Bacterial Resistance to the Tetracyclinesmmbr.asm.org/content/42/4/707.full.pdfHowever, tetracycline uptake by mem-brane vesicles wasreported to be energy inde-pendent andapparently

MICROBIOLOGICAL RIzviEws, Dec. 1978, p. 707-7240146-0749/78/0042-0707$02.00/0Copyright i 1978 American Society for Microbiology

Vol. 42, No. 4

Printed i U.S.A.

Bacterial Resistance to the TetracyclinesI. CHOPRA* AND T. G. B. HOWE

Department of Bacteriology, University of Bristol, Bristol BS8 ITD, United Kingdom

NTRODUCION ............................................................. 707MECHANISMS OF PE?NET1RATION OF THE TEITRACYCLINES INrO BAC-

TERIA ............................. 707General Features of Tetracycline Uptake ................................... 707Passage of Tetracyclines Across the Outer Membrane of Gram-NegativeBacteria and Binding to the Surface of the Inner Membrane.708

Binding of Tetracyclines to the Cytoplasmic Membrane of Gram-PositiveBacteria.. 709

Passage of Tetracyclines Across the Cytoplasmic Membranes of Gram-Posi-tive and -Negative Bacteria ............................................... 709

PLASMID-MEDIATED RESISTANCE ........... 711Resistance Phenotypes........... 711Proteins Associated with Plasmid-Determined Resistance in Eswherichia coliand Staphylococcus awu eus ................................................ 713

Function of Proteins Associated with Plasmid-Determined Tetracycline Re-sistance ........................................ 714

Membrane Architecture in Relation to the Binding and Function of ProteinsAssociated with Resistance ............................................... 715

RELATION BEIWEEN TETRACYCLINE GENE COPY NUMBER AND RE-SISTANCE LEVEL.715

ORIGIN OF TETRACYCLINE RESISTANCE GENES.717CONCLUSIONS .............................................................. 718LITERATURE CITED ................... 718

What boots it at one gate to make defense, and atanother to let in the foe?

John Milton, Samson Agonistes

INTRODUCTIONThe tetracyclines are a family of broad-spec-

trum antibiotics which inhibit protein synthesisby preventing the binding of amino acyl transferribonucleic acid to the bacterial ribosome (70).Although a large number of tetracycline deriv-atives are now known (15), relatively few havefound extensive clinical use apart from tetracy-cline itself, chlortetracycline, oxytetracycline,and, more recently, doxycycline and minocycline(for the structures of these compounds, see ref-erence 15). These tetracyclines provide safe, in-expensive, and effective treatment for many bac-terial infections (159) and are also used for an-timicrobial prophylaxis (131). However, theemergence of tetracycline-resistant bacteria hasconsiderably reduced the usefulness of these an-tibiotics (159).

Studies on the mechanism of resistance totetracycline (and antibiotics in general) may beexpected (i) to contribute toward the more ef-fective deployment of these drugs against bac-teria and (ii) to lead to the development of newderivatives with improved antibacterial spectra

and less susceptibility to bacterial resistance.Although the molecular basis of resistance tothe tetracyclines has, until recently, been poorlyunderstood, sufficient information has now ac-crued to warrant this review of the topic. Recentstudies in this area have employed many of theadvanced techniques now available in moleculargenetics and membrane biochemistry.Because resistance to the tetracyclines in the

majority of clinical isolates probably results fromacquisition of plasmids, this review will primar-ily be concerned with the mechanism of plasmid-mediated resistance to the tetracyclines. Be-cause resistance results from changes in the cellenvelope, it will be logical first to consider thepassage of tetracyclines across this structure insensitive bacteria. Subsequent sections deal withthe nature of the membrane changes, gene dos-age and expression of resistance, and, finally, thepossible origin of the resistance genes.

MECHANISMS OF PENETRATION OFTHE TETRACYCLINES INTO BACTERIAGeneral Features of Tetracycline UptakeThe energy dependence of tetracycline accu-

mulation by bacteria was first reported in 1963(4). However, as recently as 1973 Franklin (60)stated, "the active transport of the tetracycline

707

on May 17, 2018 by guest

http://mm

br.asm.org/

Dow

nloaded from

http://mmbr.asm.org/

708 CHOPRA AND HOWE

antibiotics poses many more questions about themechanism of this phenomenon than can yet beanswered." Some of the puzzles noted by Frank-lin (60) now seem closer to resolution, but thereis still no model for tetracycline accumulationwhich is consistent with all of the data. Becausethe envelope structures of gram-positive and-negative bacteria differ, the mechanisms of tet-racycline uptake by the two bacterial types mustbe considered separately. However, in both casesthe energy-dependent movement of tetracy-clines across the cytoplasmic membrane is prob-ably preceded by an energy-independent bindingto the cell surface that most likely representsadsorption to the cytoplasmic membrane.

Passage of Tetracyclines Across the OuterMembrane of Gram-Negative Bacteria andBinding to the Surface of the Inner

MembraneThe envelope of gram-negative bacteria (as

typified by Escherichia colt) consists of threedistinct regions (37): the outer membrane, theperiplasmic space, and the inner (cytoplasmic)membrane. Most gram-negative bacteria are re-sistant to many antibiotics effective againstgram-positive organisms, and this intrinsic in-sensitivity is attributable to the inability of thedrugs to penetrate the outer membrane (113).An approximate estimate of the ability of an

antibiotic to penetrate the outer membrane canbe derived from comparison of the minimumdrug concentrations required to inhibit thegrowth of E. coli and Staphylococcus aureus(Table 1). On this basis the outer membrane canmost readily exclude rifamycin, fusidic acid, andpenicillin G, but is essentially permeable to D-cycloserine and streptomycin. The ratios of in-hibitory activity for the tetracyclines are lowand in some cases approach unity (Table 1),indicating that these molecules can readily pen-etrate the E. coli outer membrane.Our understanding of the permeability of the

outer membrane of gram-negative bacteria hasrecently been considerably extended, in partic-ular by the work ofBraun, Henning, Lugtenberg,Nakae, and Nikaido (for recent reviews, see ref-erences 18 and 44). The nomenclatures used byvarious research groups to describe the majorouter membrane proteins of E. coli K-12 differ(18, 44, 77). In this article we shall refer to theproteins by the terminology of Henning (18, 44,72, 78, 80). At least three of the major outermembrane proteins of E. coli (proteins Ia, Tb,and 11*) appear to forn aqueous pores whichspan the membrane and allow diffusion of smallhydrophilic molecules into the periplasmic space(9, 18, 44, 56, 103, 110, 119, 153, 154). Although

TABLE 1. Relative efficacy ofsome antibioticsagainst E. coli

Antibiotic

Raimtotrc

Rifamycin SV ..........................

Fusidic acid ............................Penicilin G ............................IAncomycin ............................Ampicilin .............................Erythromycin ..........................Chloramphenicol.Doxycycline ............................Chlortetracycline.Tetracycline ...........................Minocycline ............................Oxytetracycline.Streptomycin ..........................r.-Cycloserine ..........................

atio of min-aum inhibi-ry concen-mtions (E.oli:S. au-reus)

o0,oooa10,000-5,000"2,O0Oc

33-255a130C

70-140a3cgd8d

3d2d2d

1-2c0.5-2.0a

a From data quoted in reference 73.b From data in reference 25 and from D. J. Sinden

(unpublished data).From data in reference 71.

d From data in reference 15.

we have been skeptical that these proteins formpores (30), there is considerable evidence to sup-port the pore model (18, 44) which, with somereservations (29), we now accept. In addition tocontroversy concerning pores, a complete under-standing of the organization of the outer mem-brane is still lacking as, for instance, there ap-pears to be insufficient space in the outer half ofthe bilayer to accommodate the outermembraneproteins (compare data in references 80 and 136).Because tetracyclines have molecular weights

of less than 500 and the pores can accommodatemolecules of up to 650 daltons (40), these anti-biotics, in principle, could diffuse through thepores. This does occur for tetracycline (7, 29, 56,113, 119), as might be expected for a hydrophilicmolecule (Table 2). Tetracycline diffuses pref-erentially through pores formed from protein Ia(29, 56, 119), but those features of the moleculethat consign it to this route are unknown be-cause protein Ia pores are used by several othercompounds (e.g., nucleotides, fi-lactams, andcopper ions) (56, 101, 119, 153, 154) which haveno apparent structural similarity to tetracycline.Both chlortetracycline and oxytetracycline alsodiffuse through protein Ia pores (56), which sug-gests that this may be the route of entry for allhydrophilic tetracyclines. Minocycline, however,is markedly more lipophilic than other tetracy-clines (Table 2) and appears to penetrate theouter membrane by diffing across the hydro-carbon interior of the bilayer (7, 29). Doxycy-

MICROBIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from


BACTERIAL RESISTANCE TO TETRACYCLINES 709

cline may also be too hydrophobic (Table 2) toenter through pores.Because the tetracyclines are acids which can

ionize in aqueous solution (Fig. 1 and Table 3),it is important to consider the nature of thespecies that diffuse across the outer membrane.Tetracycline hydrochloride and oxytetracyclinehydrochloride each have four ionic or molecularforms (Fig. 1 and Table 3). At neutral pH theforms TH2, TH- and OH2, 0H- predominate(11, 84). Because a Donnan equilibrium existsacross the outer membrane (negative chargeinside; 142), diffusion of the anionic species islikely to be severely hindered. However, theformation of antibiotic-cation coordination com-plexes (2), for example with magnesium (Fig. 1),could produce cationic chelates which wouldrespond to the Donnan equilibrium. Severalstudies (11, 84) indicate that only the productsof the first dissociation (i.e., TH2 and OH2) are

likely to penetrate gram-negative bacteria. Fur-thermore, the sensitivity of E. coli, Aerobacteraerogenes, and S. aureus to tetracycline is de-pressed in the presence of increasing concentra-tions of magnesium (11, 84, 141). This responsehas been attributed to a direct effect on the

TABLE 2. Partition coefficients of some of thetetracyclinesa

Partition coefficient

Antibiotic Chloroform/ Octanol/water water

Minocycline 30 1.1Doxycycline 0.48 0.60Tetracycline 0.09 0.036Oxytetracycline 0.007 0.025

a Based on data quoted in reference 8.

A) TH, H(H) + TH 2 (H') +K)K2

B) JM ') + 0IH,- (Hi +

"7K,

v.- IH+) -+ T2

TH K, *2

, M) +

ow K3+rs Mg"

FIG. 1. Dissociation schemes for (A) tetracyclinehydrochloritde and (B) oxytetracycline hydrochloride.Redrawn from reference 84 with permission.

structure of tetracycline itself (141) and presum-ably reflects the formation of (TH-Mg)+ com-

plexes which reduce the content of TH2 species(11, 84). Passage of tetracycline and oxytetracy-chine through protein Ia pores, therefore, pre-sumably involves diffusion of TH2 and OH2 spe-

cies down concentration gradients.A dissociation scheme for the ionizable groups

in minocycline apparently has not been re-

ported. However, in view of the Donnan equilib-rium across the outer membrane (142), minocy-cine is likely to diffuse through hydrophobicregions either in a molecular (uncharged) fornor as a cationic chelate.Because the pH in the periplasmic space is

below that in the extracellular medium (142),tetracyclines entering the periplasmic space are

likely to become protonated. Protonation maypromote binding of antibiotic molecules to thesurface of the membrane by interaction withanionic residues. These events could representthe initial energy-independent phase of uptakenoted by several authors (43, 97, 98).

