INFECTIVE HEREDITY OFMULTIPLE DRUG RESISTANCE IN [600472]

INFECTIVE HEREDITY OFMULTIPLE DRUG RESISTANCE IN
BACTERIA
TSUTOMU WATANABE
Department ofBacteriology, KeioUniversity SchoolofMedicine, Tokyo,Japan
INTRODUCTION………………………………………………………………………. 87
DRUG-RESISTANT SHIGELLAE INJAPAN ……………………………………………….. 88
DEVELOPMENT OFMULTIPLE DRUGRESISTANT SHIGELLAE ANDESCHERICHIA COLI………….. 88
EPIDEMIOLOGY OFMULTIPLE DRUGRESISTANT SHIGELLAE ANDESCHERICHIA COLI………….. 89
GENETICS OFMULTIPLE DRUGRESISTANCE …………………………………………….. 89
Transfer ofMultipleDrugResistance byConjugation ……………………………………. 89
Elimination ofResistance FactorswithAcridines ……………………………………….. 92
Transduction ofMultipleDrugResistance …………………………………………….. 93
ConceptofResistance Transfer Factor(RTF)andRFactors ……………………………… 94
ConceptofEpisomes ……………………………………………………………….. 95
Spontaneous Segregation ofResistance Factors …………………………………………. 95
Recombination ofRFactors …………………………………………………………. 97
Circular ModelsofRFactors ………………………………………………………… 97
Interactions betweenRTFandFFactorinEscherichia coliK-12………………………….. 98
Interactions between RTFType1andColicinogenic Factors inEscherichia coliK-12 …………. 100
Evidence forChromosomal Attachment ofRTF…………………………………………. 101
HighFrequency Resistance Transfer System(HFRT) ……………………………………. 101
Biological Functions andMutations ofRTF…………………………………………… 102
Comparison ofRTFwithOtherEpisomes ……………………………………………… 104
OriginofResistance Factors CarriedbyRTF………………………………………….. 105
Molecular BasisofRFactors ………………………………………………………… 106
BIOCHEMICAL MECHANISMS OFMULTIPLE DRUGRESISTANCE ……………………………… 107
SUMMARY………………………………………………………………………….. 108
LITERATURE CITED ……………………………………………………………. 109
INTRODUCTION
Incivilizedcountries, bacillary dysentery isno
longeranimportant diseasebutinareaswhere
publichealthisunderdeveloped itstillpresents a
seriousproblem. Itseemsratherparadoxical
thatinJapan,wheresanitary conditions are
considered tobefairlygood,bacillary dysentery
isstilloneofthemostimportant infectious
diseases. Thisisapparently duetothedevelop-
mentofbacterial strainshighlyresistant to
drugs.ShortlyafterWorldWarII,ahighin-
cidence ofsulfonamide-resistant shigellae ap-
peared,and,since1957,shigella strainswith
multiple drugresistance havebeenisolated with
increasing frequency eachyear.Thismultiple
drugresistance involves streptomycin (Sm),
chloramphenicol (Cm),tetracycline (Tc),and
sulfonamide (Su).Asmallproportion ofthese
Shigellastrainsareresistant toonlysomeofthe
drugsbutthemajority areresistant toall.Since
thesedrugsareourmostpowerful chemothera-
peuticsagainstdysentery, thephenomenon of
87multiple drugresistance createsaseriousproblem
inthetherapyofthisdisease.
In1959,itwasfoundbyJapanese investigators
(5,6,89)thatmultiple drugresistance canbe
easilytransferred between shigellae andEscher-
ichiacolibymixedcultivation. Thisdiscovery led
manyJapanese workers tothegeneticstudyof
multiple drugresistance. Wehavefound(115,
123,124,125,131)thatthemultiple drugresist-
ancefactorsarecarriedandtransferred byan
episome(17,47,50).Multiple drugresistance is,
therefore, anexample of"infective heredity"
(Zinder, 147;Lederberg andLederberg, 61).
Investigations ofthebiochemical mechanisms of
multiple drugresistance indicate that,atleast
withSu,Cm,andTc,theresistance isdueto
reduced permeability ofthecellstothedrugs.
