07davis.pm

J. Mol. Microbiol. Biotechnol. (2001) 3(2): 179-184.
Enterococcus faecalis MDR Transporters 179 JMMB Symposium
Enterococcus faecalis Multi-Drug Resistance
Transporters: Application for Antibiotic Discovery
Deborah vR. Davis1, James B. McAlpine1*, Christopher
class, the first members of which were discovered in 1953 J. Pazoles1 , M. Kelly Talbot1, Elisabeth A. Alder1, Abbie
(Charney et al., 1953). Virginiamycin, a naturally occurring C. White1, Brandie M. Jonas2, Barbara E. Murray2,
streptogramin A and B combination, was discovered in 1955 George M. Weinstock2, and Bruce L. Rogers1,3
(Somer et al., 1957) and has been used as a growthpromoter in animal feed until recent years. It has been 1Phytera, Inc., 377 Plantation St., Worcester, MA 01605, proposed that this use has selected for virginiamycin- resistant strains of E. faecium, which are cross-resistant 2University of Texas Medical School, 6431 Fannin St., to Synercid® (Werner et al., 1998). Although Synercid® is active against most MLS (macrolide-lincosamide- 3 Present address: Genetics Institute, 35 Cambridge streptogramin) resistant strains, the existence of a known mechanism directed at this class is cause for concern.
Moreover, E. faecalis appears to be intrinsically resistantto Synercid® and virginiamycin (Rende-Fournier et al., Abstract
1993; Maddock et al., 1999; Lewis and Jones, 2000). Inaddition, E. faecalis resistance and/or superinfection was Using bioinformatics approaches, 34 potential multi-
seen during clinical trials with Synercid® (Blumberg et al., drug resistance (MDR) transporter sequences
representing 4 different transporter families were
Linezolid® is the first approved member of a class identified in the unannotated Enterococcus faecalis
(oxazolidinones) for which the pharmacophore was database (TIGR). A functional genomics campaign
reported in 1987 (Slee et al., 1987). Despite the fact that generating single-gene insertional disruptions
this is an exceedingly simple chemical class (Figure 1) and revealed several genes whose absence confers
very amenable to modern approaches for combinatorial significant hypersensitivities to known antimicrobials.
analog synthesis, it has taken 13 years to bring this drug We constructed specific strains, disrupted in a variety
to market. The rate and extent of the development of of previously unpublished, putative MDR transporter
resistance to oxazolidinones remains to be seen. However, genes, as tools to improve the success of whole-cell
with few new antibacterials directed towards resistant Gram antimicrobial screening and discovery. Each of the
positive bacteria in the clinic, the identification of new potential transporters was inactivated at the gene level
pharmacophores continues to be of urgent importance.
and then phenotypically characterized, both with single
While many mechanisms have been identified as the disruption mutants and with 2-gene mutants built upon
cause of emergent drug resistance, the expression, or a ∆ norA deleted strain background.
overexpression, of multidrug MDR efflux pumps have beendemonstrated to contribute significantly. Antibiotic Introduction
resistance has been directly attributed to this mechanismin a number of model organisms, as well as important Gram Emerging antibiotic resistance, particularly in nosocomial positive pathogens, including S. aureus (Yoshida et al., situations, has received wide exposure even in the popular 1990; Hsieh et al., 1998), E. faecalis (Lynch et al., 1997) press (Gorman, 1997; Preston, 1999). Awareness has been and S. pneumoniae (Gill et al., 1999). In Gram positive heightened by the incidence of vancomycin-resistant pathogens in particular, pmrA has been identified as an Enteroccocus faecalis (French, 1998) and the recent efflux pump associated with fluoroquinolone resistance in emergence of a similar resistance in Staphylococcus Streptococcus pneumoniae (Gill et al., 1999) and mreA aureus (MMWR, 1997). This observation is particularly from Streptococcus agalactiae has been characterized as significant since Vancomycin was considered to be the last a novel macrolide efflux gene (Dib-Hajj et al., 1997). These line of defense against Gram positive pathogens. Withinthe last 6 months, two new Gram positive antibacterials,Synercid and Linezolid, have been approved by the FDA.
These two agents belong to classes of antibiotics not inprior human therapeutic use and as such could give riseto the hope of a battle won. However, severalconsiderations warn against complacency. Synercid®(quinupristin-dalfopristin) is a combination of twosemisynthetic compounds, belonging to the streptogramin *For correspondence. Email [email protected].
