Plasmid-Determined Resistance to Antimicrobial Toxic Metal ... · Plasmid-Determined Resistance to Antimicrobial Drugs and Toxic Metal Ions in Bacteria T. J. FOSTER Microbiology Department,

MICROBIOLOGICAL REVIEWS, Sept. 1983, p. 361-409 Vol. 47, No. 30146-0749/83/030361-49$02.00/0Copyright C 1983, American Society for Microbiology

Plasmid-Determined Resistance to Antimicrobial Drugs andToxic Metal Ions in Bacteria

T. J. FOSTER

Microbiology Department, Moyne Institute, Trinity College, Dublin 2, Ireland

INTRODUCTION........................................................... 362INACTIVATION MECHANISMS.................................................. 363

Resistance to ,B-Lactam Antibiotics ............................................... 363Plasmid-determined ,B-lactamases in gram-negative bacteria ...... ................. 363Plasmid-specified FI,anctamases in gram-positive bacteria ....... ................... 365Mechanism of hydrolysis of -lactams ............. ............................. 365Factors affecting expression of 3-lactam antibiotic resistance in

gram-positive and gram-negative bacteria ......... .......................... 365Do some plasmils determine a barrier to the penetration of j-lactam

antibiotics?........................................................... 367Regulation of f-lactamase exRression ............. .............................. 367

Resistance to Chloramphenicol.................................... ................ 367CATs in gram-negative bacteria................................................ 367CATs in gram-positive bacteria ................... ............................. 368Mechanism of Cm inactivation................................................. 368Catalytic properties of CAT................................................... 368Type I CAT confers resistnce to fusidic acid .................................... 368Regulation of CAT synthesis in E. coli.......................................... 369Regulation of CAT synthesis in S. aureus........................................ 370

Resistance to Aminoglycoside-Aminocyclitol Antibiotics ........ ..................... 370Transport of aminoglycosides in sensitive and resistant bacteria ...... .............. 371Location of aminoglycoside-modifying enzymes in the bacterial cell ..... ............ 371Classification of aminoglycoside-aminocyclitol antibiotics and their

modifying enzymes ....................................................... 372Resistance to Mercuric Ions and Organomercurials ................................. 373

Sensitivity and resistance to mercurial compounds................................ 373Spectrum of resistance to organomercurials in gram-positive and

gram-negative bacteria.................................................... 375Mercuric reductase and mechanism of reduction of mercuric ions ...... ............ 380Detoxification of organomercurial compounds.................................... 380Evidence for an Hg2+-specific transport system and its role in resistance ..... ....... 380Expression of resistance by multicopy plasmids ......... ......................... 381Genetic analysis of the mer region.............................................. 381Regulation of expression of mer genes........................................... 382

BYPASS MECHANISMS......................................................... 382Sulfonamide Resistance ......................................................... 382

Inhibitory effect of sulfonamides ................. .............................. 382Plasmid-specified Sur DHPS................................................... 382Regulation of plasmid-specified DHPS ............. ............................. 383

Trimethoprim Resistance........................................................ 384Inhibitory effects of Tp ....................................................... 384Plasmid-specified Tp-resistant DHFR ............. .............................. 384

DECREASED ACCUMULATION INVOLVING EFFLUX MECHANISMS.............. 384Resistance to Tetracyclines ...................................................... 384

Uptake of Tc by susceptible cells ............................................... 384Naturafly occurring Tc resistance determinants ......... ......................... 385Mechanism of Tc resistance................................................... 385Structure and function of the TnlO Tc resistance region........................... 386Regulation of tet gene expression ................. .............................. 387

Resistance to Cadmium ......................................................... 388Resistance to Arsenate.......................................................... 388

361

on March 24, 2021 by guest

http://mm

br.asm.org/

Dow

nloaded from

http://mmbr.asm.org/

ALTERATION TO THE ANTIBIOTIC TARGET SITE ....... ....................... 388Resistance to Antibiotics in the MLS Group ......... .............................. 388

Occurrence of MLS resistance in natural isolates................................. 389Mechanism of resistance to MLS antibiotics ......... ............................ 389Induction of MLS resistance determinants....................................... 389Regulation of expression of MLS resistance encoded by plasmid pE194:

the translational attenuation model ......................................... 389Structure of nascent transcripts and basal expression of methylase.................. 390The paradox of induction by a mechanism involving translational attenuation ........ 390

UNCHARACTERIZED RESISTANCE MECHANISMS............................... 391Resitance to Fusidic Acid in S. aureus............................................ 391Nonenzymatic Chloramphenicol Resistance in Gram-Negative Bacteria ................ 391Resistance to Fosfomycin........................................................ 392Resitance to Cadmium Ions Specified by the CadB Determinant ...... ............... 392Resistance to Arsenite and Antimony(Ii).......................................... 392Resistance to Silver Ions........................................................ 392Resistance to Other Heavy Metal Compounds...................................... 392Resistance to Ethidium Bromide ............... .................................. 392

MECHANISMS FOR INCREASING EXPRESSION OF ANTIBIOTIC RESISTANCEDETERMINANTS ....................................................... 392

Amplification by Tandem Duplication............................................. 392Acquisition of Multiple Copies of a Transposon ........ ............................ 394Increased Plasmid Copy Number................................................. 394Increased Transcription of Redsbtance Genes....................................... 394

ORIGINS AND EVOLUTION OF ANTIBIOTIC RESISTANCE DETERMINANTS ...... 394Origin of MLS Resistance Determinants........................................... 395Origin of Amnolycoside-Modifying Enzymes..................................... 395Origin of Tetracycline Resistance................................................. 396Origins of Chloramphenicol Restance............................................ 396Origins of -Lactamases........................................................ 396Orgin of Sulfonamide and Trimethoprim Resistance............................... 397

CONCLUDING REMARKS AND FUrTURE PROSPECTS ....... ..................... 397ACKNOWLEDGMENTS.......................................................... 398LITERATURE CITED ............................................................ 398

INTRODUCTION

Most clinically significant antibiotic resistanceis determined by genes located on extrachromo-somal DNA elements called plasmids (31, 108,154). Different species of bacteria harbor char-acteristic types of plasmids, some of which canmediate their own transfer by conjugation. Also,resistance genes are often incorporated into dis-crete genetic units called transposons (204),which have the capacity to transpose from oneDNA molecule to another. This has undoubtedlycontributed to the rapid dissemination of antibi-otic resistance by providing an efficient mecha-nism for incorporating resistance determinantsinto new vectors which can transfer to andstably replicate in diverse hosts.The major objectives of this review are to

discuss (i) the biochemical mechanisms of plas-mid-specified resistance to drugs and toxic metalions, (ii) the regulatory mechanisms controllingexpression of resistance genes, (iii) comparativestudies on resistance determinants, and (iv) theevolutionary origins of resistance genes.One might ask the questions, why concentrate

on plasmid-specified resistance and why consid-er resistance to drugs and metal ions? As men-tioned above, resistance determined by plasmidsis most frequently responsible for resistance inclinical situations. Resistance conferred bymutations in chromosomal genes occurs by dif-ferent biochemical mechanisms and is less com-monly encountered in nature. However, impor-tant examples of mutational changes have beenrecognized and these are discussed. The distinc-tion between plasmid-encoded and chromosom-ally encoded resistance can become blurred be-cause transposons can integrate into thechromosome and have been found in this loca-tion in clinical isolates. However, these twotypes of chromosomally specified resistance canbe readily distinguished by biochemical and ge-netic tests. It is instructive to compare thebiochemical mechanisms of resistance to metalions and drugs since most metal resistance isplasmid determined and has presumably evolvedin response to similar selective pressures.Knowledge of metal ion resistance may provehelpful in understanding antibiotic resistancemechanisms.

362 FOSTER MICROBIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from


PLASMID-DETERMINED DRUG RESISTANCE 363

There are four classic mechanisms of resist-ance specified by plasmids: (i) inactivation, (ii)impermeability, (iii) bypass, and (iv) alteredtarget site (93). These form the major headingsfor this review. Recent work allows more pre-cise definitions of some mechanisms and sug-gests that additional mechanisms might exist.For example, resistance previously described asbeing due to decreased permeability has beenshown to be caused by specific efflux systems.Also, intracellular binding now seems to be avalid mechanism for immobilizing an inhibitor.The advent of rapid DNA sequencing tech-

niques means that detailed comparisons ofresistance genes from different sources can bemade to provide insights into their origins andevolution. The notion that R-plasmid-specifiedantibiotic resistance may have evolved as self-protection mechanisms in antibiotic-producingmicroorganisms could be strengthened by thisapproach.

INACTIVATION MECHANISMSResistance to j-Lactam Antibiotics

P-Lactam antibiotics inhibit enzymes in-volved in cell wall biosynthesis and cell division(for review, see reference 317). The proteins towhich the drugs bind (penicillin-binding proteins[PBPs]) are integral components of the cytoplas-mic membrane of bacteria. Gram-negative bac-teria are intrinsically resistant to many ,B-lactamdrugs. This is partly due to the inability of thedrugs to diffuse easily through pores in the outermembrane (58, 317) and partly to low levels ofchromosomally encoded 3-lactamase (81, 284).Acquired resistance to 3-lactam drugs in clini-

cal isolates may be due to 3-lactamase activityspecified by plasmids or to mutational changesin chromosomal loci. These mutations alter theaffinity or amount of PBPs or (in the case ofgram-negative organisms) increase the imperme-ability of the outer membrane. Both types ofmutation can be selected in the laboratory (135,153, 208). Mutations altering PBPs are responsi-ble for penicillin resistance in Neisseria gonorr-hoeae (100) and Streptococcus pneumoniae(145) and for methicillin (155, 158) and cephra-dine (134) resistance in Staphylococcus aureus,and they can be a mechanism for resistance topiperacillin and carbenicillin in Pseudomonasaeruginosa (80, 135, 136, 316, 269). In somestrains a single-step mutation may be responsi-ble for resistance, whereas in others resistancemay be due to several mutations (145, 441). Insome isolates of P. aeruginosa carbenicillinresistance is due to a combination of reducedsensitivity of PBPs and increased outer mem-brane impermeability (269, 316).

Another proposed mechanism of resistance toj-lactamase-insensitive cephalosporins is theapparent ability of the periplasmic enzyme tobind to drug and prevent it from reaching itstarget (384). Also, the phenomenon of toleranceto penicillins in S. aureus appears to be due tomutations that reduce autolytic activity (411).

It is conceivable that the resistance mecha-nisms described above will become more preva-lent after the introduction of 3-lactamase inhibi-tors and j-lactamase-insensitive drugs intoclinical practice.

Plasmid-determined j-lactamases in gram-neg-ative bacteria. High-level resistance to broad-spectrum P-lactam antibiotics in gram-negativebacteria is usually due to j-lactamase activityspecified by plasmids. In addition, most, if notall, gram-negative organisms express low levelsof chromosomally encoded j-lactamase whichmay contribute to intrinsic resistance (313, 376).The question of j-lactamase nomenclature

should be mentioned at this point. j-Lactamaseshave been classified into three groups (A, B, andC) on the basis of evolutionary and mechanisticconsiderations (8). Class A includes the S. aur-eus penicillinases and the TEM P-lactamasespecified by plasmids in gram-negative bacteria.The chromosomally encoded cephalosporinasesof gram-negative bacteria are class C enzymes.Another classification system based on bio-chemical criteria places j-lactamases into fivegroups (I to V), with the chromosomal (class C)cephalosporinases in group I and the plasmid-specified penicillinases in groups III and V (313).Some authors have preferred to name plasmid-specified enzymes of gram-negative bacteria ac-cording to the bacterial host in which the en-zyme was first isolated or after the P-lactamdrug which they best hydrolyze. In the absenceof an agreed convention, this scheme is usedhere for ease of reference to published work.A number of plasmid-specified 3-lactamases

in gram-negative bacteria have been distin-guished by isoelectric focusing and inhibitorstudies and by substrate specificity (Table 1). Allare expressed constitutively and are located inthe periplasmic space. Most are predominantlypenicillinases. These may have low activityagainst cephalosporins such as cephaloridineand cephalothin. Other drugs such as cephalexinand cefuroxime are hydrolyzed poorly, if at all,by most enzymes (Table 1).The TEM enzyme was the first R-plasmid-

specified ,B-lactamase to be recognized in gram-negative bacteria (90). The enzyme was namedafter the RTEM plasmid. There are two sub-types called TEM-1 and TEM-2, which can bedistinguished by isoelectric focusing. They differin sequence by a single amino acid (9). These

VOL. 47, 1983


http://mm

br.asm.org/

Dow

nloaded from


TABLE 1. Properties of ,B-lactamases of gram-positive and gram-negative bacteriaa

Type of ,- Relative rates of hydrolysis' Inhibition by: Mol wt Refer-lactamase Pen Amp Carb Oxa Meth Cep Clox pCMB NaCI pI (103) ence(s)

TEM-1 100 106 10 5 0 76 S R R 5.4 28.5d 90 160TEM-2 100 107 10 5 0 74 S R R 5.6 28.5dSHV-1 100 212 8 0 0 56 S S/R R 7.6 17.0 251HMS-1 100 253 14 0 0 183 S S R 5.2 21.0 251ROB 100 186 25 6 24 8.1 323OXA-1 100 382 30 197 332 30 R PS S 7.4 23.3 83, 160OXA-2 100 179 15 646 23 37 R R S 7.45/7.7 44.6 83, 160OXA-3 100 178 10 336 29 44 R R S 7.1 41.2 83, 160OXA-4(PSE-2) 100 267 121 317 803 32 R S S 6.1 12.4 248CARB-1(PSE-4) 100 88 150 8 16 40 R 5.3 32.0 162CARB-2(PSE-1) 100 90 97 0 0 18 R S 5.7 28.5 248CARB-4(PSE-3) 100 101 253 10 R 6.9 12.0 326CEP-1 100 5 0 325 R R 8.1 31.8d 27CEP-2 100 low 48 108'e R R 8.1 36.2 231S. aureus A, B 100 185 4.5 1.5 10 28.8d 102, 307S. aureus C 100 1 0.6 102, 307

a Abbreviations: Pen, Benzyl penicillin; Amp, ampicillin; Carb, carbenicillin; Oxa, oxacillin; Meth, methicil-lin; Clox, cloxacillin; Cep, cephaloridine; pCMB, p-chloromercuribenzoate.

b The rates of hydrolysis are expressed as a function of the benzyl penicillin rate. The data were selected fromMatthew (248), Richmond (307), and Dyke (102) primarily to highlight comparatively the important features ofthe enzymes.

S, Sensitive; R, resistant; PS, partially sensitive.d Molecular weight deduced from DNA sequence of the gene (189, 371) or amino acid sequence or both of the

polypeptide (7, 9). The molecular weight given for CEP-1 is that of the ampC ,B-lactamase of E. coli.' Also hydrolyzes cephalothin, cefamandole, and cefazolin at rates similar to that of cephaloridine.

enzymes are widely distributed in nature interms of geographical distribution, associationwith plasmids of different incompatibilitygroups, and occurrence in different bacterialspecies (249). Undoubtedly, this widespread oc-currence is the result of the TEM P-lactamasegene being carried by transposons (161, 209).The SHV-1 (sulfhydryl variable) ,B-lactamase

has a substrate profile similar to that of TEMenzymes but differs by hydrolyzing ampicillin ata greater rate (249, 251). It has the distinctiveproperty of variable inhibition by p-chloromer-curibenzoate depending on the substrate. Thus,hydrolysis of benzyl penicillin is not affected byp-chloromercuribenzoate, whereas the reactionwith cephaloridine is inhibited.

P-Lactamase HMS-1 is very similar to SHV-1in substrate profile, with the exception that ithydrolyzes cephaloridine at a threefold greaterrate. Also, it is uniformly sensitive to p-chloro-mercuribenzoate. It is a rare enzyme, havingbeen found in a single strain of Proteus mirabilis(249, 251).Another enzyme (ROB) with a broad TEM-

like substrate specificity has been isolated fromHaemophilus influenzae type b (323). It differsfrom TEM in having a higher pl and a faster rateof ampicillin hydrolysis.

The OXA group of P-lactamases has the dis-tinctive property of hydrolyzing isoxazoyl peni-cillins (including oxacillin) and methicillin.Three different OXA enzymes were originallydescribed (83). In addition, the enzyme classi-fied as PSE-2 because it was discovered in P.aeruginosa (see below) has been renamed OXA-4 (A. M. Philippon, G. Paul, and G. A. Jacoby,Program Abstr. Intersci. Conf. Antimicrob.Agents Chemother. 21st, Chicago, Ill., abstr.no. 680, 1981). OXA-1 and OXA-4 hydrolyzemethicillin at much greater rates than do OXA-2and OXA-3. The OXA-1, -2, and -3 enzymes arespecified by plasmids of three, four, and twodifferent incompatibility groups, respectively(249). In addition, OXA-1 and OXA-3 have beenencountered in P. aeruginosa (185; A. Philip-pon, G. Paul, and G. A. Jacoby, 13th Int. Conf.Microbiol., abstr. no. P42:1, 1982). Also, PSE-2/OXA-4 has-been found in several species ofenteric bacteria (D. M. Livermore and J. D.Williams, personal communication). The OXA-1determinant has been shown to be transposable(435).

Several plasmid-specified 3-lactamases hy-drolyze carbenicillin at a greater rate than doespenicillin. They were originally discovered in P.aeruginosa and are called pseudomonas-specific



http://mm

br.asm.org/

Dow

nloaded from



enzymes (PSE; 185). These enzymes have alsobeen named after the drug which they besthydrolyze (211; A. Philippon and M. Matthew,personal communications). Thus, PSE-2 iscalled OXA-4 because it attacks isoxazoyl peni-cillins best (see above). The PSE-4/CARB-1enzyme is indistinguishable from the classicchromosomal Dalgleish enzyme (126, 162). Itis commonly specified by a transposon locatedin the P. aeruginosa chromosome (349). Thetype strain for plasmid-specified PSE-4/CARB-1(PU21, pMG19) actually carries the ,-lacta-mase gene on a similar chromosomal transposon(G. R. Jacoby, L. R. Knobel, and L. Sutton,13th Int. Conf. Microbiol., abstr. no. P21:9,1982). Indeed, it is possible that the originalDalgleish strain carries its PSE-4/CARB-1 deter-minant on such an element.The PSE-1/CARB-2 enzyme was originally a

determinant of the narrow-host-range IncP2plasmids ofP. aeruginosa (252), but has recentlybeen reported in enteric bacteria and is linked totransposons (1%, 259). CARB-4/PSE-3 is rare,being determined by a single plasmid in P.aeruginosa (249, 326). Another carbenicillin-hydrolyzing enzyme, CARB-3 (211), does notseem to be plasmid specified and is not listed inTable 1.There have been two reports of plasmid-speci-

fied P-lactamases which predominantly hydro-lyze cephalosporins. The CEP-1 determinantwas transferred from Proteus mirabilis intoEscherichia coli, where it specified a 3-lacta-mase with properties indistinguishable fromthose of the E. coli chromosomal ampC enzyme(27). It has not been demonstrated unequivocal-ly that the CEP-1 structural gene resides on theplasmid (27; M. Matthew, personal communica-tion). It is possible that the plasmid encodes aregulatory function which stimulates expressionof ampC. A true plasmid-encoded cephalospo-rin-hydrolyzing enzyme was recently discoveredin a strain of Achromobacter sp. (230, 231).

In addition to the enzymes described aboveand in Table 1, five new 1-lactamases haverecently been found in E. coli and P. aeruginosa.One has broad TEM-like substrate specificity,whereas the others are OXA-like but have pIvalues different from those of OXA-1 to -4(A. A. Medeiros and G. A. Jacoby, ProgramAbstr. Intersci. Conf. Antimicrob. Agents Che-mother. 22nd, Miami Beach, Fla., abstr. no.716, 1982).The relationships between the plasmid-speci-

fied P-lactamases have yet to be established.Immunological studies have provided some in-formation. The CARB/PSE and OXA enzymesdid not cross-react with anti-TEM serum (377).There have been confficting reports of cross-

reaction between TEM-1 and SHV-1 (377; G.Paul, A. Philippon, M. Bartelemy, R. Labia, andP. Nevet, Program Abstr. Intersci. Conf. Anti-microb. Agents Chemother. 21st, abstr. no. 681,1981). The evolutionary relationships between,B-lactamases could be profitably studied byDNA hybridization and sequencing studies.

Plasmid-specified ji-lactamases in gram-posi-tive bacteria. Penicillin resistance in S. aureus isdetermined by P-lactamases which are usuallyplasmid encoded. There are four minor variantsof staphylococcal penicillinase (types A to D),three of which (types A, B, and C) have beenstudied in some detail (307). The enzymes arepredominantly penicillinases, hydrolyzing ben-zyl penicillin and ampicillin at high rates. Theyhydrolyze methicillin, oxacillin, and cloxacillinpoorly. With the exception of the type D enzyme(320), they are expressed at high levels only afterinduction, with most of the activity being extra-cellular. It is assumed that the four penicillinasesare closely related, with a small number ofdifferences in amino acid residues being suffi-cient to cause the small but significant differ-ences in hydrolysis rates (307).Mechanism of hydrolysis of P-lactams. Recent-

ly the serine-70 residue has been shown to be theactive-site nucleophile of the class A Bacilluscereus and TEM 3-lactamases (67, 111, 112). Itis now clear that the class C ,B-lactamases arealso serine enzymes (203), although they areotherwise unrelated to class A enzymes (189).The interaction between ,-lactam drugs and1-lactamases is kinetically complex and may berepresented, in the simplest case, as

k, k2 k3[H201E + EL .--* E-L -, E + L'

where EL is a noncovalent complex between theenzyme (E) and the P-lactam (L) which mayeither dissociate (k-1) or break down (k2) toform a covalent intermediate, the acyl enzymeE-L. The acyl enzyme may hydrolyze (k3) toyield free enzyme and the acid derivative of the1-lactam (L') (Fig. 1). It appears that the rateof hydrolysis of the acyl enzyme determineswhether compounds are degraded efficiently orbehave as ,B-lactamase inhibitors (113). The acylenzyme may also undergo P-elimination to yieldenzyme-bound chromophores (113).

Factors affecting expression of (-lactam antibi-otic resistance in gram-positive and gram-nega-tive bacteria. The gram-positive bacterium S.aureus secretes copious amounts of ,B-lactamaseinto the surrounding medium. The enzyme mustinactivate the drug before it penetrates the cellwall and reaches the cytoplasmic membrane-

VOL. 47, 1983


http://mm

br.asm.org/

Dow

nloaded from


366 FOSTER

located PBP targets. There is a pronounced"inoculum effect" on the level ofresistance thatcan be attained in S. aureus. A single cell isvirtually defenseless against even low concen-trations of the drug (102). However, the resist-ance level increases by as much as 10,000-foldwhen larger inocula are used (102).The kinetic properties of the P-lactamase also

influence the level of resistance (150). For exam-ple, the S. aureus P-lactamase has a very lowaffinity for methicillin (278). In contrast, theaffinity for cephalothin is very high but the rateof hydrolysis is low (149). In both cases the rateof hydrolysis is so slow that the drug is essential-ly stable and can be used at low concentrationsto kill penicillin-resistant staphylococci. Theamount of enzyme synthesized will also be animportant determinant of resistance.Plasmid-specified P-lactamases of gram-nega-

tive bacteria are expressed constitutively andare located in the periplasmic space between the

RCONH C

A. o7N /CH3

RCONH S

B. o HN CHEnzyme

4'RCONH S

+_S\/CH3C. 0 HN /CH3

OHCOO-

FIG. 1. Inactivation of penicillin antibiotics by -lactamase. The penicillin molecule (A) forms an acylenzyme intermediate (B) with a serine residue in theactive center of the 1-lactamase. The hydrolyzedpenicilloic acid derivative (C) is released.

cytoplasmic and outer membranes. Because ofthis strategic location the cells are protected bysmaller amounts of an enzyme which generallyhas a low affinity for its substrate compared withthe gram-positive enzymes (150, 310). There isno marked inoculum effect if the 1-lactamasecan cope with the drug as it passes into theperiplasm, and single cells will be able to grow inquite high concentrations of drug. In contrast, ifthe enzyme cannot inactivate the drug quicklyenough, an inoculum effect may occur. Some ofthe cells in a large population will be killed bythe drug and will release their 3-lactamase intothe medium. This will hydrolyze the drug andallow some of the surviving cells to grow.The level of resistance to ,B-lactam antibiotics

in gram-negative bacteria is determined by theinterplay of three factors: (i) the amount ofenzyme, (ii) kinetic properties of the enzyme(i.e., rate of hydrolysis), and (iii) the ability ofthe drug to penetrate the outer membrane.

(i) Amount of enzyme. In E. coli there is adirect relationship among the number of copiesof the TEM ,B-lactamase gene, the amount ofenzyme synthesized, and the resistance levelexpressed (a gene-dosage effect) (399). Also,TEM ,-lactamase genes located on plasmids ofdifferent incompatibility groups which are indig-enous to one or a few bacterial species oftenexpress different levels of enzyme when intro-duced into the same strain of E. coli. This maybe due to different rates of transcription of thegene (436). Perhaps mutations have occurred inTEM promoters to adapt them to transcriptionin different species of gram-negative bacteria,resulting in a variation in transcription by E. coliRNA polymerase. Poor transcription may alsoexplain the observation that plasmid RP1 ex-presses 1/50 the level of P-lactamase in Proteusmirabilis compared with E. coli (313).The level of P-lactamase will also be depen-

dent on factors that influence plasmid copynumber and thus the number of gene copies percell. The copy number of plasmid Rl is higher atslower growth rates, with a corresponding gene-dose effect for TEM ,B-lactamase (105). Differ-ences in plasmid copy number may also explainthe increased synthesis of 1-lactamase by anaer-obically grown E. coli (304).

(ii) Ability of the drug to cross the outermembrane. The intrinsic resistance of plasmid-free cells is partly determined by the barrier todrug penetration posed by the outer membrane(58, 71, 310, 311). The ability of f-lactam drugsto diffuse through outer membrane pores isdependent on their charge and hydrophilic prop-erties (58). Also, the expression of resistance in1-lactamase-producing cells is in part controlledby this surface layer. For example, E. coli cells

MICROBIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from


PLASMID-DETERMINED DRUG RESISTANCE

synthesizing TEM P-lactamase are quite sensi-tive to cephaloridine despite the fact that thepurified enzyme hydrolyzes the drug at almostthe same rate as ampicillin (Table 1) (313), towhich high-level resistance is expressed. Cepha-loridine is thought to penetrate the outer mem-brane very rapidly and swamps the 3-lactamase(313). Different rates of drug penetration are alsoresponsible for the cryptic nature of 3-lactamaseactivity towards some drugs in gram-negativeorganisms. The specific activities of 1-lacta-mases are often higher in broken cell extractsthan in whole cells (313). Also, the outer mem-branes of different gram-negative bacteria varyin their ability to exclude ,3-lactams. Extremeexamples are the highly impervious P. aerugin-osa (311) and the relatively porous H. influenzaemembranes.

(iii) Kinetic properties of f-lactamase. Just asfor staphylococcal ,B-lactamases discussedabove, the affinity of the enzyme for substrateand the rate of reaction are important determi-nants of resistance (150).Do some plasmids determine a barrier to the

penetration of ,-lactam antibiotics? The sugges-tion that some 3-lactamase-specifying plasmidsalso determine a barrier that hinders the diffu-sion of drugs across the outer layers of the cellwas proposed to explain the expression of high-level carbenicillin resistance in strains with lowor undetectable enzyme activity (82, 437). Re-cently it was shown that the resistance ex-pressed by plasmids with defective ,B-lactamaseswas entirely attributable to residual P-lactamaseactivity of the mutant proteins (76, 77).

Regulation of ,-lactamase expression. (i)Gram-negative bacteria. All plasmid-specifiedP-lactamases of gram-negative bacteria are ex-pressed constitutively. Factors that can indirect-ly affect the level of ,-lactamase expression andhence the resistance level of the cell are plasmidcopy number, which may be higher at slowergrowth rates (105), and anaerobic growth (304).

In contrast, the chromosomally encoded ,B-lactamase of P. aeruginosa is expressed induci-bly (81, 284, 313), although the regulatory mech-anism involved is not understood. Thetranscription of the E. coli chromosomal ampCgene is subject to growth rate-dependent regula-tion by an attenuation mechanism (190).

(ii) Gram-positive bacteria. The type A, B,and C penicillinases specified by S. aureus plas-mids are expressed inducibly (102, 307, 308),whereas the type D enzyme is constitutive (320).Inducible penicillinase synthesis is controllednegatively by the product of the penl gene (308).Trans-recessive penI mutations caused constitu-tive expression of penicillinase (308). Also, peni-cillinase synthesis can be derepressed by incor-

poration of the amino acid analog 5-methyltryptophan into repressor protein (180).The operator binding function of the repressor isprobably inactivated.Mutants which expressed low basal levels of

wild-type penicillinase either partly or noninduc-ibly formed an active repressor in diploid cellscarrying a compatible penI mutant plasmid(309). This led to the suggestion of a secondplasmid-located regulatory gene. However, amore acceptable explanation is that the nonin-ducible phenotype is caused by a penP (super-repressor) mutation. In diploid cells carrying theputative penP mutant along with a penI- ele-ment, mixing of mutant repressor subunits couldoccur to form an active molecule by a form ofintragenic complementation. Also, the putativepenP mutants were inducible by 5-methyltrypto-phan, a finding that is difficult to reconcile with apositive regulator model (183). Another class ofmutant that expressed very low constitutivelevels of penicillinase has two mutations, a penPlesion along with a structural gene mutation(183). The simplest model for penicillinase regu-lation involves a repressor protein binding to anoperator site to regulate transcription of thestructural gene. The putative operator site hasbeen identified as a 22-base pair (bp) invertedrepeat sequence which overlaps the structuralgene promoter (254).Another factor in the regulation of penicillin-

ase in S. aureus is a chromosomal locus whichappears to affect the level of penicillinase syn-thesis (180, 181). Mutations in this gene caused ameso-inducible phenotype (180, 181). It is possi-ble that the host specifies an antirepressor whichcomplexes the plasmid repressor in the presenceof inducer (181).

