Expression Analysis of Up-Regulated Genes Responding to Plumbagin in Escherichia coli
http://www.100md.com
《细菌学杂志》
Institute of Microbiology and Immunology,Institute of Biochemistry and Molecular Biology, National Yang-Ming University,National Research Institute of Chinese Medicine,Department of Research and Development, U-Vision Biotech Inc., Taipei, Taiwan, Republic of China
ABSTRACT
Plumbagin is found in many medicinal plants and has been reported to have antimicrobial activities. We examined the molecular responses of Escherichia coli to plumbagin by using a proteomic approach to search for bacterial genes up-regulated by the drug. The protein profile obtained was compared with that of E. coli without the plumbagin treatment. Subsequent analyses of the induced proteins by mass spectroscopy identified several up-regulated genes, including ygfZ, whose function has not been defined. Analyses of the 5'-flanking sequences indicate that most of these genes contain a marbox-like stretch, and several of them are categorized as members of the mar/sox regulon. Representatives of these genes were cloned into plasmids, and the marbox-like sequences were modified by site-directed mutagenesis. It was proven that mutations in these regions substantially repressed the level of proteins encoded by the downstream genes. Furthermore, plumbagin's early effect was demonstrated to robustly induce SoxS rather than MarA, an observation distinctly different from that seen with sodium salicylate.
INTRODUCTION
Plumbagin (5-hydroxy-2-methyl-1,4-naphthoquinone) is a naphthoquinone having antibacterial (6), antifungal (5), anticancer (22), and antimutagenic activities (7). Like other redox-cycling chemicals such as paraquat and menadione, plumbagin has been used as an agent to generate superoxide or reactive oxygen species in order to study oxidative stress (14). Plumbagin has been suggested to activate SoxS by oxidizing the SoxR molecule (10, 21) or inhibiting the repression of MarR (2), the effect of which is exerted on marA, resulting in activation of the mar/sox regulon. In addition, it has been observed that the expression of some members of the mar/sox regulon in Escherichia coli, such as sodA (8), nfo (3), ribA (20), and pqi (21), is up-regulated by treatment with plumbagin.
Although plumbagin could induce excessive expression of superoxide dismutase and catalase, overexpression of sodA failed to protect E. coli (17). The toxic effect of plumbagin may not simply result from the production of reactive oxygen species. It has been reported that plumbagin inhibits NADH dehydrogenase, as well as causing respiratory arrest (17). Plumbagin has also been shown to modify the lactose carrier and inhibit its binding with galactoside; the modified carrier then becomes completely inactive (29). The above effects appeared more or less to result directly from the chemical nature of plumbagin. In this report, we focus on the responses of the bacteria to the chemical, in which multiple proteins were simultaneously induced by plumbagin treatment. We report the evaluation of the bacterial regulatory systems after treating E. coli with plumbagin.
MATERIALS AND METHODS
Bacterial strains, plasmids, and growth conditions. E. coli JM109 was used as the major experimental strain for the chemical treatments and the cloning host. Plasmids pGEM-T-easy (Promega, Madison, WI), pBluescript II SK+ (Stratagene, La Jolla, CA), and pQE60 (QIAGEN, Valencia, CA) were used as general cloning vectors. Plasmid pMH was modified from pQE60 by deleting the T5 promoter. Bacteria were cultivated in tubes containing Luria-Bertani medium. To prepare proteins for two-dimensional gel electrophoresis (2-DE), bacteria (80 ml) were grown at 37°C with vigorous shaking in 250-ml flasks. To measure the MIC, either an agar diffusion assay or the liquid broth method was used and the standard procedure recommended by the Clinical and Laboratory Standards Institute (formerly the National Committee for Clinical Laboratory Standards) was followed. Plumbagin (Sigma, St. Louis, MO) was dissolved in dimethyl sulfoxide (DMSO) or methanol with a stock concentration at 10 mg/ml without further purification.
2-DE. Bacteria from overnight cultures were diluted 100-fold into 80 ml of Luria-Bertani medium. Bacteria were grown with aeration at 37°C to an A600 of 0.2. Plumbagin was added to make a final concentration of 25 μg/ml (0.13 mM), and the culture was further agitated at 37°C for 2 h. Protein extraction was performed as previously described (12), except for a slight modification. In brief, bacteria were harvested by centrifugation and washed twice with cold 0.9% NaCl. After resuspension in 3 ml of TSD buffer (28 mM Tris HCl, 22 mM Tris, 0.3% sodium dodecyl sulfate, 200 mM dithiothreitol) and addition of 0.3 g of glass beads (0.1 to 0.25 mm in diameter), the bacteria were broken by vigorous vortexing for 10 min. Unbroken bacteria were removed by centrifugation, and the supernatant was boiled for 5 min. Chilled samples received 150 μl of TM buffer (24 mM Tris, 476 mM Tris HCl, 50 mM MgCl2) containing DNase I (1 mg/ml) and RNase A (0.25 mg/ml). The enzymatic digestion was stopped after 30 min by adding 4 volumes of ice-cold acetone, and samples were incubated overnight at –20°C. Proteins were collected by centrifugation at 11,000 x g for 15 min and then dissolved in 200 μl of isoelectric focusing (IEF) sample buffer containing 8 M urea, 4% 3-[(3-cholamidopropyl)-dimethylammonio]-1-propanesulfonate (CHAPS), and 40 mM Tris base. The first-dimension (1-D) IEF was performed with the Ettan IPGphor II IEF System, and 2-DE was carried out with an SE260 electrophoresis apparatus (Amersham Biosciences, Piscataway, NJ).
Western blotting. The bacterial culture and plumbagin treatment were similar to those used for 2-DE. The proteins were separated by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) and transferred onto a nitrocellulose membrane. To detect ectopically expressed His6-tagged proteins, a rabbit polyclonal antibody specific for the tag (Bethyl, Montgomery, TX) was used as the primary antibody. A horseradish peroxidase-conjugated goat anti-rabbit antibody (Sigma) was used as the secondary antibody. The membranes were finally developed by a chemiluminescence method (16). The images developed on films were quantitatively monitored with Personal Densitometer SI (Molecular Dynamics).
Plasmid construction. To construct pMH-mdaB, a 1,590-bp fragment covering the entire mdaB gene and its 5'-flanking region was obtained by PCR amplification with the genomic DNA of E. coli as the template and primers NmdabF (CGGGATCCCCGC TATTGCTAAGTTTG) and NmdabR (AACTGCAGTTTCCACAAGATGCTTGCG). The PCR product was digested with BamHI/PstI and ligated into pMH that had been previously digested with the same enzymes. The resulting plasmid was named pMH-mdaB, and the encoded MdaB protein had a hexahistidine extension at the C terminus. Plasmids pMH-ahpC, pMH-gatY, pMH-nfnB, pMH-nfo, pMH-tpx, pMH-talB, pMH-ygfZ, pMH-ygfZ', pMH-ygaG, pMH-marRA, and pMH-soxRS were similarly generated by the same strategy, except that the primer pairs used were NahpcF (CGGGATCCTCGTCCATCAGTTTCTC)-NahpcR (AAGTCGACTTTACCAACCAGGTCC), NgatyF (CGGGATCCTTATGCGCATCGCATTC)-NgatyR (AACTGCAGCACAATCGAATCAC), NnfnbF (GCGGATCCTGGCCCTTGAGTTACCC)-NnfnbR (GCGGATCCATTCGTTAGGTGATG), NnfoF (CAGGATCCACATCGTACACTGTGGC)-NnfoR (GGGAGGCTACCGCTTTTTCAG), NtpxF (CCCGGATCCTAAAGATGCAATTCGC C)-NtpxR (CCACTGCAGGTGCTTTCAGTACAGCC), NtalbF (TCCCGGGCTGCCAGCCAAAAGCA)-NtalbR (GGAAGATCTCAGCAGATCGCCGATC), NygfzF (GCGGATCCCAAACGCCCGTCCATCAACAAGCG)-NygfzR (GCGGATCCGGCGGTGGATCTTAGCCAGCGCC), Nygfz'F (CGGGATCCCGAAATTAATTCAC)-NygfzR, NygagF (GCGTCGAAGTGCGTTCGCTGGACATCAACCCG)-NygagR (GCGGATCCGGATATGCGGCTGGTAACCTTC), NmarraF (GGGCTGCAGGCCAATTGCTTTAAACAAATC)-NmarraR (GGGAGATCTGCTGTTGTAATGATTTAATGG), and NsoxrsF (CCTCTGCAGGCCCTTGTGGCGCTTTAG)-NsoxrsR (CCTAGATCTCAGGCGGTGGCGATAATCG), respectively.
To construct pACYC-soxRS, which expresses SoxS from a plasmid compatible with pMH, the soxRS-containing fragment was obtained by treating pMH-soxRS with BamHI and HindIII and cloned into pACYC184 that had been digested with the same enzymes.