Initially, theproblem ofmultiple drugresist-
ancereceived attention because ofitsmedical
importance, butmorerecently muchefforthas
beendevoted togeneticstudiesfromwhichthe
episomal natureoftheresponsible factorsis

BACTERIOL. REV.
11i
0q
|.8
x
N-7
4r-
–
-YEARR
FIG.1.Statistics ofbacillary dysentery inJapan
between1946and1969(72).
emerging asoneofthemostinteresting problems.
Thereaderisreferred totwoexcellent reviewson
episomes, Jacob,Schaeffer, andWollman (47)
andCampbell (17),whicharerelevant tothe
problem ofmultiple drugresistance. Thepresent
reviewwillplaceparticular emphasis ongenetic
studiesofmultiple drugresistance. Because many
ofthepaperscitedinthepresent revieware
writteninJapanese, Iwillrefertotheminmore
detailthanisusualinreviews.
DRUG-RESISTANT SHIGELLAE INJAPAN
AttheendofWorldWVarII,variousderivatives
ofSuwereintroduced inJapanforthetreatment
ofdysentery andprovedtobeveryeffective for
thefirstseveralyears,reducing theincidence
considerably (Fig.1).However, afterabout1949,
theincidence ofdysentery againincreased,
despiteextensive useofSu,andmostofthe
Shigella strainsisolated fromthosecaseswere
foundtoberesistant tothedrug.Thepeak
incidence ofdysentery occurred in1952.Sub-
sequently, newerantibiotics, suchasSm,Cm,andTc,wereemployed forthetreatment ofSu-
resistant shigellae, withinitially excellent thera-
peuticeffects.Thenumber ofdysentery patients
wasagainreduced. However, withinonly4
yearsShigella strainsresistant totheseanti-
bioticsemerged, andthenumber ofdysentery
patients hasincreased markedly since1957.On
theotherhand,themortality ratehascontinued
todecline.
DEVELOPMENT OFMULTIPLE DRUG
RESISTANT SHIGELLAE AND
ESCHERICHIA COLI
StrainsofShigellaresistant toeitherSmorTc
wereisolated asearlyas1953butstrainswith
multiple resistance werenotfounduntil1955
(Table1).Themostfrequently encountered type
ofresistance isforSm,Cm,andTc.Thesecond
mostfrequent typeisforSmandCm.Strains
resistant toeitherTcorSmalonearealsofairly
common. However, relatively fewstrains are
resistant toSmandTc,ortoCmandTc,andno
strainsresistant toCmalonehavebeenfound.
ManyofthestrainslistedinTable1areresistant
toSuaswell.Thefirstisolation ofshigellae with
multiple drugresistance wasreported inadys-
TABLE1.Statistics ofantibiotic-resistant
shigellae inJapan*
Year
1953
1954
1955
1956
1957
1958
1959
1960
TotalNo.ofstrainsItested
4,900
4,876
5,327
4,399
4,873
6,563
4,071
3,396
38,405No.ofstrainsresistant to
Sm
5
11
4
8
13
18
16
29Tc
2
0
0
4
45
20
32
36
104139Cm
0
0
0
0
0
0
0
0
0SmandCm
0
0
0
0
2
7
71
61SmandTc
0
0
0
0
2
2
0
9
14113CmandTc
0
0
0
1
0
0
0
7Sm,Cm,andTc
0
0
1
0
37
193
74
308
8613
*Onlyonestrainfromeachepidemic isin-
cludedinthistable.Thestatistics from1953to
1958arederived fromseveralindependent reports
(Mitsuhashi etal.,76).Thestatistics for1959
and1960werereported bytheMinistry ofHealth
andWelfare, Japan.Abbreviations: Sm=strep-
tomycin; Tc=tetracycline; Cm =chloram-
phenicol; Su=sulfonamide.88 WATANABE

VOL.27,1963INFECTIVE HEREDITY OFBACTERIAL DRUGRESISTANCE
enterypatientwhohadjustreturned fromHong
Kongin1955(Kitamoto etal.,52).Afterthis
report,manyepidemics ofmultipledrugresistant
shigellae wereobserved bymanyworkers
throughout Japan.Inoneofthoseepidemics,
Matsuyama etal.(68)isolated multiple drug
resistant E.colistrainsfromdysentery patients.