Resistance to Chloramphenicol

Chloramphenicol (Cm) diffuses passively intothe cytoplasm of bacterial cells, where it inter-acts with the SOS ribosome subunit to inhibit thepeptidyltransferase step in translation (130).Resistance to Cm is usually due to inactivationof the drug by acetylation (335, 336, 374). Theenzyme responsible is Cm acetyltransferase(CAT), which is usually plasmid encoded inclinical isolates. The properties of CAT haverecently been comprehensively reviewed (337).It should be noted that plasmid-determined Cmresistance in gram-negative bacteria is occasion-ally caused by another mechanism which doesnot involve inactivation (99, 128, 271).CATs in gram-negative bacteria. CATs have

been found in many different bacterial genera(337). The enzymes are tetramers of identical

367VOL. 47, 1983


http://mm

br.asm.org/

Dow

nloaded from


368 FOSTER

subunits with apparent molecular weights in therange of 23,000 to 25,000. The plasmid-encodedenzymes in gram-negative bacteria are ex-pressed constitutively, whereas those of gram-positive organisms are inducible. The CAT en-zyme variants have been classified on the basisof electrophoretic mobility, kinetics, inhibitorsusceptibility, and reactivity to antisera (114,337).There are three types of plasmid-encoded

CATs in enteric bacteria (129). The ubiquitoustype I enzyme has been studied in most detail.Its primary amino acid sequence has been com-pared with the DNA sequence of the CAT genes(6, 244, 342). The type I CAT has the unusualproperty of binding strongly to fusidic acid (406)and rosaniline dyes (302). The partial amino acidsequences of the type II and III enzymes arealso known (275, 440). The chromosomal en-zyme of Proteus mirabilis is related to the type Ivariant (440). CAT has also been found in non-enteric species of gram-negative bacteria. Theplasmid-specified enzyme of H. influenzae isrelated to type II CAT(314), as is the enzyme ofBacteroides fragilis (30, 339), whereas the Bac-teroides ochraceus enzyme is similar to type I(337). Chromosomally determined enzymeshave also been described in Agrobacterium sp.(an inducible enzyme; 143, 440), Flavobacte-rium sp. (440), and Myxococcus sp. (440).CATs in gram-positive bacteria. Five CAT

variants have been observed in staphylococci.These include the prototype A to D enzymes(93, 324, 336-338) and the newly described CATof pC194 (174). The amino acid sequence of thepC194 CAT was inferred from the DNA se-quence (174). Amino-terminal sequences of theA, B, and D variants and a more extendedpartial sequence of the type C enzyme have alsobeen determined (115).The Streptococcus agalactiae (440), S. faeca-

lis (74), and S. pneumoniae (86) CATs are relat-ed to staphylococcal enzymes. The S. pneumo-niae determinant is chromosomal, whereas theothers are plasmid located. Clostridium perfrin-gens (440), Bacillus pumilus (201), and Strepto-myces sp. (341) also specify CAT activity.Mechanism of Cm inactivation. Early studies

established that Cm inactivation resulted fromthe formation of 3-acetoxy-Cm, followed by theslower appearance of 1,3-diacetoxy-Cm (Fig. 2)(335, 374). The 3-acetoxy group is transferrednon-enzymatically by an intramolecular rear-rangement to yield 1-acetoxy-Cm, which in turnis acetylated to give the diacetoxy form (Fig. 2).All of the acetoxy derivatives of Cm are inactiveas antibiotics because they cannot bind to theribosome (343). Thus, resistant cells need onlyto acetylate Cm once to inactivate it, but are

obliged to divert a second molecule of acetylcoenzyme A (CoA) into another acetylation re-action. This may not present a problem to wild-type cells growing in low concentrations of drug.However, the lack of unlimited supplies of ace-tyl CoA is likely to be responsible for thenonlinear relationship between the number ofenzyme molecules (generated by amplifying thecat gene) and the Cm resistance level (277).

Catalytic properties of CAT. The substratespecificity of the type I and C CAT variants andpeptidyltransferase (the ribosomal target forCm) have been compared (337). All share theabsolute specificity for the D,threo isomer ofCmand a requirement for a substitution of the C-2amino group and for absence of substitution ofthe C-1 and C-3 protons. CAT is more tolerantthan peptidyltransferase of C-2 amino groupsubstitutions or para-substitutions of the phenylgroup. Fluoro substitutions of the 3-hydroxylgroup of Cm results in a drug with good antimi-crobial activity which cannot be inactivated byCAT (270). Indeed, fluoro-Cm is a competitiveinhibitor of CAT, forming a nonproductive com-plex with the enzyme and the acetyl donor.The critical first acetylation step couples the

breaking of a high-energy thiol ester bond withformation of an O-acyl derivative of Cm. Cir-cumstantial evidence originally suggested a di-rect role for a thiol group in the active center ofgram-negative CAT variants, but not for those ofgram-positive bacteria (129, 439, 440). The sen-sitivity to thiol inhibitors of the enteric CATenzymes is now thought to be due to a cysteineresidue located near the active site, which doesnot participate directly in the catalytic reaction(337). Thus, kinetic studies of the type I and Cenzymes favor the formation of a ternary com-plex whereby CAT binds both acetyl CoA andCm together at the active site rather than asequential mechanism involving an acyl enzymeintermediate with an active-site cysteine. Thethree enteric CAT variants have reactive cyste-ine residues at an equivalent position, whereasthe staphylococcal enzymes do not and are thusresistant to thiol-directed inhibitors (276, 277).A detailed model of the active center of CAT

has been proposed and the catalytic mechanismhas been discussed (337). Of central importanceis a highly reactive histidine residue (His-193 oftype I CAT) whose reactivity to alkylating andmodifying reagents is protected by Cm. Thishistidine is present at an equivalent position inall CAT variants studied (337).Type I CAT confers resistance to fusidic acid.

Gram-negative bacteria are intrinsically resis-tant to fusidic acid because this steroid antibioticcannot penetrate the outer membrane. Fusidicacid resistance specified by R plasmids of enter-

MICROBIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from



CHCI2

COI

H NH

02N O1'-C H2OH H OH

rHinRAMPHFNICnL

CHC[2CO

H NH

CAT >02N c-JC-CHZOH H O- CO-CH3

i -ACEITXLC

CHCL2 CHC12Co COI I

H NH H NH

ONcC'-C'1-is I- CAT 41. It.

~~~0 H OH0 H O- CO-CH3 pIHOO-CH3 COCH3

13-DIACETOXY CM -ACFTOXY CM,

FIG. 2. Inactivation of Cm by CAT.

ic bacteria can only be detected in strains withalterations in outer membrane structure whichcan be correlated with increased sensitivity tofusidic acid. Many R plasmids have been shownto promote resistance to fusidic acid in suchstrains (89). Genetic analysis of the r-determi-nant region of R100 showed that cat and fuswere closely linked (219, 262, 263) but appearedto be separable (96). Recently, however, cat andfus have been shown to be the same gene by theisolation of amber mutations affecting both catand fus simultaneously (406). DNA sequenceanalysis showed that one amber mutation (atGln-38) preceded the only possible internaltranslational start (at Met-77), thus eliminatingthe possibility that the fus product could betranslated from an initiation codon internal tothe cat message.

Resistance to fusidic acid is specified only bytype I CAT. The enzyme protein must mediateresistance to the antibiotic despite the lack ofstructural similarity between fusidic acid andCm and the fact that fusidic acid is not inactivat-ed or modified in any way (337). Thus, CATbinds fusidic acid very strongly to its activecenter and fusidic acid acts as a competitiveinhibitor of Cm acetylation (337). This may berelated to the finding that rosaniline dyes such as

crystal violet and carbol fuchsin also bindstrongly to CAT and are competitive inhibitorsof acetylation (302). Type I CAT seems to have abinding "pocket" or domain which readily ac-commodates hydrophobic and, in particular, ar-omatic groups of compounds which are notrelated structurally.The high concentrations ofCAT protein in the

cytoplasm combined with the strong affinity offusidic acid for the protein and its slow diffusioninto the gram-negative cell makes a "sponge" or"buffer" mechanism seem plausible, wherebyan inhibitor is rapidly sequestered by a bindingprotein before it can reach its target. A similarmechanism has been postulated for CadB-medi-ated cadmium resistance (295).

Regulation of CAT synthesis in E. coli. It hasbeen known for some time that type I CATsynthesis in E. coli is subject to catabolite re-pression. Constitutive synthesis of the enzymeoccurred at a higher rate in glycerol-grown cul-tures than in those grown in glucose (156).Mutants lacking adenylate cyclase or cataboliteactivator protein (CAP) failed to support high-level CAT synthesis, and cells lacking CAPcould not be stimulated to do so by the additionof cyclic AMP (94). Furthermore, in vitro cou-pled transcription-translation of the cat gene

369VOL. 47, 1983


http://mm

br.asm.org/

Dow

nloaded from


MICROBIOL. REV.

demonstrated the requirement for both cyclicAMP and CAP for maximum CAT synthesis(94).DNA sequence analysis of the type I CAT

gene and DNase protection experiments identi-fied two CAP sites in the promoter-proximalregion of the gene (227). One site is located 130bp upstream from the start point of transcriptionbut is not required for the stimulation of tran-scription. The CAP site which overlaps thepromoter is involved in regulating the transcrip-tion of cat. The cyclic AMP-CAP complex mayinteract with RNA polymerase and also changethe structure of the promoter.

Regulation of CAT synthesis in S. aureus.CATs are encoded by small multicopy plasmidsin S. aureus. CAT is expressed at a high levelonly after induction with Cm. Such inductionexperiments are normally complicated by thefact that the inducer is inactivated by the newlysynthesized CAT (acetoxy-Cm cannot act as aninducer) and also because the inducer is aninhibitor of the protein synthesis required forinduction to occur (340, 425). These problemscan be circumvented by using 3-deoxy-Cm,which is a gratuitous inducer (425). Apart fromthe absence of a requirement for a 3-OH group,the structural requirements for the inducer par-allel those of Cm as an antibiotic and CATsubstrate.The mechanism of regulation of CAT expres-

sion in S. aureus is poorly understood. Thefailure of repeated attempts to generate Cmrconstitutive mutants (337) would argue against asimple negative (repressor-controlled) regula-tory mechanism. Also, it is known that inductionfails to occur in the presence of inhibitors oftranscription, an observation which diminishesthe likelihood of a translational attenuationmechanism as has been proposed for the induc-tion of staphylococcal macrolide-lincosamide-streptogramin B (MLS) resistance (171, 345).Cloning and DNA sequencing studies have

provided some insight into the regulatory proc-ess. Staphylococcal CAT determinants expressCmr inducibly when cloned in an E. coli vectorplasmid (174, 424), although the 5-fold level ofCAT induction was poor compared with the 100-fold induction seen in the native host. A smallfragment of pC194, where the only open transla-tional reading frame was for CAT protein, stillexpressed inducible Cmr in E. coli. This suggest-ed that no plasmid-encoded regulatory protein(apart from CAT) could be involved in S. aureusCAT induction in E. coli. This and the presenceof a 37-bp inverted repeat sequence between thelikely start of transcription and the translationalinitiation codon for CAT led to the suggestionthat cat gene transcription is controlled autoge-

nously by the CAT protein. Two simple autoge-nous regulation models are possible: (i) negativeregulation, where CAT protein acts as a repres-sor of cat gene transcription; and (ii) positiveregulation, where CAT acts as an inducer in thepresence of Cm. Either model would need sub-stantiation by showing that CAT protein has invitro DNA binding activity and can alter the rateof labeled CAT synthesis in a coupled in vitrosystem. So far, attempts to demonstrate in vitroregulation by pure CAT protein in an E. coli-coupled system have failed (337). Evidence thatCAT is a negative autoregulator could comefrom the isolation of Cms point mutants in whichthe altered CAT protein was synthesized consti-tutively.

Resistance to Aminoglycoside-AminocyclitolAntibiotics

The inhibitory effect of aminoglycosides onsensitive bacterial cells is due to the binding ofthe drug to ribosomes, thus interfering withtranslation (130). Defining the lethal event hasbeen confused by the pleiotropic effects thatoccur after sensitive cells are exposed to thedrug and by controversy about effects of amino-glycosides on translation (for reviews, see refer-ences 130, 151, 152).

Mutations causing resistance to aminoglyco-sides either reduce drug binding to the ribosomeor impair transport across the cytoplasmic mem-brane. Single-step mutations in the rpsL gene ofE. coli K-12 cause high-level resistance to strep-tomycin by altering the structure of ribosomalprotein S12 and preventing the drug from bind-ing to its target (130). Ribosomal resistance tostreptomycin is found in some clinical isolates ofN. gonorrhoeae (241), enterococci (442), S. aur-eus (213), and P. aeruginosa (3%). Mutationswhich cause decreased accumulation of amino-glycosides due to impaired transport across themembrane, resulting from a defect in membraneenergization, can be selected in the laboratory(1, 38, 267) and also occur in clinical isolates (34,36). In addition, obligate and facultative anaer-obes growing anaerobically are intrinsically re-sistant to aminoglycosides because they cannotaccumulate the drug (35, 44).

High-level ribosomal resistance to deoxy-streptamine-containing aminoglycosides (see be-low) cannot be selected in a single step (130),probably because these drugs have several bind-ing sites on ribosomes. Low-level resistance dueto changes in ribosomal proteins have beenreported (40, 388, 432). Sometimes resistantstrains have more than one mutation, each onecontributing a small increment in resistance (1,

370 FOSTER


http://mm

br.asm.org/

Dow

nloaded from



40, 388). Thus, gentamicin resistance can be dueto a mutation in rplF affecting ribosomal proteinL6 (40), which alters drug binding to ribosomes,in combination with an unc mutation whichimpairs transport across the membrane (1).High-level resistance to the aminocyclitol spec-tinomycin can be selected (rpsE; 106, 428),whereas hygromycin B resistance is caused byimpaired uptake (1). Resistance to kasugamycinmay be by either mechanism (106).Most clinically significant resistance to amino-

glycosides is caused by R-plasmid-specifiedphosphotransferase, acetyltransferase, and ade-nyltransferase enzymes (78, 306, 333). The mod-ified antibiotics no longer bind to ribosomes andcannot inhibit protein synthesis (433). However,unlike other drug resistance mechanisms in-volving enzymatic inactivation, only a smallproportion of the drug in the medium is actuallydestroyed (93). Furthermore, less drug is accu-mulated in resistant cells (97, 169).

Transport of aminoglycosides in sensitive andresistant bacteria. To reach the ribosomes, ami-noglycoside antibiotics must penetrate the cellsurface layers. Uptake of aminoglycosides insensitive strains of both gram-positive and gram-negative bacteria occurs in three stages (36-38,97, 168).The first phase of uptake is rapid energy-

independent (EIP) binding to the cell surface. Ingram-negative bacteria this includes passive dif-fusion through outer membrane pores. Positive-ly charged aminoglycosides can readily passthrough these pores despite being larger than thesize exclusion limit for uncharged molecules(272).The second stage of uptake is the slow energy-

dependent phase I (EDPI), which representsinitial passage of the drug across the cytoplasmicmembrane. It requires the membrane to be ener-gized sufficiently and uses the At (electric po-tential) component of proton motive force (85,246, 267). Indeed, there is a direct relationshipbetween the magnitude of At! and the rate ofuptake and killing by aminoglycosides (85, 246).The cells remain viable during EDPI and poly-peptide synthesis continues (1, 267), althoughmisreading may occur. The duration of EDPIdepends on the concentration of drug in themedium and the susceptibility of the ribosomes(1). The nature of the carrier in the membrane isnot known; different authors have proposedrespiratory quinones (33), cytochromes (44), andthe combined use of several different carbohy-drate transport systems (85).The third phase of aminoglycoside transport is

energy-dependent phase II (EDPII). It is initiat-ed when sufficient drug is present in the cyto-plasm to bind to all of the ribosome particles. It

signals the onset of lethality and the cessation ofprotein synthesis. Its initiation requires thatsusceptible ribosomes in the cell be activelyengaged in protein synthesis (1, 177). The sug-gestion that EDPII results from induction of acognate polyamine transport system (168) is nowdiscounted (1, 44, 85). Indeed, EDPII may notactually be due to energy-dependent transportper se, but could result from an increase in"runoff" ribosomes from polysomes to provideunimpeded binding sites (177) or from a pleiotro-pic alteration to membrane permeability (1).

Cells resistant to aminoglycosides because ofR-plasmid-specified modifying enzymes exhibitonly the EIP and EDPI phases of transport (97,169). This is because any drug that manages topenetrate the cytoplasm during EDPI is inacti-vated and fails to induce EDPII because itcannot bind to ribosomes and inhibit proteinsynthesis (433). However, if the rate of drugtransport during EDPI exceeds the rate of druginactivation, then significant amounts of unmod-ified drug will penetrate the cytoplasm and bindto ribosomes. Thus, some strains which arephenotypically susceptible to certain drugs ex-hibit aminoglycoside-modifying activity whenassayed in vitro (25, 144, 403). Also, competitiveinhibitors of enzyme activity such as the amino-glycoside adenyltransferase inhibitor 7-hydroxy-tropolone may decrease the rate of inactivationsufficiently to cause a significant decrease inresistance level (4).

Location of amoglycoside-modifying enzymesin the bacterial cell. There has been some confu-sion about the cellular location of aminoglyco-side-modifying enzymes. Originally it was sug-gested that these enzymes were located in theperiplasmic space of gram-negative bacteria be-cause significant amounts were released by os-motic shock (91, 137, 138, 422). However, it isdifficult to imagine how enzymes that require thehigh-energy cofactors ATP or acetyl CoA couldbe active in the periplasm. Recent experimentswith an aminoglycoside 3'-phosphotransferasein E. coli showed that the enzyme is located inthe cytoplasm (290). It would be attractive tothink that modifying enzymes were attached tothe cytoplasmic membrane where they would bestrategically placed to inactivate incoming drugmolecules before they could reach the ribo-somes. However, there is no direct evidence forthese enzymes being integral membrane proteinssince purified membrane fractions from resistantbacteria have little activity. Any membrane at-tachment of enzymes must be loose and easilydisrupted.

It is now considered that osmotic shock treat-ment may result in release of some cytoplasmicenzymes (290). To establish that an enzyme is

371VOL. 47, 1983


http://mm

br.asm.org/

Dow

nloaded from


372 FOSTER

\k 6'Cl

DDI() i -HA

AAC(OAPH5 -

Ri R2 R3

NH2 H H2NH2 H C2NH2OH H CH2NH2

NH2 C<HOHCH2CH2JH2 -

OH H I

R4CH2Nil2

HH

- (No ring IV)

CHi2NH2 (No 3' OH)

FIG. 3. Structure and modification of neomycin antibiotics. The structure of the neomycin family ofantibiotics is shown. The structure of variants can be deduced from the table under the structure. The sites ofphosphorylation (APH), acetylation (AAC), and adenylylation (AAD) are indicated by arrows.

periplasmic, it is important to test for releasefrom lysozyme-EDTA-generated spheroplastsand from periplasmic-leaky mutants as well asfor release by osmotic shock.

Classfication of anogycoside-aminocyclitolantibiotics and their modifying enzymes. Amino-glycoside-aminocyclitol antibiotics are classifiedinto two major groups, those containing strepti-dine (e.g., streptomycin) and those containing 2-deoxystreptamine. The 2-deoxystreptamineantibiotics are subdivided on the basis of substi-tutions to the 2-deoxystreptamine ring: (i) 4,5substitutions as in neomycin, paromomycin, bu-tirosin, and lividomycin (Fig. 3); (ii) 4,6 substitu-tions as in kanamycins, gentamicins, tobramy-cin, and amikacin (Fig. 4 and 5). Other relateddrugs which do not fit in these groups are theaminocyclitols spectinomycin and hygromycinB, as well as kasugamycin. There are three typesof aminoglycoside-modifying enzymes: amino-glycoside-O-phosphotransferase (APH), amino-glycoside-N-acetyltransferase (AAC), and ami-noglycoside-O-nucleotidyltransferase (AAD).The reactions that they catalyze are shown inFig. 6. The phosphorylation and nucleotidyla-tion reactions use ATP as a cofactor, whereas

acetyltransferases use acetyl CoA. The site ofmodification of the drug is indicated by thenumber in parentheses. Thus, AAC(6') acety-lates the 6'-amino group on aminohexose I ofsusceptible drugs. The modification sites aresummarized in Fig. 3, 4, 5, and 7, and in Tables2, 3, and 4. It may be possible to infer the type ofenzyme present in cells from the pattern ofresistance to growth inhibition by a standard setof antibiotics (266). This could be a convenientmethod for screening strains but is no substitutefor biochemical analysis, particularly becausesome strains express more than one aminoglyco-side-modifying activity.Some enzymes have been divided into sub-

types, primarily on the basis of differences insubstrate profile (Tables 2, 3, and 4). However,this has not always been done consistently, andthere seems to have been little attempt to com-pare directly enzymes isolated in different labo-ratories. It should be emphasized that beingassigned to the same subgroup does not neces-sarily imply evolutionary relatedness. In mostcases immunological cross-reaction and DNAhybridization data are lacking. Also, it is possi-ble that some of the subgroups listed in Tables 2,

';F.OMYCIN B

NEOMYCIN CPAROMOMYC INBUTIROSIN B

LIVIDOMYCIN

MICROBIOL. REV.

AM


http://mm

br.asm.org/

Dow

nloaded from



KANAMYCINS

Ri R2 R3

KANAMYCIN A NH2 OH H

KANAMYCIN B NH2 NH2 H

KANAMYCIN C OH NH2 H

AMIKACIN NH2 OH CQCHOHCHS2H2HH2TOBRA¶MYCIN NiH2 NH2 H (3 OH MISSING)

FIG. 4. Structure and modification of kanamycin antibiotics. The structure of the kanamycin family ofantibiotics is shown. The structure of variants can be deduced from the table under the structure. The sites ofphosphorylation (APH), acetylation (AAC), and adenylylation (AAD) are indicated by arrows.

3, and 4 [e.g., some AAC(6') enzymes] are notsufficiently different in substrate profile to war-rant separate classification. Other possiblesources of confusion are the presence of morethan one enzyme in the same strain and the factsthat the same enzyme may not confer the samephenotype in different hosts and modification ofthe drug in vitro does not imply that the enzymewill confer resistance in vivo. Perhaps enzymesshould only be reported and classified afterkinetic tests have been done with purified pro-teins.

Resistance to Mercuric Ions andOrganomercurials

Sensitivity and resistance to mercurial com-pounds. Mercuric ions are toxic to bacteriabecause they bind avidly to sulfhydryl groupsand inhibit macromolecule synthesis and en-

zyme action. Many enzymes have critical thiolgroups and are sensitive to Hg2' in vitro. Tran-scription and translation are particularly sensi-tive. This may be due to inhibition of precursor

synthesis or Hg2+ binding to polynucleotides(84). Also, RNase may be activated by Hg2+(22).

Resistance to mercurials (Hg') is a commonplasmid-determined property of both gram-posi-tive and gram-negative bacteria (64, 331, 369,417, 418). This may have been caused by the use(until recently) of mercurials such as phenylmer-cury and thimerosal as hospital disinfectants(148) or by industrial and urban pollution. Adecreasing incidence of mercurial resistance inhospital strains has coincided with the discontin-uation of mercurial disinfectant usage (300).Also, Hgr is frequently specified by drug resist-ance plasmids (331) and is also common in soilpseudomonads and bacilli. It has recently beenfound in Thiobacillus ferrooxidans (286).

Bacterial resistance to Hg2e is determined byenzymatic reduction of the ion to Hgo, which ismuch less toxic (368, 369). Also, Hg° is virtuallyinsoluble in water and evaporates because of itshigh vapor pressure. The enzyme that catalyzesthe reduction of Hg2+ is the intracellular, cyto-plasmic, FAD-containing mercuric reductase

VOL. 47, 1983 373


http://mm

br.asm.org/

Dow

nloaded from


374 FOSTER

AAC(6')7/~

-AAC(1

CH-HH3C-HN

GENTAM IC INS

GENTAMICINGENTAMICINGENTAMIC INSISOMICINNETILMICIN

C1aC1C2

Rl R2 R3 R4 R5H NH2 H H HCH3 NHCH3 H H HCH3 NH2 H H fIf NH2 - H II

If N H 2 - 2 5

FIG. 5. Structure and modification of gentamicin antibiotics. The structure of the gentamicin family ofantibiotics is shown. The structure of variants can be deduced from the table under the structure. The sites ofphosphorylation (APH), acetylation (AAC), and adenylylation (AAD) are indicated by arrows.

0

AMINOGLYCOSIDE-OH + ATP APH AMINOGLYCOSIDE-0-P-OH + ADPOH

AMINOGLYCOSIDE-OH + ATP 3.-*AA AMINOGLYCOSIDE-O-ADENO3INE + PPi

AMINOGLYCOSIDE-NH + ACETYL AAC -- AMINOGLYCOSIDE-NH-CO-CH3 + COA-SH2 COA

FIG. 6. Biochemical mechanisms of aminoglycoside modification. The mechanisms of modification ofaminoglycoside antibiotics by phosphotransferases (APH), adenyltransferases (AAD), and acetyltransferases(AAC) are summarized.

MICROBIOL. REV.

00

Art,?,iI


http://mm

br.asm.org/

Dow

nloaded from



SPECTINOMYCINSTREPTnMyclYCINNH AP

H2N-C-NH /NH HO7 OH

H2N-C-HN OH

H3COH 0

HO-H2CHO NH-CH3,fOH

AADT)

21 AAD(9)HO NHCH3 /

OH

H3C-HNC

HO OH

CH3

FIG. 7. Structure and modification of streptomycin and spectinomycin. The sites of modification ofstreptomycin and spectinomycin by adenyltransferases (AAD) and phosphotransferases (APH) are indicated byarrows.

NADPH NADP

NAtP NP

H LYASE > O1+HN2 Hge

FIG. 8. Mechanism of resistance to mercuric ions.The upper part shows in diagrammatic form twopossible mechanisms for detoxification of mercuricions. The parallel lines represent the cytoplasmicmembrane. The diagram on the left indicates that thereductase protein (R) interacts directly with the mem-brane-bound transport protein (T). On the right it issuggested that another mer operon-specified protein(possibly the merC product; C) is required to bind theincoming Hg in the form of an adduct and transport itto cytoplasmic reductase (R) molecules. The lowerdiagram shows the reactions involved in detoxifyingthe organomercurial phenyl mercury. The C-Hgbond of phenyl mercury is cleaved by organomercuriallyase to form benzene and Hg2+. The Hg2+ is subse-quently converted to Hge by reductase.

(330, 369). The resistance mechanism also in-volves a plasmid-specified Hg-specific transportsystem (274). It seems to be required to directHg2+ through the cytoplasmic membrane, whereit would otherwise encounter sensitive enzymes.It is attractive to think that the reductase andtransport functions might interact physically. In-deed, some reductase protein appears to bemembrane associated in E. coli minicells (187),although this was not found in experiments withwhole-cell envelopes (330, 368). The mechanismof resistance is presented diagrammatically inFig. 8.

Strains which detoxify organomercurialsspecify a second cytoplasmic enzyme, organo-mercurial lyase (330 383). This cleaves C-Hgbonds to release Hgi+, which is in turn volatil-ized by the reductase (Fig. 8).Spectrum of resistance to organomercurials in

gram-positive and gram-negative bacteria. Twodifferent mercurial resistance phenotypes havebeen found in gram-negative bacteria (331, 418).Narrow-spectrum resistance is determined bystrains that express reductase alone, whereasthose strains that specify both reductase andlyase have the broad-spectrum resistance phe-notype. However, the phenotypes of entericbacteria and P. aeruginosa differ even when thebacteria harbor the same plasmid (64, 418). Anadditional complication is that resistance tosome organomercurials does not involve enzy-matic degradation. All staphylococci resistant to

VOL. 47, 1983


http://mm

br.asm.org/

Dow

nloaded from



rA rA~~~ (A r -

Q.o *1)r .SA2~

C.~~~~~~~~4.42C ' s rA~'0'

8 .~~~2~~s~~R~~( ro .2 i~A '

f ~~~~~~~~~~~0 0a2U ccZ '

In 4)~~~~~~~V

0 C)

C) .- :i2W.2:5~~~~~ C~ C

0 ~~ ~~~~~~~0 0 0

4)03.- '~~~~~0.o'~r 4)O OjA 0~~~~U~~~~~~~~ < z0 '

0 .2 .2kS.0 S04) .0~< ~ .4) C t~U~.o en~

'S~~~~~~~~~~~~~~~(0 0E~~~~~~~~C co_

0~~~~~~~~~~~~~~00~~~~~~~~~~~~'

In~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~C

0-4 ~ 0

00 0 ~~~~~~~~~~~~~

4)~~ ~ ) ~ n

a In4 &0i .2 .01


http://mm

br.asm.org/

Dow

nloaded from



'O..I 0I-Sr-

04)~~~~~~~~~~~~~~~~~4

7 o 2)8 s .

Gcd~~~ ~ ~ ~ ~~~~~'

E_

D < <

E~7,5 7 J

CU

to0cdCdE

0~~~~~C

4)C U~~~~~ .0 0.~~~~~ 0. 0

JD0- ~

C C

Cl) CO)

0dC.)~~~~~ ~~ ~ C)~~~L ")C-

0~ 0 A

0 0~~~~~~~~~~~~~~~~~.0.4-

~~~~~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ -

0~~~~~~~~~~~~~~~>b)

(~~~~~24 0 . 4) 4)

.~~~~b 4

4)~~~~~~~~~~~~0 4.

