated organic solvents from the bacterial membrane had to play an essential role in solvent tolerance (28). Such an efflux system was indeed identified in P. putida S12 by means of an assay based on radiolabeled toluene (9). The genes (srpABC) for this solvent efflux system were subsequently cloned, sequenced, and shown to impart the solvent-resistant phenotype to solvent-sensitive P. putida strains (12). This efflux system shows strong homology to those of the resistance-nodulation-cell division family of transporters known to be involved in the extrusion of hydrophobic antibiotics, dyes, detergents, bile salts, heavy metals, and fatty acids from the membrane (17, 18).

Recently, several researchers also reported the involvement of active organic solvent efflux in solvent-resistant strains of Pseudomonas (2, 13, 14, 22). Induction in Pseudomonas species of efflux systems involved in either multidrug resistance or solvent tolerance has not been studied in detail.

Construction of the lacZ reporter plasmid pKRZ-srp.

Plasmid isolation, restriction analysis, ligations, electroporation, sequencing, and PCRs were performed according to standard methods as described previously (12). The region of DNA encompassing the putative promoter region of the srp operon was amplified by PCR from pJD101 (12). Primers for the PCR were 5′-GGGTCGACGCTGCTCTGGCGATGACC-3′ and 5′-GGTCTAGATCTGTCTCACGGTTTGGC-3′, which amplify a 312-bp fragment corresponding to the region immediately upstream of the srpABC genes (positions −285 to +8, where 0 is the G of the GTG start codon of srpA). The primers contain added recognition sites for SalI and XbaI, respectively. The PCR fragment was cloned into the lacZ reporter plasmid pKRZ-1 (23) cut with XbaI and SalI. The resulting plasmid, containing the sequence shown in Fig. 1, was designated pKRZ-srp.

FIG. 1.

Nucleotide sequence of the PCR-amplified SalI-XbaI fragment containing the srpABC promoter region. The terminal SalI and XbaI sites were added by incorporating their cutting sites into the PCR primers. The amino acid sequences of srpA are shown extending outward on the right side of the PCR product. This fragment was inserted into pKRZ-1 so that the promoterless lacZ gene is downstream of the start codon of srpA. An asterisk indicates the stop codon inserted into srpA by the added XbaI site.

Basal levels of srp promoter activity.

P. putida S12 strains were grown to late log phase at 30°C in 200 ml of Luria-Bertani broth (24) supplemented with 50 μg of kanamycin per ml. Cells were harvested by centrifugation at 4°C (16,000 × g, 10 min) and washed twice with 100 mM potassium phosphate buffer (pH 7.0). The washed pellet was resuspended in 2 ml of the same buffer and lysed by sonication for 5 min. Cell debris was removed by centrifuging the crude cell extract at 4°C and 20,000 × g for 20 min. β-Galactosidase activity in the extracts was determined in triplicate by the method of Miller (16). Total protein content in the extracts was determined in triplicate by the bicinchoninic acid method (27). P. putida S12 containing either the promoter probe vector or pKRZ-srp was grown with no added inducers. β-Galactosidase activity for the vector only was 0.9 ± 0.1 nmol min⁻¹ mg⁻¹ while the basal activity for the clone containing the promoter region was 3.8 ± 0.1 nmol min⁻¹ mg⁻¹. Similar levels of basal activity were observed when the strains were grown on minimal medium D (3) supplemented with 50 μg of kanamycin per ml and 20 mM glucose as the sole source of carbon and energy (0.8 and 3.9 nmol min⁻¹ mg⁻¹, respectively).

Activation of the srp promoter over time.

In order to determine the time course for induction of the srp operon, β-galactosidase activity was measured as a function of time following exposure of P. putida S12 containing either the promoter probe vector or pKRZ-srp. P. putida S12(pKRZ-srp) cells were grown in LB broth, and 3 mM toluene was added when the cells reached an optical density of 0.3 (Fig. 2). Induction of srp-lacZ expression was observed 30 min after the addition of toluene and gradually increased over time. The maximum level of induction of the srp operon was observed when the cells reached stationary phase. This clearly shows that exposure to one organic solvent results in a significant increase in transcription of the srp operon.

FIG. 2.