Littleattention waspaidtothisreportuntil
epidemiological studiesofmultiple drugresistant
shigellae disclosed severalunexpected features,
whichwillbedescribed below.
EPIDEMIOLOGY OFMULTIPLE DRUG
RESISTANT SHIGELLAE AND
ESCHERICHIA COLI
Inepidemiological studiesofdysentery, several
workers repeatedly foundthatcompletely
sensitive Shigella strains canbeisolated from
somepatients, whereas thestrainsisolated from
otherpatients inthesameepidemic, andbelong-
ingtothesameserological type,showmultiple
drugresistance (4,68,86).Somepatients were
evenfoundtoexcretebothsensitive andmul-
tipledrugresistant Shigella strainsofthesame
serological type.Ontheotherhand,thead-
ministration ofasingledrug,forexample Cm,
topatients harboring sensitive Shigella strains
sometimes causedtheexcretion ofmultipledrug
resistant Shigella strains(51,53,88).This
observation cannotbeattributed tomutations,
aswillbediscussed below.Noacceptable ex-
planation ofthesecuriousfindings wasforth-
cominguntilAkiba(2)suggested thatmultiple
drugresistance maybetransferred frommultiple
drugresistant E.colitoshigellae intheintestinal
tractsofthepatients. Thistransfer wasdemon-
stratedindependently byOchiaietal.(89)and
Akibaetal.(5,6)invitro.Judgedbytheseverity
ofclinicalsymptoms, themultipledrugresistant
Shigellastrainsdidnotdifferfromthesensitive
strainsintheirpathogenicity tohumanbeings
(Ochiai,86).Thereduction inmortality seenin
Fig.1ispresumably duetoimproved methods
fortreatment ofdysentery.
GENETICS OFMULTIPLE DRUGRESISTANCE
Transfer ofMultiple DrugResistance
byConjugation
AfterOchiaietal.(89)andAkibaetal.(5,6)
succeeded indiscovering thetransferofmultiple
drugresistance invitro,theyattempted un-
successfully totransfer theresistance withcell-freefiltrates ofresistant donorcultures. Accord-
ingly,theyconcluded thatcell-to-cell contact
(orconjugation) isessential forthetransfer.
Theirresultsindicated alsothatneither the
serological northebiochemical markers, which
arestudiedroutinely indiagnostic bacteriology,
arechanged bythetransfer ofmultiple drug
resistance. Theysuggested thatmultiple drug
resistance maybetransferred bymating(recom-
bination afterconjugation). Although this
explanation provedtobeincorrect, thepreviously
mentioned unexpected epidemiological observa-
tionswereclearlyexplained bythesuccessful
transfer ofmultiple drugresistance invitro.
Subsequently, Kagiwada etal.(51)andAkida
etal.(7)provedwithhumanvolunteers that
multiple drugresistance canbetransferred from
multiple drugresistant E.colitoshigellae inthe
intestinal tract.Thetransfer ofmultiple drug
resistance invivowasalsoshownindogs
(Mitsuhashi, Harada, andHashimoto, 73)and
mice(Akibaetal.,7).
Mitsuhashi etal.(80)reported thatmultiple
drugresistance couldbetransferred amongthe
substrains ofE.coliK-12irrespective oftheir
sexuality (presence orabsence ofFfactor), in-
dicatingthattheFfactorofK-12isnotrequired
forthetransfer ofmultiple drugresistance.
Watanabe andFukasawa (116,123)foundthat
theresponsible resistance factorsaretransferred
bymixedcultivation independently ofthehost
chromosome, usingFstrainsofE.coliK-12
withvarious chromosomal markers. Wecan,
therefore, regardmultiple drugresistance asa
kindof"infective heredity" (Zinder, 147;
Lederberg andLederberg, 61).Mitsuhashi (72)
reported thathewasunabletoisolatemultiple
drugresistant mutants byexposing sensitive
Shigella andE.colistrainstosingledrugsin
vitro,andthatthesingledrugresistance thus
obtained couldnotbetransferred toothersensi-
tivestrainsbymixedcultivation. Mutants with
multiple drugresistance wereobtained invitroas
multistep mutants byexposing sensitive strains
tosingledrugs,oneafteranother. Multiple drug
resistance ofthesemutants, however, couldnot
betransferred bymixedcultivation. Watanabe
andFukasawa (116,123)reported thatthe
transfer ofmultiple drugresistance doesnot
reducethegrowthrateofcells.Theseorganisms,
therefore, differfromtheusualslow-growing
drug-resistant mutants(114,137,138).89

BACTERIOL. REV.