004)'a c

~~~ ~ ~~~~~~~0r4)O~ ~ ~ 02 02

44 z 41a.

~~ ~ ~ d

VOL. 47, 1983


http://mm

br.asm.org/

Dow

nloaded from


378 FOSTER

u^o4)002.4).4)4)

4)C

E0U

MICROBIOL. REV.

2e

4.)Z

5 .5 8 2

= Qw

W0 *-..

.0.~

>.0,4 0 0

00

0O 44.' *; . 2

.E

C)2o-U0)..5, Ecw. ,,a ^

.

.0 *_. (, 4)-,. _ 5

4.)

Cl I

.0

>.ED ,E =

E oEE EoCd

,@ b oC Cd

:.r

CD- °JDX

I

&

'4=ow E r E

0

00

m4o

(D

I--,4-

02~~~~~r

0

lb

_

02 4)

Cd (A

* -; Cd

4)4_.4

04 0-

sL, .5m

E *. :6>

4)02

0..

0.

04.0

0._

02

ow;A

4)mN

02

0

10

w(

5-

Z:

4)

0

.0

.0

,.OD!CT)

r-

.2

c)

a0

4)o

.2

.0

.0

ce

s5200.

4)2

roo

C:

2.=0-.

5-

0 .)

C30

24>4(5ow

S * .-

to 4pt tu3~3 z0 2)~in

Z.0.4)

5-

OS-

4)

.Q0

>No E


http://mm

br.asm.org/

Dow

nloaded from



mercurials express a broad-spectrum phenotypewhich differs from that of gram-negative organ-isms. These phenotypes are summarized below.

(i) Narrow-spectrum resistance in enteric bac-teria. Enteric bacterial strains volatilize Hg2'but also confer resistance to tnerbromin andfluorescein mercury acetate without hydrolysisor reduction of mercury (331, 418). A permeabil-ity barrier has been postulated (369), but thegene responsible in the mer operon has not beenidentified.

(ii) Broad-spectrum resistance in enteric bac-teria. Broad-spectrum resistance is restricted toa small number of plasmids from incompatibilitygroups A to C, L, and H2. The narrow-spectrumphenotype is extended to include phenylmer-cury and thimerosal (331, 418). Although Hgo isformed by volatilization from thimerosal, p-hydroxymercuribenzoate (pHMB), methyl mer-cury, and ethyl mercury, resistance is only con-ferred to thimerosal. This is attributed tolow-level lyase activity and the greater toxicityof ethyl and methyl mercury. The IncA-C deter-minants specify very low lyase activities and arealso sensitive to thimerosal.

(iii) Narrow-spectrum resistance in P. aeru-ginosa. Narrow-spectrum resistance in P. aeru-ginosa differs from that in E. coli in thatresistance is conferred to pHMB without volatil-ization of Hgo (64). A host factor is probablyresponsible because broad-host-range Hgr plas-mids do not confer resistance to pHMB in E. coli(64, 418).

(iv) Broad-spectrum resistance in P. aerugi-nosa. As with the narrow-spectrum P. aerugi-nosa determinants described above, resistanceto pHMB occurs without hydrolysis or volatil-ization, whereas E. coli plasmids slowly volatil-ize Hg0 from pHMB yet are sensitive to thecompound (418). These plasmids specify resist-ance (with volatilization) to phenyl, methyl, andethyl mercury (418).

(v) Resistance in S. aureus. All S. aureus merplasmids determine broad-spectrum resistance.Two phenotypic groups have been described.The first differs from the gram-negative broad-spectrum phenotype by not involving resistanceto merbromin. Resistance topHMB and fluores-cein mercury acetate is conferred without vola-tilization (417). Furthermore, these strains aresensitive to thimerosal, although slow hydroly-sis and volatilization were detected. The secondclass of determinants expresses resistance tothimerosal, merbromin, and methyl mercury.Thimerosal resistance was attributed to thiscompound being a better inducer of the meroperon as well as to the lyase having greaterhydrolytic activity (300). Also, that methylmer-cury was hydrolyzed suggests that a broader-spectrum lyase was involved.

VOL. 47, 1983

o NT-4 e)

>0

>t->

c -

*D C 04)^°C_-_4e

> ~2)s_ ;U

0 0.

U0

Cw 04) C* .E

C) C

c)0)C 1

.0

0.0

C-,

4-0

0-403 _

0

'00&.X~EC::Y *

c-

04

04

4)

4-

0u7

4-

0.00.

.0

04

0 r.

o0 =

Cd o.o-

0~-_

24)i2

cn

a-E.= >0.0U~C.2 V

'0

10U

24)2I.'-w2 4)>0 E!

L-

C.)

'4-,0o0

2be

X 00 rl..

>. C's:x w

1-

04

4.)

~o00 I-

200cE=


http://mm

br.asm.org/

Dow

nloaded from


380 FOSTER

Mercuric reductase and mechanism of reduc-tion of mercuric ions. Mercuric reductase is asoluble cytoplasmic protein (330, 369). The en-zymes specified by plasmid R100 and transpo-son TnSOI in E. coli are composed of subunits ofabout 63,000 molecular weight. Each subunithas a bound FAD moiety. The subunit structureof the enzyme from different sources seems tovary remarkably. The soil Pseudomonas sp.strain K62 enzyme is apparently a monomer of67,000 daltons (127). Two distinct types of re-ductase enzyme encoded by plasmids in gram-negative bacteria have been recognized by im-munological studies and by differences in chargeand molecular weight (121, 330). The prototypeclass I enzyme specified by Tn51 is a dimer of125,000 daltons, whereas the type II (R100)enzyme appears to be a trimer of 180,000 dal-tons. Despite these differences in subunit struc-ture, the type I and II enzymes are closelyrelated to each other in terms of amino acidcomposition, immunological cross-reaction (T.Kinscherf and S. Silver, personal communica-tion), and DNA sequence homology between themer genes (S. Silver and N. L. Brown, personalcommunications).The reductase encoded by TnSO1 has been

studied in greatest detail at the biochemical level(121). It has a redox-active cysteine at the activesite which acts as an electron acceptor. Hg2+ ispresented to the enzyme as a dithiol adduct RS-Hg-SR (121). The possible biochemical mecha-nisms involved in the reduction of mercuric ionshave been discussed before (121).

Detoxification of organomercurial compounds.The relationship between hydrolysis of organo-mercurials and resistance phenotypes in differ-ent organisms has been discussed above. Broad-spectrum organomercurial-resistant bacteria

REaLATOR I

p-X__~

contain the enzyme organomercurial lyase (330,383), which cleaves the C-Hg bonds of phenyl,ethyl, and methyl mercury and in some casesthiomerosol (Fig. 8). Two distinct lyases wereseparated from the soil Pseudomonas sp. strainK62 (383). There is also a possibility that R831specifies two different, but so far inseparable,activities in E. coli (330). Genetic studies arerequired to determine whether one or two struc-tural genes are responsible for the two activities.

Evidence for an Hg2 -specific transport systemand its role in resistance. The finding that muta-tions which inactivated mercuric reductasecaused hypersensitivity to Hg2+ was the key torevealing the Hg2+-specific transport system(119, 274). Hypersensitive strains are sevenfoldmore sensitive to Hg2+ than plasmid-free cellsand bind 203Hg more rapidly (274).The Hg2' transport system has been difficult

to characterize. It does not involve titratablethiol groups (274). It has been suggested thatenhanced uptake of Hg2+ is due to facilitateddiffusion, with Hg2+ passing more rapidly intocells and saturating the available binding sitesmore quickly (367).The gene which specifies the transport func-

tion (merT) is the closest structural gene to themer operon promoter (Fig. 9). Attempts havebeen made to assign one of the polypeptidesexpressed in E. coli minicells to merT. Thus, the15K and 14K Hg2+-inducible polfpeptidesbound to the cytoplasmic membrane fractioncould be candidates for the merT protein. TheTnSOI DNA sequence suggests that the merTproduct is a 12.5K highly hydrophobic proteinwith two cysteine residues (N. L. Brown, R. D.Pridmore, and D. C. Fritzinger, personal com-munication). To assign polypeptides unequivo-cally to genes, it will be necessary to examine

REDUCTASE58,6K 57.4K

i- a I I=__

mert

FIG. 9. Structure of the mer operon. Data obtained from genetic analysis of the R100 mer region and fromDNA sequence analysis of the closely related TnSOI element are summarized. The putative mer genes merC andmerD correspond to the open translational reading frames urf-l and urf-2 revealed in the Tn5O1 sequence. Thedirection of transcription of the merR and the merTCA genes from promoters (P) are shown by arrows. Abovethese lines are the sizes of putative mer polypeptides deduced from the TnSO1 DNA sequence. The regulatoryprotein(s) specified by the merR region acts as both a positive and a negative regulator of the merTCA operon andalso autogenously regulates merR expression.

MICROBIOL. REV.

merR rT


http://mm

br.asm.org/

Dow

nloaded from



polypeptides expressed in in vivo systems bywild-type and mapped mutations.

At first glance it might seem strange that aresistance mechanism should involve a processthat specifically transports a toxic ion into thecell. However, it can be argued that this isnecessary to prevent the highly toxic Hg2+ frombinding to sensitive sulfhydryl groups as itpasses through the membrane. Thus, a strainharboring a merT mutant plasmid which stillexpresses a near-wild-type level of reductase isphenotypically sensitive to Hg2+, despite thefact that it can rapidly volatilize the mercuryfrom the medium (N. Ni Bhriain and T. J.Foster, unpublished data).

It is reasonable to think that reductase inter-acts with the membrane-bound transport protein(Fig. 8). This might be necessary to preventHg'+ from binding to sensitive cytoplasmic en-zymes once it has been transported through themembrane. However, there is no evidence forreductase being associated with the membranefraction of lysed cells, although it was suggestedthat the 66K processed form of reductase mightbe loosely associated with the minicell mem-brane (187), possibly by loose ionic binding.Alternatively, another mer operon protein couldbe involved in binding Hg in the cytoplasm andcarrying it as an adduct to the reductase. Theprotein which could be specified by urf-l TnSOJ(which is equivalent to merC in R100; see belowand Fig. 9) has two cysteine residues (Brown etal., personal communication) and could be acandidate for such a role.

Expression of resistance by multicopy plasmids.Comparison of the volatilizing activity of E. colistrains carrying different copy number mer plas-mids revealed a cryptic gene-dosage effect forreductase (273). Strains carrying the R100 mergenes cloned in the ColEl: :TnA vector RSF2124(16 copies per chromosome) expressed an eight-fold greater reductase activity in cell extractsthan did R100 (one to two copies per chromo-some), although there was no increase in resist-ance level or volatilizing activity of intact cells.It was suggested that crypticity was due to alimitation in Hg2+ transport sites in strains car-rying multicopy mer plasmids. The rate of Hg2etransport would be expected to be the sameirrespective of the number of copies of the merTgene and the amount of reductase within the cell.Our recent experiments have indicated that

there is a gene-dosage effect for the transportfunction in E. coli (NI Bhriain and Foster, un-published data). The R100 mer genes werecloned into pBR322, a plasmid with a highercopy number than that of ColEl. Reductase-deficient merA mutants determined greaterHg2' hypersensitivity than did merA mutations

on ColEl or R100. The pBR322 mer+ plasmidalso specified a higher level of reductase thanthe other plasmids. However, it did not deter-mine a normal wild-type Hgr phenotype. Theincomplete resistance to Hg2+ could be ex-plained by overproduction of the transport func-tion to such an extent that the cells cannot copewith the flood of incoming Hg2+, despite thesimultaneous induction of large amounts of re-ductase. A more extreme manifestation of thisimbalance was observed in cells carrying thesingle-copy wild-type mer+ plasmid R100 alongwith a multicopy merA mutant plasmid. Thesecells were more sensitive to Hg2+ than multi-copy mer+ plasmid-carrying cells or plasmid-free cells, despite the production after inductionof substantial reductase activity by the R100merA+ gene (Ni Bhriain and Foster, unpub-lished data).

Genetic analysis of the mer regon. One geneticstudy of mer analyzed transposon TnA inser-tions in plasmid R100 (119). Three genes wererecognized and mapped: the regulatory gene(merR), the transport function gene (merT), andthe reductase gene (merA). An operon structurewas proposed to explain the coordinate induc-tion of merT and merA, and the approximatephysical locations of the genes on the plasmidwere deduced (119). We have been engaged in amore detailed genetic analysis of mer and haveconfirmed the main features of the model (NfBhriain and Foster, unpublished data; Fig. 9).This can now be interpreted in relation to theDNA sequence of the mer region of TnSOJ, withthe slight caveat that the two systems specifydifferent types of reductase and probably differsomewhat at the DNA sequence level.The physical coordinates of the mer genes of

R100 were deduced by mapping a number ofdifferent transposon TnS insertions. This ap-proach allows the minimum and maximum sizesof the gene products to be estimated. The regula-tory gene merR is located between the IS1sequence derived from the r-determinant regionof R100 (Fig. 9) and the merT-merA genes. ThemerR region of TnSOI has two partially overlap-ping protein-coding sequences which couldspecify polypeptides of 9K and 7.5K. It is notknown whether the merR region of R100 has thesame organization or whether the two proteinsare expressed in vivo and are required for merRfunction.Transposon insertions in merR and merT both

cause a sensitive phenotype and low constitutiveexpression of reductase. They can be distin-guished by complementation tests. Located be-tween merT and merA is a stretch ofDNA whereTnS insertions cause a hypersensitive phenotypebut allow synthesis of low constitutive reduc-

VOL. 47, 1983 381


http://mm

br.asm.org/

Dow

nloaded from


382 FOSTER

tase. This region could correspond to a new mergene which we have called merC and an openreading frame (urf-l) of TnSO. The proteinproduct of urf-l would be 9.5K (Brown et al.,personal communication). However, this regionhas not been assigned a minicell-specified pro-tein or a function in resistance.

Unexpectedly, several Tn5 insertions in themulticopy mer plasmid which caused a sensitivephenotype mapped promoter-distal to the hyper-sensitivity-conferring merA mutants (Foster andNf Bhriain, unpublished data). They expressednormal levels of a reductase protein which wasindistinguishable from wild type in terms of itsstability, catalytic properties, and molecularweight (F. D. Porter and S. Silver, personalcommunication). It seems unlikely that the mu-tations have affected the carboxy terminus of thereductase. Interestingly, Tn501 has an opentranslational reading frame (urf-2) in the corre-sponding location, which could specify a poly-peptide of 57.5K. We have assigned the TnSmutations to a new gene (merD). However,there must be some doubt about merD becausewhen the copy number of the merD::TnS inser-tions was reduced a wild-type Hgr phenotypewas expressed. Also, we have shown that merDis not responsible for merbromin resistance (seeabove). No polypeptide has so far been assignedto merD or urf-2. However, the urf-2 region ofTnS01 is probably transcribed because inductionof the mer operon stimulates expression of theadjacent transposase and resolvase genes, pre-sumably by transcription from the mer operonpromoter (203).

Regulation of expression of mer genes. Mercuryresistance is expressed only after exposure tosubtoxic levels of Hg2' (369). This correlateswith induction of volatilizing activity (330, 369)and Hg2+ hyperbinding by hypersensitive mu-tants (274). Possibly the mer genes are regulatedboth positively and negatively in a similar fash-ion to araC-mediated control of the arabinoseoperon (283). Circumstantial evidence for thistype of regulation was the difficulty in isolatingconstitutive mutants and the fact that mutationscausing temperature-sensitive induction of re-ductase have been obtained (369). In addition,insertions in merR caused a sensitive phenotypeand allowed expression of low constitutive lev-els of reductase. These mutations could be com-plemented to give a wild-type inducible pheno-type (119). Reductase was fully repressed inuninduced diploid merR merA+lmerR+ merAcells and could be induced by Hg2+, showingthat merR encodes a trans-acting regulatoryelement which acts as both a repressor and aninducer of the mer operon (Ni Bhriain andFoster, unpublished data).

The araC protein negatively regulates its owntranscription as well as controls the araBADoperon (283). We have investigated the regula-tion of merR, using merR-lac transcriptionalfusions (Foster and Nf Bhriain, unpublisheddata). The constitutive expression of 3-galacto-sidase by merR-lac fusions was repressed eight-fold by merR+ plasmids and was not elevated byinduction. Thus, merR protein negatively regu-lates the transcription of the merR gene.DNA sequence analysis of TnS01 suggests

that the structural organization of mer differsfrom that of ara. Thus, merR and merTCA aretranscribed in the same direction from well-separated promoters (Brown et al., personalcommunication; Fig. 9), whereas araC and ara-BAD are transcribed divergently from overlap-ping promoters located in a common regulatoryregion (283).

Expression of reductase also appears to besensitive to catabolite repression in E. coli.Thus, reductase induced in glucose-grown cellsis twofold lower than in glycerol-grown cells.Reductase expression was also lower in cya andcrp mutants (367).

BYPASS MECHANISMS

Sulfonamide Resistance

Inhibitory effect of sulfonamides. Sulfonamidesdiffuse passively into bacterial cells and inhibitgrowth by interfering with the biosynthesis offolic acid (32). The sensitive step is the conden-sation of p-aminobenzoic acid (of which sulfon-amides are structural analogs) with dihydropteri-dine to form dihydropteroic acid (Fig. 10). Theinhibitory action of the drug can be explainedboth by competitive inhibition of dihydropter-oate synthase (DHPS) and by draining the sup-ply of pteridine precursor into a sulfonamideanalog of dihydropteroate which is excreted intothe medium (Fig. 10). Insufficient pterin-sulfon-amide remains in the cell to inhibit dihydrofolatesynthase, although it inhibits this enzyme invitro. In addition, the pteridine precursor doesnot accumulate in inhibited cells; it is consumedat the same rate in the presence or absence ofthe drug (319). Thus, reversal of sulfonamideaction by precursor accumulation could not oc-cur.

Pamid-specified Sur DMPS. Early reportsthat plasmid-specified sulfonamide resistance(Su') was due to impermeability have not beensubstantiated. Gram-negative bacteria carryingSur plasmids specify DHPS activity which canfunction normally in the presence of high con-centrations of drug (351, 352, 375, 426). These

MICROBIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from



resistant strains thus have two different DHPSenzymes which can be separated by gel filtration(375). A convenient, but not infallible, methodfor screening for plasmid-encoded DHPS activi-ty is to test for suppression of a mutation in thechromosomal DHPS gene which directs synthe-sis of a heat-labile enzyme (351, 375).Two different types of plasmid-encoded Sur

DHPS have been reported (375; G. Swedberg,Ph.D. thesis, University of Uppsala, Uppsala,Sweden, 1982). These enzymes differ markedlyin heat stability and are encoded by DNA se-quences that have no measurable homology asassessed by Southern blot hybridization andrestriction mapping. They are slightly smallerthan the chromosomal DHPS, having a molecu-lar weight of around 40,000 (375). The type I Su'DHPS specified by IncFII plasmids such as Rland R6 and the IncW plasmid R388 is extremelythermolabile (352, 375), hence the difficulty inthe above-mentioned ts mutation suppressiontest. The earlier report of failure to detect SurDHPS activity in R388-carrying cells is nowattributed to the synthesis of small amounts of avery labile enzyme (Swedberg, Ph.D. thesis).Resistant DHPS activity was detected in cellscarrying the R388 Sur determinant cloned inpBR322, which resulted in amplification of thegene and a gene-dosage effect for the enzyme.

DIHXDR22IE2l1

FIG. 10. Inhibition of the folic acid biosyntheticpathway by Su and Tp. The folic acid biosyntheticpathway is shown in outline along with the effects ofSu and Tp in susceptible and R-plasmid-containingresistant cells. The enzymes involved are DHFR andDHPS, specified by either plasmids (PL) or the chro-mosome (CS).

The type II (stable) enzyme is specified by smallnonconjugative plasmids such as RSF1010 (141)and pJP1 (402).The biochemical properties of the E. coli

chromosomal (Su-sensitive) DHPS have beencompared with the plasmid-encoded enzymes(375). The chromosomal enzyme is about 10%larger than the plasmid-determined DHPS (375)and binds Su slightly more efficiently than thenatural substrate p-aminobenzoic acid. In con-trast, the plasmid-encoded enzymes bind Su10,000-fold less efficiently. This degree of resist-ance is achieved without sacrificing the efficien-cy with which the natural substrate is bound.Plasmid-determined DHPS enzymes must beable to distinguish very precisely the carboxylgroup of p-aminobenzoic acid from the sulfa-mido group of the inhibitor.Between 10 and 50% of the DHPS activity in

extracts of Su' plasmid-carrying E. coli cells isresistant to Su in vitro (375). The resistant cellsmust be able to generate sufficient folic acid forgrowth, despite the fact that the chromosomalenzyme will continue to divert some pteridineinto the pterin-sulfonamide analog. That pterin-sulfonamide is excreted into the medium pre-vents it from reaching a concentration whichwould inhibit dihydrofolate synthase activity.

Regulation of plasmid-specified DHPS. Verylittle has been reported on the regulation ofplasmid-encoded DHPS expression. It is gener-ally assumed that the enzymes are synthesizedconstitutively, although no systematic study ofthis has been published. One difficulty is thelability of type I DHPS, which makes accurateenzyme assays difficult.The Su' determinant of the type II DHPS-

specifying plasmid RSF1010 is expressed co-ordinately with the adjacent streptomycin re-sistance (Sm') determinant, which specifies a3"-O-phosphotransferase (164). Evidence forthis was the finding that transposon A insertionsin the Sur determinant also reduced expressionof Sm', presumably due to polarity (164). Thus,the two genes are transcribed from a singlepromoter in the order Su-Sm. In view of thereport that some aminoglycoside-modifying en-zymes are regulated by catabolite repression(93) (presumably at the level of transcription), itis possible that the type II DHPS could beregulated in a similar fashion. This hypothesishas not been directly tested.The type I DHPS enzyme-specifying Sur de-

terminants of IncFII plasmids are also closelylinked to a gene encoding a 3"-19-O-adenyltrans-ferase which modifies Sm and spectinomycin(219, 263, 389). In this case there is no evidencefor coordinate expression and the linkage maybe entirely fortuitous.

71

VOL. 47, 1983 383


http://mm

br.asm.org/

Dow

nloaded from


384 FOSTER

Trimethoprim Resistance

Inhibitory effects of Tp. Trimethoprim (Tp) is apotent inhibitor of chromosomally encoded di-hydrofolate reductase (DHFR), an enzyme in-volved in the metabolism of folic acid (Fig. 10).The drug is an analog of the natural substratedihydrofolate. It binds 1,000-fold more stronglyto the active site of the enzyme than does thenatural substrate (11). A synergistic bactericidaleffect is obtained when Su and Tp are combined;hence, the drugs are often administered togetherin treatment of infections.

Plasmid-specified Tp-resistant DHFR. Themechanism of resistance to Tp is analogous tothat of sulfonamides, with plasmid-located genesspecifying a Tpr DHFR activity (10, 289, 352,382). It has been encountered in many species ofenteric bacteria (87, 88, 191). Epidemiologicalstudies have indicated that the incidence of Tpris increasing (87, 117). Occasionally, Tpr strainscausing urinary tract infections have a mutationin the gene encoding thymidylate synthase (thyAin E. coli) (245, 358). This causes a thymine-requiring phenotype and resistance to Tp be-cause one of the major cellular requirements forreduced folic acid is eliminated (358).Three distinct Tpr DHFR enzymes have been

identified (116, 117, 289, 382). The enzymes canclearly be distinguished from one another andfrom the chromosomal DHFR by biochemicaltests such as pH optima, stability, sensitivity toinhibitors, subunit structure, and immunologicalcross-reaction. These groupings have been con-firmed by hybridization studies with cloned TprDHFR genes (117).The type I DHFR whose prototype is encoded

by Tn7 on R483 (17) is highly resistant to inhibi-tion by Tp in vitro (50%o inhibitory concentra-tion, 57 ,uM) in comparison to the chromosomalenzyme (50% inhibitory concentration, 0.007FM). It is composed of two identical subunits of18,000 molecular weight each (116). In contrast,type II enzymes (prototypes encoded by R67and R751::Tn4O2) are essentially completely re-sistant to Tp (50% inhibitory concentration,70,000 puM). In addition, the R67-specified en-zyme is structurally different from the type Ienzymes because it is composed offour identical8,500-dalton subunits (116, 356, 359). A variantof the type II enzyme specified by R388 has aslightly larger (10,500 dalton) subunit (116, 443).Both type I and type II DHFR enzymes deter-mine very high levels of Tp resistance in E. coli(minimum inhibitory concentration, 2,000,ug/ml) compared with plasmid-free strains(minimum inhibitory concentration, 0.2 p.g/ml).In contrast, plasmids specifying the recentlydiscovered and less well-characterized type IIIDHFR determine moderate increases in Tpr

(minimum inhibitory concentration, 200 Fg/ml)(394). The type III enzyme is also inhibited bylower concentrations of Tp in vitro (50% inhibi-tory concentration, 1.5 FLM) (117).

DECREASED ACCUMULATION INVOLVINGEFFLUX MECHANISMS

Resistance to TetracyclinesUptake of Tc by susceptible cells. At low con-

centrations the tetracycines (Tc) are bacterio-static antibiotics that inhibit protein synthesis bypreventing the binding of aminoacyl tRNA to theribosome A site (130). The drug also chelatescations, notably Mg2+. A Tc-Mg chelation com-plex is involved in ribosome binding (130, 395).The mechanism of entry of Tc into sensitive

cells has been discussed in recent reviews (56,61, 232). Tc's are broad-spectrum antibioticswith considerable activity against gram-negativebacteria. Passage of Tc through the outer mem-brane appears to be the rate-limiting step in drugaccumulation by gram-negative organisms (256).Tc is relatively hydrophilic and appears to dif-fuse preferentially through ompF porins (60).Mutants deficient in ompF porn are only three-fold more resistant to Tc, so the drug must beable to enter by other routes. Hydrophobicderivatives such as minocycline probably passthrough the lipid interior of the outer membrane(60, 256).

Early work (reviewed by Chopra and Howe[61]) showed that Tc uptake is energy depen-dent. More recent studies indicate that Tc up-take in E. coli is biphasic (109, 257). The initialrapid phase is energy independent and probablyrepresents binding of drug to the cell surface andpassage by diffusion through the outer layersof the cell. The second slower phase of uptake isenergy dependent and corresponds to transportacross the cytoplasmic membrane. This trans-port is inhibited by energy poisons, uncouplers,and depolarizers, but not by inhibitors of mem-brane-bound ATPase (58, 232), suggesting thatproton motive force is required. Studies on theaccumulation of Tc in membrane vesicles pre-pared from susceptible E. coli cells confirmedthat active Tc transport depends on proton mo-tive force (256).The identity of the Tc carrier in the bacterial

cytoplasmic membrane has remained elusive.Indeed, until recently formal kinetic evidencefor a carrier was lacking because of its lowaffinity for Tc (163). The earlier suggestion thatMg2+ permeases might be involved, with drugbinding to Mg2+ while being transported (124),seems unlikely because E. coli mutants defec-tive in Mg2+ transport accumulated Tc as well asthe wild type (15). Other possible candidates

MICROBIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from



such as the dicarboxylic and glutamate transportsystems are now known not to be involved (58,61).

Naturaily occurring Tc resistance determi-nants. Tc resistance is widely encountered ingram-positive and gram-negative bacteria. It isusually plasmid encoded and expressed induc-ibly, although chromosomally located and con-stitutively expressed Tcr determinants havebeen reported.