Figure 1. The chemical structure of Linezolid® Figure 2. Dose/response curves demonstrate observed hypersensitivities in disruption mutants. Each IC50 was calculated using GraphPad Prism software.
export systems can extrude a wide range of structurally A general problem facing natural product extract-based dissimilar compounds. (Paulsen et al., 1996; Lewis et al., antimicrobial screening is, that because such extracts are 1997). Significantly, MDR efflux pumps have been complex mixtures of chemicals, the concentration of any described in phylogenetically diverse organisms including one particular component is undefined, and can be very bacteria (Nikaido, 1996), yeast (Kolaczkowski et al., 1998) low. In addition, the effectiveness of screening using whole and mammals (Chen et al., 1986), suggesting their cells is further diminished by the presence of multiple fundamental importance. Many MDR efflux pumps are
members of multi-gene families with the best described
being the ATP Binding Cassette (ABC) family of membrane
proteins, the Multiple Facilitator Superfamily (MFS) of
proteins, Multidrug And Toxic compound Extrusion family
(MATE), and the Small Multidrug Resistance (SMR) family,
(Michaelis and Berkower, 1995; Paulsen et al., 1996, Brown
et al., 1999).
A traditional approach to screening for new antimicrobials typically involves the inhibition of growth ofa screening strain, preferably of the target pathogen. Suchwhole-cell antimicrobial assays have the inherentadvantage in that advanced knowledge of the mode ofaction of an antibacterial agent is not required. Hence, suchassays are capable of revealing agents with completelynovel molecular targets and/or modes of action.
Figure 3. The chemical structure of Daunorubicin and Doxorubicin.
Enterococcus faecalis MDR Transporters 181 Table 1. E. faecalis putative transporter sequences and demonstrated Table 2. Table of observed MIC’s in insertional knockout strains of E. faecalis.
homologies to known multidrug transporters from bacteria, yeasts, and The total panel of drugs tested consisted of enoxacin, ofloxacin, norfloxacin, mammals. Shaded boxes represent E value calculations generated by ciprofloxacin, tylosin, erythromycin, spiramycin, clindamycin, lincomycin, BLAST searching each candidate sequence against GenBank. Unshaded virginiamycin, Synercid, chloramphenicol, tetracycline, ampicillin, bacitracin, E values represent hypothetical MDR transporters or transporter genes novobiocin, puromycin, gentamicin, rifamycin, pentamidine, crystal violet, known to efflux metabolites, metals, or other non-antibacterial compounds.
CCCP, acriflavin, daunorubicin, doxorubicin, ethidium bromide, rhodamine123, and mithramycin.
E. faecalis
E. faecalis
E. faecalis
Sequence
sequence
sequence
MIC µg/ml
identifier
identifier
identifier
Wild-type
-gene
WT/-gene
knockout
a A total of 28 known, chemically diverse antimicrobial agents were tested.
None of the 24 other single gene disrupted strains created consistently showed more than 2 fold increased sensitivity (compared to the parent) towards any of the b Compounds indicated were not tested against all of the single-gene disruptants. They were tested only against the specific gene disruption strain that initially was susceptible to a related chemical compound. Putative MDR Transporter Identification
a The E value relates to the smallest sum probability of the number of hits one can expect to see bychance when searching a database of a certain size. In practice, an E-value threshold of 0.001-0.1 isoften considered significant.
Active efflux has previously been demonstrated in E.
faecalis by studying the effect of energy (glucose fed versusglucose starved) on the accumulation of labeled multidrug pumps in many bacteria and yeast. In particular, norfloxacin, chloramphenicol, tetracycline and benzyl extrusion by MDR pumps of antibiotics present in screening penicillin by wild-type strains. (Lynch et al., 1997). Individual samples at low concentrations significantly decreases the pumps were not characterized, nor was any genetic chances of antimicrobial discovery. Thus, there is a strong analysis part of this work. In our study, we used the amino rationale for increasing the sensitivity of whole-cell screens acid sequence and membrane topology of several known for new antibiotics. For intracellular target assays, the MDR transporters to probe, using a series of BLAST effective concentration of a test agent is a function of the searches for homologs, in the TIGR unannotated E. faecalis rates of influx and efflux and the ability of the agent to bind database (www.TIGR.org) (Altschul et al., 1990; Worely et to its target within the cell. Each of these offers opportunity al.,1995, 1998). From this, an MDR pump gene candidate for increasing screen sensitivity. For example, one list of 23 ABC superfamily homologs, 9 MF family homologs, approach has been to compromise the cell wall and/or cell 1 MATE homolog, and 1 SMR homolog was generated membrane and hence increase the rate of influx. This is, (Table 1). All candidate sequences were retrieved from the however, accompanied by an increase in the rate of passive unannotated database and examined for open reading efflux. The prevalence of active efflux via transmembrane frames. The 23 ABC transporter candidates were also pumps provides a more targeted approach to constructing analyzed for predicted transmembrane regions using supersensitive whole-cell screening strains. In unpublished SOSUI (Hirokawa et al., 1998) through the BCM search work, we have improved the success of whole-cell launcher (Smith et al., 1996) and predicted ABC family antifungal screening by using specific strains deficient in a signature sequences and ATP-binding motifs using Motif variety of novel MDR-transporter genes. These strains are finder (Hofmann et al., 1999; www.motif.genome.ad.jp).
