Pii: s0966-842x(99)01589-9

74 Prince, A.M. et al. (1992) J. Infect. Dis. 165, 438–443
78 Laskus, T. et al. (1996) Virology 220, 171–176
75 Farci, P. et al. (1992) Science 258, 135–140
79 Cane, E.J. et al. (1996) New Engl. J. Med. 334, 815–820
76 Martell, M. et al. (1994) J. Virol. 68, 3425–3436
80 Okamoto, H. et al. (1994) Hepatology 20, 1131–1136
77 Gretch, D.R. et al. (1996) J. Virol. 70, 7622–7631
81 Sullivan, D.G. et al. (1998) J. Virol. 72, 10036–10043
The mar regulon: multiple resistance to
antibiotics and other toxic chemicals
Antibiotic resistance in The chromosomal multiple antibiotic repeats – sites I and II (Ref.15)
resistance (mar) locus of Escherichia coli
and other members of the
Enterobacteriaceae controls resistance to
multiple, structurally unrelated
compounds including antibiotics,
household disinfectants, organic solvents
and other toxic chemicals. The Mar
phenotype is induced following exposure
to a variety of chemicals with aromatic
by altering the expression of M.N. Alekshun and S.B. Levy* are in the Center for Adaptation Genetics and Drug Resistance, Dept of Molecular Biology and Microbiology, Tufts University School of Medicine, Boston, MA 02111, USA; S.B. Levy is also in the Dept of Medicine, Tufts University School of Medicine, Boston, oxidative stress agents5 and household disinfec- tional activators, including MarA, Rob (an E. coli tants6,7. Since the early description of the mar locus, MarA homolog that is constitutively expressed), Fis other intrinsic regulatory mechanisms that bacteria and probably SoxS (an E. coli MarA homolog that use to resist the lethal effects of a wide range of toxic controls resistance to oxidative stress), to sites within agents have been described8,9. In this sense, the E. coli Pmar (Refs 17–19) (Fig. 1c). MarA and Rob, mem- mar locus has been a useful model.
bers of the XylS/AraC family of transcriptional acti-vators, bind to a sequence called the marbox within Pmar (Fig. 1c). Whereas MarA and Rob are major The E. coli mar locus comprises two transcriptional contributors to marRAB activation in whole cells, the units: marC and marRAB (Ref. 10) (Fig. 1a). The ex- effect of SoxS, unless induced, is negligible17. Fis, a pression of each operon is under the control of the multifunctional DNA-binding protein20, binds to a divergent promoters Pmar and Pmar , which are in a site upstream of the marbox and plays an accessory centrally located promoter–operator region (Fig. 1a).
(positive) role in marRAB expression in the presence marC encodes a putative inner membrane protein11 of of these activators17 (Fig. 1c). In addition to activating unknown function and marRAB (Ref. 10) specifies its own expression, MarA also regulates the expres- the mar repressor (MarR), activator (MarA) and a sion of many other genes – constituents of the mar small protein (MarB), the function of which has not regulon (reviewed in Ref. 10) – on the E. coli chromo- some, and ultimately induces a variety of phenotypes In E. coli, expression of the marRAB operon is in- duced by tetracycline, chloramphenicol, sodium sali- There are two genes, ychE and yhgN, specifying cylate and other unrelated compounds5,12–14. Under putative proteins of unknown function that are non-inducing conditions, MarR binds to two direct homologous to marC in E. coli. In addition to its 0966-842X/99/$ - see front matter 1999 Elsevier Science Ltd. All rights reserved. PII: S0966-842X(99)01589-9 410 VOL. 7 NO. 10 OCTOBER 1999
counterpart in Salmonella typhimurium21,marC homologs are also found in Methanococcus jannaschii, Aquifex aeoli-cus, Pyrococcus horikoshii, Archaeoglobusfulgidus and Treponema pallidum, and allhave unknown functions. The only known gene homologous to E. coli marB is thatfrom S. typhimurium21.
regulators MarR homologs are widespread among many different bacterial genera and someare believed to interact with different phenolic compounds22. For example, the invitro activity of CinR, a MarR homolog from Butyrivibrio fibrisolvens E14 that negatively regulates the expression of anenzyme involved in releasing cinnamic acids (salicylate precursors in plants) dur-ing plant cell wall catabolism, is affected bycertain products of this metabolic path- way23. Moreover, recent experiments havedemonstrated that many of the chemicals that induce mar expression in vivo antag-onize the repressor activity of MarR in vitro16. Thus, it appears that both MarRand CinR have DNA-binding and inducer-recognition properties. Data pertaining to the functional regions mutations at particular residues result in a ‘super-repressor’ protein that has increasedDNA-binding affinity24. Whether these mu- Fig. 1. Genetic organization of the Escherichia coli mar locus. (a) Expression of marCand marRAB is controlled by independent promoters Pmar and Pmar , respectively.