(i) Tc resistance in gram-negative bacteria.Five plasmid-specified Tcr determinants (TetAto TetE) have been distinguished in entericbacteria by DNA hybridization experiments andresistance phenotypes (49, 63, 120, 260, 315).Despite the failure of these elements to formhybrids, recent DNA sequence analysis of theTnlO (TetB), Tnl721 and RP1 (TetA), andpSC101 (TetC) determinants has revealed signif-icant homologies between the tet structuralgenes. This is discussed below. The TetB deter-minant is the most frequently encountered inenteric coliform bacteria (232). It is also respon-sible for Tcr in H. influenzae (188). Most strainsof Proteus mirabilis express inducible high-levelTcr from a chromosomal locus (232). The possi-bility that this is related to one of the plasmid-located determinants has not been tested. Theplasmid-linked Tcr determinant in B. fragilis isunusual- because conjugation functions andexpression of Tcr may be coordinately regulated(301).

(ii) Tc resistance in gram-positive bacteria. Tcresistance in S. aureus is usually determined bysmall, closely related, multicopy plasmids (184).Chromosomally mediated Tcr is rarer. Chromo-somal and plasmid determinants have been dis-tinguished on the basis of resistance phenotype(14, 62), but the possibility remains that they aregenetically related. Many different soil bacilliharbor small Tcr plasmids (26, 299), but nosequence homology between these and the S.aureus plasmid-linked determinants was foundby hybridization experiments (299).Tc resistance is very common in the strepto-

cocci. Three distinct Tcr determinants havebeen found by hybridization studies (42). ThetetL determinant is present on nonconjugativeplasmids, including the amplifiable element ofpAMal in Streptococcus faecalis (66, 431). ThetetM genes are found in the chromosomes of S.agalactiae, S. pneumoniae, and S. mutans (391)and are also associated with the conjugativetransposon Tn916 (123) of S. faecalis (42). Tcrstrains of Clostridium difficile also harbor achromosomal Tcr element which promotes con-jugational transfer (354), but it is not knownwhether this is related to Tn916. The tetN ele-ment is characteristic of large conjugative plas-mids (42). Hybridization tests failed to reveal

any homology between the streptococcal tetdeterminants and the TetA-D determinants ofenteric bacteria, or the tet plasmids of Staphylo-coccus aureus and B. sphaericus (42).Mechanism of Tc resistance. It is accepted that

Tc resistance is due mainly to a decrease in drugaccumulation (61, 124). A significant recent con-tribution to understanding the mechanism of Tcresistance was the discovery of a drug effluxmechanism (16, 258). Whether efflux alone issufficient to promote resistance (as seems to bethe case for cadmium and arsenate resistance)remains to be determined. It is possible that thedecrease in drug accumulation reflects both de-creased uptake and increased efflux mechanismsacting together. In addition, a ribosome protec-tion mechanism may be involved (235, 236).Two different experimental approaches have

implicated an efflux mechanism as the majorfactor in Tc resistance. Using spectrofluori-metry to measure drug uptake under conditionswhere cells remained metabolically active, itwas shown that resistant cells bound less Tc andthat accumulated Tc was rapidly lost by anenergy-dependent process when the cells weretransferred to a drug-free medium (16). In otherstudies, inverted cytoplasmic membrane vesi-cles prepared from Tcr bacteria accumulated Tcby an energy-dependent process (163, 258). TheTc transport system (which presumably deter-mines efflux from whole cells) had a high affinityfor Tc (163), was saturable and energy depen-dent, had different pH and Mg2" requirementscompared with transport into sensitive cells, andwas competitively inhibited by minocycline(258). Clearly this represents a novel plasmid-encoded Tc-specific transport system.

It has been argued that reduced Tc accumula-tion is not sufficient to explain high-level Tcresistance, since resistant cells still take up asubstantial quantity of drug (234, 236). There isno evidence for compartmentalization or inacti-vation of the drug in resistant cells (232). Severalreports have suggested that ribosomes in Tcrcells may be protected from the inhibitory ef-fects of the drug (232, 236). However, the levelof protection detected in vitro was small and thetet product involved has not yet been identified.In one study, polypeptide synthesis directed byendogenous message or by polyuridylic acid inunfractionated lysates of resistant bacteria wastwo- to threefold more resistant to Tc inhibitionthan in extracts of susceptible cells (236). Thereis some doubt about the validity of this finding.Attempts to repeat the experiments with un-washed and washed ribosomes have failed, andfurthermore, toluene treatment of Tcr cells com-pletely eliminated the resistance of polypeptidesynthesis to Tc inhibition (I. Chopra, personalcommunication).

385VOL. 47, 1983


http://mm

br.asm.org/

Dow

nloaded from


386 FOSTER

There has also been some speculation aboutthe tet repressor functioning as a resistanceprotein, possibly by binding Tc (18). The datathat led to this suggestion were obtained fromexperiments with multicopy tet plasmids.Whether there is sufficient repressor in cellscarrying a single copy of TnJO to significantlylower the Tc concentration is questionable. It ispossible that binding of Tc to repressor in unin-duced cells delays the inhibitory effect of thedrug on protein synthesis sufficiently for theresistance gene to be transcribed and translated.Also, the proposal that the repressor is a mem-brane protein needs to be confirmed. Evidencefor membrane association was presented for ahybrid repressor-p-galactosidase protein but notfor the repressor itself (18).

Structure and function of the Tn1O Tc resist-ance region. The Tcr determinant of TnJO hasbeen studied in most detail at the molecularlevel. Our current understanding of the geneticorganization and regulation of the TnJO tet genesis summarized in Fig. 11. tetR (repressor) andtetA (resistance gene encoding TET protein) ofTnJO are transcribed divergently from a commonregulatory region that contains overlapping pro-moters (166; Fig. 11). The direction of transcrip-tion was deduced by studying polypeptides syn-thesized by cloned fragments and by tet mutantplasmids in minicells (18, 69, 193, 430). The

genetic organization was confirmed by DNAsequence analysis (166). The regulation of tetgene expression is discussed below.Two Tc-inducible polypeptides specified by

TnlO have been unequivocally assigned to tetgenes. These are the 23.5K repressor proteinand the membrane-bound TET protein. TheTET protein was originally identified in E. coliminicells harboring plasmid R222, which carriesTnlO. Its molecular weight has been reported as50,000 or 36,000 depending on the gel electro-phoresis buffer system used (110, 193, 233, 438,444). DNA sequence analysis of the TnlO tetregion shows a single translational reading framethat spans the entire resistance gene region(166). This would encode a polypeptide of 43.3K(Fig. 11). The discrepancy between the molecu-lar weight measured by electrophoresis and thatdeduced from the DNA sequence could be ex-plained if the hydrophobic protein migrated ab-errantly, as has been shown for other membraneproteins (276). Post-translational processing ofthe TET protein does not seem likely becausethe predicted N-terminal amino acid sequencedoes not suggest a leader sequence and becauseTET protein synthesized in minicells and in vitromigrated at the same rate (166, 438).

It has also been suggested that a Tc-induciblepolypeptide of 15K may be specified by TnlO(438), but this protein has not been consistently

TETP- tetA

1Ak

1431( _tetB

-..-O 74w - p

tetR

! -1iQ RSNA0

I -.

RNA -

FIG. 11. Genetic organization and regulation of the Tc resistance genes of TnlO. The upper part shows theorganization of the tet region of TnlO. The tetR gene is transcribed from right to left from promoters located inthe central regulatory region. It encodes the 23K repressor protein, which binds to operator sites to preventtranscription of both tetR and the TET structural gene. DNA sequence analysis indicates that the TET resistancegene encodes a polypeptide of 43.3K which migrates with an apparent molecular weight of 36,000 in sodiumdodecyl sulfate-polyacrylamide gels. The location of mutations which form complementation groups tetA andtetB is indicated. An expanded diagram of the regulatory region is shown at the bottom. The -35 and -10 regionsof the promoters for the tet resistance gene and the tetR repressor gene are shown, as well as the transcriptionalstart points (-k) and regions of the DNA sequence protected from DNase digestion by repressor binding (-).

IL . . I a -

MICROBIOL. REV.

r%


http://mm

br.asm.org/

Dow

nloaded from



seen in minicell experiments. A protein of thissize was identified in the cytoplasmic membraneof resistant bacteria, using immunological meth-ods (Chopra, personal communication). Howev-er, there is no evidence that this is a product of atet gene required for expression of resistance.The TnJO tet DNA sequence reveals an opentranslational reading frame overlapping the pro-moter-distal part of the tetA coding sequence,which could encode a protein of 17.6K (166; Fig.11). This might be responsible for the low-molecular-weight Tc-inducible polypeptide, butany function in resistance is dubious.

Genetic complementation tests performed be-tween compatible plasmids carrying differentTc-sensitive mutations provided evidence fortwo complementation groups called tetA andtetB (68, 79) (Fig. 11). Both map within theHincII fragment known to carry the tet structur-al gene(s). Since there is a single coding se-quence spanning the tetA-tetB region (166) andthe complementation tests failed to generate awild-type high-level Tcr phenotype, it is likelythat Tc resistance was generated by intrageniccomplementation resulting from the interactionof two different mutant TET proteins. The com-plementation groups may represent two func-tional domains within the TET protein ratherthan different polypeptide products of two cis-trons. The membrane-bound TET protein maybe a homomultimer.The 25K Tc-inducible polypeptide seen in

XTnlO-infected UV-irradiated cells (444) proba-bly corresponds to the 23.5K repressor proteinidentified by studying in vitro generated recom-binant tet plasmids in minicells (18, 430). Thatthe repressor is itself inducible by Tc indicatesthat it regulates its own expression.The tet determinants of RP1 (TetA group; S.

Waters and J. Grinsted, personal communica-tion), Tnl721 (TetA group; R. Schmitt, personalcommunication), and pSC101 (TetC group; 360)have a structural organization similar to that ofthe TetB group determinant of TnlO. Further-more, significant DNA sequence homology ex-ists between these determinants in the regionspecifying the resistance proteins. These Tetdeterminants have clearly evolved from a com-mon ancestor, but their DNA sequences havediverged to an extent which prevents detectablehybridization (260). The predicted amino acidsequences of the TET proteins reveal the follow-ing homologies: RP1-TnJO, 45%; RP1-pSC101,52%; TnJO-pSC101, 25% (Chopra, personalcommunication). The conserved amino acidsoccur in localized regions which could be impor-tant in the function of the protein.The TetC determinant of pBR322 (derived

from pSC101) may be organized in a slightlydifferent fashion from the TnJO and Tnl721

elements (372). According to the DNA se-quences, the TET protein of pSC101 lacks 56 N-terminal amino acids from the correspondingregion of the TnlO TET protein (Chopra, person-al communication). However, the reported mo-lecular weight of the pSC101 TET protein is34,000 (379), which is only 2,000 smaller thanthat of the TnJO protein. The DNA sequence ofpBR322 also indicates that a small protein couldbe encoded by a promoter-proximal readingframe which partly overlaps the coding se-quence for the N-terminus of the TET protein(372). This region has some homology with thecorresponding region of TnlO and could code forone of the Tc-inducible polypeptides specifiedby pSC101 in minicells (379). A frameshift muta-tion within the coding sequence for the ancestralTET protein could account for this change.

Regulation of tet gene expression. Molecular,genetic (18, 430, 438), and DNA sequence analy-ses (166) of the TnlO tet genes show that therepressor and structural genes are transcribeddivergently from a common regulatory region.The repressor protein negatively regulates thetranscription of the tetA gene as well as that ofthe tetR gene (15, 430). DNase protection ex-periments showed that the repressor binds toseveral parts of the regulatory region, some ofwhich overlap the -35 region of tet promoters(166; Fig. 11). The Tnl721 tet genes are orga-nized and regulated in the same way (R.Schmitt, personal communication). In the pres-ence of Tc the repressor is presumably detachedfrom the operators, allowing transcription ofboth genes. Biochemical analysis of the interac-tion of purified repressor protein with Tc andwith a small DNA fragment carrying the opera-tors (165, 167) supports this interpretation.DNA sequence analysis of the TnJO and

Tnl 721 tet regions revealed two promoters capa-ble of initiating transcription of tetR, both ofwhich are active in vitro (166; W. Hillen and R.Schmitt, personal communications). The func-tion of the two tetR promoters is not known, butit is possible that one is not fully repressed andprovides a "basal" level of repressor synthesisin uninduced cells.Another aspect of the expression of Tcr by

TnWO is the reduction in resistance level when tetgenes are present in multiple copies (69). Incontrast, TetA and TetC determinants expresshigher levels of resistance when the tet genecopy number is raised (43, 328, 329, 421). Mostof the transposon TnS insertions in the codingsequence for the TET protein prevented expres-sion of high-level Tcr from a single copy of thewild-type TnWO in the chromosome (43). Thus,amino-terminal fragments of the TET proteincould still interfere with Tc resistance. Overpro-duction of the repressor was shown not to be

VOL. 47, 1983 387


http://mm

br.asm.org/

Dow

nloaded from


388 FOSTER

responsible for the negative gene-dose effect(43). One possible explanation is autogenousregulation of translation of the TET proteinwhen it is overproduced. Alternatively, overpro-duction of the TET protein might simply bedeleterious to the cell and prevent growth.

Resistance to CadmiumCadmium (Cd) ions are taken into sensitive

bacterial cells by the energy-dependent manga-nese transport systems (397, 398, 419), wherethey cause rapid cessation of respiration bybinding to sulfhydryl groups in proteins.

Resistance to Cd2+ is a common plasmid-specified function in S. aureus (103, 106, 281,355). It has not been encountered in other bacte-rial genera. Two distinct Cd resistance determi-nants have been recognized: (i) CadA, whichcauses a 100-fold increase in resistance level dueto a specific Cd2" efflux system (295, 355, 398);and (ii) CadB, which specifies a lower level ofresistance than CadA by an uncharacterizedmechanism which does not involve decreasedaccumulation (295). Cd resistance is frequentlyassociated with the large "penicillinase" plas-mids (280, 281, 355). Some plasmids (e.g.,pII147) specify both CadA and CadB resistancedeterminants, whereas others (e.g., p1258) carryonly CadA determinants (280). A small multi-copy CadA plasmid has also recently been de-scribed (106). Both CadA and CadB determi-nants are expressed constitutively and alsoconfer resistance to zinc ions (295).

It has been known for some time that high-level resistance to Cd2" (specified by CadAdeterminants) involves decreased accumulation(55, 56, 209, 419). It is now clear from thedetailed studies of Tynecka et al. (397, 398) thatCd2' resistance is caused by a plasmid-encodedefflux system. Experiments with uncouplers andionophores showed that resistant strains ex-pelled Cd2+ via a Cd2+/2H+ antiport systemwhich is energized chemiosmotically by protoncirculation across the membrane. Thus, oneCd2+ ion is exchanged for two protons and atexternal Cd2+ concentrations below 100 ,uM noinhibition of growth or respiration occurs. At-tempts to demonstrate efflux of Zn2+ have so farfailed (295).

Resistance to Arsenate

Arsenate ions enter bacterial cells via thephosphate transport systems (423). Arsenate istoxic to bacteria because it is an analog ofphosphate and can inhibit enzymes such askinases. Also, arsenylated sugars hydrolyzespontaneously, resulting in a loss of free energyin glycolysis (369).

Resistance to arsenate is determined by plas-

mids in S. aureus (280, 281, 347) and entericbacteria (159, 354). Arsenate resistance is invari-ably inducible and is linked to arsenite andantimony(III) resistance determinants in bothgroups of bacteria. Resistance to the three ionsis coordinately induced, with each ion beingcapable of acting as an inducer at subtoxicconcentrations (347). In addition, bismuth(III)can induce these resistance determinants, butresistance to bismuth is not expressed (347). Thearsenate resistance determinant is distinct fromarsenite and antimony(III) resistance. Resist-ance to arsenate is known to be due to a specificefflux mechanism (see below), whereas resist-ance to arsenite and antimony(III) is poorlyunderstood. In addition, the arsenate and arse-nite resistance determinants are genetically sep-arable (280).The arsenate resistance determinants of both

S. aureus and E. coli specify an efilux systemwhich has recently been studied in some detailby measuring the loss of 74As03- from preload-ed resistant cells (269a, 348). The efflux systemis highly specific for arsenate. Unlike the uptakesystems it does not recognize phosphate. Incontrast to the Cd efflux mechanism, arsenateefflux is not driven chemiosmotically and is notdependent on membrane potential or a pH gradi-ent across the membrane. Rather, it is driven byhydrolysis of ATP or some other high-energycofactor.

ALTERATION TO THE ANTIBIOTICTARGET SITE

Resistance to Antibiotics in the MLS GroupMLS antibiotics are a structurally diverse

group which cause a bacteristatic effect by bind-ing to the 50S ribosome subunit of susceptiblebacteria and inhibiting protein synthesis (for areview, see reference 130). Resistance to MLSdrugs in natural isolates is caused by methyl-ation of adenine residues in 23S rRNA whichprevents the drugs from binding to their target.The MLS resistance phenotype is often plasmidencoded (see below). Many gram-negative bac-teria are intrinsically resistant to these drugsbecause of their impermeable outer membranesand possibly because of the level of methylationof rRNA (71, 380). However, Bacteroides spp.are susceptible to the lincosamide drug clinda-mycin. Mutations in three chromosomal locicause higher resistance to macrolides (includingerythromycin) in E. coli. Lesions in eryA anderyB alter ribosomal proteins L4 and L22, re-spectively (427), whereas eryC mutations proba-bly affect maturation of 16S and 23S rRNA(288). The resistance phenotypes conferred bythese mutations are narrower than naturallyoccurring MLS resistance.

MICROBIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from



Occurrence of MLS resistance in natural iso-lates. Resistance to MLS antibiotics is specifiedby plasmids in the gram-positive bacteria S.aureus and Streptococcus spp. (65, 173) and inthe gram-negative Bacteroides spp. (420). Chro-mosomally located MLS resistance determi-nants have been found in Bacillus licheniformis(98) and Streptococcus pneumoniae (65) and inseveral species of Streptomyces which produceMLS antibiotics (125). The S. aureus p1258determinant is located on a transposon TnS51(279) and the Streptococcus faecalis plasmidpAD2 has an interesting Emr transposon Tn917,in which transposition is inducible by erythro-mycin (392).DNA hybridization studies indicate that there

are at least three distinct groups of MLS resist-ance determinants in gram-positive bacteria.The Emr gene of TnS51 has significant homologywith the streptococcal determinants (415) butnot with the S. aureus plasmid pE194. The Emrdeterminant of B. licheniformis does not hybrid-ize with either staphylococcal determinant (98).The relationship between these determinantsand the Bacteroides sp. and Streptomyces sp.genes has not been tested. It is interesting tonote that although homology was not detectedbetween the Emr regions of pE194 and pAM77(which is related to the pI258::TnS51 determi-nant) by Southern blotting, significant similaritywas revealed by comparing the DNA sequencesand the derived amino acid sequences of themethylase proteins (173).Mechanism of resistance to MLS antibiotics.

Resistance to MLS drugs is caused by specificN6,N6-dimethylation of adenine residues in 23SrRNA (216-218). The methylase encoded byplasmid pE194 has been purified to homogene-ity and studied in vitro (344). It is a 29,000-daltonpolypeptide which depends on S-adenosylmeth-ionine for its activity. In vitro, it will methylatefree 23S rRNA and the 50S ribosome subunit,but not the 70S ribosome. Access to the methyl-ation site on the 50S subunit may be stericallyblocked when it interacts with the 30S particle.Ribosomes with dimethylated 23S rRNA bindMLS antibiotics less effectively and are notinhibited by the drugs (218).

Induction of MLS resistance determinants.Some MLS resistance determinants in S. aureusand streptococci are inducible, whereas others[e.g., S. aureus (pl258)] are expressed constitu-tively. The number of antibiotics that can inducethe MLS resistance phenotype is limited. ThepE194 determinant can only be induced byerythromycin and oleandomycin (171, 345, 413,414), whereas other staphylococcal determi-nants can be induced by these drugs and bycarbomycin (3). This phenomenon has been de-scribed as "dissociated" resistance (48, 132) and

means that resistance to the full range of MLSantibiotics can only be expressed after inductionby one of the limited range of inducers, or whenconstitutive mutants are selected. The reasonfor this can now be explained in molecular terms(see below; 146, 171, 345).The Streptococcus pyogenes MLS resistance

determinant is induced by erythromycin andlincomycin (239, 240). However, this organismexpresses an unusual zonal pattern of resistanceto lincomycin (239, 240). Cells only grow atbelow 0.06 ,ug or between 60 and 250 ,ug oflincomycin per ml. Growth at intermediate lin-comycin concentrations occurs only after induc-tion with erythromycin or with constitutive mu-tants. At concentrations between 0.06 and 60p.g/ml, lincomycin cannot induce resistance andcell growth is inhibited.

Further variation in induction patterns andresistance phenotypes occurs among the variousStreptomyces species that produce MLS antibi-otics (125).

Regulation of expression of MLS resistanceencoded by plasmid pE194: the translational at-tenuation model. The MLS resistance determi-nant (ermC) located on S. aureus plasmid pE194has been subjected to detailed biochemical andgenetic analysis in the laboratories of Dubnau(146, 345) and Weisblum (171, 172). Induction ofermC-encoded resistance by erythromycin andoleandomycin occurs by a mechanism that re-sults in an increased rate of translation of ermCmRNA rather than by stimulating transcriptionof the gene. This induction occurs in the pres-ence of inhibitors of transcription (345) andresults in the mRNA becoming more stable,probably because of changes in its secondarystructure.The DNA sequence of the ermC regulatory

region provided important clues about the regu-latory mechanism (140, 171). The ermC messagehas two open translational reading frames, eachpreceded at the appropriate position by a ribo-some binding site showing the strong homologyto 16S rRNA typical of gram-positive bacteria(146, 171, 254). One reading frame could specifya short 19-amino acid peptide. The other deter-mines the 29K methylase. There are six repeatedsequences in this region of the message whichhave the potential to form a variety of hairpinloop structures (140, 146, 171, 172) (Fig. 12). Inuninduced cells the ribosome binding site (SD2)or the translational initiation codon (AUG2) forthe methylase or both may be masked by sec-ondary structure. The translational attenuationmodel (140, 146, 171) suggests that induction iseffected by altering the secondary structure ofthe transcript so that SD2 and AUG2 are un-masked (Fig. 12).The model suggests that an Em-sensitive ribo-

VOL. 47, 1983 389


http://mm

br.asm.org/

Dow

nloaded from


390 FOSTER

INDUCED

UNINDUCEDFIG. 12. Induction of Em resistance by translational attenuation. Possible secondary structures that could

occur at the 5' end of the ermC message are indicated. The open boxes represent the ribosome binding sites forthe 19-amino acid leader peptide (SD1) and for the methylase (SD2). The small circles depict the translationalstart signals AUG1 and AUG2. The large circles are ribosome particles. The change in secondary structure of themRNA which is thought to be induced by stalling of the ribosomes with bound Em during the course oftranslating the leader peptide is shown. This allows ribosomes to bind to AUG2 and to commence translating themethylase.

some stalls during translation of the leader pep-tide in the presence of the drug. This allows themRNA to change structure and thus exposesSD2 and AUG2. Resistant ribosomes or sensi-tive ribosomes without bound drug will thentranslate the methylase. When sufficient ribo-somes have been methylated, the stalling duringtranslation of the leader will no longer occur andthe mRNA will revert to the uninduced second-ary structure.One feature of this model is the inhibitory

activity of antibiotics that can induce resistance.Em will only bind to free ribosomes, not to thoseengaged in protein synthesis on polysomes (2%,378). The Em-bound ribosome participates inlimited translation before stalling occurs. Itseems significant that the drugs capable of in-ducing resistance are those which allow transla-tion to occur for the longest period before theribosomes stall (210, 242, 243). Also, the struc-ture of the amino acids being incorporated intothe peptide may be important determinants ofwhere the ribosomes stall. The bulky hydropho-bic side chains of the proline at position 16 andthe lysines at positions 18 and 19 in the leaderpeptide may favor stalling towards its carboxyterminus. The amino terminus has many resi-dues which are free of hydrophobic side chainsand which will continue to be incorporated intothe polypeptide by Em-bound ribosomes (47,242, 243).

If the ribosome stalls in repeat sequence 2,

there is a possibility that a conformationalchange in mRNA structure will occur and thatrepeat sequences 3 and 4 will pair (Fig. 12). Thisrearrangement will expose the AUG2 for themethylase and will allow translation of the pro-tein. It is conceivable that SD2 is occupied by aribosome in a preinitiation complex which ispoised to commence translation immediatelywhen the conformational change occurs.

Structure of nascent transcripts and basalexpression of methylase. The ermC gene is tran-scribed continuously in uninduced MLS+ cells.The kinetic trapping model (Fig. 13; 146) pro-poses that the message folds in different configu-rations as transcription proceeds. Initially, afairly stable structure (A in Fig. 13) is formedbefore the distal end of the regulatory region istranscribed. Thus, SD2 will be exposed immedi-ately as it is transcribed, allowing ribosomes tobind and a short burst of translation of methyl-ase protein to occur before the structure of themessage changes. When repeat sequences 5 and6 are transcribed, SD2 and AUG2 will becometrapped because the mRNA will rearrange into amore stable form (B and C in Fig. 13).The paradox of induction by a mechanism

involving translational attenuation. The transla-tional attenuation model proposes that Em in-duces translation of the methylase by stoppingthe translation of the leader peptide. This eventis required to stimulate the mRNA structuralchange. How can this occur when many ribo-

MICROBIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from



somes in the cell have drug bound to them?First, basal expression of methylase renders 5%of the ribosomes in uninduced cells resistant.Thus, SD2 has a 1 in 20 chance of being boundby a resistant ribosome and some transcripts willbe translated by such ribosomes immediatelyafter induction. Second, at low drug concentra-tions many sensitive ribosomes will avoid beinginactivated by Em before initiating translation ofmethylase. Once translation starts the drug can-not bind to the ribosome. Third, the ribosomesin the putative preinitiation complex at SD2 maynot be able to bind Em. Thus, transcripts may beprimed to begin translation as soon as attenua-tion occurs, ensuring that the first round oftranslation cannot be halted by the drug.