significantly more sensitive to known antifungal agents and Candidates containing 4-6 predicted transmembrane have enabled the discovery of new agents, from both regions, ABC signature sequences and ATP-binding motifs natural product and synthetic chemical libraries, where were placed at a high priority. MF candidates were also these agents would not have been discovered using wild examined for predicted transmembrane regions and were excluded if they did not meet the arbitrary criteria of havinggreater than nine of these regions.
We initiated a comprehensive functional genomics study aimed at generating site-directed insertionalmutagenesis to inactivate each individual gene listed in mutant has shown a significant hypersensitivity todaunorubicin, doxorubicin, and ofloxacin. Ethidium bromideand chloramphenicol showed a 2-fold increase insusceptibility, which given the inherent variability of an MIC-based assay can only be regarded as significant in that itwas a consistent result from several independentmeasurements. Figure 2B illustrates an approximately 60-fold increase in hypersensitivity for doxorubicin against theabc7 - strain compared to the wild-type OG1RF (Table 2).
Although conventional wisdom dictates that these MDRpumps are generally promiscuous and accept a widevariety of substrates, these data suggest that this may bea misconception arising from the large number of suchpumps present in each organism. For example, the ABC7transporter exhibits a high degree of specificity within astructural type. The increased sensitivity of strains with Figure 4. The chemical structures of Lincomycin and Clindamycin.
disabled ABC7 to the two anthracyclines, daunorubicin anddoxorubicin is significantly different (8 and 60-fold,respectively) despite the fact that they differ in structure Table 1. For these studies, a 500-700 base pair internal only by a single hydroxyl group. (Figure 3) fragment from the N-terminal end of each target gene was Similarly, the abc23- disruption strain has shown recovered from genomic DNA (E. faecalis OG1RF), cloned significant hypersensitivity to clindamycin, lincomycin, into vector pTEX4577 and the plasmids were virginiamycin and Synercid® (Table 2; Figure 2D).
electroporated (Qin et al., 1999) into OG1RF and selected Coincidentally, the lincosamides and the streptogramins for transformants on BHI agar supplemented with (virginiamycin and Synercid® exert their antibacterial kanamycin. Transformants were individually confirmed by activity by binding to the bacterial 50S ribosomal sub-unit PCR to have the targeted insertion of each plasmid. Once and inhibiting protein synthesis even though they have confirmed, each single-gene disruption mutant strain was unrelated chemical structures (Figures 4 and 5). The two tested against a panel of 28 compounds using a microbroth components of the streptogramins are thought to act dilution assay (NCCLS document M7-A5) and its synergistically by binding to two separate sites of the susceptibility compared to that of the wild-type.
bacterial 50S ribosomal subunit. Synercid® is a sterile The TIGR sequence database was generated from lyophilized formulation of two semi-synthetic pristinamycin vancomycin-resistant E. faecalis strain, V583. Although we derivatives, quinupristin (derived from pristinamycin I) and amplified candidate sequences abc9, abc12, abc20, and dalfopristin (derived from pristinamycin IIA) in the ratio of smr1 from this strain, we were unable to amplify them after 30:70 (w/w) (http://www.aventispharma-us.com).
several attempts from working strain OG1RF. The apparent Lincomycin and clindamycin also inhibit protein synthesis absence of these genes from OG1RF may be explained by blocking peptidyl transferase activity of the 50S ribosomal subunit (Gale et al., 1972).