Binding of Tetracyclines to theCytoplasmic Membrane ofGram-Positive Bacteria

Little is known about the accumulation oftetracyclines by gram-positive bacteria. Becausethe gram-positive bacterial cell wall is predomi-nantly anionic (36), chelation of tetracyclineswith magnesium in the medium may be impor-tant in the binding of antibiotic to the cytoplas-mic membrane. However, Dockter and Magnu-son (45) present evidence that unchelated anti-biotic may be involved in the energy-independ-ent binding of tetracycline to the surface of thestaphylococcal membrane.

Passage of Tetracyclines Across theCytoplasmic Membranes ofGram-Positive and -Negative

BacteriaStudies on the passage of tetracyclines across

the bacterial cytoplasmic membrane have beenconfined almost entirely to E. coli, S. aureus,

and Bacillus megaterium. Franklin (60) re-

viewed earlier work dealing with this aspect oftetracycline accumulation, which appears to be

TABLE 3. Ionization constants (K1, K2, K3) of tetracycline and oxytetracycline hydrochlorides and stabilityconstants Ks of the complexes with magnesiuma

Tetracycline Oxytetracycline

pKI pK2 p16 pKs pK, pK2 pK3 pKs

3.3-3.45 7.68-7.92 9.57-9.81 4.16-4.29 3.1-3.46 7.26-7.76 9.11-9.70 3.80-3.96a Based on data quoted in reference 84.

VOL. 42, 1978


http://mm

br.asm.org/

Dow

nloaded from


710 CHOPRA AND HOWE

a true active transport phenomenon. A simpleexplanation of the data (60) would be that trans-port of the tetracyclines requires protein carriermolecules dependent on membrane energiza-tion. We can consider this hypothesis by evalu-ating (i) data on the nature of energy couplingto transport, (ii) evidence which favors mem-brane carriers. and (iii) information on the na-ture of the carriers.Recent studies clearly establish two major

types of active transport systems in gram-nega-tive bacteria (for reviews, see references 74, 76,and 134). One type is membrane bound and canbe observed in membrane vesicles. The othertype is sensitive to osmotic shock and requiresperiplasmic binding proteins to produce activetransport. The membrane-bound systems deriveenergy from an electrochemical gradient of pro-tons, the protonmotive force, which is the directdriving force for active transport. The transportsystems involving binding proteins are more di-rectly coupled to phosphate bond energy, prob-ably as adenosine 5'-triphosphate. Althoughgram-positive bacteria have no periplasmic bind-ing proteins, they can employ both adenosine 5'-triphosphate and the protonmotive force todrive transport (76).

Tetracycline uptake by E. coli is not impairedby osmotic shock (63), which suggests that amembrane-bound system powered by the pro-ton-motive force could be involved in drug trans-port. However, tetracycline uptake by mem-brane vesicles was reported to be energy inde-pendent and apparently did not result in accu-mulation of antibiotic against a concentrationgradient (59). The results of the vesicle experi-ments do not necessarily contradict the dataobtained with whole cells. Franklin (59) did notcheck the transport capacity of his vesicles forany other solutes, and the use of Triton X-100in some experiments (59) could have preventedthe generation of a protonmotive force due tothe formation of transmembrane pores by thedetergent (129). More recently Levy et al. (98)reported that the sensitivity to tetracycline of anunc adenosine triphosphatase- mutant (strainAN120) of E. coli K-12 was similar to that of thewild-type parental strain. Because adenosine 5'-triphosphate-driven, but not respiration-driven,transport is impaired in strain AN120 (134),these data (98) could support the model thattetracycline is transported across the innermembrane by a carrier driven by the proton-motive force. However, this interpretation iscomplicated by the fact that Levy et al. (98) didnote that the kinetics of tetracycline accumula-tion differed between strain AN120 and its wild-type parent. The apparent discrepancy whenstrain AN120 was used might relate to the leak-

iness of this mutant, which can exhibit as muchas 50% of the parental adenosine triphosphataseactivity under certain conditions (134). Unfor-tunately, Levy et al. (98) did not estimate aden-osine triphosphatase activity in their experi-ments with strain AN120, so that their resultson the nature of energy coupling to tetracyclineaccumulation remain equivocal. Nothing isknown of the nature of energy coupling to tet-racycline transport in gram-positive bacteria.

If transport of tetracyclines is mediated bycarriers, then it might be expected that theinitial rates of uptake would be dependent onthe external antibiotic concentrations up to thepoint when all of the available carrier sites aresaturated with antibiotic. Although Franklinand Higginson (63) were unable to demonstratesaturation kinetics for the initial rate of uptakeby whole cells, Franklin (59) showed that uptakeby membrane vesicles was saturable. Data ob-tained more recently (45, 47, 158) provide con-vincing evidence that uptake of tetracycline andchlortetracycline by several bacteria, includingE. coli, S. aureus, and B. megaterium, demon-strates saturation kinetics. Furthermore, in B.megaterium there is evidence from biophysicalstudies that the membrane carrier for tetracy-cline is a protein (47). Data of another type alsosupport the notion of a protein carrier for chlor-tetracycline transport. Dockter and co-workers(46, 47) showed that energy-dependent chlor-tetracycline accumulation in S. aureus and B.megaterium is influenced by the fluidity ofmem-brane fatty acids. These results favor an activetransport system for chlortetracycline in whichthe movement of the protein carrier moleculesbecomes restricted when certain membranefatty acids are in a crystalline state.By analogy with the transport of other anti-

biotics and inhibitory analogs (3), the tetracy-clines are presumably accumulated by illicit useof normal carrier systems. Several suggestionshave been made concerning the identity of thepostulated carriers. Franklin (60) proposed thattetracycline could be accumulated as an anti-biotic-cation coordination complex via a mobilecarrier involved in divalent cation transport. In-deed, chlortetracycline may be transportedacross the staphylococcal membrane as an anti-biotic-magnesium complex (45). Magnesium canbe transported into E. coli by one or the otherof two separate energy-dependent carrier sys-tems (112, 116) which are presumably located inthe cytoplasmic membrane. The first system isexpressed constitutively and in addition to mag-nesium can transport other cations, such as co-balt (112, 116). The lack of total substrate spec-ificity exhibited by this system indicates that itmight transport magnesium-tetracycline com-

MICROBIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from



plexes. The second magnesium transport systemis less likely to be a candidate for tetracyclineaccumulation because it is active only at lowextracellular magnesium concentrations and issubstrate specific (112, 116). Mutants lackingthese magnesium transport systems accumulatetetracycline as readily as wild-type bacteria (7),and, therefore, at least in E. coli magnesiumtransport and tetracycline accumulation are notlikely to be directly related. Transport of anti-biotic into bacteria via the magnesium system isalso unlikely for other reasons. The selectivetoxicity of the tetracyclines toward an infectingorganism depends upon the failure of the anti-biotics to be accumulated by aian cells(58). Because mammalian cells have an energy-dependent magnesium transport system (10),they might be expected to accumulate anti-biotic-magnesium complexes. Therefore, theability of bacterial cells to accumulate antibioticpresumably cannot depend on magnesium trans-port.

Tetracycline, chlortetracycline, and oxytetra-cycline could be accumulated in E. coli and S.aureus by the C4-dicarboxylic acid transportsystem because oxaloacetate apparently com-petes with the antibiotics for uptake (108, 109,118). In E. coli this transport system is inducibleand requires cyclic adenosine 3',5'-monophos-phate as part of a positive control system for itsinduction (100). Indeed, some tetracyclines(doxycycline and chlortetracycline, but appar-ently not tetracycline, oxytetracycline, or meth-acycline) may be transported into Salmonellatyphimuriun by systems that are cyclic adeno-sine 3',5'-monophosphate dependent (3). Thesestudies in S. typhimurium (3) imply that sepa-rate carriers might be utilized by different tet-racycline derivatives. Few data are available toclarify the situation, but other studies show thatseparate carriers could be involved in the trans-port of minocycline and tetracycline across theinner membrane of E. coli (41).

Although, as mentioned above, mammaliancells are impermeable to tetracyclines (58), iso-lated mitochondria can accumulate the antibiot-ics (22, 58) in a manner which depends upon therespiratory state ofthe mitochondria (22). Theseobservations are at least consistent with the viewthat passage of tetracyclines across biologicalmembranes could depend upon transport by di-carboxylate systems because mitochondria dopossess such a transport system (86). The im-permeability of mammalian cytoplasmnic mem-branes presumably might reflect absence of di-carboxylate trnsport systems equivalent tothose found in bacteria and mitochondria.Although the preceding description favors a

model involving membrane carriers for the ac-

tive transport of tetracyclines into bacteria,some authors would probably disagree with thisinterpretation. It should be possible to clarifysome ofthe questions concerning the mechanismof accumulation of tetracyclines in bacteria bythe more extensive use of mutants impaired inenergy transduction (134) and the C4-dicarbox-ylic acid transport system (99). A model for thepassage of tetracycline across the E. coli enve-lope will be presented (Fig. 2) after a discussionof the mechanism of plasmid-mediated resist-ance.PLASMID-MEDIATED RESISTANCE

Resistance PhenotypesPlasmids which confer resistance to the tet-

racyclines have been found in at least 25 bacte-rial species, including both gram-negative andgram-positive types (19). Early studies on thenature of plasmid-determined tetracycline re-sistance in E. coli and S. aureus established thatfull expression of resistance was obtained onlyafter cells had been exposed to a sub-inhibitoryconcentration of the drug (60). These observa-tions have subsequently been confirmed for alarge number of plasmids (e.g., see references 31and 55), and the only naturally occurring plas-mid-linked tetracycline determinant to be ex-pressed constitutively is that of plmmid pAMalfound in Streptococcus faecalis (33). Expressionof inducible resistance is likely to be under neg-ative regulatory control, and proteins with re-pressor-like properties have been isolated fromextracts of E. coli containing R222 (163) andpSC101 (144).A wide variety in the characteristics of expres-

sion of resistance to the tetracyclines by natu-rally occurring R plasmids in E. coli has been

Proteins fronm Tcresistance region ofTnlO pSC1O1 Or membrane

i57< protein la pore

15K UK' <mr mebrane36K tetracycline transport protein

15K 34K*t Membrane bound polysome

FIG. 2. Model for tetracycline accumulation in E.coli and its blockade by plasmid-specified proteins.General features of envelope structure have beenomitted for clarity. The diagram shows the possiblestes of interaction in the envelope ofproteins speci-fied by the tetracycline resistance regions of TnlOandpSC101. The organization ofthe inner membranetransport protein is based on the model of Singer(135). Four proteins are specified by the resistanceregion ofpSC101, but only two have been assigned tothe resistance region of ThlO. K, Kilodalton;location at the site ofmembrane-bound polysomes istentative; _b, located in the mnicell envelope, butsite (if any) in the whole cell envelope is unknown.