UNCHARACTERIZED RESISTANCEMECHANISMS

Resistance to Fusidic Acid in S. aureusFusidic acid is a steroid antibiotic which inhib-

its protein synthesis by interfering with thetranslocation protein EFG1 (130). Chromosomalmutations which cause G factor to bind fusidicacid with decreased efficiency have been isolat-ed in E. coli (130). Chromosomal mutationscausing Fur can also be easily isolated in S.aureus (213). Despite this, Fur clinical isolates ofS. aureus usually have plasmid-bome resistancedeterminants (213, 214), possibly becausestrains with chromosomal mutations are lessvirulent (57).

3

L5

1 6

The expression of plasmid-encoded Fur in S.aureus is rapidly inducible (57). It does notprotect ribosomes and it does not cause detect-able drug modification or inactivation (57). De-spite the failure to demonstrate decreased fusi-dic acid uptake by resistant cells, it waspostulated that the resistance mechanism result-ed in decreased permeability because of achange in the ratio of phosphatidylglycerol tolysylphosphatidylglycerol in resistant cells.There is no evidence that resistance is associat-ed with CAT, as it is in enteric bacteria (406),but an intracellular binding mechanism cannotbe ruled out.

Nonenzymatic Chloramphenicol Resistance inGram-Negative Bacteria

Some Cm resistance determinants in gram-negative bacteria do not express CAT (128, 129,271). As with the fusidic acid resistance determi-nant of S. aureus, the nonenzymatic Cmr pheno-type is inducible, does not cause detectable drugmodification or inactivation, and does not pro-tect ribosomes against Cm inhibition in vitro(128). A permeability mechanism operating atthe level of the cytoplasmic membrane is sus-pected because spheroplasts still express resist-ance and also because cells carrying both enzy-matic and nonenzymatic Cm resistancedeterminants do not acetylate the drug (128).The nonenzymatic Cmr determinants of R26 andR55-1 do not confer high-level resistance to thefluorinated analogs of Cm (99).

5 6

a-7

FIG. 13. Kinetic trapping model. Changes that could occur in the secondary structure of the 5' end of theermC message during the course of transcription are shown. Ribosomes might bind transiently to SD2 to allow ashort burst of methylase synthesis (structure A) before more stable mRNA structures (B and C) can form andsequester SD2/AUG2 initiation signals upon completion of the transcription of the region.

391VOL. 47, 1983


http://mm

br.asm.org/

Dow

nloaded from


392 FOSTER

Resistance to FosfomycinFosfomycin interferes with bacterial cell wall

synthesis by inhibiting UDP-N-acetylglucosa-mine-3-0-enolpyruvyl transferase (195). Thedrug enters the E. coli cell via the L-a-glycero-phosphate (glpT) or the hexose-6-phosphate(uph) transport system (195, 317). Chromosom-ally specified resistance to fosfomycin is due todefects in one of these transport systems or tomutations in the target enzyme (195, 404).

In clinical isolates the most frequently en-countered mechanisms of resistance to fosfomy-cin are mutations which impair drug transport (J.M. Ortiz, personal communication). Plasmid-specified fosfomycin resistance in Serratiamarcescens has been reported (261). In one casea transposon was implicated (131). There was noimpairment of drug transport in resistant cells(228). Recently, evidence has been obtained foran intracellular drug inactivation mechanism (J.Leon, J. M. Garcia-Lobo, and J. M. Ortiz,personal communication).

Resistance to Cadmium Ions Specified by theCadB Determinant

Unlike the CadA determinant which promotesefflux of Cd2+ from resistant cells (398), CadBdoes not alter the rate of Cd2' accumulation inwhole cells or in vesicles. In contrast, resistantcells bind more Cd2+ than sensitive cells, whichsuggests that an intracellular binding mechanismmight be operating (295). An analogy was drawnbetween CadB resistance and metallothioneinproteins of Neurospora sp. (229), yeast (118),and blue-green algae (cyanobacteria) (285)which bind Cd2+, Zn2+, and Cu2+ strongly.

Resistance to Arsenite and Antimony(E)Arsenite and antimony(III) are toxic to bacte-

ria because they bind to cysteine residues inproteins (2). Resistance to arsenite and anti-mony(III) is linked to and coordinately inducedwith arsenate resistance (347). Little is knownabout the mechanism of resistance to arseniteand antimony(III). Excretion of a chelatingagent (e.g., dimercaptal) into the medium wasruled out (347) as was the possibility that arse-nite was oxidized to arsenate and then eliminat-ed by the arsenate efflux system (369). Anintracellular binding mechanism seems unlikelybecause no difference in binding was foundbetween resistant and sensitive cells (347).

Resistance to Silver IonsThe occurrence of plasmid-mediated resist-

ance to silver ions in clinical isolates of gram-negative bacteria probably reflects the use ofsilver salts to treat burns (12, 253). Silver is toxicto bacteria because it interferes with respiration

(29) and other cell surface-associated functions(122, 322).The silver resistance determinant is expressed

constitutively and confers a high level of resist-ance (346). Resistance to AgNO3 is dependenton the presence of halide ions. Susceptible cellscan extract Ag2' from an AgCl2 precipitate,whereas resistant cells cannot. These determi-nants also confer halide-independent resistanceto silver-sulfadiazine. Thus, silver resistanceappears to be due to a function which is proba-bly located at the cell surface and which pre-vents cells from extracting silver, which is in aninsoluble or slowly released form. The mecha-nism cannot exclude soluble silver salts.

Resistance to Other Heavy Metal CompoundsPlasmid-specified resistance to bismuth, lead,

boron, chromium, cobalt, nickel, tellurium, andzinc compounds has been described previously(364, 365, 369), but nothing is known about themechanisms involved.

Resistance to Ethidiua BromideSome strains of S: aureus are resistant to

ethidium bromide, proflavine, acriflavine, andacridine (107, 192). The determinant is oftenlocated on penicillinase plasmids. Little isknown of the resistance mechanism, but de-creased permeability has been suggested on thebasis of binding experiments (192). It is possiblethat this determinant has been selected in thestaphylococci by the use, until recently, of acri-flavine as an antiseptic.

MECHANISMS FOR INCREASINGEXPRESSION OF ANTIBIOTIC RESISTANCE

DETERMINANTSAmplification by Tandem Duplication

The activity of a genetic determinant can beincreased by elevating the number of genecopies by tandem duplication (12). Chromo-somal genes in eucaryotes (118, 197, 407) andprocaryotes (12, 104), as well as plasmid-borneantibiotic resistance determinants in bacteria(292, 328, 329, 421, 431), can be amplified by thismechanism. Tandem duplications are often initi-ated by legitimate recombination between di-rectly repeated DNA sequences which flank thegene (Fig. 14 and 15). These are promoted byhost recombination enzymes. Some plasmid-borne antibiotic resistance determinants haveevolved to facilitate duplications by having long(700 to 1,200 bp) direct repeat sequences tostimulate the recombination event. They includethe drug resistance r-determinant region ofIncFII R plasmids (50, 176, 303), the Tcr genesof Tn)721 (328, 421), and the kanamycin resist-ance determinant of Rtsl (179) (Fig. 16).

MICROBIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from



-KII-Z

1. Dimerization

C2

42 Oligomerization

FIG. 14. Model for formation of tandem duplica-tions by unequal crossing over. The open boxes repre-sent the direct repeat sequence§ that flank amplifiablegenes. The antibiotic resistance gene is indicated bythe thicker line between the two repeats. Recombina-tion is represented by the thin lines.

Amplification can be selected in the labora-tory by prolonged exponential growth in liquidmedium containing a high but subinhibitory con-centration of drug. Cells expressing higherresistance levels will outgrow those carryingnonamplified resistance genes (292, 328, 329,421, 431). Alternatively, the amplified cellswhich exist in the population can be directlyselected by plating on agar containing a concen-tration of drug which is inhibitory to nonampli-fled cells (157, 421). When the drug is removedfrom the medium the plasmid population willreturn to the nonamplified state because tandemduplications are inherently unstable (12, 292,421).Two models have been proposed to explain

how tandem duplications can occur (176, 262,303, 431) (Fig. 14 and 15). A common feature isthat the duplication is initiated by recA-depen-dent homologous (legitimate) recombination be-tween directly repeated DNA sequences whichflank the resistance gene. In model 1 (Fig. 14)the duplication is initiated during DNA replica-tion and requires a double but unequal crossoverevent. Model 2 (Fig. 15) requires the resistancegerue to be excised by a single crossover andthen to integrate into another plasmid carryingan intact resistance determinant. Model 1 pre-dicts that any small circular molecules presentwill be due to excision of the tandemly duplicat-ed sequences. Indeed, this probably leads'to lossof duplicated sequences during the reversal ofamplification that occurs during growth in drug-free medium. In model 2 the excised circular

resistance determinants are intermediates in theduplication process.

Amplification (or "transition") of the drugresistance (r-determinant) region of plasmidNR1 (Fig. 16) in Proteus mirabilis (157, 292, 321)requires the directly repeated copies of the 760-bp insertion sequence IS] which flank the resist-ance genes to provide homology for the initialduplication to take place (50, 176, 303). Thescheme proposed for transition was essentiallythe same as model 2 (Fig. 15), with the addedpossibility that independent r-determinantsmight increase their copy number by replicatingautonomously (176, 293, 303). However, it nowseems unlikely that r-determinant monomerscan replicate autonomously, at least in E. coli(52). A proposal that transition occurred byamplification of the entire plasmid populationrather than by selective outgrowth of a few cellscarrying preamplified plasmids (294) needs to besubstantiated.A more complicated type of transition was

reported for plasmid NR84, which is related toNR1 (321). Several different regions of the r-determinant were amplified in different experi-ments (321). Perhaps insertion sequences IS]and ISIO transposed to different sites in theplasmid before amplification to provide alterna-tive direct repeats to the IS1 elements that flankthe r-determinant.

6

Excision

IDimer

If

jOligomer

Q.

| Replication

QQ

FIG. 15. Model for formation of tandem duplica-tions by excision and reinsertion of the resistancegene. The open boxes represent the direct repeatsequences that flank amplifiable resistance genes. Theantibiotic resistance gene is indicated by the thickerline. Recombination events are indicated by x.

VOL. 47, 1983


http://mm

br.asm.org/

Dow

nloaded from


394 FOSTER

iDmI2,Minor Tramnoson .u-..*, I a -I.T,r

r-determinant_,_ Ha Sm Su Cm L 2

Tn9 Cm

FIG. 16. Structure of amplifiable drug resistancedeterminants. The amplifiable Tc resistance determi-nant of transposon Tnl721 is shown in the upper

diagram. The filled-in arrows represent the 38-bprepeat sequences which mark the termini of the minorand major transposons of Tn1721. The lower lineshows the position of the 1,200-bp direct duplication ofpart of the minor transposon which provides homologyfor the initiation of tandem duplications. The amplifia-ble drug resistance r-determinant region of the Rplasmid NR1 and the related transposon Tn9 areshown in the lower diagrams. The open boxes repre-sent the directly repeated IS) sequences which flankthe resistance genes. These elements confer resistanceto Hg, Cm, Sm, and Su.

Transposon Tn9 is flanked by direct repeats ofIS) (Fig. 16), and appears to be derived from aplasmid related to NR1 by deletion of a substan-tial portion of the r-determinant (178). Transpo-sons with similar structures have been isolatedfrom the r-determinant of R100, which is id'enti-cal to NRI (178). Unlike the r-determinant,these small ISI-flanked transposons can be am-plified in E. coli (51, 262). Kanamycin resistancetransposon Tn2680 from plasmid Rtsl is flankedby 800-bp direct repeats and is also amplifiable(179).The Tcr gene contained within transposons

Tnl721 and Tnl771 can be amplified by tandemduplication in E. coli (327-329, 421). Part of theregion of the transposon which encodes transpo-sition functions is repeated on the other side oftet to provide a 1,200-bp direct sequence dupli-cation (327; Fig. 16). The amplified forms (carry-ing up to nine copies of tet) were only found inRec+ strains but could be transferred to andstably maintained in a RecA- mutant (421). Thisdemonstrates that homologous recombination isrequired for both amplification and the reductionof duplications.The tet gene of Streptococcus faecalis plas-

mid pAMal is the only reported example of anamplifiable determinant in gram-positive bacte-ria (66, 431). In this case the presence of directlyrepeated sequences flanking the tet genes havebeen inferred but not demonstrated directly.

Acquisition of Multiple Copies of a TransposonCultivation of H. influenzae cells harboring a

plasmid with a Tcr determinant located on atransposon similar to TnJO in subinhibitory con-centrations of Tc resulted in the emergence ofvariants harboring plasmids with two, three, orfour copies of the transposon (188). A similarprocess of duplication by intramolecular trans-position of Tn2660 on R6K occurred when mu-tants expressing elevated levels of ampicillinresistance were selected (170). Intramoleculartransposition was at least 10-fold more frequentthan intermolecular transposition. Thus, thetransposition immunity mechanism of TnA-liketransposons only seems to apply to intermolecu-lar transposition into a replicon already harbor-ing that element and has little effect on intramo-lecular transposition (170, 410).

Increased Plasmid Copy NumberPlasmid JR66 specifies an aminoglycoside

phosphotransferase [APH(3')]-Il enzyme whichmodifies neomycin and kanamycin (Table 4). Ithas little activity on amikacin, and E. coli cellscarrying JR66 are phenotypically sensitive to thedrug. Amikacin-resistant mutants synthesizingincreased levels of enzyme which had a sub-strate proffle indistinguishable from the parentalenzyme provided sufficient amikacin-modifyingactivity to confer resistance to the drug (28). Theincreased enzyme activity was caused by a gene-dose effect due to an increase in plasmid copynumber, which was probably the effect of achromosomal mutation (28). Mutations in plas-mid-located replication control genes, whichcause an increase in plasmid copy number, alsoconferred an increased level of drug resistancethrough a gene-dose effect (277, 381, 399).

Increased Transcription of Resitance GenesA possible mechanism for increasing the

expression of an antibiotic resistance gene ismutation in the promoter, which increases theefficiency of transcription. Such mutations mayhave occurred in the TEM t-lactamase genepromoter and could be responsible for TEMgenes isolated from different host species beingtranscribed at different rates in E. coli (436).

Transcription could also be increased by dele-tions which link the structural gene to a moreefficient promoter or by insertion of a promoter-carrying insertion sequence close to the gene.

ORIGINS AND EVOLUTION OFANTIBIOTIC RESISTANCE DETERMINANTS

It has been thought for some time that R-plasmid-mediated antibiotic resistance determi-nants may have evolved before the antibiotic era

MICROBIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from



of the past 40 years in species of bacteria otherthan the pathogens and commensals, in whichresistance is often encountered today (108). Bac-teria in the genus Streptomyces have mecha-nisms to protect their potentially susceptibletargets from the antibiotics which they produce.These genes may have been transferred to otherbacteria as plasmid- or transposon-encoded drugresistance or both. Also, there is evidence thatsoil bacteria commonly harbor antibiotic resist-ance plasmids (26, 98, 108, 299, 337), presum-ably to help them combat antibiotics released intheir environment by competitors. These genescould have transferred to clinically importantbacteria either directly or via a number of inter-mediate hosts and plasmid vectors.When a gene is transferred into a new host,

changes may be selected to allow the proteinspecified to survive protease activity or becausemore efficient expression or secretion is re-quired. Genetic drift will occur if amino acidsubstitutions can be tolerated without alteringcatalytic properties. Changes in the third "wob-ble" base of the triplet codons will also occur tochange the base sequence of the gene. Semisyn-thetic antibiotics designed to evade existingresistance mechanisms impose additional selec-tive pressure which may lead to variants of theresistance determinants being selected. Suchchanges have been selected in the laboratory(147, 291).

Close relationships between resistance deter-minants can be recognized by comparative tech-niques such as immunological cross-reactionand isoelectric focusing of the resistance geneprotein products and by DNA-DNA hybridiza-tion. When substantial variation has occurred,homology may only be recognized by amino acidor DNA sequence analysis. When the nucleotidesequences of resistance genes have been com-pared, it has been possible to demonstrate evo-lutionary relatedness even though the DNAfailed to hybridize in Southern blot experiments(166, 173, 260). The Southern blotting techniquewill not detect <50% base sequence homology.The definition of homology groups will alsodepend on the stringency of the conditions usedin the hybridization reaction. Being assigned todifferent homology groups does not necessarilyimply that the determinants have evolved inde-pendently. However, if a small degree of homol-ogy confined to the active site of an enzyme isfound, it could be argued that this is due toconvergent evolution. Convergence would cer-tainly be suspected where no homology wasdetectable (8, 189).

Origin of MLS Resitance DeterminantsThe streptomycetes that synthesize MLS anti-

biotics (including S. erythreus, the producer of

erythromycin) protect their ribosomes from theinhibitory effects of the drugs by N6, N6-dimeth-ylation of adenine residues in 23S rRNA in thesame way as MLS-resistant staphylococci andstreptococci (125, 139, 386). The adenine meth-ylase of S. erythreus was shown to be a determi-nant of MLS resistance when it was transferredinto S. lividans (386, 387). DNA sequence analy-sis of Emr determinants representing the twohomology groups found in staphylococci andstreptococci revealed significant homology de-spite their failure to hybridize in Southern blotexperiments (173). It will be very interesting todiscover whether such homology exists betweenthe staphylococcal/streptococcal determinantsand that of S. erythreus to show whether the twogroups are evolutionarily related.

It is appropriate to mention here that strepto-mycetes which produce thiostreptone and relat-ed antibiotics protect their ribosomes by 2-0-pentose methylation of an adenosine residue in23S rRNA (78, 385). However, as yet thesedeterminants have not been found in other bac-teria.

Orgins of Aminoglycoside-Modifying EnzymesTwo mechanisms of resistance to autotoxic

aminoglycoside antibiotics occur in producingbacteria. (i) Ribosomal resistance has beenfound in the gentamicin producer Micromono-spora purpurea (297) and in the istamycin-pro-ducing Streptomyces tenjimaniensis (431). Inboth cases the spectrum of resistance conferredto other aminoglycosides was unique and drug-modifying activity was not detected in extracts.(ii) Resistance which is probably due to drugmodification occurs in streptomycetes that pro-duce kanamycin [AAC(6)?; 175], neomycin[APH(3'); 73, 175], streptomycin [APH(6); 175,248, 362], hygromycin [APH(4)?; 221], and bu-tirosin [APH(3'); 75]. In addition, the producersof the non-aminoglycosides viomycin and ca-preomycin protect themselves by drug phos-phorylation (350).Ribosomes from the streptomycetes are sensi-

tive to the aminoglycoside antibiotic produced,even during the phase of antibiotic synthesis (45,46, 175, 361, 362, 363). No active drug is detect-able in the cytoplasm during the phase whenantibiotic is accumulating in the medium (45).These organisms synthesize phosphotransfer-ases and acetyltransferases, which can inacti-vate the drugs in vitro (175, 207, 221, 298, 362,386, 407). In some cases the enzyme is notproduced during the exponential phase ofgrowth when the bacteria are phenotypicallysensitive to the antibiotic (298, 362). Evidencethat these drug-modifying enzymes can act asresistance determinants comes from transfer ex-periments which result in resistance being con-

VOL. 47, 1983


http://mm

br.asm.org/

Dow

nloaded from


396 FOSTER

ferred in another host (386, 387). This has beendemonstrated for S. fradiae APH(3') andAAC(6) (which were required together to conferhigh-level resistance to neomycin; 386, 387) andB. circulans APH(3') (75).An attractive hypothesis for the role of the

phosphotransferases in Streptomyces spp. isthat they are involved in the biosynthesis andexcretion of the antibiotic, as well as in deter-mining resistance to exogenous drug (207, 405).The penultimate biosynthetic precursor of ami-noglycosides may be an inactive phosphate de-rivative which is excreted from the cell by anefflux mechanism coupled to phosphatase activi-ty (237, 405). This is supported by the observa-tions mentioned in the preceding paragraph andby the finding that the expression of APH(3') inS. fradiae is coregulated with the level of neo-mycin synthesis and resistance (207).

It thus seems plausible that some R-plasmid-mediated aminoglycoside-modifying enzymesare derived from the enzyme determinants inantibiotic-producing bacteria. The R-plasmid-specified APH(3') of neomycin- and kanamycin-resistant bacteria (Table 4; Fig. 4 and 5) may berelated to the S. fradiae enzyme, the APH(6) ofstreptomycin-resistant staphylococci (Table 4;Fig. 7) may be derived from the S. griseus and S.bikiniensis determinants, and the APH(4) ofhygromycin-resistant E. coli (Table 4) (R. Rao,personal communication) may be related to theS. hygroscopicus enzyme (221). It is also possi-ble that some acetyltransferases evolved in thisway. However, there have been no reports ofadenyltransferases in antibiotic-producing bac-tena.

Early attempts to substantiate this hypothesisby DNA hybridization and immunological cross-reaction failed to show any homology among theAPH(3') determinants of gram-negative bacte-ria, staphylococci, B. circulans, and S. fradiae(73). However, recent DNA sequence analysishas revealed sufficient homology between the S.fradiae and E. coli determinants to suggest thatthey have a common origin (J. Davies, personalcommunication).

Origin of Tetracycline ResistanceThe organisms that produce Tc (Streptomyces

rimosus and S. aureofaciens) may have an in-ducible mechanism for protecting their ribo-somes from the inhibitory effects of Tc. Ribo-somes isolated from cells not engaged in Tcbiosynthesis are more sensitive to inhibition byTc in vitro than those isolated from cells activelysynthesizing the antibiotic (95, 264). Further-more, the level of ribosomal Tc resistance invitro seemed to be correlated with the level ofdrug production (95, 264). It is conceivable thatanother component of Tc resistance in Strepto-

myces spp. could involve efflux of the drug. Thiswould be analogous to the plasmid-specified Tcresistance mechanism. This could be examinedby cloning the streptomycete Tcr determi-nant(s), by elucidating the resistance mecha-nisms, and by comparing the DNA sequences.

Origins of Chloramphenicol ResceRibosomes isolated from Streptomyces vene-

zuelae (the producer of Cm) were found to besensitive to inhibition by Cm in vitro, irrespec-tive of whether they were isolated from cellsactively engaged in producing the drug (238,405). Furthermore, Cm was not detected in thecytoplasm (238, 405). This implies that the drugis actively exported from the cell. Possibly thepenultimate step in Cm biosynthesis is an inac-tive form which is activated during the exportprocess (405). In addition, the Cmr determinantcan be induced in non-antibiotic-producing cells.It seems possible that the inducible nonenzymat-ic Cmr determinants encoded by some Cmr Rplasmids (128, 271) could have originated here.The common R-plasmid-encoded Cm-inacti-

vating enzyme CAT does not occur in S. vene-zuelae, although it is specified by other strepto-mycetes (341). There is no evidence that thismechanism evolved in the drug-producing orga-nism. R-plasmid-specified CAT may haveevolved from an acetyltransferase involved inactivating sugars or amino acids for catabolicpathways.Proteus mirabilis is phenotypically sensitive

to Cm but it produces a small amount of CAT.Cm-resistant variants can be selected whichexpress an elevated level of a CAT enzymewhich is closely related to the type I enzymecommonly specified by enteric bacterial R plas-mids (336). It is possible that the type I enzymeis derived from the Proteus mirabilis determi-nant.

Origins of "-LactamaseThree evolutionary distinct types of P-lacta-

mase have been identified on the basis of bio-chemical criteria and amino acid and DNA se-quence comparisons (8, 24, 189). (i) Class Aenzymes are penicillinases which are primarilyassociated with plasmid-specified 3-lactamresistance. (ii) For class B, the only example isthe Bacillus cereus type II enzyme. (iii) Class Cincludes the chromosomally encoded cephalo-sporinases of enteric bacteria and P. aeruginosa.The R-plasmid-specified TEM P-lactamase of

gram-negative bacteria and the plasmid-speci-fied penicillinases of S. aureus have significantamino acid sequence homology and have clearlyevolved from a common ancestor (8). They maybe derived from the D-alanine carboxypeptidaseas originally postulated by Tipper and Stro-

MICROBIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from



minger (390), since amino acid homologies existaround the serine residues in their active centers(412). However, it could be argued that thislimited homology is due to convergent evolution(8).Comparison of the DNA and amino acid se-

quences of the TEM and E. coli ampC determi-nants have shown that the enzymes are notrelated in evolutionary terms (9, 189, 371), al-though they have similar catalytic mechanismsinvolving a serine residue in the active site (112,203). It will be interesting to see whether theplasmid-borne cephalosporinases CEP-1 andCEP-2 (Table 1) have homology with the chro-mosomally specified enzymes of gram-negativebacteria.

Origins of Sulfonamide and TrimethoprimResistance

The DHFR and DHPS enzymes that conferresistance to the synthetic drugs Tp and Su were

probably not widespread in soil or clinical bacte-ria before the antibiotic era. It is likely that thesedeterminants became associated with R plas-mids after the introduction of the drugs intoclinical practice. However, their origins remaina complete mystery, despite the existence (in thecase of Tp) of naturally resistant enzymes speci-fied by mammalian cells and bacteriophages(443). It is possible that mutations in chromo-somal genes encoding drug-sensitive enzymescould have resulted in drug resistance, althoughit is difficult to obtain high resistance levels inthe laboratory. Alternatively, DHFR could haveevolved from a reductase with another sub-strate. In the case of Tpr DHFR this eventoccurred at least twice because there is nodetectable homology between the genes encod-ing the type I and type II enzymes (443).

CONCLUDING REMARKS AND FUrLUREPROSPECTS

One of the most important recent develop-ments in the chemotherapy of bacterial infec-tions is the introduction of antibiotic derivativeswhich evade existing (mainly plasmid-specified)resistance mechanisms. Examples of these are,B-lactamase inhibitors, 3-lactamase-resistantcephalosporins, and new aminoglycosides. Theefficacy of these drugs may prove to be short-lived because existing resistance mechanismsmight adapt to deal with them, or new mecha-nisms of resistance might evolve.