Activation of P. putida S12 srp promoter over time. The arrow indicates the addition of 3 mM toluene. Cell density of P. putida S12(pKRZ-srp) in LB broth (•, control; ■, induced cells) and β-galactosidase activity (○, control; □, induced cells) were monitored over time. β-Galactosidase activity in the time scale experiments was determined in triplicate cultures by the method of Miller (16) by using chloroform and sodium dodecyl sulfate to permeabilize the cells.

Activation of the srp promoter by organic solvents.

The influence of toluene concentration on gene induction was determined by varying the amounts of toluene added to the cells in the early exponential phase. β-Galactosidase activity was measured in the late exponential phase of growth (optical density of 1.0 at 600 nm). As shown in Table 1, increasing the level of toluene added to the growth medium increases the level of srp-lacZ gene induction, reaching a maximum at 6.0 mM. (The saturating limit of toluene in aqueous solutions at 30°C is 6.2 mM.) Induction by toluene thus results in a 15- to 17-fold increase in induction over basal levels.

TABLE 1.

Induction of β-galactosidase expression in P. putida S12(pKRZ-srp) by solvents

Chemical	Concn (mM)	β-Galactosidase activitya (nmol min−1 mg of protein−1)	Induction (fold)
None		3.8 ± 0.1	1.0
Aromatic solvents
Toluene	1.0	13.3 ± 0.4	3.5
	2.0	36.5 ± 0.6	9.6
	3.0	48.6 ± 5.4	12.8
	4.0	57.0 ± 7.5	15.0
	5.0	59.1 ± 3.3	15.6
	6.0	64.3 ± 6.6	16.9
Benzene	3.0	36.6 ± 0.9	9.6
Styrene	3.0	32.2 ± 3.4	8.5
p-Xylene	3.0	36.0 ± 4.9	9.5
Ethylbenzene	3.0	47.3 ± 4.1	12.4
Propylbenzene	3.0	59.0 ± 2.0	15.5
Aliphatic solvents
Pentane	1.0	30.1 ± 0.8	7.9
Hexane	1.0	25.2 ± 0.7	6.6
Heptane	1.0	34.2 ± 1.2	9.0
Octane	1.0	26.8 ± 0.1	7.1
Nonane	1.0	23.7 ± 1.0	6.2
Alcohols
1-Propanol	3.0	22.2 ± 1.2	5.8
1-Butanol	3.0	25.1 ± 0.5	6.6
1-Pentanol	3.0	38.1 ± 1.2	10.0
1-Hexanol	3.0	60.3 ± 2.1	15.9
1-Heptanol	3.0	66.1 ± 0.2	17.4
1-Octanol	3.0	46.2 ± 1.2	12.2

^aP. putida S12(pKRZ-srp) cells were grown to late exponential phase (optical density of 1.0 at 600 nm) in LB broth. β-Galactosidase was determined in cell extracts by the method of Miller (16). 

Several hydrophobic organic solvents including aromatic compounds, aliphatic compounds, and aliphatic alcohols were tested for their ability to induce the srp-lacZ construct. As can be seen by the data presented in Table 1, all of the compounds tested are able to induce the srp genes. Certain aromatic compounds and aliphatic alcohols showed the highest levels of induction. The level of induction seems to correlate with increasing side chain length in the case of the alkyl-substituted aromatics and with chain length in the case of the aliphatic alcohols (up to a 15- to 17-fold induction).

Antibiotics and heavy metals.

We previously showed that the srp operon shows a high level of similarity to proton-dependent multidrug efflux systems (12), which are known to be involved in the efflux of a variety of antibiotics and heavy metals (17, 18).

Firstly, the ability of certain antibiotics to induce the srp genes was tested. In initial experiments, we determined the MIC of each antibiotic by twofold serial dilution in LB broth. The inoculum was 2% of an overnight culture, and growth was determined by measuring the optical density at 600 nm after 12 h at 30°C. The level of β-galactosidase activity was measured in the late exponential phase of cultures exposed to a level of each antibiotic resulting in approximately 50% decrease in growth rate. Growth of P. putida S12(pKRZ-srp) in the presence of 128 μg of chloramphenicol per ml, 128 μg of ampicillin per ml, and 4 μg of tetracycline per ml resulted in only a twofold increase in induction over basal levels. No increase in induction over basal levels was observed in the presence of the other antibiotics tested: 256 μg of penicillin G per ml, 256 μg of novobiocin per ml, and 4 μg of rifampin per ml.