VOL. 42, 1978


http://mm

br.asm.org/

Dow

nloaded from


712 CHOPRA AND HOWE

found (23, 55). This prompts the question ofwhether the different phenotypic classes of Rplasmid result from differences in the expressionof a common tetracycline determinant on differ-ent plasmids. The tetracycline resistance genes

of the plasmid R100-1 (a derepressed analog ofR222; 51) are contained in an element calledtransposon 10 (TnlO; 34, 54, 88). This elementcan be translocated from one replicon to anotherby a recombination process which conserves thestructure of the transposon and which occurs

independently of the recA+ product of the hostcell (see references 34 and 87 for reviews). Dis-semination of TnlO could therefore account forthe evolution of R plasmids specifying tetracy-cline resistance, and recent evidence (123) doesindeed indicate that several plasmids contain a

determinant related to TnlO. That altered tet-racycline resistance phenotypes may arise fromdifferences in the transcription of one set ofresistance genes is supported by the studies ofTait and co-workers (104, 146, 147), who founddifferences in the expression of tetracycline re-sistance in derivatives of pSC101 which wereobtained by insertion of fragments of deoxyri-bonucleic acid (DNA) into the EcoRI orHindmcleavage sites of the plasmid. The EcoRI site inpSC101 probably lies within one ofthe structuralgenes for a polypeptide involved in resistance(104, 146), and some altered phenotypes havebeen attributed to disruption ofa structural gene(104, 146). On the other hand, insertion of DNAat the EcoRI site also seems to affect transcrip-tion of an adjacent gene by altering ribonucleicacid polymerase attachment sites (147). Inser-tion of DNA at the HindIII site does not inac-tivate a structural gene but alters transcriptionbecause the HindIII site is located in a promoterfor one or more of the structural genes involvedin resistance (146).Although, as mentioned above, one group of

plasmids contains TnlO, other plasmids existwhose tetracycline resistance determinants are

either only partially related or seemingly totallyunrelated to TnlO. Chabbert and Scavizzi (23)noted that expression of plasmid-mediated re-

sistance to minocycline and tetracycline in E.coli resembled plasmid-determined macrolideresistance in S. aureus. Because macrolide re-

sistance in the latter involves ribosomal modifi-cation, Chabbert and Scavizzi (23) suggestedthat plasmids which confer resistance to tetra-cycline and minocycline might do so by provok-ing alterations at the 30S ribosomal subunitrather than the membrane changes (see below)determined by plasmids containing TnlO. Thissituation certainly implies that some tetracy-cline resistance plasmids code for products not

MICROBIOL. REV.

specified by Tnl0. The plasmid pIP69 (formerlydesignated RIP69) belongs to that group postu-lated to confer resistance solely by ribosomalmodification (23), but evidence for some degreeof base similarity between the tetracycline re-sistance regions of this plasmid and R100 (con-taining Tnl0) is provided by the finding of re-combination between tetracycline-sensitivepoint mutants of these two plasmids (T. J. Fos-ter, personal communication). Therefore, thetetracycline resistance region ofpIP69 may havearisen by microevolution from TnlO DNA. Theproducts of pSC101 involved in tetracycline re-sistance appear to be totally unrelated to thosespecified by TnlO because the resistance regionof pSC101 has no homology with that in R6 (35),which is believed to contain TnlO (123). Al-though tests (e.g., immunological cross-reactiv-ity) to compare the similarity of proteins speci-fied by the resistance regions of TnlO andpSC101 have yet to be performed, both elementscode for membrane proteins which exclude tet-racycline from the cell (see below), and it wouldtherefore seem that, although some tetracyclineresistance regions may have evolved separatelyfrom one another, in each case their productsact in a similar manner.

In S. aureus there are two phenotypic patternsof high-level resistance to the tetracyclines (5).One phenotype is determined by plasmid-lo-cated genes, and the other is determined bychromosomal genes. Because all of the plasmidswhich specify tetracycline resistance in S. au-reus have a similar size (2.7 to 2.8 megadaltons;19, 92) and confer similar levels of resistance tothe antibiotic, it has been suggested (92) thatone plasmid (i.e., a plasmid containing the sametetracycline resistance determinant) has spreadthroughout the staphylococci. The relationshipbetween the two loci determining tetracyclineresistance is unknown. However, the differentphenotypes might be explained by assuming thatthe resistance genes are contained in a transpo-son, the expression of which is altered by inte-gration into the chromosome. There is evidencethat genes for resistance to some antibiotics areparts of transposons in S. aureus (117, 130), andin gram-negative bacteria transposition of TnlOfrom a plasmid to a chromosomal locus hasoccurred under natural conditions (50).

Clearly, studies on the nature and function ofthe products of various tetracycline resistancedeterminants will contribute to the unravellingof the basis of the multiple phenotypes. Detailedstudies have as yet been confined to the E. coliplasmids pSC101 and R100-1, but a little infor-mation is also available on the nature ofplasmid-determined resistance in S. aureus.


http://mm

br.asm.org/

Dow

nloaded from



Proteins Associated with Plasmid-Determined Resistance in Escherichia coli

and Staphylococcus aureus

Wojdani et al. (160) were the first authors toclaim isolation of proteins specified by tetracy-cline resistance determinants. In E. coli theydetected an envelope protein of apparent molec-ular weight 50,000, and in S. aureus they de-tected a membrane protein of apparent molec-ular weight 32,000. The two proteins were im-munologically unrelated. It is difficult to assigneither of these proteins as products of tetracy-cline resistance genes because the authors com-pared proteins produced by bacteria carryingplasnids with those from plasmidless organisms.Thus the proteins isolated could well representplaid products other than proteins associatedwith tetracycline resistance. Furthermore, in-duction of resistance in S. aureus caused no

detectable increase in the content of the mem-brane protein (160), and the molecular weight of50,000 reported-for the E. coli protein is higherthan that of any of the proteins now identifiedas products of R-factor tetracycline resistancedeterminants (see below). The most recent stud-ies of Wojdani and his co-workers (139) still donot provide totally convincing evidence that the32,000-dalton staphylococcal protein is involvedin tetracycline resistance. The remainder of theexperiments summarized in this section eitherhave involved comparison of strains which areisogenic apart from possession of tetracyclineresistance determinants or have followedchanges in protein content after induction ofresistance. We can, therefore, be more confidentthat the proteins detected are products of thetetracycline resistance genes.The E. coli plasmid pSC101 contains genetic

information sufficient for its replication and theexpression of tetracycline resistance (35). Resist-ance is inducible and results in the increasedsynthesis of three polypeptides of molecularweights 34,000, 26,000, and 14,000 (104, 144, 146).In minicelLs these proteins are located in theenvelope (144), but their location in whole cellsis unknown. A fourth protein, of molecularweight 18,000, is also associated with pSC101-mediated resistance, but its synthesis is consti-tutive (144). A protein of molecular weight36,000 has been tentatively identified as therepressor of the pSC101 tetracycline resistanceregion (144). However, transcriptional control ofthe genes coding for the proteins of molecularweights 34,000, 26,000, and 14,000 is seeminglynot achieved by regulating the synthesis of asingle polycistronic messenger ribonucleic acidmolecule, because the structural genes con-

cerned are not contiguous (145).As noted earlier, the genes conferring induc-

ible resistance to tetracycline in the E. coli plas-mid R100-1 are contained in Tn1O. Levy and co-workers (163) detected in minicells two R222-coded proteins whose synthesis was stimulatedafter incubation in the presence of tetracycline.One ofthese proteins, ofmolecular weight 36,000(163), is preferentially located in the minicellenvelope (96, 163), whereas the second protein,of molecular weight 15,000, is located in thecytoplasmic fraction (163). These authors alsodemonstrated synthesis ofthe 36,000-dalton pro-tein in a cell-free system for coupled transcrip-tion and translation directed by R222 DNA(163).

Until recently, the location of these proteinsin whole cells of E. coli was unknown, andindeed our recent experiments (I. Chopra, P. R.Ball, and T. G. B. Howe, manuscript in prepa-ration) have been directed toward clarificationof the problem. In a strain containing a chro-mosomally integrated TnlO element we havedetected an inner membrane protein of molecu-lar weight 15,000 whose synthesis is apparentlyinducible by tetracycline and which comprisesabout 4,000 molecules per membrane. This pro-tein seems certain to be identical to that de-tected by Levy and co-workers (163) and, in viewof the specific derepression of its synthesis bytetracycline, is unlikely to be a protein coded bya region of TnlO not involved in expression ofresistance. Levy et al. (98) reported that theprotein of molecular weight 36,000 was locatedin the inner membrane of whole bacteria in-hibited with rifamycin, but we have been unableto detect this protein in the envelope of resistantbacteria grown under physiological conditions,suggesting that normally it may only be presentin very small amounts. As noted before, thetetracycline resistance region of pSC101 deter-mines the synthesis of five proteins (includingthe repressor), but to date only two proteins(molecular weights, 36,000 and 15,000) have beenidentified as probable products of the resistanceregion of TnlO. Consideration of the coding ca-pacity of the resistance region of TnlO (Table 4)suggests that one or more as-yet-unidentifiedproteins (with a total molecular weight not ex-ceeding about 27,000) could be involved in re-sistance. Indeed, studies on the location of ribo-some binding sites within the central portion ofTnlO (i.e., excluding the terminal inverted re-peat sequences) are consistent with the synthesisof a TnlO-specified polypeptide of approximatemolecular weight 24,000 (21).

Franklin and Foster (61) showed that plasmid-determined tetracycline resistance in E. coli is

VOL. 42, 1978


http://mm

br.asm.org/

Dow

nloaded from


714 CHOPRA AND HOWE

TABLE 4. Coding capacity of the region conferringtetracycline resistance in TnlO

Region of TnlO Length(base pairs)

Containing control system for expression oftetracycline resistance and the resistancegenes ................... ............ 3,125a

Containing structural gene for 36,000-daltonrepressor proteinb. .................... 960C

Containiny structural gene for 36,000-daltonprotein ............... .............. 960c

Containing structural gene for 15,000-daltonprotein ............................ 400c

Containing operator and promoter control-ling synthesis of 15,000- and 36,000-daltonproteins, and promoter for repressor .... 95'

Unassigned to any known product ........ 710a From the data of Jorgensen et al. (85).b Molecular weight of TnlO-determined repressor

unknown; calculation based on tentative value forpSC101 repressor (144).

c Calculated on the basis that a protein of molecularweight 45,000 contains 400 amino acids (103).

d Molecular weight taken from Yang et al. (163).'Based on the assumptions that the lactose opera-

tor sequence has a minimum length of 17 base pairs(6) and that a promoter involved in expression ofpSC101-mediated tetracycline resistance is 39 basepairs long (146).

impaired by subjecting bacteria to osmoticshock. Because such treatment releases peri-plasmic proteins (79), the data of Franklin andFoster are consistent with a periplasmic locationfor one or more of the proteins involved inresistance. However, we have found (I. Chopraand S. J. Eccles, Proc. Soc. Gen. Microbiol., inpress) that a protein IV (lipoprotein)-deficientmutant of E. coli, known to leak periplasmicproteins (69, 82), allows normal expression ofTnlO-mediated resistance. Therefore, none ofthe proteins involved in resistance is likely to beperiplasmic. On the other hand, expression ofresistance is reduced in mutants lacking outermembrane protein Ia or containing deep roughlipopolysaccharide (Chopra and Eccles, inpress), which possibly suggests an outer mem-brane location for a TnlO-specified protein.