Methicillin is a semisynthetic derivative ofpenicillin which is not hydrolyzed by staphylo-coccal penicillinases. It was introduced 20 yearsago to combat penicillin-resistant S. aureus

strains which produced penicillinase. Methicil-lin-resistant strains which did not inactivate thedrug soon emerged. The mechanism of methicil-

lin resistance was poorly understood until it wasrecently shown to be associated with PBP (tar-get) changes (155, 158). The response of S.aureus to methicillin may be repeated by otherbacteria becoming resistant to new P-lactams.Indeed, strains of P. aeruginosa which are resis-tant to carbencillin and cefsulodin because ofmutations altering the affinity of PBPs havealready been encountered (80, 136, 269, 316).Cephradine resistance in S. aureus also involvesPBP changes (134). In P. aeruginosa the PBPchanges may occur in combination with muta-tions which reduce the permeability of the outermembrane to P-lactams (269, 316). It is possiblethat the incidence of these mechanisms of resist-ance will increase as the new P-lactams are usedmore. Another possibility is the spread of plas-mid-specified 1-lactamases which can hydrolyzecephalosporins (e.g., CEP-1 and CEP-2; Table1).New aminoglycoside antibiotics which inhibit

bacteria synthesizing the commonly encoun-tered plasmid-specified modifying enzymes arenow available. These include the 5-epi isomersof gentamicin and sisomicin, amikacin, andapramycin. Aminoglycoside resistance may alsobe controlled by inhibitors such as 7-hydroxyto-polone, which interferes with AAD(2") activity(4). However, it seems unlikely that aminoglyco-side resistance will be completely contained inthis way. The drugs have several sites present-ing potential targets for modifying enzymes. Itmay prove impossible to block or remove allpotential modification sites without detractingfrom antibiotic activity. Also, mechanisms havebeen described which indicate how bacteria canrespond to challenge by new drugs: (i) by acqui-sition of an enzyme which is capable of modify-ing the drug; (ii) by mutations which alter thesubstrate specificity of an enzyme [derivativesof APH(3') which rapidly modify and conferresistance to amikacin and other minor sub-strates can be selected (291)]; (iii) by increasedexpression of an enzyme with poor activityagainst the new drug, which could occur byincreasing the rate of transcription or by a gene-dose effect as described above; (iv) the finding ofmutations which impair the critical EDPI phaseof drug transport by interfering with membraneenergization in clinical strains of P. aeruginosaand S. aureus (34, 267); (v) by alteration of theribosome target site. In clinical strains thismechanism has so far been confined to strepto-mycin resistance (213, 241, 3%, 442). It is possi-ble that multiple (stepwise) ribosomal changesconferring resistance to deoxystreptamine-con-taining aminoglycosides will emerge, possibly incombination with mutations affecting drug trans-port.The diversity of the mechanisms by which

VOL. 47, 1983


http://mm

br.asm.org/

Dow

nloaded from


398 FOSTER

bacteria have adapted to antimicrobial drugs andthe rapidity with which the resistance has spreadare impressive. It seems likely that bacteria willevolve resistance to most, if not all, antibioticswith which they are challenged. Coordinatedprograms of antibiotic use in clinical and veteri-nary practice will help to increase the usefulnessof drugs by reducing the rate with which resist-ance emerges.

ACKNOWLEDGMENTS

I thank the following for their invaluable help in communi-cating information prior to publication: A. Medeiros, G.Jacoby, M. Matthew, N. L. Brown, S. Silver, B. Weisblum,D. Dubnau, I. Chopra, J. Grinsted, P. Bennett, L. Elwell,W. V. Shaw, A. Philippon, 0. Skold, S. Levy, J. Davies, E.Cundliffe, S. Lerner, D. M. Livermore, R. Schmitt, W. Hillen,J. M. Ortiz, R. Rao, R. Kaback, and M. Miller. A special wordof thanks goes to Niamh Ni Bhriain, Bill Shaw, Simon Silver,Ian Chopra, and Julian Davies for suggestions and for criti-cisms of the manuscript.

LITERATURE CITED

1. Ahmad, M. H., A. Rechenmacher, and A. Bock. 1980.Interaction between aminoglycoside uptake and ribo-somal resistance mutations. Antimicrob. Agents Che-mother. 18:798-806.

2. Albert, A. 1973. Arsenicals, antimonials and mercurials,p. 392-397. In Selective toxicity, 5th ed. Chapman andHall, London.

3. Alen, N. E. 1977. Macrolide resistance in Staphylococ-cus aureus: inducers of macrolide resistance. Antimi-crob. Agents Chemother. 11:669-674.

4. Allen, N. E., W. E. Alborn, J. N. Hobbs, and H. A. Kirst.1982. 7-Hydroxytropolone: an inhibitor of aminoglyco-side-2'-adenylyltransferase. Antimicrob. Agents Che-mother. 22:824-831.

5. Altenbuchner, J., K. Schmid, and R. Schmitt. 1983.Tnl721-encoded tetracycline resistance: mapping ofstructural and regulatory genes mediating resistance. J.Bacteriol. 153:116-123.

6. Alton, N. K., and D.Vapnek. 1979. Nucleotide sequence

analysis of the chloramphenicol resistance transposonTn9. Nature (London) 282:864-869.

7. Ambler, R. P. 1975. The amino acid sequence of theStaphylococcus aureus penicillinase. Biochem. J.151:197-218.

8. Ambler, R. P. 1980. The structure of 3-lactamases.Phil. Trans. R. Soc. London Ser. B 289:321-331.

9. Ambler, R. P., and G. K. Scott. 1978. Partial amino acidsequence of penicillinase coded by Escherichia coliplasmid R6K. Proc. Natl. Acad. Sci. U.S.A. 75:3732-3736.

10. Amyes, S. G. B., and J. T. Smith. 1974. R-factor trimeth-oprim resistance mechanism: an insusceptible target site.Biochem. Biophys. Res. Commun. 58:412-418.

11. Amyes, S. G. B., and J. T. Smith. 1976. The purificationand properties of the trimethoprim-resistant dihydrofo-late reductase mediated by the R-factor, R388. Eur. J.Biochem. 61:597-603.

12. Anderson, R. P., and J. R. Roth. 1977. Tandem geneticduplications in phage and bacteria. Annu. Rev. Micro-biol. 31:473-505.

13. Annear, D. I., B. J. Mee, and M. Bailey. 1976. Instabilityand linkage of silver resistance, lactose fermentation andcolony structure in Enterobacter cloacae from burnwounds. J. Clin. Pathol. 29:441-443.

14. Asheshov, E. H. 1975. The genetics of tetracycline resist-ance in Staphylococcus aureus. J. Gen. Microbiol.88:132-140.

15. Ball, P. R, I. Chopra, and S. J. Eccles. 1977. Accumula-

tion of tetracyclines by Escherichia coli K-12. Biochem.Biophys. Res. Commun. 77:1500-1507.

16. Ball, P. R., S. W. Shales, and I. Chopra. 1980. Plasmid-mediated tetracycline resistance involves increased ef-flux of the antibiotic. Biochem. Biophys. Res. Commun.93:74-81.

17. Barth, P. T., N. Datta, R. W. Hedges, and N. J. Grlnter.1976. Transposition of a deoxyribonucleic acid sequenceencoding trimethoprim and streptomycin resistance fromR483 to other replicons J. Bacteriol. 125:800-810.

18. Beck, C. F., R. Mutzel, J. Barbe, and W. Muller. 1982. Amultifunctional gene (tetR) controls TnlO-encoded tetra-cycline resistance. J. Bacteriol. 150:633-642.

19. BenvenLste, R., and J. Davies. 1971. Enzymatic acetyla-tion of aminoglycoside antibiotics by Escherichia colicarrying an R factor. Biochemistry 10:1787-17%.

20. Benveniste, R., and J. Davies. 1971. R-factor mediatedgentamicin resistance: a new enzyme which modifiesaminoglycoside antibiotics. FEBS Lett. 14:293-296.

21. Benveniste, R., and J. Davies. 1973. Aminoglycosideantibiotic-inactivating enzymes in actinomycetes similarto those present in clinical isolates of antibiotic-resistantbacteria. Proc. Natl. Acad. Sci. U.S.A. 70:2276-2280.

22. Beppu, T., and K. Arima. 1969. Induction by mercuricion of extensive degradation of cellular ribonucleic acidin Escherichia coli. J. Bacteriol. 98:888-897.

23. Berg, D. E., J. Davies, B. Allet, and J. D. Rochatx. 1975.Transposition of R factor genes to bacteriophage X. Proc.Natl. Acad. Sci. U.S.A. 72:3628-3632.

24. Bergstrom, S., 0. Olson, and S. Normark. 1982. Com-mon evolutionary origin of chromosomal beta-lactamasegenes in enterobacteria. J. Bacteriol. 150:528-534.

25. Beddcombe, S., M. Haas, J. DavIes, G. H. Miller, D. F.Rane, and P. J. L. Daniels. 1976. Enzymatic modificationof aminoglycoside antibiotics: a new 3-N-acetylatingenzyme from a Pseudomonas aeruginosa isolate. Anti-microb. Agents Chemother. 9:951-955.

26. Bingham, A. H. A., C. J. Bruton, and T. Atkinson. 1979.Isolation and partial characterization of four plasmidsfrom antibiotic-resistant thermophilic bacilli. J. Gen-.Microbiol. 114:401-408.

27. Bobrowski, M. M., M. Matthew, P. T. Barth, N. Datta,N. J. Grinter, A. E. Jacob, P. Kontomichalou, J. W. Dale,and J. T. Smith. 1976. Plasmid-determined -lactamaseindistinguishable from the chromosomal 3-lactamase ofEscherichia coli. J. Bacteriol. 125:149-157.

28. Bongaerts, G. P. A., and G. H. P. Kaptijn. 1981. Amino-glycoside phosphotransferase II-mediated amikacinresistance in Escherichia coli. Antimicrob. Agents Che-mother. 20:344-350.

29. Bragg, P. D., and D. J. Ralnnie. 1974. The effect of silverions on the respiratory chain of Escherichia coli. Can. J.Microbiol. 20:883-889.

30. Britz, M. L., and R. G. Wilnso. 1978. Chlorampheni-col acetyltransferase of Bacteriodesfragilis. Antimicrob.Agents Chemother. 14:105-111.

31. Broda, P. 1979. Plasmids. W. H. Freeman & Co.,London.

32. Brown, G. M. 1962. The biosynthesis of folic acid. II.Inhibition of sulfonamides. J. Biol. Chem. 237:536-540.

33. Bryan, L. E. 1980. Mechanisms of plasmid mediated drugresistance, p. 57-81. In C. Stuttard and K. R. Rozee(ed.), Plasmids and transposons. Environmental effectsand maintenance mechanisms. Academic Press, Inc.,New York.

34. Bryan, L. E., R. Haraphongse, and H. M. van den Elzen.1976. Gentamicin resistance in clinical isolates of Pseu-domonas aeruginosa associated with decreased gentami-cin accumulation and no detectable enzymatic modifica-tion. J. Antibiot. 29:743-753.

35. Bryan, L. E., and S. Kwan. 1981. Mechanisms of amino-glycoside resistance of anaerobic bacteria and facultativebacteria grown anaerobically. J. Antimicrob. Che-mother.8(Suppl. D):1-8.

36. Bryan, L. E., T. Nicas, B. W. Holloway, and C.

MICROBIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from



Crowther. 1980. Aminoglycoside-resistant mutation ofPseudomonas aeruginosa defective in cytochrome C552and nitrate reductase. Antimicrob. Agents Chemother.17:71-79.

37. Bryan, L. E., and H. M. van den Elzen. 1976. Streptomy-cin accumulation in susceptible and resistant strains ofEscherichia coli and Pseudomonas aeruginosa. Antimi-crob. Agents Chemother. 9:928-938.

38. Bryan, L. E., and H. M. van den Elzen. 1977. Effects ofmembrane energy mutations and cations on streptomy-cin and gentamicin accumulation by bacteria: a model forentry of streptomycin and gentamicin in susceptible andresistant bacteria. Antimicrob. Agents Chemother.12:163-177.

39. rska, M., R. Benveniste, J. Davies, P. J. L. Daniels,and J. Weinstein. 1972. Gentamycin resistance in strainsof Pseudomonas aeruginosa mediated by enzymatic N-acetylation of the deoxystreptamine moiety. Biochemis-try 11:761-766.

40. Buckel, P., A. Buchberger, A. Bock, and H. G. Wittman.1977. Alteration of ribosomal protein L6 in mutants ofEscherichia coli resistant to gentamicin. Mol. Gen. Gen-et. 158:47-54.

41. Bmchall, J. J., and G. H. Hltchings. 1965. Inhibitorbinding analysis of dihydrofolate reductases from vari-ous species. Mol. Pharmacol. 1:126-136.

42. Burdett, V., J. Inamlne, and S. Rajagopalau. 1982. Het-erogeneity of tetracycline resistance determinants inStreptococcus. J. Bacteriol. 149:995-1004.

43. Cabelo, F., K. N. Tlmnis, and S. N. Cohen. 1976.Replication control in a composite plasmid constructedby in vitro linkage of two distinct replicons. Nature(London) 259:285-290.

44. Campbell, B. D., and R. J. Kadner. 1980. Relation ofaerobiosis and ionic strength to the uptake of dihydro-streptomycin in Escherichia coli. Biochim. Biophys.Acta593:1-10.

45. Cela, R., and L. C. Vining. 1974. Action of streptomycinon the growth of Streptomyces griseus. Can. J. Micro-biol. 20:1591-1597.

46. Cela, R., and L. C. Vining. 1974. Resistance to strepto-mycin in a producing strain of Streptomyces griseus.Can. J. Microbiol. 21:463-472.

47. Cerna, J., I. Rychllk, and P. Pulkarbek. 1969. The effectof antibiotics on the coded binding of peptidyl-tRNA tothe ribosome and on the transfer of the peptidyl residueto puromycin. Eur. J. Biochem. 9:27-35.

48. Chabbert, Y. 1956. Antagonisme in vitro entrel'erythro-mycine et la spririmycine. Ann. Inst. Pasteur (Paris)90:787-790.

49. Chabbert, Y. A., and M. R. Scavizz. 1976. Chelocardin-inducible resistance in Escherichia coli bearing R plas-mids. Antimicrob. Agents Chemother. 9:36-41.

50. Chandler, M., B. Allet, E. Gallay, E. Boy dela Tour, andL. Caro. 1977. Involvement of IS) in the dissociation ofthe r-determinant and RTF components of the plasmidR100.1. Mol. Gen. Genet. 153:289-295.

51. Chandler, M., E. Boy de la Tour, D. Willems, and L.Caro. 1979. Some properties of the chloramphenicolresistance transposon Tn9. Mol. Gen. Genet. 176:221-231.

52. Chandler, M., L. Silver, J. Frey, and L. Caro. 1977.Suppression of an Escherichia coli dnaA mutation by theintegrated R factor R100.1: generation of small plasmidsafter integration. J Bacteriol. 130.303-311.

53. Chevareau, M., P. J. L. Daniels, J. Davies, and F. LeGoffic. 1974. Aminoglycoside resistance in bacteria me-diated by gentamycin acetyltransferase II, an enzymemodifying the 2'-amino group of aminoglycoside antibi-otics. Biochemistry 13:598-603.

54. Chiang, S. J., and R. C. Clowes. 1980. Intramoleculartransposition and inversion in plasmid R6K. J. Bacteriol.142:668-682.

55. Chopra, I. 1970. Decreased uptake of cadmium by aresistant strain of Staphylococcus aureus. J. Gen. Micro-

biol. 63:265-267.56. Chopra, I. 1975. Mechanism of plasmid-mediated resist-

ance to cadmium in Staphylococcus aureus. Antimicrob.Agents Chemother. 7:8-14.

57. Chopra, I. 1976. Mechanism of resistance to fusidic acidin Staphylococcus aureus. J. Gen. Microbiol. 96:229-238.

58. Chopra, I., and P. R. Ball. 1982. Transport of antibioticsinto bacteria. Adv. Microb. Physiol. 23:183-240.

59. Chopra, I., P. R. Ball, S. J. Eccles, and S. W. Shales.1981. TnlO encoded proteins that mediate tetracyclineresistance in E. coli, p. 592. InS. B. Levy, R. C. Clowes,and E. L. Koenig (ed.), Molecular biology, pathogenicityand ecology of bacterial plasmids. Plenum Press, NewYork.

60. Chopra, I., and S. J. Eccles. 1978. Diffusion of tetracy-cline across the outer membrane of Escherichia coli K-12: involvement of protein Ia. Biochem. Biophys. Res.Commun. 83:550-557.

61. Chopra, I., and T. G. B. Howe. 1978. Bacterial resistanceto the tetracyclines. Microbiol. Rev. 42:707-724.

62. Chopra, I., R. W. Lacey, and J. Connolly. 1974. Bio-chemical and genetic basis of tetracycline resistance inStaphylococcus aureus. Antimicrob. Agents Chemother.6:397-404.

63. Chopra, I., S. Shales, and P. Ball. 1982. Tetracyclineresistance determinants from groups A to D vary in theirability to confer decreased accumulation of tetracyclinederivatives in Escherichia coli. J. Gen. Microbiol.128:689-692.

64. Clark, D. L., A. A. Weiss, and S. Silver. 1977. Mercuryand organomercurial resistances determined by plasmidsin Pseudomonas. J. Bacteriol. 132:186-196.

65. Ciewell, D. B. 1981. Plasmids, drug resistance, and genetransfer in the genus Streptococcus. Microbiol. Rev.45:409-436.

66. Clewell, D. B., Y. Yagi, and B. Bauer. 1975. Plasmid-determined tetracycline resistance in Streptococcusfae-calis: evidence for gene amplification during growth inthe presence of tetracycline. Proc. Natl. Acad. Sci.U.S.A. 72:1720-1724.

67. Cohen, S. A., and R. F. Pratt. 1980. Inactivation ofBacillus cereus 3-lactamase I by 6-bromopenicillanicacid: mechanism. Biochemistry 19:3996-4003.

68. Coleman, D. C., I. Chopra, S. W. Shales, T. G. B. Howe,and T. J. Foster. 1983. Analysis of tetracycline resistanceencoded by transposon TnlO: deletion mapping of tetra-cycline-sensitive point mutations and identification oftwo structural genes. J. Bacteriol. 153:921-929.

69. Coleman, D. C., and T. J. Foster. 1981. Analysis of thereduction in expression of tetracycline resistance deter-mined by transposon TnJO in the multicopy state. Mol.Gen. Genet. 182:171-177.

70. Coombe, R. G., and A. M. George. 1981. New plasmid-mediated aminoglycoside adenyltransferase of broadsubstrate range that adenylates amikacin. Antimicrob.Agents Chemother. 20:75-80.

71. Costerton, J. W., and K. J. Cheng. 1975. The role of thebacterial cell envelope in antibiotic resistance. J. Antimi-crob. Chemother. 1:363-377.

72. Courvaln, P., and J. Davies. 1977. Plasmid-mediatedaminoglycoside phosphotransferase of broad substraterange that phosphorylates amikacin. Antimicrob. AgentsChemother. 11:619-624.

73. Courvalln, P., M. Friandt, and J. Davies. 1978. DNArelationships between genes coding for aminoglycoside-modifying enzymes from antibiotic-producing bacteriaand R plasmids, p. 262-266. In D. Schlessinger (ed.),Microbiology-1978. American Society for Microbiolo-gy, Washington, D.C.

74. Courvalln, P. M., W. V. Shaw, and A. E. Jacob. 1978.Plasmid-mediated mechanisms of resistance toaminogly-coside-aminocyclitol antibiotics and to chloramphenicolin group D streptococci. Antimicrob. Agents Chemother.13:716-725.

VOL. 47, 1983


http://mm

br.asm.org/

Dow

nloaded from


400 FOSTER

75. Courvaln, P., B. Welablum, and J. Davies. 1977. Amino-glycoside-modifying enzymes of an antibiotic producingbacterium acts as a determinant of antibiotic resistancein Escherichia coli. Proc. Natl. Acad. Sci. U.S.A.74:999-1003.

76. Crowlesmkth, I., and T. G. B. Howe. 1980. Characteriza-tion of P-lactamase-deficient (bla) mutants of the Rplasmid RI in Escherichia coli K-12 and comparison withsimilar mutants of RP1. Antimicrob. Agents Chemother.18:667-674.

77. Crowlemith, I., and T. G. B. Howe. 1980. Quantitativecorrelation between penicillin resistance and -lacta-mase activity specified by the R plasmids Rl, Rl bla.45,and RP1 in Escherichia coli K-12. Antimicrob. AgentsChemother. 18:675-679.

78. Condlfe, E., and J. Thompson. 1981. The mode ofaction of nosiheptide (multhiomycin) and the mechanismof resistance in the producing organism. J. Gen. Micro-biol. 126:185-192.

79. Curial, M. S., and S. B. Levy. 1982. Two complementa-tion groups mediate tetracycline resistance determinedby Tn1O. J. Bacteriol. 151:209-215.

80. Curtss, N. A. C., C. Brown, M. Boxall, and M. G.Boulton. 1978. Modified pertidoglycan transpeptidaseactivity in a carbenicillin-resistant mutant of Pseudomo-nas aeruginosa 18S. Antimicrob. Agents Chemother.14:246-251.

81. Curtis, N. A. C., D. Orr, M. G. Boulton, and G. W.Rosa. 1981. Penicillin binding proteins of Pseudomonasaeruginosa. Comparison of two strains differing in theirresistance to P-lactam antibiotics. J. Antimicrob. Che-mother. 7:127-136.

82. Curtiss, N. A. C., and M. H. RIchmond. 1974. Effect ofR-factor-mediated genes on some surface properties ofEscherichia coli. Antimicrob. Agents Chemother. 6:666-671.

83. Dak, J. W., and J. T. Smith. 1974. R-factor-mediated ,Blactamases that hydrolyze oxacillin: evidence for twodistinct groups. J. Bacteriol. 119:351-356.

84. Dale, R. M. K., and D. C. Ward. 1975. Mercuratedpolynucleotides: new probes for hybridization and selec-tive polymer fractionation. Biochemistry 14:2458-2469.

85. Damper, P. D., and W. Epstein. 1981. Role of membranepotential in bacterial resistance to aminoglycoside antibi-otics. Antimicrob. Agents Chemother. 20:803-808.

86. Daug-Van, A., G. Tiraby, J. F. Acar, W. V. Shaw, andD. H. Bonanchamd. 1978. Chloramphenicol resistance inStreptococcus pneumoniae: enzymatic acetylation andpossible plasmid linkage. Antimicrob. Agents Che-mother. 13:557-583.

87. Datta, N., S. Dacey, V. Hulges, S. Knight, H. Richards,G. Wllams, M. Casewel, and K. P. Shannon. 1980.Distribution of genes for trimethoprim and gentamicinresistance in bacteria and their plasmids in a generalhospital. J. Gen. Microbiol. 118:495-508.

88. Datta, N., and R. W. Hedges. 1972. Trimethoprim resist-ance conferred by W plasmids in Enterobacteriaceae. J.Gen. Microbiol. 72:349-355.

89. Datta, N., R. W. Hedges, D. Becker, and J. Davie. 1974.Plasmid-determined fusidic acid resistance in the Entero-bacteriaceae. J. Gen. Microbiol. 83:191-196.

90. Datta, N., and P. Kontomklcao. 1965. Penicillinasesynthesis controlled by infectious R-factors in Entero-bacteriaceae. Nature (London) 208:239-241.

91. Davies, J., and R. E. Benvenlste. 1974. Enzymes thatinactivate antibiotics in transit to their targets. Ann.N.Y. Acad. Sci. 235:130-136.

92. Davies, J., and S. O'Connor. 1978. Enzymatic modifica-tion of aminoglycoside antibiotics: 3-N-acetyltransferasewith broad substrate specificity that determines resist-ance to the novel aminoglycoside apramycin. Antimi-crob. Agents Chemother. 14:69-72.

93. Davies, J., ad D. I. Smith. 1978. Plasmid-determinedresistance to antimicrobial agents. Annu. Rev. Micro-biol. 32:469-518.

94. de Crombrugghe, B., I. Psan, W. V. Sbaw, and J. L.Rone. 1973. Stimulation by cyclic AMP and ppGpp ofchloramphenicol acetyltransferase synthesis. Nature(London) New Biol. 241:237-239.

95. Demaln, A. L. 1974. How do antibiotic-producing micro-organisms avoid suicide? Ann. N.Y. Acad. Sci. 235:601-612.

96. Dempsey, W. B., and N. S. Willetts. 1976. Plasmid co-integrates of prophage lambda and R factor R100. J.Bacteriol. 126:166-176.

97. Dcklie, P., L. E. Bryan, and M. A. Pckard. 1978. Effectof enzymatic adeylylation on dihydrostreptomycin accu-mulation in Escherichia coli carrying an R-factor: modelexplaining aminoglycoside resistance by inactivatingmechanisms. Antimicrob. Agents Chemother. 14:569-580.

98. Dochety, A., G. Grandi, R. Grandi, T. J. Gryczan, A. G.ShlvWmar, and D. Dubsau. 1981. Naturally occurringmacrolide-lincosamide-streptogramin B resistance in Ba-cillus lichenformis. J. Bacteriol. 145:129-137.

99. Dwrman, C. J., and T. J. Foster. 1982. Nonenzymaticchloramphenicol resistance determinants specified byplasmids R26 and R55-1 in Escherichia coli K-12 do notconfer high-level resistance to fluorinated analogs. Anti-microb. Agents Chemother. 22:912-914.

100. Dougherty, T. J., A. E. Koller, and A. Tom_as 1980.Penicillin binding proteins of penicillin-susceptible andintrinsically resistant Neisseria gonorrhoeae. Antimi-crob. Agents Chemother. 18:730-737.

101. Dowding, J. E. 1977. Mechanisms of gentamicin resist-ance in Staphylococcus aureus. Antimicrob. AgentsChemother. 11:47-50.

102. Dyke, K. H. G. 1979. f3-Lactamases of Staphylococcusaureus, p. 291-310. In J. M. T. Hamilton-Miller andJ. T. Smith (ed.), Beta-lactamases. Academic Press,London.

103. Dyke, K. H. G., M. T. Parker, and M. H. Rkhmond.1970. Penicillinase production and metal ion resistancein Staphylococcus aureus cultures isolated from hospitalpatients. J. Med. Microbiol 3:125-136.

104. Edlund, T., T. Grndtrom, and S. Nrmark. 1979.Isolation and characterization ofDNA repetitions carry-ing the chromosomal ,B-lactamase gene of Escherichiacoli K-12. Mol. Gen. Genet. 173:115-125.

105. Engberg, B., and K. Nordstm. 1975. Replication of R-factor RI in Escherichia coli K-12 at different growthrates. J. Bacteriol. 123:179-186.

106. El Solh, N., and S. D. Ehrlich. 1982. A small cadmiumresistance plasmid isolated from Staphylococcus aureus.Plasmid 7:77-84.

107. Erics, C. 1969. Resistance to acriflavine and cadmium,and changed phage reactions-markers of a new staphy-lococcal penicillinase plasmid. Acta Pathol. Microbiol.Scand. 76:333.

108. Falkow, S. 1975. Infectious multiple drug resistance.Pion Ltd., London.

109. Fayolle, F., G. Privatera, and M. Sebald. 1980. Tetracy-cline transport in Bacteroides fragilis. Antimicrob.Agents Chemother. 18:502-505.

110. Ferra, D., and S. B. Levy. 1980. Biochemical andimmunological characterization of an R plasmid-encodedprotein with properties resembling those of major cellu-lar outer membrane proteins. J. Bacteriol. 144:149-158.

111. Fisher, J., J. G. Belasco, S. Khosda, and J. R. Knowles.1980. 1-Lactamase proceeds via an acyl-enzyme inter-mediate. Interaction of the Escherichia coli RTEM en-zyme with cefoxitin. Biochemistry 19:2895-2901.

112. FIsher, J., R. L. Chars, S. M. Bradley, and J. R.Knowles. 1981. Inactivation of the RTEM 3-lactamasefrom Escherichia coli: interaction of penam sulphoneswith enzyme. Biochemistry 20:2726-2731.

113. FIher, J. F., and J. R. Knowles. 1980. The inactivationof P-lactamase by mechanism based reagents, p. 183-207. In M. Sandler (ed.), Enzyme inhibitors as drugs.Macmillan, London.

MICROBIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from



114. Fltton, J. E., L. C. Packman, S. Harford, Y. Zaldezaig,and W. V. Shaw. 1978. Plasmids and the evolution ofchloramphenicol resistance, p. 249-252. In D. Schles-singer (ed.), Microbiology-1978. American Society forMicrobiology, Washington, D.C.

115. Fltton, J. E., and W. V. Shaw. 1979. Comparison ofchloramphenicol acetyltransferase variants in staphylo-cocci. Biochem. J. 177:575-582.

116. Fling, M. E., ad L. P. Elwell. 1980. Protein expressionin Escherichia coli miniceils containing recombinantplasmids specifying trimethoprim-resistant dihydrofolatereductases. J. Bacteriol. 141:779-785.

117. Fling, M. E., L. Walton, and L. P. Elwell. 1982. Monitor-ing of plasmid-encoded, trimethoprim-resistant dihydro-folate reductase genes: detection of a new resistantenzyme. Antimicrob. Agents Chemother. 22:882-885.

118. Fogel, S., ad J. W. Welch. 1982. Tandem gene amplifi-cation mediates copper resistance in yeast. Proc. Natl.Acad. Sci. U.S.A. 79:5342-5346.

119. Foster, T. J., H. Nakabara, A. A. Weiss, and S. Silver.1979. Transposon A-generated mutations in the mercuricresistance genes of plasmid R100-1. J. Bacteriol.140:167-181.

120. Foster, T. J., and A. Walsh. 1974. Phenotypic character-ization of R-factor tetracycline resistance determinants.Genet. Res. 24:333-343.

121. Fox, B., and C. T. Wals. 1982. Mercuric reductase-purification and characerization of a transposon-encodedflavoprotein containing an oxidation-reduction-activedisulphide. J. Biol. Chem. 257:2498-2503.

122. Fox, C. L., and S. M. Modak. 1974. Mechanism of silversulfadiazine action on burn wound infections. Antimi-crob. Agents Chemother. 5:582-588.

123. Franke, A. E., and D. B. Clewell. 1981. Evidence for achrmosome-borne resistance transposon (Tn916) inStreptococcus faecalis that is capable of "conjugal"transfer in the absence of a conjugative plasmid. J.Bacteriol. 145:494-502.

124. Frankln, T. J. 1973. Antibiotic transport in bacteria.Crit. Rev. Microbiol. 2:253-272.

125. Fujisawa, Y., and B. Welsbbum. 1981. A family of r-determinants in Streptomyces spp. that specifies induc-ible resistance to macrolide, lincosamide, and strepto-gramin type B antibiotics. J. Bacteriol. 146:621-631.

126. Furth, A. J. 1975. Purification and properties of constitu-tive P-lactamase from Pseudomonas aeruginosa strainDalgleish. Biochim. Biophys. Acta 377:431-443.

127. Frukawa, K., and K. Tonomura. 1971. Enzyme systeminvolved in the decomposition of phenylmercuric ace-tate by mercury-resistant Pseudomonas. Agric. Biol.Chem. 35:604-610.

128. Gaffney, D. F., E. Cundliffe, and T. J. Foster. 1981.Chloramphenicol resistance that does not involve chlor-amphenicol acetyltransferase encoded by plasmids fromgram-negative bacteria. J. Gen. Microbiol. 125:113-121.

129. Gafney, D. F., T. J. Foster, and W. V. Shaw. 1978.Chloramphenicol acetyltransferases determined by Rplasmids from gram-negative bacteria. J. Gen. Micro-biol. 109:351-358.

130. Gale, E. F., E. Cundliffe, P. E. Reynols, M. H. Rkh.nond, and M. J. Waring. 1981. The molecular basis ofantibiotic action. Wiley Interscience, London.

131. Garda-Lobo, J. M., nd J. M. Ortlz. 1982. Tn2921, atrapsposon encoding fosfomycin resistance. J. Bacteriol.151:477-479.

132. Garrod, L. P. 1957. The erythromycin group of antibiot-ics. Br. Med. J. 2:57-63.

133. Gayda, R. C., J. H. Tanabe, K. M. Knigge, and A.Markovltz. 1979. Identification by deletion analysis of aninducible protein required for plasmid pSC101-mediatedtetracycline resistance. Plasmid 2:417-425.

134. Geogopapadakou, N. H., S. A. SmIt, and D. P. Bonner.1982. Penicillin-binding proteins in a Staphylococcusaureus strain resistant to specific ,-lactam antibiotics.Antimicrob. Agents Chemother. 22:172-175.

135. Godfrey, A. J., and L. E. Bryan. 1982. Mutation ofPseudomonas aeruginosa specifying reduced affinity forpenicillin G. Antimicrob. Agents Chemother. 21:216-223.