tations affect specific or non-specific DNA– MarR (144 residues) and MarA (127 residues) encode the Mar repressor and acti- protein interactions is currently unknown. vator. MarB (72 residues) has an unknown function, as does MarC, a 221 amino acid, putative inner-membrane protein with multiple transmembrane-spanning helices. (b) MarR negatively regulates marRAB expression by binding to two sites in the oper-ator marO. (c) Following inactivation of MarR, marRAB expression is constitutive, but its level of expression depends on the interaction of MarA, Rob and SoxS (when members, as these proteins are severely un- induced) with a sequence (the marbox) proximal to the Ϫ35/Ϫ10 hexamers. Fis plays an accessory role in this positive control17.
coworkers recently solved the 3-D struc-ture of a MarA–DNA co-crystal, providing the first agents is conferred by increased expression of cyto- direct proof for the existence of two helix-turn-helix plasmic enzymes that counteract the damaging effects (HTH) DNA-binding motifs26. Both HTH motifs in this structure appear to participate in DNA binding Inactivation of the AcrAB homolog in Haem- and are organized in such a manner that the helices ophilus influenzae confers hypersusceptibility to many responsible for DNA recognition extend from the structurally diverse chemicals31. Whether the func- same side of the protein. As MarA is assumed to tion of this multidrug efflux system is positively function as a monomer18, this assembly allows the regulated by Ya52, the sole MarA homolog in H. in-activator to bind to one face of its target DNA26.
fluenzae32, remains to be determined.
Antibiotic resistance arises in E. coli mar mutants The mar locus was originally identified by its ability both by decreasing influx and increasing efflux of to confer multiple antibiotic resistance2. mar mutants toxic chemicals from the cell. The former is accom- show enhanced saturable active efflux of tetracycline plished partially by downregulating the synthesis of and chloramphenicol2,33. These findings were extended OmpF (Ref. 27) and, possibly, by altering the expres- with the discovery that expression of the AcrAB ef- sion of other membrane proteins4. Increased efflux in flux system28, in addition to putative drug-specific both E. coli and S. typhimurium is achieved primarily efflux pumps10, was increased in mar mutants and by increased synthesis of the AcrAB–TolC multidrug that its removal led to increased susceptibility to efflux system4,28,29. Resistance to oxidative stress tetracycline, norfloxacin, chloramphenicol, ampicillin, 411 VOL. 7 NO. 10 OCTOBER 1999
(e.g. chloroxylenol, a phenol, andalkyl dimethyl benzyl ammoniumchloride, a quaternary amine) isgreater in strains lacking AcrAB use of the so-called non-specificbiocides, including triclosanTM, closanTM, which has been de-scribed as a broad-spectrum anti- E. coli35. E. coli clinical isolates that overexpress either MarA orSoxS are 50% less susceptible to triclosanTM (Ref. 7) than wild type.
Fig. 2. The mar regulon. The expression of MarA is increased directly by the inactivation of toxic to bacteria. Wild-type E. coli MarR16, or indirectly by the inducing compounds that enter through outer membrane porins or by diffusion (dotted arrow). Once overexpressed, MarA further activates AcrAB/TolC (see Refs 4,28,29) of n-hexane (its index solvent, i.e. efflux, represses the synthesis of OmpF (Ref. 27), the point of entry for some antibiotics, and alters the expression of other membrane proteins4. Protection in the cytoplasm is conferred by the increased expression of various cytoprotective enzymes10. Whether the induced increase inmarRAB mRNA is through a direct effect on MarR (dashed line) or via another mechanism is still cells that overexpress MarA, SoxSor Rob are cyclohexane toler-ant3,4. As observed for the other puromycin, nalidixic acid and rifampicin28. AcrB is mar-mediated phenotypes, organic solvent tolerance in an inner membrane protein that transports drugs out E. coli is contingent upon a correctly functioning of the cell, possibly through TolC, an outer mem- brane channel9. AcrB and TolC are presumablylinked by AcrA, a membrane-fusion protein9. The clinical importance of mar and mar-like Aside from the well-characterized antibiotic resis- tances, cells that overexpress MarA display a variety The mar locus is present in other members of the Entero- of other resistance phenotypes (reviewed in Ref. 10).