Function of Proteins Associated withPlasmid-Determined Tetracycline

Resistance

In view of their location in the cell envelopeand earlier circumstantial evidence (60), the pro-teins could be directly involved in exclusion oftetracyclines from the cell. This view is consist-ent with the observation that plasmid-deter-mined tetracycline resistance is associated with

decreased accumulation of the antibiotic in E.coli (60) and other species (90, 132, 140, 150) andwith the lack of evidence of enzymatic destruc-tion of antibiotic by resistant bacteria (60, 140).Franklin (60) suggested that the data demon-strating reduced permeability in resistant cellscould be explained if the cells actively excretedantibiotic. Young and Hubball (164) tested thishypothesis by studying tetracycline efflux fromE. coli carrying a plasmid expressing tempera-ture-conditional resistance to the drug. The ef-flux rates of tetracycline from these cells atpermissive (300C) and nonpermissive (420C)temperatures were similar to or less than theefflux rates from preloaded plasmidless cells.The envelope proteins therefore do not promotedrug efflux, but probably prevent influx.

Precise definition of the mode of action of theenvelope proteins specified by tetracycline re-sistance determinants in E. coli is currently im-possible. The protein of molecular weight 14,000specified by pSC101 may block the energy-de-pendent accumulation of tetracycline (144). Theinner membrane location of the 15,000-daltonprotein specified by TnlO is consistent with asimilar function, and indeed the energy-depend-ent phase of tetracycline accumulation is alsoknown to be modified in bacteria carrying R222(97). These proteins could prevent binding oftetracycline to the uptake system by specificassociation with carrier molecules. The proteinsof molecular weights 34,000 and 18,000 specifiedby pSC101 apparently block a final energy-in-dependent phase of uptake (144) that could rep-resent binding of antibiotic to membrane-asso-ciated ribosomes. Energy-independent bindingof tetracycline may also be inhibited by a prod-uct of R222, although in this case a primaryrather than final stage of uptake appears to beaffected (97).Levy and McMurry (97) have presented data

which tentatively suggest that at least one of theR222 (TnlO)-coded proteins, probably that ofmolecular weight 15,000, acts at the level of theribosome. Because we have detected this proteinin the inner membrane, we are left with theintriguing possibility that this protein in vivomay be associated with both ribosomes and themembrane. The distribution ofthis protein uponcell fractionation might therefore resemble thatof the E. coli phosphatidylserine synthetasewhich probably associates with both the mem-brane and ribosomes in vivo (94) and yet is notreleased from ribosomes during their fractiona-tion from crude cell lysates (120). Evidence isgradually accumulating that at least some pro-teins in E. coli (and other bacteria) are synthe-sized on membrane-bound polysomes (121, 137,138). Because protein synthesis on membrane-

MICROBIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from



bound polysomes may be particularly suscepti-ble to inhibition by tetracycline (81), the 15,000-dalton protein may specifically protect theseribosomes from inhibition by the antibiotic.Comparison of the polypeptide composition ofmembrane-bound and free ribosomes in tetra-cycline-resistant bacteria might well provide aprofitable approach toward an understanding ofthe role of this protein in resistance.The principal difficulty in establishing the

function of any of the proteins associated withtetracycline resistance stems from lack of de-tailed knowledge of the mechanism of tetracy-cline accumulation by sensitive bacteria. How-ever, a scheme for the possible interaction in theE. coli envelope of the plasmid-determined pro-teins associated with resistance is presented inFig. 2.

Membrane Architecture in Relation to theBinding and Function of Proteins

Associated with ResistanceThe membrane proteins must presumably in-

tegrate with and assume specific configurationswithin the envelope in order to function as re-sistance proteins. Several studies have suggestedthat fatty acid and phospholipid metabolism arealtered in E. coli harboring plaids that specifytetracycline resistance (48, 91, 125, 143). Suchchanges could be related to the formation ofnewlipid domains to accommodate the envelope pro-teins. However, we have been unable to findevidence to support this hypothesis either in E.coli (26, 28) or S. aureus (24). Thus, resistanceis inducible in the absence of lipid synthesis (24,26), and no differences in phospholipid or fattyacid content were noted in E. coli strains iso-genic apart from possession of TnlO (26, 28).The reported changes in the lipid content oftetracycline-resistant strains may therefore bedue to plasmid products other than those of thetetracycline resistance regions. Although newlipid domains are not synthesized to accommo-date the proteins, the physical state of the lipidsat the sites of attachment to the envelope ap-pears to influence the binding or function of theproteins. Thus, resistant organisms held at 50Ccan no longer exclude the antibiotic (132), whichpresumably reflects the transition from a fluidto a crystalline lipid phase in those membraneregions occupied by the proteins.

Minicelis derived from cultures of E. coli ex-pressing high-level resistance to tetracycline arethemselves only partially resistant to the anti-biotic, as determined by inhibition of minicellprotein synthesis (62, 95). This may indicatepreferential localization of the plasmid-specifiedproteins in the equatorial regions of the wholecell envelope because minicells are derived from

the polar regions of their parental cells (57). InS. aureus expression of high-level resistance totetracycline is associated with a decrease in thecontent of membrane proteins of molecularweight about 22,000 (31), which could representcompetition for insertion of plasmid-directedproteins and integral proteins into the samemembrane domains.

Protein synthesis in minicells containing tet-racycline resistance plasmids is more susceptibleto inhibition by the antibiotic than is synthesisin whole cells containing the same R factors (62,95). Although the sensitivity of protein synthesisin minicells could result from lack of synthesisof one or more resistance proteins (see above),it is clear that several of the proteins involved inresistance can be inserted into the minicell en-velope (96, 144, 163). This implies that the activ-ity of these proteins depends markedly on mem-brane architecture. Lack of expression of resist-ance in minicells could reflect (as noted above)absence of material from the equatorial regionsof the cell envelope. Altematively, because min-icells do not replicate (95), effective placementof the envelope proteins could require integra-tion into a growing cell membrane. However, asnoted above, concomitant lipid and protein syn-thesis is not required for inducible expression ofresistance in whole cells (24, 26). Little is knownabout the distribution of envelope componentsfrom parental cells to minicells, but researchdirected toward this problem might permit iden-tification of those integral membrane compo-nents apparently required for some or all of theplasmid-specified proteins to function.

RELATION BETWEEN TETRACYCLINEGENE COPY NUMBER AND

RESISTANCE LEVELWhere R-plasmid resistance genes determine

the synthesis of drug-metabolizing enzymes,there is usually a correlation between gene copynumber (gene dosage) and enzyme activity percell and, thence, resistance level (114, 151). Uhlinand Nordstrom (152) found that an increase ingene dosage in copy mutants of Rldrd-19 wasparalleled by an increase both in resistance toampicillin and in beta-lactamase activity up tolevels at least 10-fold higher than those charac-teristic of the unmutated plasmid. A similarrelation between chloramphenicol acetyltrans-ferase and streptomycin adenylyltransferase andgene dosage was found, although the levels ofresistance to chloramphenicol and streptomycinreached plateaus of two- and threefold higher,respectively, when gene copy numbers were in-creased.

Efforts to demonstrate whether a similar re-lationship exists in the case of tetracycline re-

VOL. 42, 1978


http://mm

br.asm.org/

Dow

nloaded from


716 CHOPRA AND HOWE

sistance have been hampered both by difficultyin the interpretation of tests for minimum inhib-itory concentration of tetracycline (23) and bythe lack, until recently, of suitable assay systemsfor the products of the resistance region. Fur-thermore, the interpretation of data relating totetracycline gene copy number and resistancelevel is complicated by the association of theplasmid products with the cell envelope. Thus,an anticipated increase in resistance may not berealized if the number of binding sites within themembrane that can accommodate the plasmid-specified proteins is limited. This type of situa-tion probably applies to the E. coli outer mem-brane protein II*, where there is also evidencethat free or nascent protein II* may inhibit itsown synthesis (38). On the other hand, severalE. coli envelope proteins do show gene dosageresponses (17, 122, 149) which can lead to verysubstantial enrichment of the envelope in spe-cific proteins (149).The first workers to consider the possibility of

a tetracycline resistance gene dosage effect wereFranklin and Rownd (64), who studied R100-mediated expression in E. coli and Proteus mi-rabilis. They found that a markedly higher levelof resistance to chloramphenicol, streptomycin,and spectinomycin in P. mirabilis when com-pared with E. coli was not paralleled by anincrease in tetracycline resistance level. This canbe ascribed to the well-known dissociation ofR100 into two separate DNA species in Proteus,in which the replicon that includes the tetracy-cline resistance region does not undergo thesame increase in copy number as does that com-prising the other resistance genes.

Foster and Walsh (55), in a survey of resist-ance levels determined by a range ofR plasmids,made several resistance measurements onstrains carrying two compatible plasmids. Com-binations of R6-S in separate experiments withR69, RP1, and R46 led to resistance levels in E.coli J5-3 no higher than that conferred by onetype of plasmid alone. Reeve and Robertson(124), however, obtained a conflicting result ina study of the R6-S plus R57 combination, find-ing that the growth rate of induced cells carryingboth R plasmids, on challenge with 160 ,ug oftetracycline per ml, was 58% that of an inducedbut unchallenged control, compared with a valueof 39% for one R plasmid alone. Although thisdiscrepancy might be explained by differences inthe host strains and plasmids used, neither groupmade any measurement of plasmid DNA con-tent and thus there are no data on the gene copynumbers in the strains compared. Where suchmeasurements have been made, there is strongevidence for a correlation between plasmid copynumber and resistance level. E. coli CR34

(pSC101) contains five copies of the tetracyclineresistance-determining plasmid pSC101 and isresistant to 25 ,ug of tetracycline per ml on solidmedium; when the same host strain containspSC134, constructed by in vitro linkage ofpSC101 to ColEl at the EcoRI cleavage sites,thus bringing the tetracycline resistance regionunder the relaxed replication control of ColEl,the plasmid copy number rises to 16, and theresistance level rises to 80 jig/ml. (20).A similar relationship between resistance level

and gene copy number is exhibited by pAMalin S. faecalis (32, 33, 161, 162). S. faecalis(pAMal) is resistant to 250 ,ug of tetracyclineper ml in liquid medium, but when the cells aregrown in a sub-inhibitory concentration of thedrug (150 i.g/ml), the resistance level rises from250 to 500 itg/ml and the 6-megadalton pAMalmolecule is concomitantly replaced by a largerspecies of DNA; both properties return to thewild-type level on withdrawal of the drug. Het-eroduplex studies and EcoRI digest preparationsshow that the enlarged DNA contains repeatedsequences 2.65 megadaltons in length, each con-sisting of a tetracycline resistance determinantflanked by a small "recombination sequence"380 nucleotide pairs in length which is presum-ably responsible for the reversible gene amplifi-cation. The ratio of plasmid to chromosomalDNA was similar in both states, indicating thatthe increase in resistance determinants per cellwas balanced partially by a reduction in plasmidcopy number. Likewise, the plasmid pRSD1 inE. coli develops multiple copies of a 3.5-mega-dalton sequence coding for tetracycline resist-ance when host cells are challenged with thedrug (R. Mattes, H. J. Burkardt, and R. Schmitt,submitted for publication).A much more complex situation is revealed by

studies on a round of replication mutant ofNR1(105, 106, 148; H. Hashimoto, Y. Ike, C. F. Mor-ris, and R. H. Rownd, submitted for publication;NR1 is probably identical to R100 and R222[51]). Plasmid pRR12 (also called R12) is a copymutant of NR1 that confers raised levels ofresistance to streptomycin and chloramphenicolbut lowered resistance to tetracycline relative toNR1. Most E. coli cells that contain pRR12,when grown on tetracycline-containing agar, arefound to contain pRR12 and one or more smallerplasmids; the latter are incompatible withpRR12 itself and can thus be recovered fromclones that have lost pRR12. Such clones arefully sensitive to tetracycline. It can be shownby heteroduplex and restriction endonucleaseanalysis that pRR12 contains all of the se-quences present in these small plasmids, andthey are therefore presumed to originate from it.All contain the same origin of replication as the

MICROBIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from



resistance transfer factor of pRR12; yet, al-though they carry no tetracycline resistance de-terminant, they confer an increased level of re-sistance when present with pRR12 itself. Severalother copy mutants of NR1 behave similarly

(148). It has been suggested (Hashimoto et aL.,submitted for publication) that the decreasedresistance conferred by pRR12 (3 1g/ml) relativeto NR1 (100 ug/ml) in E. coli results from ahigher level of tetracycline gene repressor owingto a repressor gene dosage effect in the copymutant, and a similar increase in the F-plasnidtwansfer inhibition and entry exclusion has beenexplained in the same way. If this is so, it indi-cates that comparison of resistance levels be-tween strains supposedly isogenic except for thenumber of tetracycline resistance genes must beinterpreted with caution, unless tests are in-cluded to detect differences in the genetic re-gions controlling resistance expression (seeabove).These effects may also influence the interpre-

tation of experiments designed to test the dom-inance of particular alleles of the tetracyclineresistance region, because such tests are effec-tively gene dosage experiments involving twogenetic regions. Foster (52) constructed strainsof E. coli heterozygous for the resistance region,in which wild-type TnlO was present on thechromosome and a revertible single-site tetra-cycline-sensitive mutant of TnlO was present onR100; some ofthe heterozygotes failed to expressresistance fully upon induction, but loss of thedominant sensitive allele by deletion restoredthe normal resistance leveL These data are con-sistent with a loss of ability to synthesize resist-ance products, but not ofrepressor, in the single-site mutants. Similar tests on a series of consti-tutive mutants in combination with the induc-ible wild-type TnlO showed dominance of theconstitutive over the wild-type allele, suggestingthat each mutant examined was operator con-stitutive (53). Tests on strains carrying two dif-ferent wild-type plnamids have shown a distinc-tion between inducible plasmids such as R57,R6, and R100 on the one hand and, on the other,RP1 and perhaps R46 and R199, which form alower-level resistance group in which both re-

pressor and operator are postulated to differfrom those characteristic of the group that in-cludes R100 (123).We have exploited the transposability ofTnlO

to compare the resistance levels both of wild-type TnlO and of a constitutive allele (frompDU301; 53) when located in the chromosomeand in a plasmid. Homozygotes in which a copy

of the same allele of TnlO is present in bothlocations simultaneously are no more resistantthan are strains carrying the chromosomal allele

alone (Howe and Chopra, unpublished data).This is surprising in view of the clear relationbetween gene dosage and resistance noted byCabello et al. (see above and reference 20) inpSC101 and may reflect a difference in the prod-ucts of pSC101 and R100 (see above).There is evidence for a tetracycline gene dos-

age effect in S. aureus (128), where strains of S.aureus RN450 have been constructed containinga gene for tetracycline plus minocycline resist-ance and a gene for tetracycline resistance alone,the latter transduced from strain PS84. On chal-lenge with tetracycline, the strain carrying bothgenes showed a higher resistance level thanstrains carrying either gene singly. S. aureus 649Tete Mir carries 31 to 47 tetracycline resistanceplasmids per cell (27, 93); on growth in thepresence of 10 ,g oftetracycline per ml, however,there was no evidence for increase in the tetra-cycline resistance plamid copy number, norwere resistant derivatives of strain 649 Tetr MirYisolated by incubation in the presence of thedrug, in contrast to the finding of gene amplifi-cation in S. faecalis (see above).

ORIGIN OF TETRACYCLINERESISTANCE GENES

The finding ofsome antibiotic-inactivating en-zymes both in resistant bacteria and in anti-biotic-producing streptomycetes (12, 65, 155)prompts one to ask whether the tetracyclineresistance genes discussed above might haveoriginated from soil organisms. The ability ofsuch genes from at least one group of naturallyoccurring plasmids to translocate from one rep-licon to another (TnlO; 54, 88) is consistent withthis suggestion. Among those streptomycetesable to synthesize tetracycline or derivatives ofit, the soil organism Streptomyces rimosus hasbeen genetically analyzed in some detail (1, 66).This organism contains a fertility plasmid(SCP1; 67); the frequency with which the pro-ducer strain LST-118 gives rise to nonrevertingoxytetracycline-sensitive variants is greatly in-creased by growth in the presence of acridines,suggesting that resistance is plasmid mediated(16). It is clear that tetracycline-producing strep-tomycetes must have some means of protectingthemselves against their own product; with ri-bosomes from one strain of Streptomyces aureo-faciens that produces 2 mg of tetracycline perml, protein synthesis is 50% inhibited in vitro by50 Ig of the drug per ml, whereas systems fromlow-level tetracycline producers are more sensi-tive, implicating a ribosomal resistance mecha-nism in this producer organism (42, 107).The ability of a tetracycline resistance gene to

be expressed is not restricted to the species inwhich the gene was isolated. A plasmid deter-

VOiL 42, 1978


http://mm

br.asm.org/

Dow

nloaded from


718 CHOPRA AND HOWE

mining tetracycline resistance (pBC16; molecu-lar weight, 2.8 x 106) has been isolated fromBacillus cereus GP7 (14); hybrid plasmids de-rived from pBC16, capable of replication in bothBacillus subtilis and E. coli, can confer resist-ance to tetracycline in both hosts, althoughother resistance genes present (for ampicillin,kanamycin, and chloramphenicol) are unable tobe expressed in B. subtilis (89). One S. aureusplasmid coding for tetracycline resistance(pT127) has been introduced by transformationinto B. subtilis, in which it continues to replicateand confer drug resistance (49). The promis-cuous plasmid RP4 from Pseudomonas aerugi-nosa, which determines resistance to tetracy-cline and other antibiotics, is capable of conjugaltransfer to a range ofgram-negative bacteria (13,39, 115). Such plasmids offer a wide variety ofpossibilities for spread of an ancestral geneamong different bacteria, especially when con-sidered together with the capacity of TnlO totranspose. Further progress in understandingthe origin and spread of resistance genes, how-ever, will depend on whether tetracycline resist-ance in all organisms can be shown to resultfrom an element related to TnlO or whether (aspossibly for pSC101) some completely differentdeterminant is also found.

CONCLUSIONSFranklin (60), writing in 1973 on antibiotic

transport in bacteria, commented that "thereare exceptional cases where an antibiotic suffi-ciently resembles a normal nutrient of the cellso as to be absorbed by the nutrient transportsystem." In fact, the majority of antimetabolitesmay prove fortuitously to use normal transportsystems. For instance, it is already known thatcycloserine uses the D-alanrne transport system(156), phosphonomycin uses the glycerol-3-phosphate system (133), arsenate uses phos-phate systems (127), a variety of heavy metalcations use divalent cation transport systems(111, 112, 116, 157; A. A. Weiss, S. Silver, and T.G. Kinscherf, submitted for publication), andalbomycin binds to the E. coli outer membranetonA protein used in ferrichrome uptake (18).Aminoglycoside accumulation by E. coli pro-vides a slightly different pattem of antimetabo-lite accumulation. By binding to ribosomes,some aminoglycosides are able to induce a nor-mal polyamine transport system which fortui-tously promotes further aminoglycoside accu-mulation (83). In this case, however, the mech-anism by which the aminoglycosides initiallyenter the cell is unknown (83). Passage of tetra-cycline across the bacterial cytoplasmic mem-brane may well prove to fit the general patternof fortuitous transport; if the substrate it mimics

can be identified, it may be possible to synthesizetetracyclines with improved properties of pene-tration by covalent linkage of antibiotic to thesubstrate.

High-level resistance to most of the agentslisted above can result from chromosomal mu-tations that alter or lead to loss of the respectivetransport system or receptor. However, single-step chromosomal mutations which confer highlevels of resistance to tetracycline do not occur,at least not in either E. coli (126) or S. aureus(92). The tetracycline transport system might,therefore, comprise part ofa membrane complexessential for viability, which would explain theevolution of plasmid genes whose products spe-cifically reduce the affinity of the transport sys-tem for tetracycline, yet do not disrupt mem-brane integrity. In contrast to this, some recep-tor proteins may have evolved as high-affinitybinding components of transport systems for theuptake ofsubstrates present in very low amountsin the environment (68, 75); dispensable struc-tures of this kind could then be utilized as recep-tors by noxious agents such as bacteriophagesand bacteriocins, and mutational loss of the re-ceptors would confer bacteriophage or bacterio-cin resistance without proving lethal for the cell.

ACKNOWLEDGMENTSThis work was supported by program grant G970

107B to M. H. Richmond and project grants G9751004C and G978 67SB to I. C. from the MedicalResearch Council.We thank S. Falkow, M. H. Richmond, T. J. Foster,

P. R. Ball, and D. J. Sinden for helpful discussions andmany other colleagues for the provision of data beforepublication.

LITERATURE CITED1. Alacevic, M. 1976. Recent advances in Strepto-

myces rimosus genetics, p. 513-519. In K. D.Macdonald (ed.), Second international sym-posium on the genetics of industrial microor-ganisms. Academic Press Inc., London.

2. Albert, A. 1953. Avidity of terramycin and au-reomycin for metallic cations. Nature (London)172:201.

3. Alper, M. D., and B. N. Ames. 1978. Transportof antibiotics and metabolite analogs by sys-tems under cyclic AMP control: positive selec-tion of Salmonella tphimurium cya and crpmutants. J. Bacteriol. 133:149-157.

4. Arima, K., and K. Izaki. 1963. Accumulation ofoxytetracycline relevant to its bactericidal ac-tion in the cells of Escherichia coli. Nature(London) 200:192-193.

5. Asheshov, E. H. 1975. The genetics of tetracy-cline resistance in Staphylococcus aureus. J.Gen. Microbiol. 88:132-140.

6. Bahl, C. P., R. Wu, J. Stawinsky, and S. A.Narang. 1977. Minimal length of the lactoseoperator sequence for the specific recognition

MICROBIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from



by the lactose repressor. Proc. Natl. Acad. Sci.U.S.A. 74:966-970.

7. Ball, P. R., I. Chopra, and S. J. Eccles. 1977.Accumulation of tetracyclines by Escherichiacoli K-12. Biochem. Biophys. Res. Commun.77:1500-1507.

8. Barza, M., R. B. Brown, C. Shanks, C. Gam-ble, and L Weinsutein. 1975. Relation be-tween lipophilicity and pharmacological be-havior of minocycline, doxycycline, tetracy-cline, and oxytetracycline in dogs. Antimicrob.Agents Chemother. 8:713-720.

9. Bavoil, P., H. Nikaido, and K. von Meyen-burg. 1977. Pleiotropic transport mutants ofEscherichia coli lack porin, a major outermembrane protein. Mol. Gen. Genet. 158:2333.

10. Beauchamp, R. S., S. Silver, and J. W. Hop-kins. 1971. Uptake of Mg2e by KB cells.Biochim. Biophys. Acta 225:71-76.

11. Benbough, J. E., and G. A. Morrison. 1967.The molecular basis of an inhibition by tetra-cyclines. J. Gen. Microbiol. 49:469-485.