Inaddition tothemultiple drugresistance,
variouscombinations ofresistance factorsfrom
naturally occurring drug-resistant Shigella and
E.colistrainswerefoundtobetransmissible in
mixedcultivation byOchiaietal.(89),Akiba
etal.(5,6),andMitsuhashi etal.(76,77,81).
Thoseare(Su,Sm,Cm),(Su,Sm,Tc),(Cm,Tc),
(Su,Sm),(Tc),(Sm),and(Cm).Thedrugs
indicated intheparentheses represent thecom-
binations ofresistance factorsthatcanbetrans-
ferredtogether byconjugation. (Cm)wasiso-
latedasaspontaneous segregant (seebelow)
from(Cm,Tc)invitroandhasneverbeeniso-
latedfromnaturalspecimens. Transmissible (Su)
hasneverbeenobtained eitherfromnatural
specimens orexperimentally. Manyofthestudies
ontransmissible drugresistance havebeencon-
cernedwiththe(Su,Sm,Cm,Tc)complex
because ofitsmedicalandgeneticimportance.
Transfer ofmultiple drugresistance togenera
otherthanShigella andEscherichia. Nakaya,
Nakamura, andMurata (84)andHaradaetal.
(32,33)extended thefindingsofOchiaietal.(89)
andAkibaetal.(5,6)andfoundthatmultiple
drugresistance canbetransferred toalmost
everygenusofthefamilyEnterobacteriaceae.
Ochiai(87)reported thatmultiple drugresist-
ancecannotbetransferred tostaphylococci and
Vibriocomma.Itwasrecentlyreported, however,
thatmultiple drugresistance canbetransferred
toSerratiamarcescens (Falkow etal.,22)andalso
toV.comma(BaronandFalkow, 15)bymixed
cultivation.
Thefrequencies oftransfer ofmultiple drug
resistance differconsiderably fromdonorto
donorandalsofromrecipient torecipient. F
strainsofE.coliK-12arethebestrecipients so
farencountered (Watanabe andFukasawa, 122),
thefrequencies oftransfer in1hrrangingfrom
about10-2tolessthan10-7perdonorcell,de-
pending onthedonorstrainsemployed. Some
ofthedonorstrainswerefoundtoproducecolicins
andphages, whichreducethefrequency of
transferbykillingtherecipient cells.Similarly,
donorsmaybekilledbycolicinsandphages
produced byrecipients. Whenvariousrecipients
wereusedforthesamedonor,Salmonella
strains werefoundtodevelop multiple drug
resistance onlywithverylowfrequencies, namely,
lessthan10-7perdonorcellevenwhengood
donorswereused(Watanabe andFukasawa, 116,
123).Itisnotknownwhysalmonellae arepoor
recipients.Effectsofphysiological conditions onthetransfer
ofmultiple drugresistance. Various effectsofen-
vironmental conditions onthetransferofmultiple
drugresistance havebeenreported butsomeof
themlackclear-cut interpretations andrequire
further elucidation. Koyama andAkiba(55)
reported thattheaddition ofsodiumdeoxycholate
tothemediainhibits thetransfer ofmultiple
drugresistance fromE.colitoshigellae. Since
thiscompound didnotinhibitthetransfer of
multiple drugresistance fromShigellatoShigella,
andsinceitisknownthatE.coliismoresensitive
todeoxycholate thanshigellae, areasonable
interpretation isthatsodiumdeoxycholate, even
insublethal concentrations, exertsdrasticphys-
iological effectsonE.colicells.
YokotaandAkiba(145)reported thatsodium
dodecylsulfate, ananionicdetergent, suppresses
thetransfer ofmultiple drugresistance inmixed
cultivation. Twootheranionicdetergents, caproic
acidandxylenesulfonic acid,didnotshowthis
inhibitory effect.Theyalsoreported thatmito-
mycinC,whichspecifically inhibitsthesynthesis
ofdeoxyribonucleic acid(DNA;Sekiguchi and
Takagi,102),alsoinhibitsthetransfer ofmultiple
drugresistance.