Resistance to streptogramins can develop by multiple Novel Transporters and Observed Substrates
mechanisms including; 1) modification of the drug target,2) inactivation of the drug, and 3) impaired permeability The observed phenotypes of these strains functionally from active efflux or production of altered permeases implicated several genes as responsible for drug efflux in (Quinupristin/Dalfopristin Drug Monograph, 1998). The E. faecalis. In particular, the abc7 - single-gene disruption most commonly observed of these is the modification of Figure 5. Synercid® consists of two semisynthetic pristinamycin derivatives (quinupristin and dalfopristin) in a 30:70 (w/w) ratio.
Enterococcus faecalis MDR Transporters 183 the target, which is mediated by the erm gene. This gene Conclusions
encodes an RNA methylating enzyme that results inreduced binding of macrolides, lincosamides, and A total of 34 potential genes coding for efflux pumps were streptogramin B antibiotics (MLSB) (Quinupristin/ identified in Enterococcus faecalis by mining the TIGR Dalfopristin Drug Monograph, 1998). Group A database for motifs from published MDR transporter amino streptogramin antibiotics such as dalfopristin are not acid sequences and transmembrane domains. For 30 of affected by this type of resistance. In E. faecium, resistant these, we were able to detect and inactivate by insertional isolates have been associated with the presence of vat(E) mutagenesis the corresponding sequence in E. faecalis (satG) or vat(D) (satA) genes responsible for hydrolysis or OG1RF. The phenotypic susceptibilities of the disruption acetylation of quinupristin and dalfopristin, and studies mutants to a battery of 28 structurally diverse antimicrobial suggest that this is not the only mechanism (Soltani, et al., agents were characterized and strains abc7-, abc11-, 2000). It is likely that the abc23 gene identified in E. faecalis abc16-, and abc23- were significantly more sensitive to at is responsible for transporter-mediated resistance in this least one of the agents, demonstrating that these gene organism and could explain why E. faecalis, in general, is products are indeed MDR transporters. Significantly, these intrinsically resistant to streptogramins.
four ABC insertional mutants revealed large fold-differencesthat have not been observed in similar studies with other Multiple Transporter Knockouts
Gram positive organisms. For example, with L. lactis LmrA,only minor phenotypic differences in susceptibility were S. aureus mediates resistance to norfloxacin by expression observed when knockouts were analyzed (Ian Paulsen, of the norA gene which codes for an MDR transporter. The personal communication). For the remaining gene insertion OG1RF strain of E. faecalis was constructed in which a mutants, no significant drug susceptibility was detected homologous gene had been removed and examined for against the battery of compounds tested. More MDR its susceptibility to a panel of antimicrobial agents. In E.
transporters might be identified if the panel of compounds faecalis, the ∆ norA strain was only 2 to 3-fold more were expanded. It is interesting to note that our work in E.
susceptible to norfloxacin (Figure 2A) and it showed little faecalis indicates that the dominant efflux pumps present increased sensitivity to ciprofloxacin. Similarly, in S. aureus, in this organism appear to be ABC transporters, whereas a norA- strain was only 4-fold more sensitive than the wild- ad hoc studies in other bacteria have mainly implicated type to norfloxacin (Hsieh et al., 1998). Nonetheless, we proton motive force dependent transporters (i.e. QacA, reasoned that NorA may still recognize and transport Bmr, MexAB, and AcrAB) (Paulsen et al., 1996).
agents, but if these were substrates for other transporters, Transporters were identified that play a major role in the effect of deleting norA on the intracellular concentration determining the susceptibility of the organism to could be negligible. If a drug is the substrate for multiple streptogramins and lincosamides. NorA, although present transporters, maximal intracellular concentration can only in E. faecalis, seems to have only a modest effect, even be achieved by inactivating all pumps recognizing the drug on the susceptibility to norfloxacin. It is possible that other and, a priori, one would expect that sequential inactivation transporters may also pump norfloxacin and may of the transporters would be synergistic, with the greatest compensate for the deletion of norA. MDR disruption phenotypic effect observed when the last transporter is mutants hold considerable promise as supersensitive removed. Starting with the norA deletion strain, we created screening strains for the discovery of novel antibacterial a series of 2-gene mutants employing the methodology pharmacophores. However the large number of MDR used above. The sensitivity of these to a panel of transporters present in E. faecalis and our results with 2- antimicrobials exactly matched the susceptibility profile of gene disruptants suggest that several of these transporters each of the single gene disruptants described in Table 2.
will have to be inactivated to achieve a strain that is This result suggests that disruption of 2 transporters may hypersensitive to a broad structural variety of antibacterial be insufficient to produce a significant change in drug sensitivity. Lee and co-workers have studied the effects ofdifferential multiple transporter expression in Gram Acknowledgements
negatives, (E. coli and P. aeruginosa). (Lee et al., 2000).