Site-directed mutagenesis. The 5'-flanking region of mdaB was PCR amplified in a similar way from E. coli genomic DNA with primer NmdabF paired with NmdabMR (CAAAATGCGCAAAGTCTTTTG) or NmdabMF (CAAAAGACTTTGCGCATTTTG) paired with NmdabR. The two PCR products separately amplified were mixed. Since NmdabMR and NmdabMF embedding the designated point mutations are complementary in sequence, the two DNA products were annealed at one end and extended by Vent polymerase and deoxynucleoside triphosphates. After 10 reaction cycles, NmdabF and NmdabR were added to the mixture and PCR amplification was continued for an additional 30 cycles. The PCR product was then digested with BamHI and PstI and cloned into pMH, resulting in plasmid pMH-mdaBm. Mutation-bearing plasmids pMH-nfom and pMH-ygfZ'm were constructed by a similar strategy; the paired primers used accordingly for the mutagenesis were NnfoMF (GTGATTCAAAGCGTCATTCT AGAAACCAC)-NnfoMR (GCAAGATGTAGTGGTTTCTAGAATGACGC) and NygfzMF (GTCGCCTCGCAGAAGGTACCGATCGCGTAGTG)-NygfzMR (AGTGCTTTCACTACGCGATCGGTACCTTCTGCG), respectively.
Q-TOF mass spectroscopy. In-gel tryptic digestion was performed as previously described (30), except for a slight modification. In brief, Coomassie blue dye-stained protein spots excised from a distilled-water-washed gel were cut into small pieces. Gel pieces were dehydrated and destained with three washes with 50 mM ammonium bicarbonate mixed with 50% (vol/vol) acetonitrile. After dehydration and in-gel digestion with trypsin, the peptides recovered were analyzed with a quadrupole time-of-flight (Q-TOF) mass spectrometer (Waters, Milford, MA). Data from the mass profiling were used to search the NR database with Mascot (http://www.matrix-science.com).
Construction of a soxS mutant strain. To delete soxS from the chromosome by homologous recombination (4), E. coli strain JM110 was used as the parental strain since strain JM109 lacks recA and has a low frequency of recombination. To perform the deletion, both the 5'- and 3'-flanking regions of soxS were PCR amplified from E. coli genomic DNA. The primer pairs used were 5'NsoxSF (GCGGATCCGCATCAACACCAACCGGAACC)-5'NsoxSR (GCGAATTCGTCAATCCATGCGATAAGATCC) and 3'NsoxSF (GCGAATTCCTGTAATTTTATTGCCCGCGCG)-3'NsoxSR (GCGGATCCTTCGACAAAACGCCCAATTGC), respectively. The PCR products were then cloned into pGEM-T Easy and confirmed after sequencing. These fragments were then excised from the plasmids with BamHI/EcoRI and three-way ligated with BamHI-restricted pKO3 (25), yielding pKO3-soxS. pKO3-soxS was then transformed into strain JM110, and the resulting transformants were sequentially selected in chloramphenicol- and sucrose-containing media. The soxS mutant strain obtained was confirmed by PCR amplification of the region spanning the target site, followed by restriction enzyme analysis and sequence confirmation.
RESULTS
Proteins whose expression is up-regulated in E. coli by plumbagin. To investigate the response of the bacteria, E. coli was treated with plumbagin at a subinhibitory concentration. Plumbagin was first determined to have an MIC against E. coli JM109 of 50 μg/ml. Thus, the plumbagin treatment was held at 25 μg/ml for 2 h and the total bacterial proteins were extracted and separated by 2-DE. Proteins on the gel were stained with Coomassie blue dye and compared with those prepared from bacteria without plumbagin treatment. Twelve protein spots whose expression was obviously increased by plumbagin treatment (Fig. 1) were cut out and digested with trypsin. The tryptic peptides eluted had their amino acid sequences deduced by Q-TOF mass spectroscopy. The peptides so deduced were used to search for matching proteins derived from the E. coli genome. Figure 2A shows a typical result of three tryptic peptides matched to the E. coli alkyl hydroperoxide reductase (AhpC). Results from similar identifications are listed in Fig. 2B.
Plumbagin-responding genes. Among the 12 deduced proteins, SodA was massively induced by plumbagin and could simply be identified by 1-D SDS-PAGE, followed by Coomassie blue dye staining. Among the remaining 11 protein spots, although the 2-D gel profiling of the proteins was reproducible, misleading identification could not be completely excluded due to the possibility of overlapping proteins. Therefore, an approach to augment the plumbagin induction signal information and to offer an alternative method of confirming the responsiveness of the genes was explored. Theoretically, when a gene is cloned into a plasmid with its promoter, the responsiveness of the encoded proteins would be amplified due to the multiple-copy property of the plasmid. Therefore, the genes identified above together with a 5' region, ranging from 425 bp to 1,100 bp upstream of the open reading frame, were PCR amplified. The obtained DNA fragments were separately cloned into the pMH vector, in which every single target protein is tagged with His6 at the C terminus. Proteins displayed on the 1-D system could be simply compared by Western blotting with anti-His6 tag antibody.
A bacterial transformant harboring plasmid pMH-mdaB treated with plumbagin was compared with those treated with DMSO alone. The MdaB protein increased about 2.5-fold (Fig. 3A) in the presence of plumbagin, given that the total proteins were loaded in comparable amounts (OmpC was used as an internal control; data not shown). Gene mdaB is known as "modulator of drug activity B" and has been recently proposed to function as an NADPH quinone reductase inducible by menadione (15, 19). Our results further demonstrated that mdaB also responded to plumbagin. The nfo gene encodes endonuclease IV and participates in the repair of DNA lesions induced by H2O2 (9). nfo's response to plumbagin (Fig. 3B) is consistent with observations previously reported (3). An unknown protein encoded by ygfZ was identified as being upregulated for the first time (Fig. 3C). Expressions of AhpC protein (Fig. 3D), a peroxiredoxin (alkyl hydroperoxide reductase) offering antioxidant protection, and NfnB, an oxygen-insensitive NAD(P)H nitroreductase (Fig. 3E), were both confirmed to be up-regulated. Protein derived from ahpC was observed with an extra product which was twice as large as expected (Fig. 3D). This fact suggests that AhpC may form a dimer structure or associate with another molecule(s) when bacteria encounter oxidative stress. The induced protein pattern detected with NfnB was striking. At least five additional high-molecular-weight bands above the expected product were observed (Fig. 3E). These high-molecular-weight aggregates of NfnB have not been reported before. Multiple high-molecular-weight products were also seen when gatY was examined in the same manner in the presence of plumbagin (data not shown).
The positive responsiveness of the genes fldA, encoding flavodoxin I, and yggX, encoding a protein involved in Fe(II) trafficking to minimize DNA damage (13), was also confirmed in a similar way (data not shown). In contrast, three other genes listed in Fig. 2B, namely, talB, tpx, and ygaG, could not be confirmed and were therefore excluded from further investigation.
Marbox sequence contributed to plumbagin responsiveness. To explore the mechanism of genes' up-regulation upon plumbagin treatment, the 5'-flanking regions of these genes were compared. A stretch of sequence containing a putative marbox was observed in the regions upstream of fldA, mdaB, nfnB, nfo, sodA, and yggX. These sequences have been proposed to be the controlling elements, but so far only that of yggX has been characterized by binding of SoxS to a promoter-containing fragment and those of mdaB and nfnB have been mapped by fragment deletion in a promoter assay (28, 32). It is worth noting that those characterizations were not based upon the effects of plumbagin. Instead, they were based on observations with paraquat, 4,4'-dipyridyl, or salicylate.
To confirm the effects of plumbagin treatment on gene expression, a mutation was introduced into the mapped marbox sequence upstream of mdaB (28, 32). In the pMH-mdaB construct, the putative core element GCAC of the marbox (Fig. 4A) was replaced with GCGC, resulting in plasmid pMH-mdaBm. MdaB expression was then compared. The basal expression level of MdaB from pMH-mdaBm was fivefold decreased relative to that from pMH-mdaB (compare lanes 1 and 3 in Fig. 4B). Upon addition of plumbagin, the expression of MdaB from pMH-mdaB was 2.5-fold higher than when the solvent was added alone (lane 1 versus lane 2). In contrast, the increase in MdaB from the pMH-mdaBm transformant was limited to 1.3-fold when plumbagin was added (lanes 3 and 4). Therefore, a single base mutation of the marbox sequence drastically decreased the basal level of MdaB expression, as well as the plumbagin induction effect. In a similar approach, the proposed marbox of nfo (28, 32) was mutated (Fig. 4A) and the core-containing element CGCAT in pMH-nfo was mutated to TCTAG in pMH-nfom. The basal expression level of Nfo from pMH-nfom was lower than that derived from pMH-nfo (compare lanes 1 and 3 in Fig. 4C), and a 2.4-fold decrease was observed. On treatment with plumbagin, the induction effect of plumbagin was reduced from a 3.5-fold difference in the wild type to a 1.1-fold difference in the mutant (compare lanes 1 and 2 with lanes 3 and 4).