Akibaetal.(7)foundthatthetransfer of
multiple drugresistance doesnottakeplacein
theintestinal tractaseasilyasitdoesinvitro.
Kato(personal communication) showedthatthe
addition offecalfiltrates totheculturemedia
reducesthefrequency oftransfer. Byfractiona-
tion,theactivesubstances inthefecalfiltrates
wereidentified asfattyacidsanddeoxycholate.
However, theconcentration ofdeoxycholate
foundwastoolowtoaccount fortheobserved
effect.
Iijima(40)reported thattreatment ofresist-
antdonorcellswithsodiumperiodate transiently
deprives themoftheabilitytotransfermultiple
drugresistance byconjugation. Heassumed that
thisinhibition mayresultfromthedestruction of
"mating substance" ofpolysaccharide nature.
SneathandLederberg (106),inasimilarexperi-
ment,foundthattreatment ofF+andHfrcells
withsodiumperiodate temporarily devirilizes the
malebacteria.
YokotaandAkiba(142)reported thatmultiple
drugresistance canbetransferred inthecom-
binations ofpenicillin-spheroplast donorand
intactrecipient, intactdonorandpenicillin-
spheroplast recipient, andalsoofpenicillin-
spheroplasts ofbothdonorandrecipient. These90 WATANABE

VOL.27,1963 INFECTIVE HEREDITY OFBACTERIAL DRUGRESISTANCE
findings arecomparable tothoseofLederberg
andSt.Clair(62)andHagiwara (26)onthe
matingofE.coliK-12,andofIijima(39)onthe
transferofcolicinogenic factorEl,indicating that
theintactcellwallsofdonorandrecipient cells
arenotessential forthetransfer. Ontheother
hand,ifthe"mating substances" areessential
forthetransfer ofmultiple drugresistance, F
factor,andhostchromosome, theymustbe
retained bythepenicillin spheroplasts.
Watanabe andFukasawa (129)andIwahara
andAkiba(43)foundthatcultures inthestation-
aryphaseprovide somewhat morecompetent
donorsthanthoseinthelogarithmic phase.The
reasonforthisdifference isnotknown.Inaddi-
tion,Watanabe andFukasawa (129)observed
thatcellswhichhavejustreceived themultiple
drugresistance factorcantransfer itwithex-
ceedingly highfrequencies. Thisfindingwillbe
discussed later.
AkibaandIwahara (3)reported thatthe
optimal pHandtemperature forthegrowthof
donorandrecipient strainsarealsooptimal for
thetransfer ofmultiple drugresistance. Similar
findings havebeenobtained byHayes(35)in
thematingofE.coliK-12.
Egawaetal.(19)reported thatthetransferof
multiple drugresistance doesnottakeplacein
saline,orinasynthetic mediumlackingacarbon
source,orunderanaerobic conditions inacom-
pletemedium. Theyobserved alsothatmonoiodo-
acetate,sodiumfluoride, andfluoroacetate have
nosignificant effectonthetransfer ofmultiple
drugresistance, whereas potassium cyanide,
malonate, sodiumarsenite, and2,4-dinitrophenol
areinhibitory. Theyconcluded thattheenergy
provided byoxidative phosphorylation isessential
forthetransfer. Asimilarconclusion hadbeen
reachedalreadybyFisher(23,24)instudieson
zygoteformation ofE.coliK-12.Furthermore,
Fisherfoundthattheenergyisrequired uniquely
bytheHfrparent,andonlyduringtheperiodof
chromosomal transfer.
Asindicated above,manyreportshavedealt
withtheeffectsofvariouschemical andphysical
conditions onthetransfer ofmultiple drug
resistance, butfewhaveattempted todiffer-
entiatebetween effectsondonorsandrecipients.
Thisfactmakesinterpretation ofthefindings
difficult.