This work was supported by SBIR Phase I Grant# 1R43 AI44543-01. We Here the situation is more complex in that there are two would like to thank Aventis for a sample of Synercid®. Preliminary sequence types of transporters; single component transporters which data was obtained from The Institute for Genomic Research website athttp://www.tigr.org. Sequencing of E. faecalis was accomplished with support extrude substrates from the cytoplasm to the periplasm from the National Institute of Allergy and Infectious Diseases. In addition, and multi-component pumps which take substrates all the we wish to thank John Cryan for technical assistance.
way from the cytoplasm to outside the cell. They concludedthat transporters of different structural types were References
synergistic in their effect, whereas transporters of the same Altschul, S., Gish, W., Miller, W., Myers, E., and Lipman, D. 1990. Basic type are essentially additive and explained this with the local alignment search tool. J. Mol. Biol. 215: 403-410.
two sequential pool model of the Gram negative cell. Under Blumberg, E.A., Mondler, H.D., Fuchs, A.E., Moldavi, J., Reboli, A.C., and Wood, C.A. 1996. Efficiency and toxicity of Synercid (Quinupristin/ this model one might assume that the situation for multiple Dalfopristin) [SYN(Q/D)] for vancomycin-resistant Enterococcus faecium gene inactivations in Gram positives would approximate (VREF) infections. 36th ICAAC, New Orleans, Paper LM32.
the effects seen there from transporters of the same type.
Brown, M.H., Paulsen, I.T., and Skurray, R.A. 1999. The multidrug efflux Work to produce strains lacking several transporters and protein NorM is a prototype of a new family of transporters. Molec.
Microbiol. 31: 394.
examine the effects of this in E. faecalis is ongoing.
Charney, J., Fisher, W. P., Curran, C., Machlowitz, R. A., and Tytell, A. A.
1953. Streptogramin, a new antibiotic. Antibiotics and Chemotherapy. 3: Chen, C.J., Chin, J.E., Ueda, K.,Clark, D.P., Pastan, I., Gottesman, M.M.
Soltani, M., Beighton, D., Philpott-Howard, J., and Woodford, N. 2000.
and Roninson, I.B. 1986. Internal duplication and homology with bacterial Mechanisms of Resistance to Quinupristin-Dalfopristin among isolates transport proteins in mdr1 (p-glycoprotein) gene from multi-resistant of Enterococcus faecium. Antimicrob Agents Chemother. 44: 433-436.
Somer, P. and Van Dijck, P., 1955. A Preliminary Report on Antiobiotic Dib-Hajj, C.J., Petitpas, J.W., and Yuan, W. 1997. Cloning and Number 899, a Streptogramin-like substance. Antibiotics Chemotherapy characterization of a novel macrolide efflux gene, mreA, from Streptococcus agalactiae. Antimicrob. Agents Chemother. 43: 999-989.
Werner, G., Klare, I., and Witte, W. 1998. Association between quinupristin/ French, G.L. 1998. Enterococci and vancomycin resistance. Clin. Infect.
dalfopristin resistance in glycopeptide-resistant E. faecium and the use of additives in animal feed. Eur. J. Clin. Microbiol. Infect. Dis. 17: 401- Gale, E., Cundliffe, E., Reynolds, P., Richmond, M., and Waring, M. 1972.
The Molecular Basis of Antibiotic Action. Second addition. Wiley and Sons.
Worely, K., Wiese, B., and Smith, R. 1995. BEAUTY: an enhanced BLAST- based search tool that integrates multiple biological information resources Gill, M.J., Brenwald, N.P., and Wise, R. 1999. Identification of an efflux into sequence similarity search results. Genome Research 5: 173-184 pump gene, pmrA, associated with fluoroquinolone resistance in Worley, K.C., Culpepper, P., Wiese, B.A., and Smith, R.F. 1998. BEAUTY- Streptococcus pneumoniae. Antimicrob. Agents Chemother. 43: 187-189.
X: enhanced BLAST searches for DNA queries. Bioinformatics 14: 890- Gorman, C. 1997. Germ warfare. A drug-resistant staph strain has doctors Yoshida, H., Bogaki, M., Nakamura, S., Ubukata, K. and Konno, M. 1990.