12. Benveniste, R., and J. Davies. 1973. Amino-glycoside antibiotic-inactivating enzymes in ac-tinomycetes similar to those present in clinicalisolates of antibiotic-resistant bacteria. Proc.Natl. Acad. Sci. U.S.A. 70:2276-2280.

13. Beringer, J. E. 1974. R factor transfer in Rhi-zobium legumimosarum. J. Gen. Microbiol. 84:188-198.

14. Bernhard, K., H. Schrempf, and W. Goebel.1978. Bacteriocin and antibiotic resistanceplasmids in Bacillus cereus and Bacillus sub-tilis. J. Bacteriol. 133:897-903.

15. Blackwood, R. K., and A. R. English. 1970.Structure-activity relationships in the tetracy-cline series. Adv. Microb. Physiol. 13:237-266.

16. Boronin, A. M., and L. G. Sadovnikova. 1972.Elimination by acridine dyes of oxytetracyclineresistance in Actinomyces rimosus. Genetika8:174-176.

17. Bradbeer, C., M. L Woodrow, and L. L Khal-ifah. 1976. Transport of vitamin B12 in Esche-richia coli: common receptor system for vi-tamin B12 and bacteriophage BF23 on the outermembrane of the cell envelope. J. Bacteriol.125:1032-1039.

18. Braun, V. 1978. Structure-function relationshipsof the gram-negative bacterial cell envelope, p.111-138. In R. Y. Stanier, H. J. Rogers, and B.J. Ward (ed.), Twenty-eighth symposium ofthe Society for General Microbiology. Cam-bridge University Press, Cambridge.

19. Bukhari, A. I., J. L Shapiro, and S.LAdhya.1977. DNA insertion elements, plasmids andepisomes. Cold Spring Harbor Laboratory,Cold Spring Harbor, N.Y.

20. Cabello, F., K. Timmis, and S. N. Cohen. 1976.Replication control in a composite plasmidconstructed by in vitro linkage of two distinctreplicons. Nature (London) 259:285-290.

21. Calame, K., D. Nakada, and G. Iler. 1978.Location of ribosome-binding sites on the tet-racycline resistance transposon ThlO. J. Bac-

teriol. 135:668-674.22. Caswell, A. H., and J. D. Hutchison. 1971.

Visualization of membrane bound cations by afluorescent technique. Biochem. Biophys. Res.Commun. 42:43-49.

23. Chabbert, Y. A., and M. R. Scavizzi. 1976.Chelocardin-inducible resistance in Esche-richia coli bearing R plasmids. Antimicrob.Agents Chemother. 9:36-41.

24. Chopra, L. 1975. Induction of tetracycline resist-ance in Staphylococcus aureus in the absenceof lipid synthesis. J. Gen. Microbiol. 91:433-436.

25. Chopra, . 1976. Mechanisms of resistance tofusidic acid in Staphylococcus aureus. J. Gen.Microbiol. 96:229-238.

26. Chopra, L. 1978. Plasmid-determined tetracy-cline resistance in Escherichia coli K12: lackof evidence that resistance is related to changesin lipid metabolism. Biochem. Soc. Trans. 6:431-433.

27. Chopra, I., P. M. Bennett, and R. W. Lacey.1973. A variety of staphylococcal plasmidspresent as multiple copies. J. Gen. Microbiol.79:343-345.

28. Chopra, I., and S. J. Eccles. 1977. Tetracyclineresistance in Escherichia coli K12 is not asso-ciated with a decrease in cyclopropane fattyacid content. J. Gen. Microbiol. 103:393-396.

29. Chopra, I., and S. J. Eccles. 1978. Diffusion oftetracycline across the outer membrane ofEscherichia coli K-12: involvement of proteinIa. Biochem. Biophys. Res. Commun. 83:550-557.

30. Chopra, I., T. G. B. Howe, and P. R. Ball.1977. Lysozyme-promoted association of pro-tein I molecules in the outer membrane ofEscherichia coli. J. Bacteriol. 132:411-418.

31. Chopra, L, RK W. Lacey, and J. Connolly.1974. Biochemical and genetic basis of tetra-cycline resistance in Staphylococcus aureus.Antimicrob. Agents Chemother. 6:397-404.

32. Clewell, D. B., and Y. Yagi. 1977. Amplificationof the tetracycline resistance determinant onplasmid pAMal in Streptococcus faecalis, p.235-246. In A. I. Bukhari, J. A. Shapiro, and S.L. Adhya (ed.), DNA insertion elements, plas-mids, and episomes. Cold Spring Harbor Lab-oratory, Cold Spring Harbor, N.Y.

33. Clewell, D. B., Y. Yagi, and B. Bauer. 1975.Plasmid-determined tetracycline resistance inStreptococcus faecalis evidence for gene am-plification during growth in presence of tetra-cycline. Proc. Natl. Acad. Sci. U.S.A. 72:1720-1724.

34. Cohen, S. N. 1976. Transposable genetic ele-ments and plasmid evolution. Nature (London)263:731-738.

35. Cohen, S. N., and A. C. Y. Chang. 1977. Re-vised interpretation of the origin of the pSC101plWsmid. J. Bacteriol. 132:734-737.

36. Costerton, J. W., and K. J. Cheng. 1975. Therole of the bacterial cell envelope in antibioticresistance. J. Antimicrob. Chemother. 1:363-377.

VOL. 42, 1978


http://mm

br.asm.org/

Dow

nloaded from


720 CHOPRA AND HOWE

37. Costerton, J. W., J. M. Ingram, and K.-J.Cheng. 1974. Structure and function of the cellenvelope of gram-negative bacteria. Bacteriol.Rev. 38:87-110.

38. Datta, D. B., C. Kramer, and U. Henning.1976. Diploidy for a structural gene specifyinga major protein ofthe outer cell envelope mem-brane from Escherichia coli K-12. J. Bacteriol.128:834-841.

39. Datta, N., and R. W. Hedges. 1972. Host rangesof R factors. J. Gen. Microbiol. 70:453-460.

40. Decad, G. M., and H. Nikaido. 1976. Outermembrane of gram-negative bacteria. XII. Mo-lecular-sieving function of cell wall. J. Bacte-riol. 128:325-336.

41. Del Bene, V. E., and M. Rogers. 1975. Com-parison of tetracycline and minocycline trans-port in E8cherichia coli. Antimicrob. AgentsChemother. 7:801-806.

42. Demain, A. L. 1974. How do antibiotic-produc-ing microorganisms avoid suicide? Ann. N.Y.Acad. Sci. 236:601-612.

43. de Zeeuw, J. R. 1968. Accumulation of tetracy-chines by Escherichia coli. J. Bacteriol. 96:498-506.

44. Di Rienzo, J. M., K. Nakamura, and M. In-ouye. 1978. The outer membrane proteins ofgram-negative bacteria: biosynthesis, assem-bly, and functions. Annu. Rev. Biochem. 47:481-532.

45. Dockter, M. E., and J. A. Magnuson. 1974.Characterization of the active transport ofchlorotetracycline in Staphylococcus aureusby a fluorescence technique. J. Supramol.Struct. 2:32-44.

46. Dockter, M. E., and J. A. Magnuson. 1975.Membrane phase transitions and the transportof chlorotetracycline. Arch. Biochem. Biophys.168:81-88.

47. Dockter, M. E., W. R. Trumble, and J. A.Magnuson. 1978. Membrane lateral phaseseparations and chlorotetracycline transportby Bacillus megaterium. Proc. Natl. Acad. Sci.U.S.A. 76:1319-1323.

48. Dunnick, J. K., and W. M. O'leary. 1970.Correlation of bacterial lipid composition withantibiotic resistance. J. Bacteriol. 101:892-900.

49. Ehrlich, S. D. 1977. Replication and expressionof plasmids from Staphylococcus aureus inBacillus subtilis. Proc. Natl. Acad. Sci. U.S.A.74:1680-1682.

50. Elwell, L. P., J. R. Saunders, M. H. Rich-mond, and S. Falkow. 1977. Relationshipsamong some R plasmids found in Haemophilusinfluenzae. J. Bacteriol. 131:356-362.

51. Falkow, S. 1975. The genetic properties of Rfactors, p. 76-97. In Infectious multiple drugresistance. Pion Ltd., London.

52. Foster, T. J. 1976. R factor-mediated tetracy-cline resistance in Escherichia coli K12. Dom-inance of some tetracycline sensitive mutantsand relief of dominance by deletion. Mol. Gen.Genet. 143:339-344.

53. Foster, T. J. 1977. Isolation and characterisationof mutants of R100-1 which express tetracy-

cline resistance constitutively. FEMS Micro-biol. Lett. 2:271-274.

54. Foster, T. J., T. G. B. Howe, and K. M. V.Richmond. 1975. Translocation of the tetra-cycline resistance determinant from R100-1 tothe Escherichia coli K-12 chromosome. J. Bac-teriol. 124:1153-1158.

55. Foster, T. J., and A. Walsh. 1974. Phenotypiccharacterization of R-factor tetracycline resist-ance determinants. Genet. Res. 24:333-343.

56. Foulds, J., and T.-J. Chai. 1978. New majorouter membrane protein found in an Esche-richia coli tolF mutant resistant to bacterio-phage TuIb. J. Bacteriol. 133:1478-1483.

57. Fox, C. F. 1972. Membrane assembly, p. 345-385.In C. F. Fox and A. D. Keith (ed.), Membranemolecular biology. Sinauer Associates Inc.,Sunderland, Mass.

58. Franklin, T. J. 1966. Mode of action of thetetracyclines, p. 192-212. In B. A. Newton andP. E. Reynolds (ed.), Sixteenth symposium ofthe Society for General Microbiology. Cam-bridge University Press, Cambridge.

59. Franklin T. J. 1971. Uptake of tetracycline bymembrane preparations from Escherichia coli.Biochem. J. 123:267-273.

60. Franklin T. J. 1973. Antibiotic transport inbacteria. Crit. Rev. Microbiol. 2:253-272.

61. Franklin T. J., and S. J. Foster. 1971. Effectof osmotic shock on tetracycline resistance inEscherichia coli bearing an R-factor. Biochem.J. 121:287-292.

62. Franklin, T. J., and S. J. Foster. 1974. Expres-sion of R factor-mediated resistance to tetra-cycline in Escherichia coli minicells. Antimi-crob. Agents Chemother. 6:194-195.

63. Franklin, T. J., and B. Higginson. 1970. Activeaccumulation of tetracycline by Escherichiacoli. Biochem. J. 116:287-297.

64. Franklin, T. J., and R. Rownd. 1973. R-factor-mediated resistance to tetracycline in Proteusmirabilis. J. Bacteriol. 115:235-242.

65. Freeman, R. F., and D. A. Hopwood. 1978.Unstable naturally occuring resistance to an-tibiotics in Streptomyces. J. Gen. Microbiol.106:377-381.

66. Friend, E. J., and D. A. Hopwood. 1971. Thelinkage map of Streptomyces rimosus. J. Gen.Microbiol. 68:187-197.

67. Friend, E. J., M. Warren, and D. A. Hop-wood. 1978. Genetic evidence for a plasmidcontrolling fertility in an industrial strain ofStreptomyces rimosus. J. Gen. Microbiol. 106:201-206.