Phenotypic expression ofresistance factors.The
resistance factorsreceived bytherecipient cells
areratherrapidlyexpressed phenotypically. Inthesubstrains ofE.coliK-12,resistance toCm
andTciscompletely expressed within10min
afterdonorandrecipient aremixed(Watanabe
andFukasawa, 126).Thisisincontrast tothe
phenotypic expression oftheusual(chromosomal)
Smresistance, whichisconsidered toberecessive
toitssensitive allele(Lederberg, 57;Watanabe
andWatanabe, 137).Although phenotypic ex-
pression ofthisrecessive Smresistance marker
requires segregation fromitssensitive alleles,
mostofthefactorsinvolved inmultiple drug
resistance donotrequirecelldivision fortheir
phenotypic expression. OnlytheSmresistance
factorapparently requires longerincubation for
itsphenotypic expression. However, withcertain
recipient cells,Smresistance isexpressed pheno-
typically beforethefirstcelldivision, suggesting
thatitisnotrecessive. Thedifference between the
Smresistance factorandtheotherfactorsin-
volvedinmultiple drugresistance ispossibly
duetotherapidbactericidal actionofSm.The
otherdrugsarebacteriostatic andactrather
slowly.Furthermore, aswillbediscussed later,
thestrainswithmultiple drugresistance were
foundtodecompose Cm(Miyamura, 82;Yokota
andAkiba,144)andpossiblyTc(Watanabe etal.,
unpublished data).Whenthemixedcultureof
donorandrecipient isplatedonselective media
containing CmorTc,withouteliminating donor
cells,thedonorcellsdecompose thedrugsand
reducetheiractiveconcentrations. Thenthe
recipient cells,whichhavebeenunabletodivide,
startdividingslowly.Inthismanner,thetransfer
andphenotypic expression ofmultiple drug
resistance takesplaceonselective mediawhich
initiallycontained highenoughconcentrations of
CmorTctoinhibittherecipient.
Inmediacontaining Su,thepotency ofSuis
presumably reducedbytheproduction ofantago-
nistsbythedonorcells.Therefore, todetermine
accurately thefrequency oftransfer ofmultiple
drugresistance, itisessential toeliminate the
donorcellsbeforeplatingthemixedcultureon
theselective media.Highconcentrations ofSm
(1,000,gperml)andhightitersofphageT6,to
whichdonorcellsaresensitive andrecipient
cellsareresistant, havebeensuccessfully used
foreliminating donorcells(Watanabe and
Fukasawa, 126).TheSmresistance factorin-
cludedinthemultiple drugresistance endowsE.
coliwithonlyalowlevelofSmresistance (be-
tween10and25,ugperml),aswillbedescribed
subsequently.91

BACTERIOL. REV.
Thelevelsofresistance differconsiderably
amongstrainsmanifesting multiple drugresist-
ance(Watanabe andFukasawa, 116,123).
Smresistance isveryhighinshigellae (morethan
1,000jsgperml)butratherlowinE.coli(between
10and25gAgperml).Tcresistance islowin
Salmonella strains(about10/Agperml)but
higherinshigellae andE.coli(between 100and
250jugperml).Evenamongthesubstrains of
E.coliK-12withmultiple drugresistance, the
levelsofCmandSuresistance maydiffer.Thus
multiple drugresistance provides aninteresting
technique forstudying "nucleocytoplasmic"
relationships.
Transferred multiple drugresistance canbe
further transferred toothersensitive strains
indefinitely. Thereisnoevidence thattheresist-
ancefactors aremodified byhostcellsduring
theirtransfer fromstraintostrain.
Kinetics oftransfer ofmultiple drugresistance.
ItwasfoundbyWatanabe andFukasawa (119,
123)thatthetransfer ofmultiple drugresistance
canbeinterrupted bytreatment withablender,
indicating thatthetransfer isactually causedby
conjugation. Bydiluting themixture ofdonor
andrecipient atvarious timestoreducethe
chances ofconjugation, theyshowed further
thattheconjugation startstakingplacealmost
instantaneously afterdonorandrecipient are
mixed.Byinterrupting theconjugation with
phageT6(Watanabe andFukasawa, 126),itwas
foundthatthetransferofmultipledrugresistance
takesplacewithin1minaftermixingofdonor
andrecipient. Nosegregational transfer of
multiple drugresistance couldbefound,even
whentheconjugating mixture wastreatedwitha
blenderorwithphageT6atanearlystage.