bacteriaceae36 and is well conserved at the DNA se- In contrast to the plasmid-mediated antiseptic/ quence level in S. typhimurium21. Expression of disinfectant resistance of Gram-positive and Gram- E. coli MarA in Mycobacterium smegmatis mc2155 negative organisms6, E. coli mutants resistant to pine produces a mar phenotype37, suggesting that MarA oil or formulations containing this natural product are can activate cognate promoters in a heterologous mar mutants6. Inactivation of AcrAB in these mutants host, and that a mar-like system(s) exists in M. smeg- increases susceptibility to compounds containing matis. Exposure of Burkholderia (Pseudomonas)cepacia38 and Staphylococcus aureus39 to salicylate,as well as serial exposure of Pseudomonas aeruginosa40to fluoroquinolones, confers resistance to multiple antibiotics. Whether there is a chromosomal region in • Is MarR activity or expression controlled by other genes in these organisms that is homologous to the E. coli mar locus and has a similar genetic organization is not yet • How do tetracycline and chloramphenicol increase expression of known. However, many Pseudomonas spp. contain several MarA homologs and at least one well-charac- • Do mar-deleted mutants survive less well than isogenic wild-type terized MarR homolog (MexR)10. Regulation of the bacteria in the natural environment? Do they survive less well expression of non-contiguous genes by these proteins than isogenic wild-type bacteria in the intestinal tract? would be unsurprising. Recent data have addressed • Is there a MarA functional homolog in Mycobacterium tuberculosis? the possibility that the AcrAB-type efflux system in • Is there a Mar-like response in Gram-positive bacteria? H. influenzae might facilitate the development of 412 VOL. 7 NO. 10 OCTOBER 1999
resistance31. Inducible resistance had been suggested 5 Ariza, R.R. et al. (1994) J. Bacteriol. 176, 143–148
by earlier studies that described chloramphenicol and 6 Moken, M.C., McMurry, L.M. and Levy, S.B. (1997)
tetracycline resistance in certain H. influenzae isolates Antimicrob. Agents Chemother. 41, 2770–2772 only after prior contact with tetracycline41. 7 McMurry, L.M., Oethinger, M. and Levy, S.B. (1998) FEMS
Most first-step mar mutants that are resistant to 8 George, A.M. (1996) FEMS Microbiol. Lett. 139, 1–10
low levels of several antibiotics are still treatable 9 Nikaido, H. (1998) Curr. Opin. Microbiol. 1, 516–523
within a clinical setting. However, for other first-step 10 Alekshun, M.N. and Levy, S.B. (1997) Antimicrob. Agents
mar mutants, clinical resistance to some antibiotics, namely tetracycline and rifampicin, is readily achieved10.
11 Goldman, J.D., White, D.G. and Levy, S.B. (1996) Antimicrob.
Moreover, mar mutants synergistically enhance the Agents Chemother. 40, 1266–1269 resistance mediated by plasmids, for example, tetra- 12 Hächler, H., Cohen, S.P. and Levy, S.B. (1991) J. Bacteriol. 173,
cycline on plasmid R222 (S.B. Levy, unpublished).
Bacteria in which the mar locus is constitutively ex- 13 Cohen, S.P. et al. (1993) J. Bacteriol. 175, 7856–7862
pressed are protected from the bactericidal effects of 14 Seoane, A.S. and Levy, S.B. (1995) J. Bacteriol. 177, 3414–3419
the fluoroquinolones11 and mutate more easily to 15 Martin, R.G. and Rosner, J.L. (1995) Proc. Natl. Acad. Sci. U. S. A.
confer resistance to a higher fluoroquinolone con- 16 Alekshun, M.N. and Levy, S.B. (1999) J. Bacteriol. 181,
centration [Ref. 42; V. Hullen et al. (1998) 38th International Conference on Antimicrobial Agents 17 Martin, R.G. and Rosner, J.L. (1997) J. Bacteriol. 179,
and Chemotherapy, San Diego, CA, USA, Abstr. C-187]. Certain clinical isolates of fluoroquinolone- 18 Martin, R.G. et al. (1996) J. Bacteriol. 178, 2216–2223
resistant E. coli are mar mutants43,44 and the mar 19 Jair, K-W. et al. (1995) J. Bacteriol. 177, 7100–7104
locus contributes to the resistance phenotype43,44.