Secondly, the ability of heavy metals to induce the srp-lacZ fusion was determined. Cells of P. putida S12(pKRZ-srp) were grown in LB broth in the presence of six different heavy metals (added as chloride salts in a final concentration of 1 mM). Zinc, chromium, cobalt, nickel, and copper had no detectable effect on the srp promoter, while cadmium resulted in only a 1.6-fold increase in induction.

Environmental factors.

General stress conditions such as NaCl, ethanol, and stationary phase are known to induce the AcrAB efflux system in Escherichia coli (15). On this basis, it was suggested that a general regulatory mechanism exists in E. coli to prevent hydrophobic compounds from entering the bacterial cell. Such a general response was also observed in P. putida S12 in the case of the induction of cis- to trans-isomerization of the membrane unsaturated fatty acids by environmental stress such as pH and heavy metals (6). This change in fatty acid profile coincided with an increased resistance to organic solvents. In order to investigate whether the srp promoter activity was induced by these environmental factors, cells of P. putida S12(pKRZ-srp) were grown in LB broth under different conditions. Varying the growth temperature between 15 and 37°C and varying the pH between 6.0 and 8.0 did not result in a change in srp promoter induction. High levels of inorganic ions (50 g of NaCl per liter) did not affect srp-lacZ expression, although the growth of P. putida S12(pKRZ-srp) under the conditions tested was severely inhibited by high levels of NaCl.

Weber et al. (29) previously showed that incubating P. putida S12 with high acetic acid concentrations increased the survival of the strain after these cells were exposed to organic solvents. In order to determine if this was due to induction of the srp operon, the ability of acetate to induce this active efflux system for organic solvents was tested. P. putida S12(pKRZ-srp) was grown in LB broth in the presence of 20, 40, and 60 mM acetic acid (pH 6.5). Under each concentration of acetate, only a twofold induction of the srp-lacZ construct was observed. These results suggest that the enhanced survival of acetic-acid-adapted cells is not due to activation of the SrpABC efflux system.

Conclusions.

The data in the present paper clearly shows that the srpABC operon for the efflux of organic solvents is induced by lipophilic aromatic and aliphatic solvents and alcohols. The data suggests that neither aromaticity nor charge is required for organic solvents to act as an inducer. Unlike other efflux systems, the srpABC operon is not induced by environmental stress or heavy metals, demonstrating that the genes are specifically induced and are not the result of a general regulatory mechanism as described elsewhere for E. coli (15).

The specific induction by solvents is underlined by the observation that hydrophobic antibiotics do not induce the SrpABC efflux system. P. putida S12 only becomes more resistant to hydrophobic antibiotics by preculturing cells in the presence of toluene, while P. putida S12 pregrown in the presence of antibiotics does not elicit the solvent-tolerant phenotype (10). These observations suggest that hydrophobic antibiotics are removed from the membrane in solvent-induced cells, while these hydrophobic antibiotics do not induce SrpABC-mediated efflux of hydrophobic organic solvents.

Interestingly, Ramos et al. very recently reported the existence of at least two efflux pumps in the solvent-resistant P. putida DOT-T1E. One system apparently was expressed constitutively, while a second system was inducible (22).

Although the natural function of these resistance-nodulation-cell division-type efflux systems has not been clarified yet, our data suggests that the SrpABC system plays a protective role in resistance to a wide variety of structurally unrelated hydrophobic compounds. Multidrug efflux systems in P. aeruginosa are not induced by known substrates of these efflux pumps (21), suggesting that the extrusion by these pumps of hydrophobic compounds is a nonspecific action. At present, the natural substrate(s) of these systems remains unknown. The SrpABC efflux system apparently does not have a natural function in excreting hydrophobic antibiotics because (i) solvent-sensitive mutants of P. putida S12 have normal levels of antibiotic resistance (11) and (ii) the srpABC operon is induced solely by solvent stress (this work).