68. Frost, G. E., and H. Rosenberg. 1975. Rela-tionship between the tonB locus and iron trans-port in Escherichia coli. J. Bacteriol. 124:704-712.

69. Fung, J., T. J. MacAlister, and L. L. Rothfield.1978. Role ofmurein lipoprotein in morphogen-esis of the bacterial division septum: pheno-typic similarity of IkyD and lpo mutants. J.Bacteriol. 133:1467-1471.

70. Gale, E. F., E. Cundliffe, P. E. Reynolds, M.H. Richmond, and M. J. Waring. 1972. The

MICROBIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from



molecular basis of antibiotic action. John Wiley& Sons, London.

71. Garrod, L. P., and F. O'Grady. 1968. Antibiot-ics and chemotherapy. E. and S. LivingtoneLtd., Edinburgh.

72. Garten, W., L Hindennach, and U. Henning.1975. The major proteins of the Escherichiacoli outer cell envelope membrane. Character-isation of proteins II and HI, comparison ofall proteins. Eur. J. Biochem. 59:215-221.

73. Gottlieb, D., and P. D. Shaw. 1967. Antibiotics,voL 1, Mechanism of action. Springer-Verlag,Berlin.

74. Hamilton, W. A. 1977. Energy coupling in sub-strate and group translocation, p. 185-216. InB. A. Haddock and W. A. Hamilton (ed.),Twenty-seventh symposium of the Society forGeneral Microbiology, Cambridge UniversityPress, Cambridge.

75. Hantke, K., and V. Braun. 1975. Membranereceptor dependent iron transport in Esche-richia col FEBS Lett. 49:301-305.

76. Harold, F. M. 1977. Membranes and energytransduction in bacteria. Curr. Top. Bioenerg.6:83-149.

77. Hasegawa, Y., H. Yamada, and S. Mizu-shima. 1976. Interactions of outer membraneproteins 0-8 and 0-9 with peptidoglycan sac-culus of Eswherichia coli K12. J. Biochem. 80:1401-1409.

78. Henning, U., W. Schmidmayr, andL Hinden-nach. 1977. Major proteins of the outer cellenvelope membrane of Escherichia coli K-12:multiple species of protein L. Mol. Gen. Genet.154:293-298.

79. Heppel, L. A. 1971. The concept of periplasmicenzymes, p. 223-247. In L I. Rothfield (ed.),Structure and function of biological mem-branes. Academic Press Inc., New York.

80. Hindennach, I., and U. Henning. 1975. Themajor proteins of the Eswherichia coli outercell envelope membrane. Preparative isolationof all major membrane proteins. Eur. J. Bio-chem. 59:207-213.

81. Hirashima, A., G. Childs, and M. Inouye.1973. Differential inhibitory effects of antibiot-ics on the biosynthesis of envelope proteins ofEwcherichia coli. J. MoL BioL 79:373-389.

82. Hirota, Y., H. Suzui, Y. Nishimura, and S.Yasuda. 1977. On the process of cellular divi-sion in Escherichia coli: a mutant of E. colilacking a murein-lipoprotein. Proc. Natl. Acad.Sci. U.S.A 74:1417-1420.

83. Holtje, J.-V. 1978. Streptomycin uptake via aninducible polyamine transport system in Esch-erichia coli. Eur. J. Biochem. 86:345-351.

84. Jones, J. G., and G. A. Morrison. 1962. Thebacteriostatic actions of tetracycline and oxy-tetracycline. J. Pharm. PharmacoL 14:808-824.

85. Jorgensen, R., D. Berg, and W. Reznikoff.1978. Genetic organization in the transposabletetracycline resistance determinant TnlO, p.181-183. In D. chessinger (ed.), Microbiol-ogy-1978. American Society for Microbiology,Washington, D.C.

86. Kadenbach, B., H. Freitag, and H. Kolbe.1978. On the mechanism of phosphate anddicarboxylate transport in mitochondria.FEBS Lett. 89:161-164.

87. K:leckner, N. 1977. Tanslocatable elements inprocaryotes. Cell 11:11-23.

88. Kleckner, N., R. K. Chan, B.-K. Tye, and D.Botstein. 1975. Mutagenesis by insertion of adrug-resistance element carrying an invertedrepetition. J. MoL Biol. 97:561-575.

89. Kreft, J., K. Bermhard, and W. Goebel. 1978.Recombinant plasnids capable of replicationin B. subtilis and E. coli. Mol. Gen. Genet.162:59-67.

90. Kuck, N. A., and M. Forbes. 1973. Uptake ofminocycline and tetracycline by tetracycline-susceptible and -resistant bacteria. Antimicrob.Agents Chemother. 3:662-664.

91. Kuinna, Z. A., L L Belousova, and I. M. Ter-eshin. 1977. Lipid composition of tetracycline-sensitive and -ristant Escherichia coli cells.Microbiology 46:210-214.

92. Lacey, R. W. 1975. Antibiotic resistance plas-mids of Staphylococcus aureus and their clin-ical importance. BacterioL Rev. 39:1-32.

93. Lacey, R. W., and L Chopra. 1974. Geneticstudies of a multi-resistant strain of Staphylo-coccus aureus. J. Med. Microbiol. 7:285-297.

94. Larson, T. J., and W. Dowhan. 1976. Ribo-somal-associated phosphotidylserine synthe-tase from Escherichia coli: purification by sub-strate-specific elution from phosphocelluloseusing cytidine-5'-diphospho-1,2-diacyl-sn-glyceroL Biochemistry 15:5212-5218.

95. Levy, S. B. 1975. The relation of a tetracycline-induced R factor membrane protein to tetra-cycline resistance, p. 215-225. In S. Mitsuhashi(ed.), Drug inactivating enzymes and antibioticresistance. Avicenum, Prague.

96. Levy, S. B., and L. MeMurry. 1974. Detectionof an inducible membrane protein associatedwith R-factor-mediated tetracycline resistance.Biochem. Biophys. Res. Commun. 56:1060-1068.

97. Levy, S. B., and L. MeMurry. 1978. Probingthe expression of plasmid-mediated tetracy-cline resistance in Escherichia coli, p. 177-180.In D. Schlesinger (ed.), Microbiology-1978,American Society for Microbiology, Washing-ton, D.C.

98. Levy, S. B., L McMurry, P. Onigman, and R.M. Saunders. 1977. Plasmid-mediated tetra-cycline resistance in Escherichia coli, p.181-203. In J. Drews and G. Hogenauer (ed.),Topics in infectious diseas", vol. 2. Springer-Verlag, New York.

99. Lo, T. C. Y. 1977. The molecular mechanism ofdicarboxylic acid transport in Escherichia coliK-12. J. Supramol. Struct. 7:463-480.

100. Lo, T. C. Y., L K. Rayman, and B. D. Sanwal.1972. Transport of succinate in Escherichiacoli. I. Biochemical and genetic studies oftransport in whole celLs. J. BioL Chem. 247:6323-631.

101. Luktenhaus, J. F. 1977. Role of a major outer

VOL. 42, 1978


http://mm

br.asm.org/

Dow

nloaded from


722 CHOPRA AND HOWE

membrane protein in Escherichia coli. J. Bac-teriol. 131:631-637.

102. McQuillen, K. 1966. The physical organizationof nucleic acid and protein synthesis, p.

134-158. In M. R. Pollock and M. H. Richmond(ed.), Fifteenth symposium of the Society forGeneral Microbiology. Cambridge UniversityPress, Cambridge.

103. Manning, P. A., A. P. Pugsley, and P. Reeves.1977. Defective growth functions in mutants ofEscherichia coli K-12 lacking a major outermembrane protein. J. Mol. Biol. 116:285-300.

104. Meagher, R. B., R. C. Tait, M. Betlach, andH. W. Boyer. 1977. Protein expression in E.coli minicells by recombinant plasmids. Cell10:521-536.

105. Mickel, S., and W. Bauer. 1976. Isolation, bytetracycline selection, of small plasmids de-rived from R-factor R12 in Escherichia coliK-12. J. Bacteriol. 127:644-655.

106. Mickel, S., E. Ohtsubo, and W. Bauer. 1977.Heteroduplex mapping of small plasmids de-rived from R-factor R12: in vivo recombinationoccurs at ISI insertion sequences. Gene 2:193-210.

107. Mikulik, K., J. Karnetova, N. Quyen, M.Blaumerova, I. Komersova, and Z. Vanek.1971. Interaction of tetracycline with proteinsynthesizing system of Streptomyces aureofa-ciens. J. Antibiot. 24:801-809.

108. Mylnikova, S. L, and V. K. Plakunov. 1973.Oscillatory character of antibiotic transportregulation in a chlorotetracycline resistantstrain of Staphylococcus aureus 209. Antibi-otiki (Moscow) 18:293-298.

109. Mylnikova, S. I., and V. K. Plakunov. 1976.Significance of structural elements of the chlo-rotetracycline molecule for the induction ofchlorotetracycline resistance. Microbiology 45:897-902.

110. Nakae, T. 1976. Identification of the outer mem-brane protein of Escherichia coli that pro-duces transmembrane channels in reconsti-tuted vesicle experiments. Biochem. Biophys.Res. Commun. 71:877-884.

111. Nelson, D. L., and E. P. Kennedy. 1971. Mag-nesium transport in Escherichia coli. Inhibi-tion by cobaltous ion. J. Biol. Chem. 246:3042-3049.

112. Nelson, D. L., and E. P. Kennedy. 1972. Trans-port of magnesium by a repressible and non-

repressible system in Escherichia coli. Proc.Natl. Acad. Sci. U.S.A. 69:1091-1093.

113. Nikaido, H. 1976. Outer membrane of Salmo-nella typhimurium. Transmembrane diffusionofsome hydrophobic substances. Biochim. Bio-phys. Acta 433:118-132.

114. Nordstrom, K., L C. Ingram, and A. Lund-back. 1972. Mutations in R factors of Esche-richia coli causing an increased number of R-factor copies per chromosome. J. Bacteriol.110:562-569.

115. Olsen, R. H., and P. Shipley. 1973. Host rangeand properties of the Pseudomonas aerugi-nosaR factor R1822. J. Bacteriol. 113:772-780.

116. Park, M. H., B. B. Wong, and J. E. Lusk. 1976.Mutants in three genes affecting transport ofmagnesium in Escherichia coli: genetics andphysiology. J. Bacteriol. 126:1096-1103.

117. Pattee, P. A., N. E. Thompson, D. Haubrich,and R. P. Novick. 1977. Chromosomal maplocations of integrated plasmids and relatedelements in Staphylococcus aureus. Plasmid1:38-51.

118. Plakunov, V. K. 1973. Relation between chem-ical structure of tetracyclines and their ana-logues and specific features of transport andmechanism of antibacterial effect. Antibiotiki(Moscow) 18:1069-1073.

119. Pugsley, A. P., and C. A. Schnaitman. 1978.Outer membrane proteins of Escherichia coli.VII. Evidence that bacteriophage-directed pro-tein 2 functions as a pore. J. Bacteriol. 133:1181-1189.

120. Raetz, C. R. H., and E. P. Kennedy. 1974.Partial purification and properties of phospho-tidylserine synthetase from Escherichia coli.J. Biol. Chem. 249:5038-5045.

121. Randall, L. L., and S. J. S. Hardy. 1977. Syn-thesis of exported proteins by membrane-bound polysomes from Escherichia coli. Eur.J. Biochem. 73:43-53.