Incubation ofasmallnumberofmultiple drug
resistant cellswithalargenumber ofsensitive
cells,underconditions whichpreclude chromo-
somaltransfer, causedamajority ofthesensitive
cellstoacquiremultiple drugresistance rather
quickly, indicating thattheresistance factors
responsible forthetransfer replicate fasterthan
thehostchromosome (Watanabe andFukasawa,
116,123).Thisfinding,together withthewide
hostrangefortransferofmultiple drugresistance
andtheindependence fromhostchromosomal
markers, suggests thattheresistance factors
replicate inthecytoplasm independently ofthe
hostchromosome (17,47).Itshouldbeempha-
sizedherethatallfourresistance factors aretransferred together asaunit,suggesting close
linkagebetween them.
Elimination ofResistance Factors
withAcridines
Ephrussi, Hottinguer, andChimenenes (20)
foundthatcytoplasmic factorsofyeastcanbe
eliminated byacriflavine, converting thecellsto
"petitecolonie" mutants. HirotaandIijima(37)
andHirota(36)reported thatacriflavine con-
vertsF+cellsofE.coliK-12toF-byeliminating
theFfactor.Watanabe andFukasawa (118,124)
foundthattreatment ofmultiple drugresistant
shigellae andE.coliwithacriflavine andacridine
orange converts thecellstodrugsensitivity,
although withlowfrequencies, indicating that
theresistance factorsareinanautonomous state.
Intheelimination ofresistance factorswith
acridines, nosegregated elimination wasnoted.
Furthermore, acridine orangewaslesspotent
thanacriflavine ineliminating theresistance
factors, unlikethesituation withtheFfactor
reported byHirota(36).Elimination ofmultiple
drugresistance withacriflavine wasalsodemon-
stratedbyMitsuhashi, Harada, andKameda
(78,79).
Wehavenoclear-cut explanation forthislow
frequency ofelimination atthepresenttime.It
mightresultfromprevalence ofintegrated (or
chromosome-attached) factorsofmultiple drug
resistance, aswillbediscussed below.Hirota(36)
foundthattheFfactorinHfrstrains,whichis
considered tobeintegrated, cannotbeeliminated
withacridines. Thefactthattreatment ofultra-
violet-irradiated resistant cellswithacridines
resultsinmoreefficient elimination ofmultiple
drugresistance (Watanabe andFukasawa, 118,
124)mightbeinterpreted aspromoting "induc-
tion"ofintegrated resistance factorsbyultra-
violet,resembling theultraviolet-induction of
prophages (Lwoff,66).
Cells,whichhavelosttheresistance factors
bytreatment withacridines, werefoundto
acceptsubsequent transfer ofresistance factors
aboutasreadily asdidtheoriginally sensitive
recipient. Aswillbediscussed subsequently,
recipients already possessing partoftheresist-
ancefactorscanonlyreceivetheotherresistance
factorswithareduced frequency. Hence,the
sensitive cellsobtained bytreatment with
acridines areassumed tohavenomemory ofthe
multiple drugresistance, whichmightconfer
"immunity" uponthemagainst reinfection.92 WATANABE

VOL.27,1963INFECTIVE HEREDITY OFBACTERIAL DRUG RESISTANCE
Watanabe andFukasawa (118,124)reported a
strainofShigella withmultiple drugresistance
whichgivesrisetoSu-resistant cellsaftertreat-
mentwithacridines. Because thisSuresistance is
nottransmissible, andbecausetheoriginalstrain
cantransfer multiple drugresistance including
Suresistance, theyconcluded thattheoriginal
multiple drugresistant Shigellapossessed twoSu
resistance factors,onecytoplasmic andtheother
chromosomal.
Transduction ofMultiple DrugResistance
Ochiaieta].(89),Akibaetal.(5,6),
Mitsuhashi, Harada, andHashimoto (74),and
Harada etal.(30)didnotbelievethattrans-
duction ortransformation wereinvolved inthe
transfer ofmultiple drugresistance, because
cell-free filtrates oftheresistant cultures were
incapable ofconverting thesensitive recipient
cellstoresistance. Theirconclusions werecorrect
asfarastheirsystems wereconcerned. However,
Nakaya, Nakamura, andMurata(84)andWa-
tanabeandFukasawa (117,125)latersucceeded
intransducing multiple drugresistance inE.coli
K-12withphagePlkc,andWatanabe and
Fukasawa (117,125)alsoeffected transduction
ofmultiple drugresistance inSalmonella typhi-
murium LT-2withphageP-22.Nakaya,
Nakamura, andMurata (84)didnotinvestigate
thetransduction ofSuresistance, although they
employed adonorresistant tofourdrugs.