To examine the element conferring the plumbagin induction effects on the uncharacterized ygfZ gene, the sequence upstream of ygfZ was first analyzed with the Emboss Fuzznuc program (http://www.hgmp.mrc.ac.uk/Software/EMBOSS), and several marbox-like sequences were found in this region. To facilitate the identification of the plumbagin-responsive element, pMH-ygfZ was reconstructed to generate a shortened 5'-flanking region in pMH-ygfZ' (Fig. 5A). Compared to pMH-ygfZ, the plumbagin-induced expression of ygfZ was retained in pMH-ygfZ' (lanes 1 to 4, Fig. 5C). This result suggested that the plumbagin-responsive element exists in the 151-bp region upstream of the ygfZ-coding sequence. Analysis of this sequence in detail found a zwf-like class I marbox promoter structure (27, 36) that is characterized by a space of 7 bp between a putative marbox and a putative –35 hexamer. Therefore, the core-containing stretch, GCACA, in the putative marbox sequence was replaced with GTACC (see sequence alignment in Fig. 5B), and the expression levels of YgfZ before and after the sequence alteration were analyzed (Fig. 5C). The basal expression level of YgfZ from pMH-ygfZ'm was suppressed (compare lanes 3 and 5) by this two-base mutation compared to that from pMH-ygfZ', and the plumbagin induction effect was abolished as well (compare lanes 5 and 6). The total effect is even obvious when lanes 4 and 6 are compared.
SoxS is involved in early plumbagin induction. Previous studies suggested that two marbox-binding proteins, SoxS and MarA, might be induced directly by treatment with plumbagin (2, 10). To differentiate their contributions, two plasmids containing their intrinsic control elements together with the appropriate marA and soxS genes were separately constructed. In pMH-marRA, the complete marRA locus was inserted into pMH and the product of marA was fused with a His6 tag. The expressed MarA protein was then followed by Western blotting with anti-His6 as in Fig. 3. Similarly, pMH-soxRS was constructed to detect the expression of SoxS. Bacteria transformed with pMH-marRA and pMH-soxRS, respectively, were treated with plumbagin, and the kinetics of MarA and SoxS expression were followed by Western blotting. Figure 6A shows that SoxS was robustly induced and strongly detected as early as 5 min after plumbagin treatment (Fig. 6A). In contrast, the induction of MarA was negligible over the first 10 min and slowly increased thereafter. Figure 6B, with a graphic illustration, further sketches out the concept that SoxS is induced first in the early phase of the bacterial response to plumbagin. After 30 min, the level of SoxS reached a plateau while the level of MarR had a limited increase. With the MarA level induced by sodium salicylate (33) as a reference, the 60-min plumbagin treatment only reached about 60% of the level induced by salicylate. This slow MarA induction is unique to plumbagin treatment since the same bacteria gave a rapid plateau in response to sodium salicylate treatment for less than 30 min (Fig. 6C).
SoxS is involved in the plumbagin induction of some genes but not all. The robust induction of SoxS by plumbagin suggests a role for SoxS in the activation of these gene. Since our proteomic analysis was based upon bacteria after a 2-h induction, gene activation conferred by other late events may also be possible. To distinguish the genes controlled by SoxS from those controlled by other proteins, we constructed a soxS deletion mutant and repeated the experiments involving treatment with plumbagin. Figure 7A shows that a soxS deletion mutant gave a basal level of Nfo similar to that found with the parental strain (compare lanes 1 and 3). However, when encountering plumbagin, the soxS mutant totally lost the response (Fig. 7A, compare lanes 2 and 4). After complementation with a SoxS-expressing plasmid, pACYC-soxRS, the plumbagin responsiveness was completely restored (Fig. 7A, lanes 3 to 6). Therefore, plumbagin's effect on nfo was conceivably through a direct involvement of SoxS. Similar to nfo, ygfZ and sodA lost their reactions toward plumbagin when bacterial soxS was deleted (data not shown). Although not tested, fldA and yggX are presumably in the same category since they have been known to be regulated by SoxS (32, 37).
On the other hand, mdaB, nfnB, and ahpC substantially differed from nfo. Their products remained plumbagin inducible when soxS was deleted. A typical example with MdaB is shown in Fig. 7B.
DISCUSSION
The antibacterial effect of plumbagin against E. coli was observed under aerobic conditions. When plumbagin was present at one-half of the MIC, the bacteria continued to grow, but at a slow rate, and many gene products of E. coli were up-regulated. Twelve proteins significantly elevated have been identified by proteomic analysis, and nine of the genes involved have been reconfirmed by an alternative approach. Intriguingly, more than one-half of plumbagin-responding genes could be functionally correlated with antioxidation, physiological reduction, and DNA lesion repair (1, 8, 11, 23, 24, 32, 34). Considering that plumbagin has no obvious effect on E. coli growth in the absence of oxygen and that there was no recognizable difference between the 2-DE protein profiles of E. coli cells treated with and without plumbagin (data not shown), the toxic effect of plumbagin on E. coli must be reconsidered.
The elevated expression of the genes fldA, mdaB, nfnB, nfo, sodA, and yggX caused by plumbagin in our study was also observed in a microarray study when bacteria were treated with paraquat (31). These genes have been defined as members of the mar/sox regulon, and their regulation has been proposed to occur through marbox sequences that share conserved cores with otherwise degenerated sequences (28). We have mutated the marbox sequences in front of mdaB and nfo and proved that such mutations indeed decreased the basal level of the proteins expressed, as well as their responses to plumbagin. We rationalize that fldA, nfnB, sodA, and yggX may be regulated by similar cis elements. As to trans factors, the plumbagin induction effects on nfo, fldA, sodA, and yggX were found to be exclusively dependent upon SoxS. Additional regulation at the posttranscriptional level could not be completely excluded. On the other hand, the elevated MdaB and NfnB protein levels caused by plumbagin were apparently not affected regardless of the presence or absence of soxS.
In our study, gatY was also up-regulated, and this induction has not been observed with paraquat (28). Since the expression of genes involved in sugar transport has been elevated when E. coli encounters superoxide stress (28, 31) and since gatY has been known to encode a protein involved in carbohydrate utilization, it is not surprising to see the up-regulation of gatY by plumbagin. However, no marbox-like sequence was found proximal to gatY previously (28) and we were also unable to identify a related sequence. Therefore, how this gene is molecularly regulated remains to be explored.
A gene whose function is unknown that was identified in our plumbagin treatment is ygfZ; the crystal structure of the protein it encodes has been reported, and a folate-dependent regulatory role in one-carbon metabolism has been hypothesized. However, its physiological role and real function remain unclear (35). We found that the ygfZ gene contains a zwf-like class I marbox structure, and mutation of this sequence did decrease the basal protein level, as well as its response to plumbagin, a reaction that was also soxS dependent. Therefore, these new findings suggest that ygfZ may also be involved in releasing oxidative stress.
ahpC has been reported to be regulated by OxyR (34), and it is also stimulated by plumbagin. This fact strongly suggests that plumbagin not only generates superoxide stress but also provokes the production of peroxide ions to trigger a mixed set of responses in E. coli. In our soxS deletion mutant, the plumbagin effect with an increasing AhpC level remained evident, a fact suggesting that SoxS does not act as a mediator. It is then suggested that the readily induced MarA protein, albeit at a slow response, and Rob (18), which is known to bind to sites with similar sequences and was not monitored here, may contribute to the increase in AhpC. The same reasoning may be valid with the findings on MdaB and NfnB, on neither of which the plumbagin effect was SoxS dependent. Whether the above conclusions found with E. coli are applicable to bacteria that are sensitive to plumbagin under both aerobic and anaerobic conditions remains to be explored.
Under our plumbagin treatment conditions, bacterial proteins whose levels were increased appeared to outnumber those being suppressed (Fig. 1). These studies of up-regulated proteins do not necessarily imply that those that are down-regulated are not important for bacterial survival. As proteomic investigations increase both protein resolution and staining sensitivity, it is expected that more protein spots on either the up-regulated or the down-regulated side will be unveiled. By grouping responsive genes and gene products, how bacteria use different mechanisms to resolve the same chemical stress may be better understood.
ACKNOWLEDGMENTS
We thank J. C. W. Lio for the gift of pMH, J. Y. Ho for operating the mass spectrometer, and C. M. Tzeng of U-Vision Biotech for continuous encouragement. We also thank R. Kirby for critically reading the manuscript.
This research was supported in part by grant 89-B-FA22-2-4 (Program for Promoting Academic Excellence of Universities) from the Ministry of Education and grant NSC 94-2320-B-010-034 from the National Science Council, Taiwan, Republic of China.