Hirokawa, T., Boon-Chieng, S., and Mitaku, S. 1998. SOSUI: classification Nucleotide sequence and characterization of the Staphylococcus aureus and secondary structure prediction system for membrane proteins.
norA gene, which confers resistance to quinolones. J. Bacteriol. 172: 6942- Hofmann K., Bucher P., Falquet L., Bairoch A. 1999. The PROSITE database, its status in 1999. Nucleic Acids Res. 27: 215-219.
Hooper, D.C. 1995. Quinolone mode of action. Drugs 49: Suppl 2: 10-15.
Hsieh, PC., Siegel, S., Rogers, B., Davis, D., and Lewis, K. 1998. Bacteria lacking a multidrug pump: a sensitive tool for drug discovery. Proc. Natl.
Acad. Sci. USA 95: 6602-6606.
Kolaczkowski, M., Kolaczkowska, A., Luczynski. J., Witek, S. and Goffeau, A. 1998. In vivo characterization of the drug resistance profile of the majorABC transporters and other components of the yeast pleiotropic drugresistance network. Microbial Drug Resistance 41: 143-158.
Lee, A., Mao, W., Warren, M., Mistry, A., Hoshino, K., Okumura, H. and Lomovskaya, O. 2000, Interplay between efflux pumps may provide eitheradditive or multiplicative effects on drug resistance. J. Bacteriol. 182: 3142-3150.
Lewis, K., Hooper, D., and Ouellette, M. 1997. Microbial multidrug resistance pumps: New developments and clinical significance. ASM News 63: 605-610.
Lewis, M.T., and Jones, R. N. 2000. Activity of macrolides, lincosamides, streptogramins and fluoroquinolones against Streptococcus pneumoniaeand enterococci isolates from the western hemisphere. Braz. J. Infect.
Dis. 4: 15-21.
Linden, P., Pasculle, A.W., McDevitt, D., and Dotterwiech, L. 1996.
Emergence of superinfection with Enterococcus faecalis during RP 59500treatment of Vancomycin-resistant Enterococcus faecalis infection (VREF).
36th ICAAC, New Orleans, Paper LM31.
Lynch, C., Courvalin, P. and Nikaido, H. 1997. Active efflux of antimicrobial agents in wild-type strains of Enterococci. Antimicrob. Agents Chemother.
41: 869-871.
Maddock, S., Gordon, S., Chen, S., and Mitchell, D. 1999. In vitro activities of Linezolid, Trovofloxacin, and Quinupristin-Dalfopristin against sensitiveand multiresistant Gram positive organisms. 39th ICAAC, SanFrancisco,Paper 1254.
MMWR. 1997. Reduced susceptibility of Staphylococcus aureus to vancomycin-Japan, 1996. Morb. Mortal. Wkly. Rep. 46: 624-626.
NCCLS. Methods for Dilution Antimicrobial Susceptibility Tests for Bacteria that Grow Aerobically-Fifth Edition; Approved Standard. NCCLS documentM7-A5.
Michaelis, S, and Berkower, C. 1995. Sequence comparision of yeast ATP- binding cassette proteins. Cold Spring Harb Symp. Quant. Biol. 60: 291-307 Nikaido, H. 1996. Multidrug efflux pumps of gram-negative bacteria. J.
Paulsen, I.T., Brown, M.H. and Skurray, R.A. 1996. Proton-dependent multidrug efflux systems. Microbiol. Rev. 60: 575-608.
Preston, R. 1999. What new things are going to kill me? Time. 154: 86-87.
Qin, X., Teng, F., Xu, Y., Singh, K., Weinstock, G., and Murray, B. 1999.
Targeted mutagenesis of enterococcal genes. Meth. Cell Sci. 20: 21-33.
Quinupristin/Dalfopristin (Synercid®) Drug Monograph. UHC Consortion.
Rende-Fournier, R., Leclercq. R., Galimand, M., Duval, J., and Courvalin, P. 1993. Identification of the satA gene in E. faecium BM4145. Antimicrob.
Agents Chemother. 37: 2119-2125 Slee, A.M., Wuonola, M.A., Ripley, R.J., Zajac, I., Zawada, M.J., Bartholomew, P.T., Gregory, W.A., and Forbes, M. 1987. Oxazolidinones,a new class of synthetic antibacterial agents. Antimicrob. AgentsChemother. 3: 1791-1797.
Smith R.F., Wiese, B.A., Wojzynski, M.K., Davison, D.B., and Worley, K.C.
1996. BCM Search Launcher: An integrated interface to molecular biologydata base search and analysis services available on the World Wide Web.

Source: http://www.ibs.fr/IMG/pdf/Efaecalis_MDR_Davis_et_al_JMMB2001.pdf