20 Finkel, S.E. and Johnson, R.C. (1992) Mol. Microbiol. 6,
Control of mar expression: possible therapeutics 21 Sulavik, M.C., Dazer, M. and Miller, P.F. (1997) J. Bacteriol.
One strategy for controlling mar expression would be 22 Sulavik, M.C., Gambino, L.F. and Miller, P.F. (1995) Mol. Med.
using rational drug design. In theory, it should be possible to design a new antibiotic(s) that interferes 23 Dalrymple, B.P. and Swadling, Y. (1997) Microbiology 143,
with the functioning of MarA, its homologs or the genetic loci (e.g. acrAB) regulated by this protein.
24 Alekshun, M.N. and Levy, S.B. (1999) J. Bacteriol. 181,
Alternatively, White et al.45 have demonstrated the feasibility of using mar antisense DNA analogs to 25 Egan, S.M. and Schleif, R.F. (1994) J. Mol. Biol. 243, 821–829
increase the sensitivities of E. coli mar mutants to
26 Rhee, S. et al. (1998) Proc. Natl. Acad. Sci. U. S. A. 18,
antibiotics. These synthetic oligonucleotides not only decrease the expression of a marORA lacZ fusion but 27 Cohen, S.P. et al. (1989) Antimicrob. Agents Chemother. 33,
are able to restore the bactericidal activity of nor- 28 Okusu, H., Ma, D. and Nikaido, H. (1996) J. Bacteriol. 178,
floxacin, a fluoroquinolone, in a constitutive mar 29 Fralick, J.A. (1996) J. Bacteriol. 178, 5803–5805
From what is now known about the control of mar 30 Demple, B. (1991) Annu. Rev. Genet. 25, 315–337
gene expression in E. coli, the idea that a single mu- 31 Sanchez, L. et al. (1997) J. Bacteriol. 179, 6855–6857
tation that results in overexpression of a transcrip- 32 Fleischmann, R.D. et al. (1995) Science 269, 469–512
tional activator (MarA) can reduce antibiotic suscep- 33 McMurry, L.M., George, A.M. and Levy, S.B. (1994)
tibility is important to consider. It is anticipated that Antimicrob. Agents Chemother. 38, 542–546 these mechanisms might undermine treatment and 34 McDonnell, G. and Russell, A.D. (1999) Clin. Microbiol. Rev.
serve to thwart chemotherapy. Evidence for a role of the E. coli mar locus as an important stepping stone 35 McMurry, L.M., Oethinger, M. and Levy, S.B. (1998) Nature
towards clinically significant levels of resistance, result- 36 Cohen, S.P., Yan, W. and Levy, S.B. (1993) J. Infect. Dis. 168,
ing from mutations elsewhere on the chromosome, has begun to emerge. Nevertheless, its fundamental 37 McDermott, P.F. et al. (1998) J. Bacteriol. 180, 2995–2998
role could still be to ensure survival of its host under 38 Burns, J.L. and Clark, D.K. (1992) Antimicrob. Agents
39 Gustafson, J.E. et al. (1999) Antimicrob. Agents Chemother. 43,
40 Zhanel, G.G. et al. (1995) Antimicrob. Agents Chemother. 39,
We thank L.M. McMurry for critical reading of the manuscript. Our research is supported by NIH grant GM 51661.
41 Marshall, B. et al. (1984) J. Infect. Dis. 149, 1028–1029
42 Cohen, S.P. et al. (1988) Antimicrob. Agents Chemother. 32,
1 Gold, H.S. and Moellering, R.C., Jr (1996) New Engl. J. Med.
43 Maneewannakul, K. and Levy, S.B. (1996) Antimicrob. Agents
2 George, A.M. and Levy, S.B. (1983) J. Bacteriol. 155,
44 Oethinger, M. et al. (1998) Antimicrob. Agents Chemother. 42,
3 White, D.G. et al. (1997) J. Bacteriol. 179, 6122–6126
45 White, D.G. et al. (1997) Antimicrob. Agents Chemother. 41,
4 Aono, R. (1998) Extremophiles 2, 239–248
413 VOL. 7 NO. 10 OCTOBER 1999

Source: http://bmg.fc.ul.pt/Disciplinas/Regula_e_Diferenc/artigos/Alekshun99.pdf