136. Godfrey, A. J., L. E. Bryan, and H. R. Rabin. 1981. ,B-Lactam-resistant Pseudomonas aeruginosa with modi-fied penicillin binding proteins emerging during cysticfibrosis treatment. Antimicrob. Agents Chemother.19:705-711.

137. Goldman, P. R., and D. B. Northrop. 1975. Preparationof stable gentamicin adenylyltransferase of high specificactivity. Biochem. Biophys. Res. Commun. 66:1408-1413.

138. Goldman, P. R. and D. B. Northrop. 1976. Purificationand spectrophotometric assay of neomycin phospho-transferase II. Biochem. Biophys. Res. Commun.69:230-236.

139. Graham, M. Y., and B. Weisblum. 1979. 23S ribosomalribonucleic acid of macrlide-producing streptomycetescontains methylated adenine. J. Bacteriol. 137:1464-1467.

140. Gryczan, T. J., G. Grandi, J. Hahn, and D. Dubnau.1980. Conformational alteration of mRNA structure andthe posttranscriptional regulation of erythromycin-in-duced drug resistance. Nucleic Acids Res. 8:6081-6097.

141. Guerry, P., J. van Embden, and S. Falkow. 1974. Molec-ular nature of two nonconjugative plasmids carrying drugresistance genes. J. Bacteriol. 117:619-630.

142. Guiney, D. B., and C. E. Davis. 1982. Incompatibility andhost range of pGD10 from Capnocytophaga ochraceus,formerly Bacteriodes ochraceus. Plasmid 7:196-198.

143. Haag, R., R. Sussmuth, and F. Lingens. 1976. Thechloramphenicol resistance of a chloramphenicol-de-grading soil bacterium. FEBS Lett. 63:62-64.

144. Hans, M., S. Blddlecombe, J. Davies, C. E. Luce, andP. J. L. Daniels. 1976. Enzymatic modification of amino-glycoside antibiotics: a new 6'-N-acetylating enzymefrom a Pseudomonas aeruginosa isolate. Antimicrob.Agents Chemother. 9:945-950.

145. Hackenbeck, R., M. Tarpay, and A. Tomasz. 1980.Multiple changes of penicillin-binding proteins in penicil-lin-resistant clinical isolates of Streptococcus pneumo-nia. Antimicrob. Agents Chemother. 17:364-371.

146. Hahn, J., G. Grandi, T. J. Gryzczan, and D. Dubnau.1982. Translational attenuation of ermC: a deletion anal-ysis. Mol. Gen. Genet. 186:204-216.

147. Hall, A., and J. R. Knowles. 1976. Directed selectivepressure on a 1-lactamase to analyze molecular changesinvolved in development of enzyme function. Nature(London) 264:803-04.

148. Hall, B. M. 1970. Mercury resistance of Staphylococcusaureus J. Hyg. 68:121-129.

149. HamIlton-Mller, J. M. T. 1966. A novel method forevluating K/Km and its application to the competitiveinhibition of staphylococcal penicillinase by cephalospo-rins. Biochem. J. 101:40C-42C.

150. Hamilton-MIler, J. M. T. 1982. ,-Lactamases and theirclinical significance. J. Antimicrob. Chemother. 9(Suppl.B):11-19.

151. Hancock, R. E. W. 1981. Aminoglycoside uptake andmode of action-with special reference to streptomycinand gentamicin. I. Antagonists and mutants. J Antimi-crob. Chemother. 8:249-276.

152. Hancock, R. E. W. 1981. Aminoglycoside uptake andmode of action-with special reference to streptomycinand gentamycin. II. Effects of aminoglycosides on cells.J. Antimicrob. Chemother. 8:429-445.

153. Harder, K. J., H. Nikaldo, and M. Matsuhashi. 1981.Mutants of Escherichia coli that are resistant to certainbeta-lactam compounds lack the ompF porin. Antimi-crob. Agents Chemother. 20:549-552.

154. Hardy, K. 1981. Bacterial plasmids. Nelson, London.155. Hartman, B., and A. Tomas. 1981. Altered penicillin-

binding proteins in methicillin-resistant strains of Staph-ylococcus aureus. Antimicrob. Agents Chemother.

VOL. 47, 1983


http://mm

br.asm.org/

Dow

nloaded from


402aaFOSTER MICOBIL. REV.

19:726-735.156. Harwood, J., and D. H. Smith 1971. Catabolite repres-

sion of choramphenicol acetyl transferase synthesis inE. coli K12. Biochem. Biophys. Res. Commun.42:57-62.

157. Heshi_tow, H., _ad R. H. Iid. 1975. Transition of theR factor NR1 in Proteus mirabilis: level of drug resist-ance of nontransitioned and transitioned cells. J. Bacter-iol. 123:56-68.

158. Hayes, M. V., N. A. C. Curd, A. W. Wyke, an J. B.Ward. 1981. Decreased affinity of a peniclin-bindingprotein for h-lactam antibiotics in a clinical isolate ofStaphylococcus aureus resistant to methicillin. FEMSLett. 10:119-122.

159. Hodges, R. W., and S. Bauberg. 1973. Resistance toarsenic compounds conferred by a plasind transmissiblebetween strains of Escherichia coli. J. Bacteriol.115:459-460.

160. Hedges, R. W., N. Dat, P. Ko t au, ad J. T.Sm1t. 1974. Molecular specificities of R factor-deter-mined beta-lactamass: correlation with plasmid com-patibility. J. Bacteriol. 117:56-62.

161. Hedges, R. W., and A. E. Jacob. 1974. Transposition ofampicillin resistance from RP4 to other replicons. Mol.Gen. Genet. 132:31-40.

162. Hedgs, R. W, an M. Mattew. 1979. Acquisition byEscherichia coli of plasmid-bore p-lactames nonmallyconfined to Pseudonionas app. Plasmid 2:169-178.

163. Heddrom, R. C., B. P. Crder, and R. G. lagS. 1982.Compason of kinetics of active tetracycline uptake andactive tetracycline efflux in sensitive and plasmid RP4-containing Pseudomonas putida. J. Bacteriol. 152:255-259.

164. Hu-es, F., C. Rubes, ad S. Falkow. 1975. Transloca-tion ofa plasmid DNA sequence which medites ampicil-lin resistance: molecular nature and specificity of inser-tion. Proc. Natl. Acad. Sci. U.S.A. 72:3623-3627.

165. Hiel, W., G. KIlock, L.IC_r, L. V. Wray, andW. S. Ro . 1982. Purification of the TET repressorand TET operator from the transposon TnlO and charac-terization of their interaction. J. Biol. Chem. 257:6605-6613.

166. Hgills, W., nd LSh. ir. 1983. Nucleotide se-quence of the TnlO encoded tetracycline resistance gene.Nucleic Acids Res. 11:525-529.

167. HI,, W., andB. Ungwer. Binding of four repressors todouble-stranded tet opertor region stabilizes it againstthermal denaturation. Nature (London) 297:700-702.

168. Holke, J. V.1978. Streptomycin uptake via an induciblepolyaniine transport system in Escherichia coli. Eur. J.Biochem. 86:345-351.

169. Holde, J. V. 1979. Induction of streptomycin uptake inresistant strains of Escherichua coli. Antimicrob. AgentsChemother. 15:177-181.

170. Holn, P. L., sdRL C. Cbwes 1979. Transposition ofa duplicate antibiotic resistance gene and generation ofdeletions in plasmid R6K. J. Bacteriol. 137:977-989.

171. Horlnch, S., an 3. Weiblnm. 1960. Posttranscrip-tional modification ofmRNA conformation: mechanismthat regulates erythromycin-induced resistance. Proc.Natl. Acad. Sci. U.S.A. 77:7079-7083.

172. H oehl, S., 4.d 3. Wdesbm. 1981. The controlregion for erythromycin resistance: free energy changesrelated to induction and mutation to constitutive expres-sion. Mol. Gen. Genet. 182:341-348.

173. Ho dochi, S., _d B. Wdeium. 1982. Amino acidsequence conseation in two MLS resistance determi-nants,p. 159-161. In D. Schlessinger (ed.), Microbiolo-gy-1982. American Society forMicrobiology, Washing-ton, D.C.

174.H l, S., adB. WeIbum. 1982. Nucleotide se-quence and functional map of pC194, a plasmid thatspecifies inducible chloramphenicol resistance. J. Bac-teriol. 1Sk815-825.

175. Hdta,K., H. Ya,mamot Y. Okami, andH. Unezawa.1981. Resistance mechanisms of kanamycin-, neomycin-

and streptomycin-producing streptomycetes to amino-glycoside drug. J. Antibiot. 34:1175-1182.

176. Ha, S4.E. Obhubo, N. Daylas, an H. Saer. 1975.Electron microscope heteroduplex studies of sequencereltions among bacterial plasmids: identification andmapping of the insertion sequences IS1 and IS2 in F andR plasmids. J. Bacteriol. 122:764-775.

177. HurwIb, C., C. B. Brana, and C. L. Rosi. 1981. Roleof ribosome recycling in uptake of dihydrostrptomycinby sensitive and resistant Escherichia coli. Biochim.Biophys. Acta 652:168-176.

178. MIda, S., ad W. Arber. 1977. Plaque-forming specializedtransducing phage P1: isolation of P1CmSmSu, a precur-sor of PlCm. Mol. Gen. Genet. 153:259-269.

179. flda, S., J. Meyer, P. Lnder, N. Goto, R. Naaya, H.RBet, ad W. Arber. 1982. The kanamycin resistancetransposon Tn2680 derived from plasmid Rtsl and car-ried by phage PlKm has flanking 0.8-kb-long directrepeats. Plasmid 8:187-198.

180. l_sane, J. 1973. Repressor and antirepressor in theregulation of staphylococcal penicillinase synthesis. Ge-netics 75:1-17.

181. Insande, J. 1978. Genetic regulation of penicillinasesynthesis in gram-positive bacteria. Microbiol. Rev.42:67-83.

182. Iun_ade, J., and J. L. Llleold. 1976. Characterizationof mutations in the penicilinase operontof Staphylococ-cus aureus. Mol. Gen. Genet. 147:23-27.

183. I_wadt, J., and J. Liehm. 1977. Nature of theplasmid-linked penicillinase regulatory region in Staphy-lococcus aureus. Mol. Gen. Genet. 153:153-157.

184. lordanc, S., M. Srdna, P. Deaf Lata, and R. P.Nolck. 1978. Incompatibility and moleadar relation-ships between small staphylococcal plasmids carryingthe same resistance marker. Plasmid 1:468-479.

185. Jacoby, G. A., and M. Matthew. 1979. The distribution off-lactamases on Pseudomonas plasmids. Plasmids 2:41-47.

186. Jackas., W. J., and A. 0. Su_e. 1982. Polypeptidesencoded by the mer operon. J. Bacteriol. 1491479-487.

187. Jackoas, W. J., and A. 0. S _m . 1982. Biochemicalcharacterization of HgClrinducible polypeptides en-coded by the mer operon of plasmid R100. J. Bacteriol.151:962-970.

188. Jain, G., R. Lau, P. M. Kulfers, and H. Koleda.1979. Molecular nature of two Haemophilus i!jluenzaeR factors containing resistances and the multiple integra-tion of drug resistance transposons. J. Bacteriol.138:584-597.

189. Jaurl, B., and T. Grn__trom. 1981. ampC cephalo-sporinase of ETcherichia coli K-12 has a different evolu-tionary origin from that of -lactamases of the penicillin-ase type. Proc. Natl. Acad. Sci. U.S.A. 7&.4897-4901.

190. Jaun, B.,T. Grudtom, T. Edlhd, a S. Normark.1981. The E. coli 1-lactamase attenuatr ndiatesgrowth rate-dependent regulation. Nature (London)290-.221-225.

191.J-b_nutr, R. S., and N. Daft. 1974. Trimethoprimresistance factors in enterobacteria from clinical spefi-mens. J. Med. Microbiol. 7:169-177.

192. J_ota, L. H., an K. H. G. Dyke. 1969. Ethidiumbromide resistance, a new marker on the staphylococcalpenicillinase plsmid. J. Bacteriol. 10&1413-1414.

193. Jergen, R. A.,and W. S. Ra_dkef. 1979. Orgpniza-tion of structural and regulatory genes that mdiatetetracycline resistance in transposon TnlO. J. Batteriol.138:705-714.

194. Kab1n, S., C. Nathas, ad S. Cobs.. 1974. Gentamicinadeylyltransferase activity as a cause of gentamicinresistance in clinical isolates of Pseudomonas aerugin-osa. Antimicrob. Agents Chemother. 5:565-570.

195. Kahs F. M.,J.S. Kahan, P. J. Camd, d H. Kropp.1974. The mechanism of action of fosfomycin. Ann.N.Y. Acad. Sci. 235:364-386.

196. Katau,K., M.Inoue, and S.Mth. 1982. Transposi-

MICROBIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from



tion of the carbenicillin-hydrolyzing beta-lactamasegene. J. Bacteriol. 150:483-489.

197. Kaufman, R. J., P. C. Brown, and R. T. Schimke. 1979.Amplified dihydrofolate reductase genes in unstablymethotrexate-resistant cells are associated with doubleminute chromosomes. Proc. Natl. Acad. Sci. U.S.A.76:5669-5673.

198. Kawabe, H., F. Kobayashi, M. Yamaguchi, R. Utara, andS. Mitsuhashi. 1971. 3"-phosphoryldihydrostreptomycinproduced by the inactivating enzyme of Pseudomonasaeruginosa. J. Antibiot. 24:651-652.

199. Kawabe, H., S. Kondo, H. Umezawa, and S. Mitsuhashi.1975. R factor-mediated aminoglycoside antibiotic resist-ance in Pseudomonas aeruginosa: a new aminoglycoside6'-N-acetyltransferase. Antimicrob. Agents Chemother.7:494-499.

199a.Kawabe, H., T. Naito, and S. Mitsuhashi. 1975. Acetyla-tion of amikacin, a new semi-synthetic antibiotic, byPseudomonas aeruginosa carrying an R factor. Antimi-crob. Agents Chemother. 7:50-54.

200. Kayser, F. H.,M. Devaud, and J. Biber. 1976. Aminogly-coside 3'-phosphotransferase IV: a new type of amino-glycoside phosphorylating enzyme found in staphylo-cocci. Microbios Lett. 3:63-68.

201. Keggins, K. M., P. S. Lovett, and E. J. Duvall. 1978.Molecular cloning of genetically active fragments ofBacillus DNA in Bacillus subtilis and properties of thevector plasmid pUB110. Proc. Natl. Acad. Sci. U.S.A.75:1423-1427.

202. Kida, M., T. Asako, M. Yoneda, and S.Mltsuhashi. 1975.Phosphorylation of dihydrostreptomycin by Pseudomo-nas aeruginosa, p. 441-448. In S. Mitsuhashi and H.Hashimoto (ed.), Microbiol drug resistance. UniversityPark Press, Tokyo.

203. Kitts, P., L. Symlngton, M. Burke, R. Reed, and D.Sherratt. 1982. Transposon-specified site specific recom-bination. Proc. Natl. Acad. Sci. U.S.A. 79:46-50.

204. Kleckner, N. 1981. Transposable elements in prokary-otes. Annu. Res. Genet. 15:341-404.

205. Knott-Hunziker, V., S. Petursson, S. G. Waley, B. Jaurin,and T. Grundstrom. 1982. The acyl-enzyme mechanismof 1-lactamase action. The evidence for class C -lactamases. Biochem. J. 207:315-322.

206. Kobayashi, F., T. Koshi, J. Eda, Y. Yoshimura, and S.Mltsuhashi. 1973. Lividomycin resistance in staphylo-cocci by enzymatic phosphorylation. Antimicrob.Agents Chemother. 4:1-5.

207. Komatsu, K., J. Leboul, S. Harford, and J. Davies. 1981.Studies of plasmids in neomycin-producing Streptomy-cesfradiae, p. 384-387. In D. Schlessinger (ed.), Micro-biology-1981. American Society for Microbiology,Washington, D.C.

208. Komatsu, Y., K. Murakami, and T. Nidlkawa. 1981.Penetration of moxalactam into its target proteins inEscherichia coli K-12: comparison of a highly moxalac-tam-resistant mutant with its parent strain. Antimicrob.Agents Chemother. 20:613-619.

209. Kondo, I., T. Ishikawa, and H. Nakahara. 1974. Mercuryand cadmium resistance mediated by the penicillinaseplasmid in Staphylococcus aureus. J. Bacteriol. 117:1-7.

210. Kubota, K., A. Okuyama, and N. Tanaka. 1972. Differen-tial effects of antibiotics on peptidyl transferase reac-tions. Biochem. Biophys. Res. Commun. 47:1196-1202.

211. LabIa, R., M. Guionie, M. Barthl6my, and A. Philippon.1981. Properties of three carbenicillin-hydrolysing,B-lactamases (CARB) from Pseudomonas aeruginosa:identification of a new enzyme. J. Antimicrob. Che-mother. 7:49-56.

212. Lacey, R. W. 1975. Antibiotic resistance plasmids ofStaphylococcus aureus and their clinical importance.Bacteriol. Rev. 39:1-32.

213. Lacey, R. W., and I. Chopra. 1972. Evidence for muta-tion to streptomycin resistance in clinical strains ofStaphylococcus aureus. J. Gen. Microbiol. 73:175-180.

214. Lacey, R. W., and J. Grinsted. 1972. Linkage of fusidic

acid resistance to the penicillinase plasmid in Staphylo-coccus aureus. J. Gen. Microbiol.73:501-508.

215. Lacey, R. W., and V. T. Rosdahl. 1973. An unusual"penicillinase plasmid" in Staphylococcus aureus; evi-dence for its transfer under natural conditions. J. Med.Microbiol. 7:1-9.

216. Lai, C. J., J. E. Dahlberg, and B. Welsblum. 1973.Structure of an inducibly methylatable nucleotide se-quence in 23S ribosomal ribonucleic acid from erythro-mycin-resistant Staphylococcus aureus. Biochemistry12:457-460.

217. Lai, C. J., and B. Welsblum. 1971. Altered methylationof ribosomal RNA in an erythromycin-resistant strain ofStaphylococcus aureus. Proc. Natl. Acad. Sci. U.S.A.68:856-860.

218. Lai, C. J., B. Weisblum, S. R. Fahnestock, and M.Nomura. 1973. Alteration of 23S ribosomal RNA anderythromycin-induced resistance to lincomycin and spir-amycin in Staphylococcus aureus. J. Mol. Biol. 74:67-72.

219. Lane, D., and M. Chandler. 1977. Mapping of the drugresistance genes carried by the r-determinant of theR100-1 plasmid. Mol. Gen. Genet. 157:17-23.

220. LeBlanc, D. J., and L. N. Lee. 1982. Characterization oftwo tetracycline resistance determinants in Streptococ-cusfaecalis JH1. J. Bacteriol. 150:835-843.

221. LeBoul, J., and J. Davies. 1982. Enzymatic modificationof hygromycin B in Streptomyces hygroscopicus. J.Antibiot. 35:527-528.

222. LeGoffic, F., M. Mareau, S. Slegrist, F. W. Goldstein,and J. Acar. 1977. La resistance plasmidique de Haemo-philus sp. aux antibiotiques aminoglycosidique isolementet etude d'une nouvelle phosphotransferase. Ann. Mi-crobiol. (Inst. Pasteur) 128A:388-391.

223. LeGofllc, F., and A. Martel. 1974. La R6sistance auxaminoglycosides provoqu6e par une isoenzyme la kana-mycine acetyltransferase. Biochimie 56:893-897.

224. LeGoffic, F., A. Martel, M. L. Capmau, B. Baca, P.Goebel, H. Chardon, C. J. Soussy, J. Duval, and D. H.Bouanchaud. 1976. New plasmid-mediated nucleotidyla-tion of aminoglycoside antibiotics in Staphylococcusaureus. Antimicrob. Agents Chemother. 10:258-264.

225. LeGofllc, F., A. Martel, M. Moreau, M. L. Capmau, C. J.Soussy, and J. Duval. 1977. 2"-O-phosphorylation ofgentamicin components by a Staphylococcus aureusstrain carrying a plasmid. Antimicrob. Agents Che-mother. 12:26-30.

226. LeGoffic, F., A. Martel, and J. Witchlitz. 1974. 3-N-enzymatic acetylation of gentamicin, tobramicin, andkanamycin by Escherichia coli carrying an R factor.Antimicrob. Agents Chemother. 6:680-684.

227. Le Grice, S. F. J., H. Matzura, R. Marcoli, S. Ilda, andT. A. BIckle. 1982. The catabolite-sensitive promoter forthe chloramphenicol acetyltransferase gene is precededby two binding sites for the catabolite gene activatorprotein. J. Bacteriol. 150:312-318.

228. Leon, J., J. M. Garcda-Lobo, and J. M. Ortiz. 1982.Fosfomycin resistance plasmids do not affect fosfomycintransport into Escherichia coli. Antimicrob. Agents Che-mother. 21:608-612.

229. Lerch, K. 1980. Copper metallothionein, a copper-bind-ing protein from Neurospora crassa. Nature (London)284:368-370.

230. Levesque, R., and P. H. Roy. 1982. Mapping of a plasmid(pLQ3) from Achromobacter and cloning of its cephalo-sporinase gene in Escherichia coli. Gene 18:69-75.

231. Levesque, R., P. H. Roy, R.Letarte, and J. C. Pechere.1982. Plasmid-mediated cephalosporinase from Achro-mobacter species. J. Infect. Dis. 145:753-761.

232. Levy, S. B. 1981. The tetracyclines: microbial sensitivityand resistance,p. 27-44. In G. G. Grassi and L. D. Sabath(ed.), New trends in antibiotics: research and therapy.Elsevier/North-Holland, Amsterdam.

233. Levy, S. B., and L. McMurry. 1974. Detection of aninducible membrane protein associated with R-factor-

VOL. 47, 1983


http://mm

br.asm.org/

Dow

nloaded from


404 FOSTER MICROBIOL. REVS

mediated tetracycline resistance. Biochem. Biophys.Res. Coun. 56:1060-1068.

234. Levy, S. B., am L. McMnrry. 1978. Plasmid-determinedtetracycline resistance involves new transport systemsfor tetracycline. Nature (London) 276.-90-92.

235. Levy, S. B., and L. MMiurry. 1978. Probing the expres-sion of plasmid-mediated tetracycline resistance in Esch-erichia coil, p. 177-180. In D. Schlessinger (edj, Micro-biology-1978. American Society for Microbiology,Washington, D.C.

236. Levy, S. B., L. McMurry, r. Ongn, aid R. M.Sauundei. 1977. Plasmidmediated tetracycline resist-ance in E. coli, p. 181-203. In J. Drews and G. Hogen-auer (ed.), R-factors: their properties and possible con-trol. Springer-Verlag, Vienna.

237. Majundar, M. K., andSLK. Majudar. 1971. Relation-ship between alkaline phosphatase and neomycin forma-tion in Streptomycesfradiae. Biochem. J. 122:397-404.

238. Mdlk, V. S., an L. C. Ving. 1972. Chloramphenicolresistance in a chloramphenicol-producing Streptomy-ces. Can. J. Microbiol. 18:583-590.

239. Mike, H. 1978. Zonal-pattern resistance to lincomycinin Streptococcus pyogenes: genetic and physical studies,p. 142-145. In D. Schlessinger (ed.), Microbiology-1978. American Society for Microbiology, Washington,D.C.

240. Mdek, H., W. Ruihrdt, M. Hartman, and F. Walter.1981. Genetic study of plasmid-associated Zdwal resist-ance to lincomycin in Streptococcus pyogenes. Antimi-crob. Agents Chemothler. 19:91-100.

241. Manes, M. J., G. C. Faster, and P. F. SparIng. 1974.Ribosomal resistance to streptomycin and spectinomycinin Neisseria gonoffhoeae. J. Bacteriol. 120k1293-1299.

242. Mao, J. C. H., ad E. E. tobthaw. 1971. Effects ofmacrolides on peptide-bond formation and translocation.Biochemistry 1N:2054-2061.

243. Mao, J. C. M., and E.E. Roblahaw. 1972. Erythromycin,a peptidyltrsfeas effector. Biochemistry 11:4864-4872.

244. Marcel, R., S.1b, and T. A. Bickle. 1980. The DNAsequence of an lsl-flanked transposon coding for resist-ance to chloramphenicol and fusidic acid. FEBS Lett.110:11-14.

245. Makdl, It., 0. A. Okubndeo, R. H.Paye, ad L. Fed.1978. Human infections with thymine requiring bacteria.J. Mod. Microbiol. 11:33-45.

246. Mates, S. M., E. S. Elienberg, L J. Mandel, L. Patel,H. R. Kabick and M. H. Mer. 1982. Membrane poten-tial and gentamicin uptake in Staphylococcus aureus.Proc. Natl. Acad. Sci. U.S.A. 79:6693-7.

247. Mathsmh, Y., M. Yagawa, S. Kosd, T. Tekudl,nd H. U_nawa. 1975. Aminoglycoside 3'-phospho-

transferases I and H in Pseudomonas aeruginosa. J.Antibiot. 28:442447.

248. Matthw, M. 1978. Properties of the 1-lactamase speci-fied by the Pseudomonas plasmid R1S1. FEMS Micro-biol. Lett. 4:241-244.

249. Matthew, M. 1979. Plasmid-modiated P-actamaes ofgram-neg ive bateria: properties and distribution. J.Antimicrob. Chemother. 5:349-358.

250. Matthew, M., A. M. Hasl, M. J. Marhll, ad G. W.R11. 1975. The use of analytical isoelectric focusing fordetection and identification of P-lactamases. J. Gen.Microbiot. 86:169-178.

251. Mathew, M., R. W. Hudgs, and J. T. Snuth. 1979.Types of i-lactamase detatmined by plasmids in gram-negatve bacteria. J. Bacteriol. 138:657-662.

252. Matthw, M., adR. B. Sykes. 1977. Properties of thebeta-lactamase specified by the Pseudomonas plasmidRP1I. J. Bacteriol. 132:341-345.

253. Mdell , G. L.,R. C. Moellerlag, C. C. H8#Ukle, adM. N. Swrt. 1975. Salmonella typhimurium resistant tosilver nitrte chloramphenicol and ampicillin. Lancet1:235-240.

254. McLngH, J. R., C.R. Mumy, and J. C.Rowix.

1981. Unique features in the ribosome bin4ing site se-quence of the gram-positive Staphylococcus aureus -lactamase gene. J. Biot. Chem. 25fo11283-11291.

255. McMurry, L. M., J. C. C _Hnane, ad S. B. Levy. 1982.Trnsport of lipophilic analog minocycline difers fromthat of tetracycline in susceptible and resistant Esche-richia coli strains. Antimicrob. Agents Chemother.22:791-799.

256. MeMury, L. M., J. C. Cune, R. E. Pbcdl, andS. B. Lev. 1981. Active uptake of tetracycline by mem-brane vesicles from susceptible Escherichia coil. Antimi-crob. Agents Chemother. 2f:307-313.

257. MMwrry, L., and S. B. Levy. 1978. Two transportsystems for tetracycline in sensitive Escherichia coli:critical role for an initial rapid uptake system insensitiveto energy inhibitors Antimicrob. Agents Chemother.14:201-209.

258. MeMurry, L., R. E. Pdiled, and S. B. Levy. 1980.Active efflux of tetracycline encoded by four geneticallydifferent tetracycline resistance determinats in Esche-richia coli. Proc. Natl. Acad. Sci. U.S.A. 77:3974-3977.

259. Medelrs, A. A., R. W. Hedgen, ad G. A. Janoby. 1982.Spread of a "Pseudomonas-speciflc" m4actase toplasmids of enteroacteria. J. Bacteriol. 149:700-707.

260. M_ndes, B., C.Tala, and S.D. Ley. 1980. Hetero-geneity of tetracycline resistmce determinants. Plasmid3:99-108.

261. Meadoza, M. C., J. M. Garda, J. I_aea, F. J. Medez,C. Hardlm, nd j. M. Ortk. 1980. Plasmid-determinedresistance to fosfomyin in Serratia marcescens. Antimi-crob. Agents Chemother. 18:215-219.

262. Meyer, J., ad S. itds. 1979. Amplification of chloram-phenicol resistance transposons carried by phage PlCmin Escherichia coli. Mol. Gen. Genet. 176:209-219.

263. MIk, T., A. M. Easte', an R. H.R. 1978. Mappingof the resistance genes of the R plasmid NRI. Mol. Gen.Genet. 158:217-224.

264. M l.U, K., J. KarIetva,,N. Q.ye, M. aer, I.Komerova, and Z. Vanek. 1971. Interaction of tetracy-cline with protein synthesizing system of Streptomycesaureofaclens. J. Antibiot. 24:8014809.

265. Mle, A. L., ad J. B. Wilker. 1970. Accumulation ofstreptomycin-pbospbate-in cultures ofstreptomycin pro-ducers grown on a hi-phosphate medium. J. Bacteriol.1W:8-12.

266. Mie, G. H., F. J. Sab9tul, R. S. are,andJ. A. Waltz.1980. Survey of aminoglycoside resistance patterns.Dev. Ind. Microbiol. 21:91-104.

267. Miller, M. H., S. C. Edbeg, L. J. M_di, C. F. Diesa,ad N. H. Sgiobd. 1980. Gentamicin uptake in wild-type and aminoglycoside-resistant sall-colony mutantsof Staphylococcus aureus. Antimicrob. Agents Che-mothei.18:722-729.

268. Meh&ew, D.H., R. K.H ,J.P. Sanfrd, ad C.R.Bater. 1974. Transfable resistance to tobramycin inKIebsiella pneumoniae and Enterobacter cloacac associ-ated with enymatic acetyation of tobrmycin. Antimi-crob. Agents chemother. 6:492-497.

269. Iembin, D., Y. N amwwltz, ad13. R- Itt. 1981.Insensitivity of pepddoglycan biosynthedc reactions toP-lactam antibiotics in a clinical isolate of Pseudomonasaeruglnosa. Antinicrob. Agents Chemother. 19:687-695.

269a.Mebly, H. L. T., ad B. P. leRoo. 1982. Energetics ofplasmid-mediated arenate effiux in Escherlchia colProc. Natl. Acad. Sci. U.S.A. 79:6119-6122.

270.N__ a, T. L., D. Kn m, H. Tdal, W. N.Ter, ad G. H. Mle. 1980. Novel clss of chloram-phenicol analogs with activity agpinst chkoramphenicol-resistant and chloramphenicol-susceptible orgnisms, p.442443. In J. D. Nelson and G. Gtassi (ed.), Currentchemothrpy and infectious dease. Amerin Societyfor Microbiology, Washington, D.C.

271. Nap, Y., and S. Mlbnhhl. 1972. New types of Rfactors incapable of inactivating chloramphenicol. J.Bacteriol. 109:1-7.

MICROBIOL. RE£v.


http://mm

br.asm.org/

Dow

nloaded from



272. Nakae, R., and T. Nakae. 1982. Diffusion of aminoglyco-side antibiotics across the outer membrane of Escherich-ia coli. Antimicrob. Agents Chemother. 22:554-559.

273. Nakabara, H., T. G. Kinswerf, S. Silver, T. Mild, A. M.Easton, and R. H. Rownd. 1979. Gene copy numbereffects in the mer operon of plasmid NR1. J. Bacteriol.138:284-287.

274. Nakahara, H., S. Silver, T. Miki, and R. H. Rownd. 1979.Hypersensitivity to Hg2+ and hyperbinding activity asso-ciated with cloned fragments of the mercurial resistanceoperon of plasmid NR1. J. Bacteriol. 140:161-166.

275. Nltzan (Za i , Y., and S. Gozhansky. 1980. Chlor-amphenicol binding site of an fi- R-factor-specified vari-ant of chloramphenicol acetyltransferase. Arch. Bio-chem. Biophys. 201:15-120.

276. Noel, D., K. Nikaido, and F.-L. Ames. 1979. A singleamino acid substitution in a histidine transport proteindrastically alters its mobility in sodium dodecyl sulphatepolyacrylamide gel electrophoresis. Biochemistry18:4159-4165.

277. Nordstrom, K., L. C. Ingram, and A. Lundback. 1972.Mutations in R factors of Escherichia coli causing anincreased number of R-factor copies per chromosome. J.Bacteriol. 110:562-569.

278. Novick, R. P. 1962. Staphylococcal penicillinase and thenew penicillins. Biochem. J. 83:229-235.

279. Novick, R. P., I. Edelman, M. D. Schwednger, A. D.Grus, E. C. Swanson, and P. A. Pattee. 1979. Genetictranslocation in Staphylococcus aureus. Proc. Natl.Acad. Sci. U.S.A. 76:400-404.

280. Novk, R. P., E. Murphy, T. J. Gryczan, E. Baron, andI. Edehuan. 1979. Penicillinase plasmids of Staphylococ-cus aureus restriction-deletion maps. Plasmid 2:109-129.

281. Novick, R. P., and C. Roth. 1968. Plasmid-linked resist-ance to inorganic salts in Staphylococcus aureus. J.Bacteriol. 95:1335-1342.

282. Nugent, M. E., and R. W. Hedges. 1979. The nature ofthe genetic determinant for the SHV-1 ,B-lactamase. Mol.Gen. Genet. 175:239-243.

283. Ogden, S., D. Haggerty, C. M. Stoner, D. Kolodrubetz,and R. Schef. 1980. The Escherichia coli L-arabinoseoperon: binding sites of the regulatory proteins and amechanism of positive and negative regulation. Proc.Natl. Acad. Sci. U.S.A. 77:3346-3350.

284. Ohmoli, H., A. Azuma, Y. Suzuld, and Y. Hshlmoto.1977. Factors involved in beta-lactam antibiotic resist-ance in Pseudomonas aeruginosa. Antimicrob. AgentsChemother. 12:537-539.

285. Olafson, R. W., K. Abel, and R. S. SIm. 1979. Prokaryot-ic metailothionein: preliminary characterization of ablue-green alga heavy metal-binding protein. Biochem.Biophys. Res. Commun. 8-.36-43.

286. Olson, G. J., W. P. Iverson, and F. E. Brinkman. 1981.Volatilization of mercury by Thiobacillus ferrooxidans.Curr. Microbiol. 5:115-118.

287. Ozanne, B., R. Benveniste, D. Tipper, and J. Davies.1969. Aminoglycoside antibiotics: inactivation by phos-phorylation in Escherichia coli carrying R factors. J.Bacteriol. 100:1144-1146.

288. Pardo, D., and R. Rousett. 1977. A new ribosomalmutation which affects the two ribosomal subunits inEscherichia coli. Mol. Gen. Genet. 153:199-204.

289. Patlihall, K. H., J. Acar, J. J. Burchall, F. W. Goldstein,and R. J. Harvey. 1977. Two distinct types of trimetho-prim-resistant dihydrofolate reductase specified by R-plasmids of different compatibility groups. J. Biol.Chem. 252:2319-2323.