REFERENCES
Adams, M. A., and Z. Jia. 2005. Structural and biochemical evidence for an enzymatic quinone redox cycle in Escherichia coli: identification of a novel quinol monooxygenase. J. Biol. Chem. 280:8358-8363.
Alekshun, M. N., and S. B. Levy. 1999. Alteration of the repressor activity of MarR, the negative regulator of the Escherichia coli marRAB locus, by multiple chemicals in vitro. J. Bacteriol. 181:4669-4672.
Chan, E., and B. Weiss. 1987. Endonuclease IV of Escherichia coli is induced by paraquat. Proc. Natl. Acad. Sci. USA 84:3189-3193.
Chiu, H. J., and W. J. Syu. 2005. Functional analysis of EspB from enterohaemorrhagic Escherichia coli. Microbiology 151:3277-3286.
Curreli, N., F. Sollai, L. Massa, O. Comandini, A. Rufo, E. Sanjust, A. Rinaldi, and A. C. Rinaldi. 2001. Effects of plant-derived naphthoquinones on the growth of Pleurotus sajor-caju and degradation of the compounds by fungal cultures. J. Basic Microbiol. 41:253-259.
de Paiva, S. R., M. R. Figueiredo, T. V. Aragao, and M. A. Kaplan. 2003. Antimicrobial activity in vitro of plumbagin isolated from Plumbago species. Mem. Inst. Oswaldo Cruz 98:959-961.
Edenharder, R., and X. Tang. 1997. Inhibition of the mutagenicity of 2-nitrofluorene, 3-nitrofluoranthene and 1-nitropyrene by flavonoids, coumarins, quinones and other phenolic compounds. Food Chem. Toxicol. 35:357-372.
Fee, J. A. 1991. Regulation of sod genes in Escherichia coli: relevance to superoxide dismutase function. Mol. Microbiol. 5:2599-2610.
Galhardo, R. S., C. E. Almeida, A. C. Leitao, and J. B. Cabral-Neto. 2000. Repair of DNA lesions induced by hydrogen peroxide in the presence of iron chelators in Escherichia coli: participation of endonuclease IV and Fpg. J. Bacteriol. 182:1964-1968.
Gaudu, P., N. Moon, and B. Weiss. 1997. Regulation of the soxRS oxidative stress regulon. Reversible oxidation of the Fe-S centers of SoxR in vivo. J. Biol. Chem. 272:5082-5086.
Gaudu, P., and B. Weiss. 2000. Flavodoxin mutants of Escherichia coli K-12. J. Bacteriol. 182:1788-1793.
Giard, J. C., A. Rince, H. Capiaux, Y. Auffray, and A. Hartke. 2000. Inactivation of the stress- and starvation-inducible gls24 operon has a pleiotrophic effect on cell morphology, stress sensitivity, and gene expression in Enterococcus faecalis. J. Bacteriol. 182:4512-4520.
Gralnick, J. A., and D. M. Downs. 2003. The YggX protein of Salmonella enterica is involved in Fe(II) trafficking and minimizes the DNA damage caused by hydroxyl radicals: residue CYS-7 is essential for YggX function. J. Biol. Chem. 278:20708-20715.
Hassan, H. M., and I. Fridovich. 1979. Intracellular production of superoxide radical and of hydrogen peroxide by redox active compounds. Arch. Biochem. Biophys. 196:385-395.
Hayashi, M., K. Hasegawa, Y. Oguni, and T. Unemoto. 1990. Characterization of FMN-dependent NADH-quinone reductase induced by menadione in Escherichia coli. Biochim. Biophys. Acta 1035:230-236.
Hsu, S. C., H. P. Lin, J. C. Wu, K. L. Ko, I. J. Sheen, B. S. Yan, C. K. Chou, and W. J. Syu. 2000. Characterization of a strain-specific monoclonal antibody to hepatitis delta virus antigen. J. Virol. Methods 87:53-62.
Imlay, J., and I. Fridovich. 1992. Exogenous quinones directly inhibit the respiratory NADH dehydrogenase in Escherichia coli. Arch. Biochem. Biophys. 296:337-346.
Jair, K. W., X. Yu, K. Skarstad, B. Thony, N. Fujita, A. Ishihama, and R. E. Wolf, Jr. 1996. Transcriptional activation of promoters of the superoxide and multiple antibiotic resistance regulons by Rob, a binding protein of the Escherichia coli origin of chromosomal replication. J. Bacteriol. 178:2507-2513.
Jorgensen, M. A., M. A. Trend, S. L. Hazell, and G. L. Mendz. 2001. Potential involvement of several nitroreductases in metronidazole resistance in Helicobacter pylori. Arch. Biochem. Biophys. 392:180-191.
Koh, Y. S., J. Choih, J. H. Lee, and J. H. Roe. 1996. Regulation of the ribA gene encoding GTP cyclohydrolase II by the soxRS locus in Escherichia coli. Mol. Gen. Genet. 251:591-598.
Koh, Y. S., and J. H. Roe. 1995. Isolation of a novel paraquat-inducible (pqi) gene regulated by the soxRS locus in Escherichia coli. J. Bacteriol. 177:2673-2678.
Krishnaswamy, M., and K. K. Purushothaman. 1980. Plumbagin: a study of its anticancer, antibacterial and antifungal properties. Indian J. Exp. Biol. 18:876-877.
Levin, J. D., A. W. Johnson, and B. Demple. 1988. Homogeneous Escherichia coli endonuclease. IV. Characterization of an enzyme that recognizes oxidative damage in DNA. J. Biol. Chem. 263:8066-8071.
Lightfoot, R. T., D. Shuman, and H. Ischiropoulos. 2000. Oxygen-insensitive nitroreductases of Escherichia coli do not reduce 3-nitrotyrosine. Free Radic. Biol. Med. 28:1132-1136.
Link, A. J., D. Phillips, and G. M. Church. 1997. Methods for generating precise deletions and insertions in the genome of wild-type Escherichia coli: application to open reading frame characterization. J. Bacteriol. 179:6228-6237.
Martin, R. G., W. K. Gillette, S. Rhee, and J. L. Rosner. 1999. Structural requirements for marbox function in transcriptional activation of mar/sox/rob regulon promoters in Escherichia coli: sequence, orientation and spatial relationship to the core promoter. Mol. Microbiol. 34:431-441.
Martin, R. G., W. K. Gillette, and J. L. Rosner. 2000. Promoter discrimination by the related transcriptional activators MarA and SoxS: differential regulation by differential binding. Mol. Microbiol. 35:623-634.
Martin, R. G., and J. L. Rosner. 2002. Genomics of the marA/soxS/rob regulon of Escherichia coli: identification of directly activated promoters by application of molecular genetics and informatics to microarray data. Mol. Microbiol. 44:1611-1624.
Neuhaus, J. M., and J. K. Wright. 1983. Chemical modification of the lactose carrier of Escherichia coli by plumbagin, phenylarsinoxide or diethylpyrocarbonate affects the binding of galactoside. Eur. J. Biochem. 137:615-621.
Phan-Thanh, L., and F. Mahouin. 1999. A proteomic approach to study the acid response in Listeria monocytogenes. Electrophoresis 20:2214-2224.
Pomposiello, P. J., M. H. Bennik, and B. Demple. 2001. Genome-wide transcriptional profiling of the Escherichia coli responses to superoxide stress and sodium salicylate. J. Bacteriol. 183:3890-3902.
Pomposiello, P. J., A. Koutsolioutsou, D. Carrasco, and B. Demple. 2003. SoxRS-regulated expression and genetic analysis of the yggX gene of Escherichia coli. J. Bacteriol. 185:6624-6632.
Sulavik, M. C., L. F. Gambino, and P. F. Miller. 1995. The MarR repressor of the multiple antibiotic resistance (mar) operon in Escherichia coli: prototypic member of a family of bacterial regulatory proteins involved in sensing phenolic compounds. Mol. Med. 1:436-446.
Tartaglia, L. A., G. Storz, and B. N. Ames. 1989. Identification and molecular analysis of oxyR-regulated promoters important for the bacterial adaptation to oxidative stress. J. Mol. Biol. 210:709-719.
Teplyakov, A., G. Obmolova, E. Sarikaya, S. Pullalarevu, W. Krajewski, A. Galkin, A. J. Howard, O. Herzberg, and G. L. Gilliland. 2004. Crystal structure of the YgfZ protein from Escherichia coli suggests a folate-dependent regulatory role in one-carbon metabolism. J. Bacteriol. 186:7134-7140.
Wood, T. I., K. L. Griffith, W. P. Fawcett, K. W. Jair, T. D. Schneider, and R. E. Wolf, Jr. 1999. Interdependence of the position and orientation of SoxS binding sites in the transcriptional activation of the class I subset of Escherichia coli superoxide-inducible promoters. Mol. Microbiol. 34:414-430.