122. Rao-Movva, N., E. Katz, P. L. Asdourian, Y.Hirota, and M. Inouye. 1978. Gene dosageeffects of the structural gene for a lipoproteinof the Escherichia coli outer membrane. J.Bacteriol. 133:81-84.

123. Reeve, E. C. R. 1978. Evidence that there aretwo types of determinant for tetracycline re-sistance among R-factors. Genet. Res. 31:75-84.

124. Reeve, E. C. R., and J. M. Robertson. 1975.The characteristics of eleven mutants of R-factor R57 constitutive for tetracycline resist-ance, selected and tested in Escherichia coliK12. Genet. Res. 25:297-312.

125. Ringrose, P. S., and J. E. Higgins. 1974. Theinterrelationship of tetracycline resistance, de-cynoyl-N-acetyl cysteamine and membranefatty acid composition in Escherichia coli. J.Antibiot. 27:833-837.

126. Robertson, J. M., and E. C. R. Reeve. 1972.Analysis of the resistance mediated by severalR-factors to tetracycline and minocycline. Ge-net. Res. 20:239-252.

127. Rosenberg, H., R. G. Gerdes, and K. Cheg-widden. 1977. Two systems for the uptake ofphosphate in Escherichia coli. J. Bacteriol.131:505-511.

128. Schaefler, S., W. Francois, and C. L. Ruby.1976. Minocycline resistance in Staphylococ-cus aureus: effect on phage susceptibility. An-timicrob. Agents Chemother. 9:600-613.

129. Schlieper, P., and E. de Robertis. 1977. TritonX100 as a channel-forming substance in artifi-cial lipid bilayer membranes. Arch. Biochem.Biophys. 184:204-208.

130. Schweisinger, M. J. 1977. Additive recombina-tion in bacteria. Bacteriol. Rev. 41:872-902.

131. Scoy, R. E. V. 1977. Prophylactic use of antimi-

MICROBIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from



crobial agents. Mayo Clin. Proc. 52:701-703.132. Shipley, P. L, and R H. Olsen. 1974. Charac-

teristics and expression of tetracycline resist-ance in gram-negative bacteria carrying thePseudomonas R factor RP1. Antimicrob.Agents Chemother. 6:183-190.

133. Silhavy, T. J., I. Hartig-Beecken, and W.Boos. 1976. Periplasmic protein related to thesn-glycerol-3-phosphate transport system ofEscherichia coli. J. Bacteriol. 126:951-958.

134. Simoni, R. D., and P. W. Postma. 1975. Theenergetics of bacterial active transport. Annu.Rev. Biochem. 43:523-554.

135. Singer, S. J. 1974. The molecular organizationof membranes. Annu. Rev. Biochem. 43:805-833.

136. Smit, J., Y. Kamio, and H. Nikaido. 1975.Outer membrane of Salmonella typhimurium:chemical analysis and freeze-fracture studieswith lipopolysaccharide mutants. J. Bacteriol.124:942-958.

137. Smith, W. P., P.-C. Tai, and B. D. Davis. 1978.Nascent peptide as sole attachment of poly-somes to membranes in bacteria. Proc. Natl.Acad. Sci. U.S.A. 75:814-817.

138. Smith, W. P., P.-C. Tai, R. C. Thompson, andB. D. Davis. 1977. Extracellular labelling ofnascent polypeptides traversing the membraneof Escherichia coli. Proc. Natl. Acad. Sci.U.S.A. 74:2830-2834.

139. Sompolinsky, D., I. Hammerman, 0. Assaf,and A. Wojdani. 1978. Tetracycline resistanceantigen from Staphylococcus aureus. FEMSMicrobiol. Lett. 4:23-26.

140. Sompolinsky, D., and J. Krausz. 1973. Actionof 12 tetracyclines on susceptible and resistantstrains of Staphylococcus aurew. Antimicrob.Agents Chemother. 4:237-247.

141. Sompolinsky, D., and Z. Samra. 1972. Influ-

ence of magnesium and manganese on some

biological and physical properties of tetracy-cline. J. Bacteriol. 110:468-476.

142. Stock, J. B., B. Rauch, and S. Roseman. 1977.Periplasmic space in SalmoneUa typhimuriumand Escherichia coli. J. Biol. Chem. 252:7850-7861.

143. Suling, W. J., and W. M. O'Leary. 1977. Lipidsof antibiotic-resistant and -susceptible mem-

bers of the Enterobacteriaceae. Can. J. Micro-biol. 23:1045-1051.

144. Tait, R. C., and H. W. Boyer. 1978. On thenature of tetracycline resistance controlled bythe plasmid pSC101. Cell 13:73-81.

145. Tait, R. C., and H. W. Boyer. 1978. Restrictionendonuclease mapping of pSC101 and pMB9.Mol. Gen. Genet. 164:285-288.

146. Tait, R C., H. L, Heyneker, R. L. Rodriguez,F. Bolivar, A. Covarrubias, M. Betlach,and H. W. Boyer. 1978. Tetracycline resist-ance conferred by pSC101, pMB9, and theirderivatives, p. 174-176. In D. Schlessinger(ed.), Microbiology-1978. American Societyfor Microbiology, Washington, D.C.

147. Tait, R. C., R L. Rodriguez, and H. W. Boyer.1977. Altered tetracycline resistance in pSC101

recombinant plasmids. Mol. Gen. Genet. 151:327-331.

148. Taylor, D. P., J. Greenberg, and R. H.Rownd. 1977. Generation of miniplasmidsfrom copy number mutants of the R plasmidNR1. J. BacterioL 132:986-995.

149. Teather, R. M., B. Miuller-Hll, U. Abrutsch,G. Aichele, and P. Overath. 1978. Amplifi-cation of the lactose carrier protein in Esche-richia coli using a plasmid vector. Mol. Gen.Genet. 159:239-248.

150. Tseng, J. T., and L. E. Bryan. 1973. Mecha-nisms of R factor R931 and chromosomal tet-racycline resistance in Pseudomonas aerugi-nosa. .Antimicrob. Agents Chemother. 3:638-641.

151. Uhlin, B. E., and K. Nordstrom. 1975. Plasmidincompatibility and control of replication: copymutants of the R-factor Rl in Escherichia coliK-12. J. Bacteriol. 124:641-649.

152. Uhlin, B. E., and K. Nordstr6m. 1977. R plas-mid gene dosage effects in Escherichia coli K-12: copy mutants of the R plasmid Rldrd-19.Plasmid 1:1-7.

153. van Alphen, W., R. van BoxteL N. van Selm,and B. Lugtenberg. 1978. Pores in the outermembrane of Escherichia coli K-12. Involve-ment of proteins b and c in the permeation ofcephaloridine and ampicillin. FEMS Microbiol.Lett. 3:103-106.

154. van Alphen, W., N. van Selm, and B. Lugten-berg. 1978. Pores in the outer membrane ofEscherichia coli K12. Involvement of proteinsb and e in the functioning of pores for nucleo-tides. Mol. Gen. Genet. 159:75-83.

155. Walker, M. S., and J. B. Walker. 1970. Strep-tomycin biosynthesis and metabolism. Enzy-matic phosphorylation of dihydrostreptobios-amine moieties of dihydrostreptomycin-(strep-tidino) phosphate and dihydrostreptomycin byStreptomyces extracts. J. Biol. Chem. 245:6683-6689.

156. Wargel, R. J., C. A. Shadur, and F. C. Neu-haus. 1971. Mechanism of D-cycloserine ac-tion: transport mutants for D-alamine, D-cyclo-serine, and glycine. J. Bacteriol. 105:1028-1035.

157. Webb, M. 1970. Interrelationships between theutilization of magnesium and the uptake ofother bivalent cations by bacteria. Biochim.Biophys. Acta 222:428-439.

158. Weckesser, J., and J. A. Magnuson. 1976.Light-induced tetracycline accumulation byRhodopseudomonas sphaeroides. J. Supra-mol. Struct. 4:515-520.

159. Wilson, W. R 1977. Antimicrobial agents: tet-racyclines, chloramphenicol, erythromycin andclindamycin. Mayo Clin. Proc. 52:635-640.

160. Wojdani, A., RK R Avtalion, and D. Sompo-linsky. 1976. Isolation and characterization oftetracycline resistance proteins from Staphy-lococcus aureus and Escherichia coli. Anti-microb. Agents Chemother. 9:526-534.

161. Yagi, Y., and D. B. Clewell. 1976. Plasmid-determined tetracycline resistance in Strepto-

VOL. 42, 1978


http://mm

br.asm.org/

Dow

nloaded from


724 CHOPRA AND HOWE

coccus faecalis: tandemly repeated resistancedeterminants in amplified forms of pAMalDNA. J. Mol. Biol. 102:583-600.

162. Yagi, Y., and D. B. Clewell. 1977. Identificationand characterization of a small sequence lo-cated at two sites on the amplifiable tetracy-cline resistance plasmid pAMal in Streptococ-cus faecalis. J. Bacteriol. 129:400-406.

163. Yang, H. L., G. Zubay, and S. B. Levy. 1976.Synthesis of an R plasmid protein associated

MICROBIOL. REV.

with tetracycline resistance is negatively regu-lated. Proc. Natl. Acad. Sci. U.S.A. 73:1509-1512.

164. Young, T. W., and S. J. Hubball. 1976. R-factor mediated resistance to tetracycline inEscherichia coli K-12. An R-factor with a mu-tation to temperature sensitive tetracycline re-sistance. Biochem. Biophys. Res. Commun. 70:117-124.

STATEh'ENT OF OWNERSHIP. MANAGEMENT AND CIRCULATION

I, ..1T O U8CTO_ UiCTO;. l)e

MICROSIOLOGICAL REVIES/ R t1 /19/;Ouarterly 4 ie.S8 n.namr. S30*..o1IwoosotcFnictie C." C-0_f. 8- _ zl C-,Nol_

1913 I St., W oWshington, C 20006

1913 I St, W, Washington, X 20006; tloOt SotiotyS Ao COlItlloo 00iSy S O 90140. 0. hg.... DC 200T06

Anerican Society for Mticrobiologyr, 1913 I St., NW, Washington, DC 2000D6

t. G. 0. Murray, University of Western Ontario. L1,doo. Ontario, CANADA

Robert A. Day, 1913 I St., NW, Washington, DC 20006

A_rican Society for Mictrobiology 1913 1 St., VW, V.oblogton, DC 20006

IL Po. COWRTS*YO. eT IOZe1A%-IL TO WGC ^L "AM O IJl.ISJ-2. T) j

T__--_F____________________ 1645 171VSCs^-m-O- O._v

R.o,... ..D ..WO.. T.... MIA D1645 174=T.11

..P- 6O

16459 17416

~~~16453 1 741

.........X. . ...4s .....Ot 5

,h.Wez-D rC D 16459 17416

3910 2577

*.Te -LNone ________e

.A) 20369 1 993

I~~~~~~~~~~~ecordfydmao*wftmm *F6 by t ................ ;s ..

"O- e co &WsreO, a a I .- s .1 4 _9 aI -.. E

00 3O26Pw002 1

........

1"' Ay.


http://mm

br.asm.org/

Dow

nloaded from


Documents

Bacterial Resistance to the Tetracyclinesmmbr.asm.org/content/42/4/707.full.pdfHowever, tetracycline uptake by mem-brane vesicles wasreported to be energy inde-pendent andapparently