290. Perlin, M. H., and S. A. Lerner. 1981. Localization of anamikacin 3'-phosphotransferase in Escherichia coli. J.Bacteriol. 147:320-325.

291. PerHn, M. H., and S. A. Lerner. 1982. Decreased suscep-tibility to 4'-deoxy-6'-N-methylamikacin (BB-K 311)conferred by a mutant plasmid in Escherichia coli.Antimicrob. Agents Chemother. 22:78-82.

292. Perlman, D., and R. H. Rownd. 1975. Transition of R

factor NRI in Proteus mirabilis: molecular structure andreplication of NRI deoxyribonucleic acid. J. Bacteriol.123:1013-1034.

293. Perlman, D., and R. H. Rownd. 1976. Two origins ofreplication in composite R plasmid DNA. Nature (Lon-don) 259:281-284.

294. Penman, D., and R. Stckgold. 1977. Selective amplifica-tion of genes on the plasmid, NRI, in Proteus mirabilis:an example of the induction of selective gene amplifica-tion. Proc. Natl. Acad. Sci. U.S.A. 74:2518-2522.

295. Perry, R. D., and S. Silver. 1982. Cadmium and manga-nese transport in Staphylococcus aureus membrane vesi-cles. J. Bacteriol. 150:973-976.

296. Pestka, S. 1974. Antibiotics as probes of ribosomesstructure: binding of chloramphenicol and erythromycinto polyribosomes; effect of other antibiotics. Antimicrob.Agents Chemother. 5:255-267.

297. Plendl, W., and A. Bock. 1982. Ribosomal resistance inthe gentamicin-producing organism Micromonosporapurpurea. Antimicrob. Agents Chemother. 22:231-236.

298. Plwowarski, J. M., and P. D. Shaw. 1979. Streptomycinresistance in a streptomycin-producing microorganism.Antimicrob. Agents Chemother. 16:176-182.

299. Polak, J., and R. P. Novick. 1982. Closely related plas-mids from Staphylococcus aureus and soil bacilli. Plas-mid 7:152-162.

300. Porter, F. D., S. Silver, C. Ong, and H. Nakahara. 1982.Selection for mercurial resistance in hospital settings.Antimicrob. Agents Chemother. 22:852-858.

301. Privatera, G., M. Sebald, and F. Fayolle. 1979. Commonregulatory mechanism of expression and conjugativeability of a tetracycline resistance plasmid in Bacteroidesfragilis. Nature (London) 278:657-659.

302. Proctor, G. N., and R. H. Rownd. 1982. Rosanilins:indicator dyes for chloramphenicol-resistant enterobac-teria containing chloramphenicol acetyltransferase. J.Bacteriol. 150:1375-1382.

303. Ptashe, K., and S. N. Cohen. 1975. Occurrence of inser-tion sequence regions on plasmid deoxyribonucleic acidas direct and invertible nucleotide sequence duplica-tions. J. Bacteriol. 122:776-781.

304. Rahtchian, A., R. Nouravarsani, G. R. Miller, and E. H.Gerlach. 1979. Increased production of beta-lactamaseunder anaerobic conditions in some strains of Escherich-ia coli. Antimicrob. Agents Chemother. 16:772-775.

305. Reeve, E. C. R. 1978. Evidence that there are two typesof determinant for tetracycline resistance among R-factors. Genet. Res. 31:75-4.

306. Reynolds, A. V., and J. T. Smith. 1979. Enzymes whichmodify aminoglycoside antibiotics, p. 165-181. In D. S.Reeves and A. Geddes (ed.), Recent advances in infec-tion. Churchill Livingstone, Edinburgh.

307. RImd, M. H. 1965. Wild-type variants of exopenicil-linase from Staphylococcus aureus. Biochem. J. 94:584-593-

308. RIhmond, M. H. 1965. Dominance of the inducible statein strains of Staphylococcus aureus containing two dis-tinct penicillinase plasmids. J. Bacteriol. 90:370-374.

309. RIhmond, M. H. 1977. A second regulatory regioninvolved in penicillinase synthesis in Staphylococcusaureus. J. Mol. Biol. 26:357-360.

310. Richmond, M. H. 1978. Factors influencing the antibac-terial action of P-lactam antibiotics J. Antimicrob. Che-mother. 4(Suppl. B):1-14.

311. Rchmond, M. H., D. C. Clark, and S. Wotton. 1976.Indirect method for assessing the penetration of beta-lactamase-nonsuspectible penicillins and cephalosporinsin Escherichia coli strains. Antimicrob. Agents Che-mother. 10:215-218.

312. RImd, M. H., and N. A. C. Curts. 1974. Theinterplay of P-lactamases and intrinsic factors in theresistance of Gram-negative bacteria to penicillins andcephalosporins. Ann. N.Y. Acad. Sci. 235:553-568.

313. Richmond, M. H., and R. B. Sykes. 1973. The beta-lactamases of gram-negative bacteria and their possible

VOL. 47, 1983


http://mm

br.asm.org/

Dow

nloaded from


406 FOSTER

physiological role. Adv. Microb. Physiol. 9:31-88.314. Roberts, M., A. Corney, and W. V. Shaw. 1982. Molecu-

lar characterization of three chloramphenicol acetyl-transferases isolated from Haemophilus irfluenzae. J.Bacteriol. 151:737-741.

315. Robertson, J. M., and E. C. R. Reeve. 1972. Analysis ofthe resistance mediated by several R factors to tetracy-cline and minocycline. Genet. Res. 20:239-252.

316. Rodriguez-Tabar, A., F. Rojo, D. Danaso, and D. Vas-quez. 1982. Carbeniciflin resistance of Pseudomonasaeruginosa. Antimicrob. Agents Chemother. 22:255-261.

317. Rogers, H. J., H. R. Perkins, and J. B. Ward. 1980.Microbial cell walls and membranes. Chapman and Hall,London.

318. Rogers, W. H., W. Springer, and F. E. Young. 1982.Cloning and expression of a Streptomyces fradiae neo-mycin resistance gene in Escherichia coli. Gene 18:133-141.

319. Roland, S., R. Ferone, R. J. Harvey, V. L. Styles, and R.W. MorrIson. 1979. The characteristics and significanceof sulfonamides as substrates for Escherichia coli dihy-dropteroate synthase. J. Biol. Chem. 254:10337-10345.

320. Rosedahl, V. T. 1973. Naturally occurring constitutive ,B-lactamase of novel serotype in Staphylococcus aureus. J.Gen. Microbiol. 77:229-231.

321. Rownd, R. H., T. Miki, E. R. Appelbaum, J. R. Miller,M. Finkeldein, and C. R. Barton. 1978. Dissociation,amplification, and reassociation of composite R-plasmidDNA, p. 33-77. In D. Schlessinger (ed.), Microbiology-1978. American Society for Microbiology, Washington,D.C.

322. Rozencranz, H. S., and H. S. Carr. 1972. Silver sulfadia-zine: effect on the growth and metabolism of bacteria.Antimicrob. Agents Chemother. 2:367-372.

323. Rubin, L. G., A. A. Medeiros, R. H. Yolken, and E.Moxon. 1981. Ampicillin treatment failure of apparently(-lactamase-negative Haemophilus influenzae type bmeningitis due to novel P-lactamase. Lanceti:1008-1010.

324. Sands, L. C., and W. V. Shaw. 1973. Mechanism ofchloramphenicol resistance in staphylococci: characteriza-tion and hybridization of variants of chloramphenicol ace-tyltransferase. Antimicrob. Agents Chemother. 3:299-305.

325. Santanam, P., and F. H. Kayser. 1978. Purification andcharacterization of an aminoglycoside inactivating en-zyme from Staphylococcus epidermidis FK109 that nu-cleotidylates the 4' and 4'-hydroxyl groups of the amino-glycoside antibiotics. J. Antibiot. 31:343-351.

326. Sawada, Y., S. Yaginuma, M. Tad, S. Iyobe, and S.Mltsuhashl. 1974. Resistance to 1-lactam antibiotics inPseudomonas aeruginosa, p. 391-397. In S. Mitsuhashiand H. Hashimoto (ed.), Microbial drug resistance.University of Tokyo Press, Tokyo.

327. Schmitt, R., J. Altenbuchner, K. Wiebauer, W. Arnold,A. Publer, and F. SchoE. 1981. Basis of transpositionand gene amplification by Tnl721 and related tetracy-cline resistance transposon. Cold Spring Harbor Symp.Quant. Biol. 45:59-65.

328. Schmitt, R., E. Bernhard, and R. Mattes. 1979. Charac-terization of TnJ721, a new transposon containing tetra-cycline resistance genes capable of amplification. Mol.Gen. Genet. 172:53-65.

329. Schofil, F., and A. Puhler. 1979. Intramolecular amplifi-cation of the tetracycline resistance determinant of tran-sposon Tnl771 in Escherichia coli. Genet. Res. 33:253-260.

330. Schottel, J. L. 1978. The mercuric and organomercurialdetoxifying enzymes from a plasmid-bearing strain ofEscherichia coli. J. Biol. Chem. 253:4341-4349.

331. Schottel, J., A. Mandal, D. Clark, S. Silver, and R. W.Hedges. 1974. Volatilisation of mercury and organomer-curials determined by inducible R-factor systems inenteric bacteria. Nature (London) 251:335-337.

332. Shales, S. W., I. Chopra, and P. R. Ball. 1980. Evidencefor more than one mechanism of plasmid-determinedtetracycline resistance in Escherichia coli. J. Gen. Mi-

crobiol. 121:221-229.333. Shannon, K., and I. PhIllips. 1982. Mechanisms of resist-

ance to aminoglycosides in clinical isolates. J. Antimi-crob. Chemother. 9:91-102.

334. Shapiro, J. A., and P. Sporn. 1977. Tn4O2: A newtransposable element determining trimethoprim resist-ance that inserts in bacteriophage lambda. J. Bacteriol.129:1632-1635.

335. Shaw, W. V. 1967. The enzymatic acetylation of chlor-amphenicol by extracts of R factor-resistant Escherichiacoli. J. Biol. Chem. 242:687-693.

336. Shaw, W. V. 1971. Comparative enzymology of chioram-phenicol resistance. Ann. N.Y. Acad. Sci. 182:234-242.

337. Shaw, W. V. 1983. Chloramphenicol acetyltransferase.Enzymology and molecular biology. Crit. Rev. Biochem.14:1-46.

338. Shaw, W. V., D. W. Bentley, and L. Sands. 1970.Mechanism of chloramphenicol resistance in Staphylo-coccus epidermidis. J. Bacteriol. 104:1095-1105.

339. Shaw, W. V., D. H. Bouanchaud, and F. W. Goldstein.1978. Mechanism of transferable resistance to chloram-phenicol in Haemophilus parainfluenzae. Antimicrob.Agents Chemother. 13:326-330.

340. Shaw, W. V., and R. F. Brodsky. 1968. Characterizationof chloramphenicol acetyltransferase from chloramphen-icol-resistant Staphylococcus aureus. J. Bacteriol.95:28-36.

341. Shaw, W. V., and D. A. Hopwood. 1976. Chlorampheni-col acetylation in Streptomyces. J. Gen. Microbiol.94:159-166.

342. Shaw, W. V., L. C. Packman, B. D. Burlelgh, A. Dell,H. R. Morris, and B. S. Hartley. 1979. Primary structureof a chloramphenicol acetyltransferase specified by Rplasmids. Nature (London) 282-870-872.

343. Shaw, W. V., and J. Unowsky. 1968. Mechanism of Rfactor-mediated chloramphenicol resistance. J. Bacter-iol. 95:1976-1978.

344. Shlvakumar, A. G., and D. Dubmau. 1981. Characteriza-tion of a plasmid-specified ribosome methylase associat-ed with macrolide resistance. Nucleic Acids Res. 9:2549-2562.

345. Shivakimar, A. G., J. Hahn, G. Grandi, Y. Kozbov, andD. Dubnau. 1980. Posttranscriptional regulation of anerythromycin resistance protein specified by plasmidpE194. Proc. Natl. Acad. Sci. U.S.A. 77:3903-3907.

346. Silver, S. 1981. Mechanisms of plasmid-determinedheavy metal resistances, p. 179-189. In S. B. Levy, R. C.Clowes, and E. L. Koenig (ed.), Molecular biology,pathogenicity and ecology of bacterial plasmids. PlenumPress, New York.

347. SIhver, S., K. Budd, K. M. Leahy, W. V. Show, D.Hammond, R. P. Novick, G. R. Willsky, M. H. Malamy,and H. Rosenberg. 1981. Inducible plasmid-determinedresistance to arsenate, arsenite and antimony (III) inEscherichia coli and Staphylococcus aureus. J. Bacte-riol. 146:983-996.

348. Silver, S., and D. Keach. 1982. Energy-dependent arse-nate efflux: the mechanism of plasmid mediated resist-ance. Proc. Natl. Acad. Sci. U.S.A. 79:6114-6118.

349. Sinclair, M. I., and B. W. Holoway. 1982. A chromosom-ally located transposon in Pseudomonas aeruginosa. J.Bacteriol. 151:569-579.

350. Skinner, R. H.,sad E. Cundlffe. 1980. Resistance to theantibiotics viomycin and capreomycin in the Streptomy-ces which produce them. J. Gen. Microbiol. 120:95-104.

351. Skold, 0. 1976. R-factor-mediated resistance to sulfon-amides by a plasmid-borne, drug resistant dihydropter-oate synthase. Antimicrob. Agents Chemother. 9:49-54.

352. Skold,O., and A. Wldh. 1974. A new dihydrofolatereductase with low trimethoprim sensitivity induced byan R factor mediating high resistance to trimethoprim. J.Biol. Chem. 249:4324-4325.

353. Smith, C. J., S. M. Markowitz, and F. L. Macrlna. 1981.Transferable tetracycline resistance in Clostridiumd&ffl-cile. Antimicrob. Agents Chemother. 19:997-1003.

MICROBIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from



354. Smith, H. W. 1978. Arsenic resistance in enterobacteria:its transmission by conjugation and by phage. J. Gen.Microbiol. 109:49-56.

355. Smith, K., and R. P. Novick. 1972. Genetic studies onplasmid-linked cadmium resistance in Staphylococcusaureus. J. Bacteriol. 112:761-772.

356. Smith, S. L., D. Stone, P. Novak, D. P. Baccanari, andJ. J. Burchall. 1979. R plasmid dihydrofolate reductasewith subunit structure. J. Biol. Chem. 254:6222-6225.

357. Sompollnay, D., I. Hammerman, 0. Assaf, and A.Wodjani. 1978. Tetracycline resistance antigen fromStaphylococcus aureus. FEMS Microbiol. Lett. 4:23-26.

358. Stacey, K. A., and E. Simpson. 1965. Improved methodfor the isolation of thymine-requiring mutants of Esche-richia coli. J. Bacteriol. 90:554-555.

359. Stone, D., and S. L. Smith. 1979. The amino acidsequence of the trimethoprim-resistant dihydrofolate re-ductase specified in Escherichia coli by R-plasmid R67.J. Biol. Chem. 254:10857-10861.

360. Stuber, D., and H. Bujard. 1981. Organization of tran-scriptional signals in plasmids pBR322 and pACYC184.Proc. Natl. Acad. Sci. U.S.A. 78:167-171.

361. Suglyama, M., H. Kobayashi, 0. Nimi, and R. Nomi.1980. Susceptibility of protein synthesis to streptomycinin streptomycin-producing Streptomyces griseus. FEBSLett. 110:250-252.

362. Suglyama, M., H. Mochlzuki, 0. Nimi, and R. Nomi.1981. Mechanism of protection of protein synthesisagainst streptomycin inhibition in a producing strain. J.Antibiot. 34:1183-1188.

363. Suglyami, M., 0. Nimi, and R. Nomi. 1980. Susceptibilityof protein synthesis to neomycin in neomycin-producingStreptomycesfradiae. J. Gen. Microbiol. 121:477-478.

364. Summers, A. O., and G. A. Jacoby. 1977. Plasmid-determined resistance to tellurium compounds. J. Bac-teriol. 129:276-281.

365. Summers, A. O., and G. A. Jacoby. 1978. Plasmid-determined resistance to boron and chromium com-pounds in Pseudomonas aeruginosa. Antimicrob.Agents Chemother. 13:637-640.

366. Summers, A. O., and L. Kight-Oiliff. 1980. TnI generatedmutants in the mercuric ion reductase of the Inc Pplasmid, R702. Mol. Gen. Genet. 180:91-97.

367. Summers, A. O., L. Kight-Olilif, and C. Slater. 1982.Effect of catabolite repression on the mer operon. J.Bacteriol. 149:191-197.

368. Summers, A. O., and S. Silver. 1972. Mercury resistancein a plasmid-bearing strain of Escherichia coli. J. Bacte-riol. 112:1228-1236.

369. Summers, A. O., and S. Silver. 1978. Microbiol transfor-mation of metals. Annu. Rev. Microbiol. 32:637-672.

370. Summers, A. O., and L. I. Sugarman. 1974. Cell-freemercury (II)-reductase activity in a plasmid-bearingstrain of Escherichia coli. J. Bacteriol. 119:242-249.

371. Sutce, J. G. 1978. Nucleotide sequence of the ampicil-lin resistance gene of Escherichia coli plasmid pBR322.Proc. Natl. Acad. Sci. U.S.A. 75:3737-3741.

372. Sutdle, J. G. 1978. Complete nucleotide sequence ofthe Escherichia coli plasmid pBR322. Cold Spring Har-bor Symp. Quant. Biol. 43:77-90.

373. Suzuki, I., N. Takahashi, S. Shirato, H. Kawabe, and S.Mltauhash. 1975. Adenylylation of streptomycin inStaphylococcus aureus: a new streptomycin adenylyl-transferase, p. 463-473. In S. Mitsuhashi and H. Hashi-moto (ed.), Microbial drug resistance. University ParkPress, Tokyo.

374. Suzuki, Y., and S. Okamoto. 1967. The enzymatic acety-lation of chloramphenicol by the multiple drug-resistantEscherichia coli carrying R factor. J. Biol. Chem.242:4722-4730.

375. Svedberg, G., and 0. Skold. 1980. Characterization ofdifferent plasmid-borne dihydropteroate synthases medi-ating bacterial resistance to sulfonamides. J. Bacteriol.142:1-7.

376. Sykes, R. B., and M. Matthew. 1976. The P-lactamases of

gram-negative bacteria and their role in resistance to 1B-lactam antibiotics. J. Antimicrob. Chemother. 2:115-157.

377. Sykes, R. B., and M. Matthew. 1979. Detection, assayand immunology of 1-lactamases. In J. M. T. Hamilton-Miller and J. T. Smith (ed.), 3-Lactamases. AcademicPress, London.

378. Tai, P. C., B. J. Wallace, and B. D. Davis. 1974. Selectiveaction of erythromycin on initiating ribosomes. Bio-chemistry 13:4653-4659.

379. Tait, R. C., and H. W. Boyer. 1978. On the nature oftetracycline resistance controlled by the plasmidpSC101. Cell 13:73-81.

380. Tanaka, T., and B. Weisblum. 1975. Systematic differ-ence in the methylation of ribosomal ribonucleic acidfrom gram-positive and gram-negative bacteria. J. Bac-teriol. 123:771-774.

381. Taylor, D. P., J. Greenberg, and R. H. Rownd. 1977.Generation of miniplasmids from copy number mutantsof the R plasmid NR1. J. Bacteriol. 132:986-995.

382. Tennhammar-Ekman, B., and 0. Skold. 1979. Trimetho-prim resistance plasmids from different origin encodedifferent drug-resistant dihydrofolate reductases. Plas-mid 2:334-346.

383. Tezuka, T., and K. Tonomura. 1978. Purification andproperties of a second enzyme catalyzing the splitting ofcarbon-mercury linkages from mercury-resistant Pseu-domonas K-62. J. Bacteriol. 135:138-143.

384. Then, R. L., and P. Angehrn. 1982. Trapping of nonhy-drolyzable cephalosporins by cephalosporinases in En-terobacter cloacae and Pseudomonas aeruginosa as apossible resistance mechanism. Antimicrob. AgentsChemother. 21:711-717.

385. Thompson, J., and E. Cundlffe. 1980. Resistance tothiostrepton, siomycin, and sporangiomycin in actino-mycetes that produce them. J. Bacteriol. 142:455-461.

386. Thompson, C. J., R. H. Skinner, J. Thompson, J. M.Ward, D. A. Hopwood, and E. Cundliffe. 1982. Biochemi-cal characterization of resistance determinants clonedfrom antibiotic-producing Streptomyces. J. Bacteriol.151:678-685.

387. Thompson, C. J., J. M. Ward, and D. A. Hopwood. 1980.DNA cloning in Streptomyces: resistance genes fromantibiotic-producing species. Nature (London) 286:525-527.

388. Thorbjarnardottir, S. H., R. A. Magnusdottir, G. Eg-gertson, S. A. Kagan, and 0. S. Andresson. 1978.Mutations determining generalized resistance to amino-glycoside antibiotics in Escherchia coli. Mol. Gen. Gen-et. 161:89-98.

389. TImmis, K. N., F. Cabello, and S. N. Cohen. 1978.Cloning and characterization of EcoRI and HindIII re-striction endonuclease-generated fragments of antibioticresistance plasmids R6-5 and R6. Mol. Gen. Genet.162:121-137.

390. Tipper, D. J., and J. L. Strominger. 1965. Mechanism ofaction of penicillins: a proposal based on their structuralsimilarity to acyl-D-alanyl-D-alanine. Proc. Natl. Acad.Sci. U.S.A. 54:1133-1141.

391. Toblan, J. A., and F. L. Macrina. 1982. Helper plasmidcloning in Streptococcus sanguis: cloning of a tetracy-cline resistance determinant from the Streptococcus mu-tans chromosome. J. Bacteriol. 152:215-222.

392. Tomkch, P. K., F. Y. An, and D. B. Clewell. 1980.Properties of an erythromycin-inducible transposonTn917 in Streptococcus faecalis. J. Bacteriol. 141:1366-1374.

393. Towner, K. J. 1981. A clinical isolate of Escherichia coliowing its trimethoprim resistance to a chromosomally-located trimethoprim transposon. J. Antimicrob. Che-mother. 7:157-162.

394. Towner, K. J., and P. A. Pinn. 1981. A transferableplasmid conferring only a moderate level of resistance totrimethoprim. FEMS Microbiol. Lett. 10:271-272.

395. Trltton, T. R. 1977. Ribosome-tetracycline interactions.

VOL. 47, 1983 407


http://mm

br.asm.org/

Dow

nloaded from


408 FOSTER

Biochemistry 16:4133-4138.396. Tsen, J. L., L. E. Bry, and H. M. vadenElen. 1972.

Mechanisms and spectrum of streptomycin resistance ina natural population of Pseudomonas aerutinosa. Anti-microb. Agents Chemother. 2:136-141.

397. Tynecka, Z., Z. Gos, and J. Zajac. 1981. Reducedcadmium transport determined by a resistance plasmid inStaphylococcus aureus. J. Bacteriol. 147:305-312.

398. Tynecka, Z., Z. Goo, and J. ZaJac. 1981. Energy-depen-dent effilux of cadmium coded by a plasmid resistancedeterminant in Staphylococcus aureus. J. Bacteriol.147:313-319.

399. UhIl, B. E., and K. Nodstrom. 1977. R plasmid genedosage effects in Escherichia coli K-12: copy mutants ofthe R plasmid Rldrd-19. Plasmid 1:1-7.

400. Umeawa, H., M. Okanlh, S. Kondo, K. Hanana, R.Uthar, K. a, and S. MtAuhashi_. 1967. Phosphory-lative inactivation of aminoglycoside antibiotics by Es-cherichia coli carrying R factor. Science 157:1559-1561.

401. Umnzawa, Y., M. Yagisawa, T. Sawn, T. Takeuchi, H.Um_a, H. Ma*s_nol, and T. Tazak. 1975. Amino-glycoside 3'-phosphotransferase III, a new phospho-transferase resistance mechanism.-J. Antibiot. 28:845-853.

402. vwan Treek, U., F. Schmidt, and B. Wed a. 1981.Molecular nature of a streptomycin and sulfonamideresistance plasmid (pBP1) prevalent in clinical Esche-richia coil strains and integration of an ampicillin resist-ance transposon (TnA). Antimicrob. Agents Chemother.19:371-380.

403. Vasto,s. P., J. Altacha, ad S. Harford. 1980. 5-epi-Sisomicin snd 5-epi-gentamicin B: substrates for amino-glycoside-modifying enzymes that retain activity againstamnoglycoside-resistant bacteria. Antimicrob. AgentsChemother. 17:79-802.

404. Veka_tswra, P. S., an&H. C. Wu. 1972. Isolation andcharacterization of a phsophonomycin-resistant mutantof Escherichta coli K-12. J. Bacteriol. 110:935-9.

405. Ving, L. C. 1979. Antibiotic tolerance in producerorganisms, Adv. Appl. Microbiol. 25:147-168.

406. Volker, T. A., S. L{da, and T. A. Bickle. 1982. A singlegene coding fbr resistance to both fusidic acid andchloramphenicol. J. Mol. Biol. 154:417-425.

407. Wahl, G. M., R. A. Pdgett, ad G. R. Strwk. 1979. Geneamplification causes overproduction of the first threeenzymes of UMP synthesis in N-(phosphonacetyl)-L-aspartate resistant hamster cells. J. Biol. Chem.2s4:8679-8689.

408. Walker, J. B., an M. Skorvaga. 1973. Phosphorylationof streptomycin and dihydrostreptomycin by Streptomy-ces. J. Biol. Chem. 248:2435-2440.

409. Walker, M. S., and J. B. Walker. 1970. Streptomycinbiosynthesis and metabolism. J. Biol. Chem. 245:6683-6689.

410. Walla, L. J., J. M. Ward, and M. H. Rkchmnd. 1981.The location of sequences of TnA required for theestablishment of transposition immunity. Mol. Gen.Genet. 184:8046.

411. Watanakuakor , C. 1978. Antibiotic-tolerant Staphylo-coccus aureus. J. Antimicrob. Chemother. 4:561-568.

412. Wana, D. J., and J. L. S-o e. 1980. Sequence ofactive site peptides from the penicillin-sensitive D-ala-nine carboxypeptidase of Bacillus subtilis. J. Biol.Chem. 255:3964-3976.

413. WebUm, B. 1975. Altered methylation of ribonucleicacid in erythromycin-resistant Staphylococcus aureus,p. 199-206. In D. Schiessinger (ed.), Microbiology-1974. American Society for Microbiology, Washington,D.C.

414. Weshum, B., ad V. Delehn. 1969. Erythromycin-inducible resistance in Staphylococcus aureus: survey ofantibiotic classes involved. J. Bacteriol. 986:447-452.

415. Weblbin, B., S. B. Holder, and S. M. Halng. 1979.Deoxyribonucleic acid sequence common to staphylo-coccal and streptococcal plasmids which specify erythro-

mycin resistance. J. Bacteriol. 138:990-998.416. WI_*n, B., C. d , C. J. La, an V. De_n.

1971. Erythromycin-inducible resistance in Staphylococ-cus aureus: requirements for induction. J. Bacteriol.106:835-847.

417. Webs, A. A., S. D. Murphy, ad S. Slver. 1977. Mercuryand organomercurial resistances determined by plasmidsin Staphylococcus aureus. J. Bacteriol. 132:197-208.

418. Webs, A. A., J. L. Schottel, D. L. Clark, R. G. Beier,andS. Siver. 1978. Mercury and organomercut resist-ance with enteric, staphylococcal, and pseudomonadplamids, p. 121-124. In D. Schlessinger (ed.), Microbi-ology-1978. American Society for Microbiology, Wah-ington, D.C.

419. Webs A. A., S. Silver, and T. G. IUh. 1978. Cationtransport alteration association with 1asmid-de edresistance to cadmium in Staphylococcus aureus. Anti-microb; Agents Chemother. 14:856-865.

420. Wdeh, R. A., K. R. Jon, ad F. L. Msan. 1979.Transferable lincosamide-macrolide resistance in Bacte-roides. Plasmid 2:261-268.

421. Wlebaer, K., S. Schrandm S. Shale, and R. Scmitt.1981. Tetracycline resistance transposon Tn)nl:recA-dependent gene amplification and expression of tetracy-cline resistance. J. Bacteriol. 147:851-859.

422. WiDMoa, J. W., and D. B. Northep. 1976. Purificationand properties of gentamicin acetyltransferse I. Bio-chemistry 15:125-131.

423. WIsky, G. R., and M. H. Mamny. 1980. Effect ofarsenate on inorganic phosphate transport in Escherichiacoli. J. Bacteriol. 144:366-374.

424. WoIes, C. R., S. E. Skinner, and W. V. Sbaw. 1981.Analysis of two chloramphenicol resistance plasmidsfrom Staphylococcus aureus: insertional incivation ofCm resistance, mapping of restriction sites, and con-struction of cloning vehicles. Plasmid 5:245-258.

425. Wiedsl, }., aWn W. V. Skew. 1969. Kinetics of induc-tion and purification of chloramphenicol acetyltransfer-ase from chloramphenicol-resistant Staphylococcus aur-eus. J. Bacteriol. 96:1248-1257.

426. WIs, E. M., ad M. M. Abos-DogI. 1975. Splfonmideresistance mechanism in Escherichia coli: R-phsmidscan determipp sulfonamide-resistant dihydropterostesynthasis. Proc. Natd. Acad. Sci. U.S.A. 72:2621-2625.

427. Wttmna, H. G., G. StoSer, D. Apios, L. Ros, K.Tanak, M. Tamad, R. Takata, S. Dddo, E. Ot, mdS. Osawa. 1973. Biochemical and genetic studies of twodiffErent types of erythromycin resistant mutants ofEscherchia coli with altered ribosomal proteins. Mol.Gen. Genet. 127:175-189.

428. WiLtman-LIeb , B., and B. Grew. 1978. The primarystrcture of protein S5 from the small subunit of theE&cherichia cali ribosome. FEBS Lett. 95:91-98.

429. Wojian, A., R. M. Avtbon, a_ D.S _pd. 1976.lation and characerization of tetracyche remstance

proteins from Staphylococcus aureus and Eacherichlacoil. Antimicrob. Agents. Chemother. 9:526-534.

430. Wray, L. V., R. A. Jormen, md W. S R _. 1981.Identification of the tetracycline resistance pmmder andrepressor in trnsposon TnIO. J. Bacteriol. 147:2-304.

431. Yag, Y., and D. B. Clbrel. 1976. Plasmid4etrminedtetracycline resistance in Streptococcus faecalis: tan-demly repeated resistance deteminants in amplifiedforms of pAMml DNA. J. Mol. Biol. 102:583-600.

432. Yagachl, M., H. G. Wittman, T. Cabmon, M. deWide,R. Vllrold, A. Heruog, and A. BoSeon. 1976. Alterationof ribosomal protein S17 by mutation linked to neamineresistance in Escherichia coli. II. ocalization of theamino acid replacement in protein S17 from a neaAmutant. J. Mol. Biol. 104:617-620.

433. Y _mda, T., D. Tiper, and J. DavIs. 1968. Enzymaticinativation of streptomycin by R-factor resistant Esche-richia coli. Nature (London) 219:228-291.

434. Yamamoto, H., K. Hotta, Y. Okami, ad H. Umzwa.

MICROBIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from



1981. Ribosomal resistance of an istamycin producer,Streptomyces tenjimariensis, to aminoglycoside antibiot-ics. Biochem. Biophys. Res. Commun. 100:1396-1401.

435. Yamamoto, T., M. Tanaka, C. Nobara, Y. Fukunag, andS. Yamagah. 1981. Transposition of the oxacillin-hydro-lyzing penicillinase gene. J. Bacteriol. 145:808-813.

436. Yamamoto, T., S. Yamapta, K. Horl, and S. Yamagishi.1982. Comparison of transcription of flactamase genesspecified by various ampicillin transposons. J. Bacteriol.150:269-276.

437. Yamomoto, T., and T. Yokota. 1977. Beta-lactamase-directed barrier for penicillins of Escherichia coli carry-ing R plasmids. Antimicrob. Agents. Chemother. 11:936-940.

438. Yang, H. L., G. Zubay, and S. B. Levy. 1976. Synthesisof an R plasmid protein associated with tetracyclineresistance is negatively regulated. Proc. Nat]. Acad. Sci.U.S.A. 73:1509-1512.

439.Z , Y., J. E. Fiton, L. C. Packman, and W. V.Shaw. 1979. Characterization and comparison of chlor-amphenicol acetyltransferase variants. Eur. J. Biochem.

100:609-618.440. Za7denzig, Y., and W. V. Shaw. 1978. The reactivity of

sulphdryl groups at the active site of an R factor-specified variant of chloramphenicol acetyltransferase.Eur. J. Biochem. 83:553-562.

441. Zigelboim, S., and A. Tomaz. 1980. Penicillin-bindingproteins of multiply antibiotic-resistant South Africanstrains of Streptococcus pneumoniae. Antimicrob.Agents Chemother. 17:434-442.

442. Zimmerman, R. A., R. C. Moellering, and A. N. Wein-berg. 1971. Mechanism of resistance to antibiotic syner-gism in entercocci. J. Bacteriol. 105:873-879.

443. ZoIg, J. W., and U. J. Hlnang. 1981. Characterization ofanR plasmid-associated, trimethoprim-resistant dihydrofo-late reductase and determination of the nucleotide se-quence of the reductase gene. Nucleic Acids Res. 9:697-709.

444. Zupandc, T. J., S. R. King, K. L. Pogue-Geile, and S. R.Jaskunas. 1980. Identification of a second tetracycline-inducible polypeptide encoded by TnlO. J. Bacteriol.144:346-355.

VOL. 47, 1983


http://mm

br.asm.org/

Dow

nloaded from


Documents

Plasmid-Determined Resistance to Antimicrobial Toxic Metal ... · Plasmid-Determined Resistance to Antimicrobial Drugs and Toxic Metal Ions in Bacteria T. J. FOSTER Microbiology Department,