Zheng, M., B. Doan, T. D. Schneider, and G. Storz. 1999. OxyR and SoxRS regulation of fur. J. Bacteriol. 181:4639-4643.(Jenn-Wei Chen, Chang-Ming)
ABSTRACT
Plumbagin is found in many medicinal plants and has been reported to have antimicrobial activities. We examined the molecular responses of Escherichia coli to plumbagin by using a proteomic approach to search for bacterial genes up-regulated by the drug. The protein profile obtained was compared with that of E. coli without the plumbagin treatment. Subsequent analyses of the induced proteins by mass spectroscopy identified several up-regulated genes, including ygfZ, whose function has not been defined. Analyses of the 5'-flanking sequences indicate that most of these genes contain a marbox-like stretch, and several of them are categorized as members of the mar/sox regulon. Representatives of these genes were cloned into plasmids, and the marbox-like sequences were modified by site-directed mutagenesis. It was proven that mutations in these regions substantially repressed the level of proteins encoded by the downstream genes. Furthermore, plumbagin's early effect was demonstrated to robustly induce SoxS rather than MarA, an observation distinctly different from that seen with sodium salicylate.
INTRODUCTION
Plumbagin (5-hydroxy-2-methyl-1,4-naphthoquinone) is a naphthoquinone having antibacterial (6), antifungal (5), anticancer (22), and antimutagenic activities (7). Like other redox-cycling chemicals such as paraquat and menadione, plumbagin has been used as an agent to generate superoxide or reactive oxygen species in order to study oxidative stress (14). Plumbagin has been suggested to activate SoxS by oxidizing the SoxR molecule (10, 21) or inhibiting the repression of MarR (2), the effect of which is exerted on marA, resulting in activation of the mar/sox regulon. In addition, it has been observed that the expression of some members of the mar/sox regulon in Escherichia coli, such as sodA (8), nfo (3), ribA (20), and pqi (21), is up-regulated by treatment with plumbagin.
Although plumbagin could induce excessive expression of superoxide dismutase and catalase, overexpression of sodA failed to protect E. coli (17). The toxic effect of plumbagin may not simply result from the production of reactive oxygen species. It has been reported that plumbagin inhibits NADH dehydrogenase, as well as causing respiratory arrest (17). Plumbagin has also been shown to modify the lactose carrier and inhibit its binding with galactoside; the modified carrier then becomes completely inactive (29). The above effects appeared more or less to result directly from the chemical nature of plumbagin. In this report, we focus on the responses of the bacteria to the chemical, in which multiple proteins were simultaneously induced by plumbagin treatment. We report the evaluation of the bacterial regulatory systems after treating E. coli with plumbagin.
MATERIALS AND METHODS
Bacterial strains, plasmids, and growth conditions. E. coli JM109 was used as the major experimental strain for the chemical treatments and the cloning host. Plasmids pGEM-T-easy (Promega, Madison, WI), pBluescript II SK+ (Stratagene, La Jolla, CA), and pQE60 (QIAGEN, Valencia, CA) were used as general cloning vectors. Plasmid pMH was modified from pQE60 by deleting the T5 promoter. Bacteria were cultivated in tubes containing Luria-Bertani medium. To prepare proteins for two-dimensional gel electrophoresis (2-DE), bacteria (80 ml) were grown at 37°C with vigorous shaking in 250-ml flasks. To measure the MIC, either an agar diffusion assay or the liquid broth method was used and the standard procedure recommended by the Clinical and Laboratory Standards Institute (formerly the National Committee for Clinical Laboratory Standards) was followed. Plumbagin (Sigma, St. Louis, MO) was dissolved in dimethyl sulfoxide (DMSO) or methanol with a stock concentration at 10 mg/ml without further purification.
2-DE. Bacteria from overnight cultures were diluted 100-fold into 80 ml of Luria-Bertani medium. Bacteria were grown with aeration at 37°C to an A600 of 0.2. Plumbagin was added to make a final concentration of 25 μg/ml (0.13 mM), and the culture was further agitated at 37°C for 2 h. Protein extraction was performed as previously described (12), except for a slight modification. In brief, bacteria were harvested by centrifugation and washed twice with cold 0.9% NaCl. After resuspension in 3 ml of TSD buffer (28 mM Tris HCl, 22 mM Tris, 0.3% sodium dodecyl sulfate, 200 mM dithiothreitol) and addition of 0.3 g of glass beads (0.1 to 0.25 mm in diameter), the bacteria were broken by vigorous vortexing for 10 min. Unbroken bacteria were removed by centrifugation, and the supernatant was boiled for 5 min. Chilled samples received 150 μl of TM buffer (24 mM Tris, 476 mM Tris HCl, 50 mM MgCl2) containing DNase I (1 mg/ml) and RNase A (0.25 mg/ml). The enzymatic digestion was stopped after 30 min by adding 4 volumes of ice-cold acetone, and samples were incubated overnight at –20°C. Proteins were collected by centrifugation at 11,000 x g for 15 min and then dissolved in 200 μl of isoelectric focusing (IEF) sample buffer containing 8 M urea, 4% 3-[(3-cholamidopropyl)-dimethylammonio]-1-propanesulfonate (CHAPS), and 40 mM Tris base. The first-dimension (1-D) IEF was performed with the Ettan IPGphor II IEF System, and 2-DE was carried out with an SE260 electrophoresis apparatus (Amersham Biosciences, Piscataway, NJ).
Western blotting. The bacterial culture and plumbagin treatment were similar to those used for 2-DE. The proteins were separated by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) and transferred onto a nitrocellulose membrane. To detect ectopically expressed His6-tagged proteins, a rabbit polyclonal antibody specific for the tag (Bethyl, Montgomery, TX) was used as the primary antibody. A horseradish peroxidase-conjugated goat anti-rabbit antibody (Sigma) was used as the secondary antibody. The membranes were finally developed by a chemiluminescence method (16). The images developed on films were quantitatively monitored with Personal Densitometer SI (Molecular Dynamics).
Plasmid construction. To construct pMH-mdaB, a 1,590-bp fragment covering the entire mdaB gene and its 5'-flanking region was obtained by PCR amplification with the genomic DNA of E. coli as the template and primers NmdabF (CGGGATCCCCGC TATTGCTAAGTTTG) and NmdabR (AACTGCAGTTTCCACAAGATGCTTGCG). The PCR product was digested with BamHI/PstI and ligated into pMH that had been previously digested with the same enzymes. The resulting plasmid was named pMH-mdaB, and the encoded MdaB protein had a hexahistidine extension at the C terminus. Plasmids pMH-ahpC, pMH-gatY, pMH-nfnB, pMH-nfo, pMH-tpx, pMH-talB, pMH-ygfZ, pMH-ygfZ', pMH-ygaG, pMH-marRA, and pMH-soxRS were similarly generated by the same strategy, except that the primer pairs used were NahpcF (CGGGATCCTCGTCCATCAGTTTCTC)-NahpcR (AAGTCGACTTTACCAACCAGGTCC), NgatyF (CGGGATCCTTATGCGCATCGCATTC)-NgatyR (AACTGCAGCACAATCGAATCAC), NnfnbF (GCGGATCCTGGCCCTTGAGTTACCC)-NnfnbR (GCGGATCCATTCGTTAGGTGATG), NnfoF (CAGGATCCACATCGTACACTGTGGC)-NnfoR (GGGAGGCTACCGCTTTTTCAG), NtpxF (CCCGGATCCTAAAGATGCAATTCGC C)-NtpxR (CCACTGCAGGTGCTTTCAGTACAGCC), NtalbF (TCCCGGGCTGCCAGCCAAAAGCA)-NtalbR (GGAAGATCTCAGCAGATCGCCGATC), NygfzF (GCGGATCCCAAACGCCCGTCCATCAACAAGCG)-NygfzR (GCGGATCCGGCGGTGGATCTTAGCCAGCGCC), Nygfz'F (CGGGATCCCGAAATTAATTCAC)-NygfzR, NygagF (GCGTCGAAGTGCGTTCGCTGGACATCAACCCG)-NygagR (GCGGATCCGGATATGCGGCTGGTAACCTTC), NmarraF (GGGCTGCAGGCCAATTGCTTTAAACAAATC)-NmarraR (GGGAGATCTGCTGTTGTAATGATTTAATGG), and NsoxrsF (CCTCTGCAGGCCCTTGTGGCGCTTTAG)-NsoxrsR (CCTAGATCTCAGGCGGTGGCGATAATCG), respectively.
To construct pACYC-soxRS, which expresses SoxS from a plasmid compatible with pMH, the soxRS-containing fragment was obtained by treating pMH-soxRS with BamHI and HindIII and cloned into pACYC184 that had been digested with the same enzymes.
Site-directed mutagenesis. The 5'-flanking region of mdaB was PCR amplified in a similar way from E. coli genomic DNA with primer NmdabF paired with NmdabMR (CAAAATGCGCAAAGTCTTTTG) or NmdabMF (CAAAAGACTTTGCGCATTTTG) paired with NmdabR. The two PCR products separately amplified were mixed. Since NmdabMR and NmdabMF embedding the designated point mutations are complementary in sequence, the two DNA products were annealed at one end and extended by Vent polymerase and deoxynucleoside triphosphates. After 10 reaction cycles, NmdabF and NmdabR were added to the mixture and PCR amplification was continued for an additional 30 cycles. The PCR product was then digested with BamHI and PstI and cloned into pMH, resulting in plasmid pMH-mdaBm. Mutation-bearing plasmids pMH-nfom and pMH-ygfZ'm were constructed by a similar strategy; the paired primers used accordingly for the mutagenesis were NnfoMF (GTGATTCAAAGCGTCATTCT AGAAACCAC)-NnfoMR (GCAAGATGTAGTGGTTTCTAGAATGACGC) and NygfzMF (GTCGCCTCGCAGAAGGTACCGATCGCGTAGTG)-NygfzMR (AGTGCTTTCACTACGCGATCGGTACCTTCTGCG), respectively.
Q-TOF mass spectroscopy. In-gel tryptic digestion was performed as previously described (30), except for a slight modification. In brief, Coomassie blue dye-stained protein spots excised from a distilled-water-washed gel were cut into small pieces. Gel pieces were dehydrated and destained with three washes with 50 mM ammonium bicarbonate mixed with 50% (vol/vol) acetonitrile. After dehydration and in-gel digestion with trypsin, the peptides recovered were analyzed with a quadrupole time-of-flight (Q-TOF) mass spectrometer (Waters, Milford, MA). Data from the mass profiling were used to search the NR database with Mascot (http://www.matrix-science.com).
Construction of a soxS mutant strain. To delete soxS from the chromosome by homologous recombination (4), E. coli strain JM110 was used as the parental strain since strain JM109 lacks recA and has a low frequency of recombination. To perform the deletion, both the 5'- and 3'-flanking regions of soxS were PCR amplified from E. coli genomic DNA. The primer pairs used were 5'NsoxSF (GCGGATCCGCATCAACACCAACCGGAACC)-5'NsoxSR (GCGAATTCGTCAATCCATGCGATAAGATCC) and 3'NsoxSF (GCGAATTCCTGTAATTTTATTGCCCGCGCG)-3'NsoxSR (GCGGATCCTTCGACAAAACGCCCAATTGC), respectively. The PCR products were then cloned into pGEM-T Easy and confirmed after sequencing. These fragments were then excised from the plasmids with BamHI/EcoRI and three-way ligated with BamHI-restricted pKO3 (25), yielding pKO3-soxS. pKO3-soxS was then transformed into strain JM110, and the resulting transformants were sequentially selected in chloramphenicol- and sucrose-containing media. The soxS mutant strain obtained was confirmed by PCR amplification of the region spanning the target site, followed by restriction enzyme analysis and sequence confirmation.
RESULTS
Proteins whose expression is up-regulated in E. coli by plumbagin. To investigate the response of the bacteria, E. coli was treated with plumbagin at a subinhibitory concentration. Plumbagin was first determined to have an MIC against E. coli JM109 of 50 μg/ml. Thus, the plumbagin treatment was held at 25 μg/ml for 2 h and the total bacterial proteins were extracted and separated by 2-DE. Proteins on the gel were stained with Coomassie blue dye and compared with those prepared from bacteria without plumbagin treatment. Twelve protein spots whose expression was obviously increased by plumbagin treatment (Fig. 1) were cut out and digested with trypsin. The tryptic peptides eluted had their amino acid sequences deduced by Q-TOF mass spectroscopy. The peptides so deduced were used to search for matching proteins derived from the E. coli genome. Figure 2A shows a typical result of three tryptic peptides matched to the E. coli alkyl hydroperoxide reductase (AhpC). Results from similar identifications are listed in Fig. 2B.
Plumbagin-responding genes. Among the 12 deduced proteins, SodA was massively induced by plumbagin and could simply be identified by 1-D SDS-PAGE, followed by Coomassie blue dye staining. Among the remaining 11 protein spots, although the 2-D gel profiling of the proteins was reproducible, misleading identification could not be completely excluded due to the possibility of overlapping proteins. Therefore, an approach to augment the plumbagin induction signal information and to offer an alternative method of confirming the responsiveness of the genes was explored. Theoretically, when a gene is cloned into a plasmid with its promoter, the responsiveness of the encoded proteins would be amplified due to the multiple-copy property of the plasmid. Therefore, the genes identified above together with a 5' region, ranging from 425 bp to 1,100 bp upstream of the open reading frame, were PCR amplified. The obtained DNA fragments were separately cloned into the pMH vector, in which every single target protein is tagged with His6 at the C terminus. Proteins displayed on the 1-D system could be simply compared by Western blotting with anti-His6 tag antibody.
A bacterial transformant harboring plasmid pMH-mdaB treated with plumbagin was compared with those treated with DMSO alone. The MdaB protein increased about 2.5-fold (Fig. 3A) in the presence of plumbagin, given that the total proteins were loaded in comparable amounts (OmpC was used as an internal control; data not shown). Gene mdaB is known as "modulator of drug activity B" and has been recently proposed to function as an NADPH quinone reductase inducible by menadione (15, 19). Our results further demonstrated that mdaB also responded to plumbagin. The nfo gene encodes endonuclease IV and participates in the repair of DNA lesions induced by H2O2 (9). nfo's response to plumbagin (Fig. 3B) is consistent with observations previously reported (3). An unknown protein encoded by ygfZ was identified as being upregulated for the first time (Fig. 3C). Expressions of AhpC protein (Fig. 3D), a peroxiredoxin (alkyl hydroperoxide reductase) offering antioxidant protection, and NfnB, an oxygen-insensitive NAD(P)H nitroreductase (Fig. 3E), were both confirmed to be up-regulated. Protein derived from ahpC was observed with an extra product which was twice as large as expected (Fig. 3D). This fact suggests that AhpC may form a dimer structure or associate with another molecule(s) when bacteria encounter oxidative stress. The induced protein pattern detected with NfnB was striking. At least five additional high-molecular-weight bands above the expected product were observed (Fig. 3E). These high-molecular-weight aggregates of NfnB have not been reported before. Multiple high-molecular-weight products were also seen when gatY was examined in the same manner in the presence of plumbagin (data not shown).
The positive responsiveness of the genes fldA, encoding flavodoxin I, and yggX, encoding a protein involved in Fe(II) trafficking to minimize DNA damage (13), was also confirmed in a similar way (data not shown). In contrast, three other genes listed in Fig. 2B, namely, talB, tpx, and ygaG, could not be confirmed and were therefore excluded from further investigation.
Marbox sequence contributed to plumbagin responsiveness. To explore the mechanism of genes' up-regulation upon plumbagin treatment, the 5'-flanking regions of these genes were compared. A stretch of sequence containing a putative marbox was observed in the regions upstream of fldA, mdaB, nfnB, nfo, sodA, and yggX. These sequences have been proposed to be the controlling elements, but so far only that of yggX has been characterized by binding of SoxS to a promoter-containing fragment and those of mdaB and nfnB have been mapped by fragment deletion in a promoter assay (28, 32). It is worth noting that those characterizations were not based upon the effects of plumbagin. Instead, they were based on observations with paraquat, 4,4'-dipyridyl, or salicylate.
To confirm the effects of plumbagin treatment on gene expression, a mutation was introduced into the mapped marbox sequence upstream of mdaB (28, 32). In the pMH-mdaB construct, the putative core element GCAC of the marbox (Fig. 4A) was replaced with GCGC, resulting in plasmid pMH-mdaBm. MdaB expression was then compared. The basal expression level of MdaB from pMH-mdaBm was fivefold decreased relative to that from pMH-mdaB (compare lanes 1 and 3 in Fig. 4B). Upon addition of plumbagin, the expression of MdaB from pMH-mdaB was 2.5-fold higher than when the solvent was added alone (lane 1 versus lane 2). In contrast, the increase in MdaB from the pMH-mdaBm transformant was limited to 1.3-fold when plumbagin was added (lanes 3 and 4). Therefore, a single base mutation of the marbox sequence drastically decreased the basal level of MdaB expression, as well as the plumbagin induction effect. In a similar approach, the proposed marbox of nfo (28, 32) was mutated (Fig. 4A) and the core-containing element CGCAT in pMH-nfo was mutated to TCTAG in pMH-nfom. The basal expression level of Nfo from pMH-nfom was lower than that derived from pMH-nfo (compare lanes 1 and 3 in Fig. 4C), and a 2.4-fold decrease was observed. On treatment with plumbagin, the induction effect of plumbagin was reduced from a 3.5-fold difference in the wild type to a 1.1-fold difference in the mutant (compare lanes 1 and 2 with lanes 3 and 4).
To examine the element conferring the plumbagin induction effects on the uncharacterized ygfZ gene, the sequence upstream of ygfZ was first analyzed with the Emboss Fuzznuc program (http://www.hgmp.mrc.ac.uk/Software/EMBOSS), and several marbox-like sequences were found in this region. To facilitate the identification of the plumbagin-responsive element, pMH-ygfZ was reconstructed to generate a shortened 5'-flanking region in pMH-ygfZ' (Fig. 5A). Compared to pMH-ygfZ, the plumbagin-induced expression of ygfZ was retained in pMH-ygfZ' (lanes 1 to 4, Fig. 5C). This result suggested that the plumbagin-responsive element exists in the 151-bp region upstream of the ygfZ-coding sequence. Analysis of this sequence in detail found a zwf-like class I marbox promoter structure (27, 36) that is characterized by a space of 7 bp between a putative marbox and a putative –35 hexamer. Therefore, the core-containing stretch, GCACA, in the putative marbox sequence was replaced with GTACC (see sequence alignment in Fig. 5B), and the expression levels of YgfZ before and after the sequence alteration were analyzed (Fig. 5C). The basal expression level of YgfZ from pMH-ygfZ'm was suppressed (compare lanes 3 and 5) by this two-base mutation compared to that from pMH-ygfZ', and the plumbagin induction effect was abolished as well (compare lanes 5 and 6). The total effect is even obvious when lanes 4 and 6 are compared.
SoxS is involved in early plumbagin induction. Previous studies suggested that two marbox-binding proteins, SoxS and MarA, might be induced directly by treatment with plumbagin (2, 10). To differentiate their contributions, two plasmids containing their intrinsic control elements together with the appropriate marA and soxS genes were separately constructed. In pMH-marRA, the complete marRA locus was inserted into pMH and the product of marA was fused with a His6 tag. The expressed MarA protein was then followed by Western blotting with anti-His6 as in Fig. 3. Similarly, pMH-soxRS was constructed to detect the expression of SoxS. Bacteria transformed with pMH-marRA and pMH-soxRS, respectively, were treated with plumbagin, and the kinetics of MarA and SoxS expression were followed by Western blotting. Figure 6A shows that SoxS was robustly induced and strongly detected as early as 5 min after plumbagin treatment (Fig. 6A). In contrast, the induction of MarA was negligible over the first 10 min and slowly increased thereafter. Figure 6B, with a graphic illustration, further sketches out the concept that SoxS is induced first in the early phase of the bacterial response to plumbagin. After 30 min, the level of SoxS reached a plateau while the level of MarR had a limited increase. With the MarA level induced by sodium salicylate (33) as a reference, the 60-min plumbagin treatment only reached about 60% of the level induced by salicylate. This slow MarA induction is unique to plumbagin treatment since the same bacteria gave a rapid plateau in response to sodium salicylate treatment for less than 30 min (Fig. 6C).
SoxS is involved in the plumbagin induction of some genes but not all. The robust induction of SoxS by plumbagin suggests a role for SoxS in the activation of these gene. Since our proteomic analysis was based upon bacteria after a 2-h induction, gene activation conferred by other late events may also be possible. To distinguish the genes controlled by SoxS from those controlled by other proteins, we constructed a soxS deletion mutant and repeated the experiments involving treatment with plumbagin. Figure 7A shows that a soxS deletion mutant gave a basal level of Nfo similar to that found with the parental strain (compare lanes 1 and 3). However, when encountering plumbagin, the soxS mutant totally lost the response (Fig. 7A, compare lanes 2 and 4). After complementation with a SoxS-expressing plasmid, pACYC-soxRS, the plumbagin responsiveness was completely restored (Fig. 7A, lanes 3 to 6). Therefore, plumbagin's effect on nfo was conceivably through a direct involvement of SoxS. Similar to nfo, ygfZ and sodA lost their reactions toward plumbagin when bacterial soxS was deleted (data not shown). Although not tested, fldA and yggX are presumably in the same category since they have been known to be regulated by SoxS (32, 37).
On the other hand, mdaB, nfnB, and ahpC substantially differed from nfo. Their products remained plumbagin inducible when soxS was deleted. A typical example with MdaB is shown in Fig. 7B.
DISCUSSION
The antibacterial effect of plumbagin against E. coli was observed under aerobic conditions. When plumbagin was present at one-half of the MIC, the bacteria continued to grow, but at a slow rate, and many gene products of E. coli were up-regulated. Twelve proteins significantly elevated have been identified by proteomic analysis, and nine of the genes involved have been reconfirmed by an alternative approach. Intriguingly, more than one-half of plumbagin-responding genes could be functionally correlated with antioxidation, physiological reduction, and DNA lesion repair (1, 8, 11, 23, 24, 32, 34). Considering that plumbagin has no obvious effect on E. coli growth in the absence of oxygen and that there was no recognizable difference between the 2-DE protein profiles of E. coli cells treated with and without plumbagin (data not shown), the toxic effect of plumbagin on E. coli must be reconsidered.
The elevated expression of the genes fldA, mdaB, nfnB, nfo, sodA, and yggX caused by plumbagin in our study was also observed in a microarray study when bacteria were treated with paraquat (31). These genes have been defined as members of the mar/sox regulon, and their regulation has been proposed to occur through marbox sequences that share conserved cores with otherwise degenerated sequences (28). We have mutated the marbox sequences in front of mdaB and nfo and proved that such mutations indeed decreased the basal level of the proteins expressed, as well as their responses to plumbagin. We rationalize that fldA, nfnB, sodA, and yggX may be regulated by similar cis elements. As to trans factors, the plumbagin induction effects on nfo, fldA, sodA, and yggX were found to be exclusively dependent upon SoxS. Additional regulation at the posttranscriptional level could not be completely excluded. On the other hand, the elevated MdaB and NfnB protein levels caused by plumbagin were apparently not affected regardless of the presence or absence of soxS.
In our study, gatY was also up-regulated, and this induction has not been observed with paraquat (28). Since the expression of genes involved in sugar transport has been elevated when E. coli encounters superoxide stress (28, 31) and since gatY has been known to encode a protein involved in carbohydrate utilization, it is not surprising to see the up-regulation of gatY by plumbagin. However, no marbox-like sequence was found proximal to gatY previously (28) and we were also unable to identify a related sequence. Therefore, how this gene is molecularly regulated remains to be explored.
A gene whose function is unknown that was identified in our plumbagin treatment is ygfZ; the crystal structure of the protein it encodes has been reported, and a folate-dependent regulatory role in one-carbon metabolism has been hypothesized. However, its physiological role and real function remain unclear (35). We found that the ygfZ gene contains a zwf-like class I marbox structure, and mutation of this sequence did decrease the basal protein level, as well as its response to plumbagin, a reaction that was also soxS dependent. Therefore, these new findings suggest that ygfZ may also be involved in releasing oxidative stress.
ahpC has been reported to be regulated by OxyR (34), and it is also stimulated by plumbagin. This fact strongly suggests that plumbagin not only generates superoxide stress but also provokes the production of peroxide ions to trigger a mixed set of responses in E. coli. In our soxS deletion mutant, the plumbagin effect with an increasing AhpC level remained evident, a fact suggesting that SoxS does not act as a mediator. It is then suggested that the readily induced MarA protein, albeit at a slow response, and Rob (18), which is known to bind to sites with similar sequences and was not monitored here, may contribute to the increase in AhpC. The same reasoning may be valid with the findings on MdaB and NfnB, on neither of which the plumbagin effect was SoxS dependent. Whether the above conclusions found with E. coli are applicable to bacteria that are sensitive to plumbagin under both aerobic and anaerobic conditions remains to be explored.
Under our plumbagin treatment conditions, bacterial proteins whose levels were increased appeared to outnumber those being suppressed (Fig. 1). These studies of up-regulated proteins do not necessarily imply that those that are down-regulated are not important for bacterial survival. As proteomic investigations increase both protein resolution and staining sensitivity, it is expected that more protein spots on either the up-regulated or the down-regulated side will be unveiled. By grouping responsive genes and gene products, how bacteria use different mechanisms to resolve the same chemical stress may be better understood.
ACKNOWLEDGMENTS
We thank J. C. W. Lio for the gift of pMH, J. Y. Ho for operating the mass spectrometer, and C. M. Tzeng of U-Vision Biotech for continuous encouragement. We also thank R. Kirby for critically reading the manuscript.
This research was supported in part by grant 89-B-FA22-2-4 (Program for Promoting Academic Excellence of Universities) from the Ministry of Education and grant NSC 94-2320-B-010-034 from the National Science Council, Taiwan, Republic of China.
REFERENCES
Adams, M. A., and Z. Jia. 2005. Structural and biochemical evidence for an enzymatic quinone redox cycle in Escherichia coli: identification of a novel quinol monooxygenase. J. Biol. Chem. 280:8358-8363.
Alekshun, M. N., and S. B. Levy. 1999. Alteration of the repressor activity of MarR, the negative regulator of the Escherichia coli marRAB locus, by multiple chemicals in vitro. J. Bacteriol. 181:4669-4672.
Chan, E., and B. Weiss. 1987. Endonuclease IV of Escherichia coli is induced by paraquat. Proc. Natl. Acad. Sci. USA 84:3189-3193.
Chiu, H. J., and W. J. Syu. 2005. Functional analysis of EspB from enterohaemorrhagic Escherichia coli. Microbiology 151:3277-3286.
Curreli, N., F. Sollai, L. Massa, O. Comandini, A. Rufo, E. Sanjust, A. Rinaldi, and A. C. Rinaldi. 2001. Effects of plant-derived naphthoquinones on the growth of Pleurotus sajor-caju and degradation of the compounds by fungal cultures. J. Basic Microbiol. 41:253-259.
de Paiva, S. R., M. R. Figueiredo, T. V. Aragao, and M. A. Kaplan. 2003. Antimicrobial activity in vitro of plumbagin isolated from Plumbago species. Mem. Inst. Oswaldo Cruz 98:959-961.
Edenharder, R., and X. Tang. 1997. Inhibition of the mutagenicity of 2-nitrofluorene, 3-nitrofluoranthene and 1-nitropyrene by flavonoids, coumarins, quinones and other phenolic compounds. Food Chem. Toxicol. 35:357-372.
Fee, J. A. 1991. Regulation of sod genes in Escherichia coli: relevance to superoxide dismutase function. Mol. Microbiol. 5:2599-2610.
Galhardo, R. S., C. E. Almeida, A. C. Leitao, and J. B. Cabral-Neto. 2000. Repair of DNA lesions induced by hydrogen peroxide in the presence of iron chelators in Escherichia coli: participation of endonuclease IV and Fpg. J. Bacteriol. 182:1964-1968.
Gaudu, P., N. Moon, and B. Weiss. 1997. Regulation of the soxRS oxidative stress regulon. Reversible oxidation of the Fe-S centers of SoxR in vivo. J. Biol. Chem. 272:5082-5086.
Gaudu, P., and B. Weiss. 2000. Flavodoxin mutants of Escherichia coli K-12. J. Bacteriol. 182:1788-1793.
Giard, J. C., A. Rince, H. Capiaux, Y. Auffray, and A. Hartke. 2000. Inactivation of the stress- and starvation-inducible gls24 operon has a pleiotrophic effect on cell morphology, stress sensitivity, and gene expression in Enterococcus faecalis. J. Bacteriol. 182:4512-4520.
Gralnick, J. A., and D. M. Downs. 2003. The YggX protein of Salmonella enterica is involved in Fe(II) trafficking and minimizes the DNA damage caused by hydroxyl radicals: residue CYS-7 is essential for YggX function. J. Biol. Chem. 278:20708-20715.
Hassan, H. M., and I. Fridovich. 1979. Intracellular production of superoxide radical and of hydrogen peroxide by redox active compounds. Arch. Biochem. Biophys. 196:385-395.
Hayashi, M., K. Hasegawa, Y. Oguni, and T. Unemoto. 1990. Characterization of FMN-dependent NADH-quinone reductase induced by menadione in Escherichia coli. Biochim. Biophys. Acta 1035:230-236.
Hsu, S. C., H. P. Lin, J. C. Wu, K. L. Ko, I. J. Sheen, B. S. Yan, C. K. Chou, and W. J. Syu. 2000. Characterization of a strain-specific monoclonal antibody to hepatitis delta virus antigen. J. Virol. Methods 87:53-62.
Imlay, J., and I. Fridovich. 1992. Exogenous quinones directly inhibit the respiratory NADH dehydrogenase in Escherichia coli. Arch. Biochem. Biophys. 296:337-346.
Jair, K. W., X. Yu, K. Skarstad, B. Thony, N. Fujita, A. Ishihama, and R. E. Wolf, Jr. 1996. Transcriptional activation of promoters of the superoxide and multiple antibiotic resistance regulons by Rob, a binding protein of the Escherichia coli origin of chromosomal replication. J. Bacteriol. 178:2507-2513.
Jorgensen, M. A., M. A. Trend, S. L. Hazell, and G. L. Mendz. 2001. Potential involvement of several nitroreductases in metronidazole resistance in Helicobacter pylori. Arch. Biochem. Biophys. 392:180-191.
Koh, Y. S., J. Choih, J. H. Lee, and J. H. Roe. 1996. Regulation of the ribA gene encoding GTP cyclohydrolase II by the soxRS locus in Escherichia coli. Mol. Gen. Genet. 251:591-598.
Koh, Y. S., and J. H. Roe. 1995. Isolation of a novel paraquat-inducible (pqi) gene regulated by the soxRS locus in Escherichia coli. J. Bacteriol. 177:2673-2678.
Krishnaswamy, M., and K. K. Purushothaman. 1980. Plumbagin: a study of its anticancer, antibacterial and antifungal properties. Indian J. Exp. Biol. 18:876-877.
Levin, J. D., A. W. Johnson, and B. Demple. 1988. Homogeneous Escherichia coli endonuclease. IV. Characterization of an enzyme that recognizes oxidative damage in DNA. J. Biol. Chem. 263:8066-8071.
Lightfoot, R. T., D. Shuman, and H. Ischiropoulos. 2000. Oxygen-insensitive nitroreductases of Escherichia coli do not reduce 3-nitrotyrosine. Free Radic. Biol. Med. 28:1132-1136.
Link, A. J., D. Phillips, and G. M. Church. 1997. Methods for generating precise deletions and insertions in the genome of wild-type Escherichia coli: application to open reading frame characterization. J. Bacteriol. 179:6228-6237.
Martin, R. G., W. K. Gillette, S. Rhee, and J. L. Rosner. 1999. Structural requirements for marbox function in transcriptional activation of mar/sox/rob regulon promoters in Escherichia coli: sequence, orientation and spatial relationship to the core promoter. Mol. Microbiol. 34:431-441.
Martin, R. G., W. K. Gillette, and J. L. Rosner. 2000. Promoter discrimination by the related transcriptional activators MarA and SoxS: differential regulation by differential binding. Mol. Microbiol. 35:623-634.
Martin, R. G., and J. L. Rosner. 2002. Genomics of the marA/soxS/rob regulon of Escherichia coli: identification of directly activated promoters by application of molecular genetics and informatics to microarray data. Mol. Microbiol. 44:1611-1624.
Neuhaus, J. M., and J. K. Wright. 1983. Chemical modification of the lactose carrier of Escherichia coli by plumbagin, phenylarsinoxide or diethylpyrocarbonate affects the binding of galactoside. Eur. J. Biochem. 137:615-621.
Phan-Thanh, L., and F. Mahouin. 1999. A proteomic approach to study the acid response in Listeria monocytogenes. Electrophoresis 20:2214-2224.
Pomposiello, P. J., M. H. Bennik, and B. Demple. 2001. Genome-wide transcriptional profiling of the Escherichia coli responses to superoxide stress and sodium salicylate. J. Bacteriol. 183:3890-3902.
Pomposiello, P. J., A. Koutsolioutsou, D. Carrasco, and B. Demple. 2003. SoxRS-regulated expression and genetic analysis of the yggX gene of Escherichia coli. J. Bacteriol. 185:6624-6632.
Sulavik, M. C., L. F. Gambino, and P. F. Miller. 1995. The MarR repressor of the multiple antibiotic resistance (mar) operon in Escherichia coli: prototypic member of a family of bacterial regulatory proteins involved in sensing phenolic compounds. Mol. Med. 1:436-446.
Tartaglia, L. A., G. Storz, and B. N. Ames. 1989. Identification and molecular analysis of oxyR-regulated promoters important for the bacterial adaptation to oxidative stress. J. Mol. Biol. 210:709-719.
Teplyakov, A., G. Obmolova, E. Sarikaya, S. Pullalarevu, W. Krajewski, A. Galkin, A. J. Howard, O. Herzberg, and G. L. Gilliland. 2004. Crystal structure of the YgfZ protein from Escherichia coli suggests a folate-dependent regulatory role in one-carbon metabolism. J. Bacteriol. 186:7134-7140.
Wood, T. I., K. L. Griffith, W. P. Fawcett, K. W. Jair, T. D. Schneider, and R. E. Wolf, Jr. 1999. Interdependence of the position and orientation of SoxS binding sites in the transcriptional activation of the class I subset of Escherichia coli superoxide-inducible promoters. Mol. Microbiol. 34:414-430.
Zheng, M., B. Doan, T. D. Schneider, and G. Storz. 1999. OxyR and SoxRS regulation of fur. J. Bacteriol. 181:4639-4643.(Jenn-Wei Chen, Chang-Ming)