Watanabe andFukasawa (117,125)studiedthe
transduction ofthefourresistance factorsin-
volvedandalsothetransferability ofthetrans-
ducedresistance factorsbyconjugation. The
resultsarequitedifferent inthesetwosystems.
(i)WithPlkc,Su,Sm,Cm,andTcresistance
factorsaretransduced together mostfrequently;
segregated transductions areratherrare,the
patterns ofsegregation sofarobserved beingthe
Tc-resistant typeandtheSm,Cm,Tc-resistant
type.(ii)Allthesetransductants withPlkccan
transfertheirresistance factor(s) byconjugation.
(iii)WithP-22,theresistance factorsarein-
variably segregated, thepatterns ofsegregation
beingSu,Sm,Cmresistance, Tcresistance, and,
veryrarely,Su,Smresistance. (iv)Onlyaminor-
ityofthemultiple-resistant andtheTc-resistant
transductants withP-22cantransfer their
resistance factor(s) byconjugation. Ultraviolet
irradiation oftheothertransductants didnot
inducesuchabilitytotransfer. Byusingspon-
taneous segregants (seebelow)derived fromamultiple drugresistant strain,Tcresistance was
transferred toanSu,Sm,Cm-resistant trans-
ductant ofLT-2,andSu,Sm,Cmresistance was
transferred toaTc-resistant transductant ofLT-2
withnoabilitytotransfer theTcresistance by
conjugation. Thefour-drug resistant strainsthus
obtained, however, couldtransfer byconjugation
onlytheresistance factor(s) received bycon-
jugation andnoneofthosefactorsinitially re-
ceivedbvtransduction.
Fromtheseresults,Watanabe andFukasawa
(120,125)presented ahypothesis ofepisome-
mediated transfer, according towhichthedrug
resistance factorsarecarriedandtransferred by
anepisome, whichtheydesignated as"resistance
transfer factor" (RTF). Theypostulated the
participation ofanepisome, because theresist-
ancefactors inthetransductants ofLT-2,
whichcannotbetransferred byconjugation, were
suspected tobeinanintegrated state.They
considered thatunlesstheresistance factorsare
integrated, thosesegregated fromtheepisome
shouldbeunabletoreplicate autonomously and
wouldassumethestateofabortive transduction
(Stocker, 108;Lederberg, 59;Ozeki,94).Such
abortivetransductants shouldfailtoformcolonies
ondrug-containing media.Infact,however, the
resistant transductants ofLT-2inherittheir
drugresistance quitestably.
Haradaetal.(34)succeeded inreproducing the
resultsoftransduction inS.typhimurium LT-2
withP-22obtained byWatanabe andFukasawa
(117,125),usinginsteadS.anatumwithphage
E15(IsekiandSakai,41,42;Uetake,Nakagawa,
andAkiba,113).Thecharacteristics ofthe
resistant transductants werequitesimilartothose
ofLT-2withP-22.Thetransduced resistance
factorscouldalsobetransduced byanother
phage, -34(Harada, 28,29;Uetake, Luria,and
Burrous, 112).Inthissecond-round transduction
by04,Su,Sm,andCmresistance factors were
againtransduced together.
Harada etal.(31)recently foundthatthe
transferofFfactortoaTC-resistant transductant
ofS.anatuminducesthetransferoftheresistance
factorbyconjugation tootherstrainsinde-
pendently ofthehostchromosomes. Therecipient
cells,whichreceived bothFfactorandthere-
sistancefactor,wereableinturntotransfer the
resistance factor,butthosewhichreceived only
theresistance factorcouldnolongertransferitby
conjugation. Theseresults suggest thatthe
resistance factorintheTC-resistant trans-93

BACTERIOL. REV.
ductant ofsalmonellae maybeinanautono-
mousstatebutcannotbetransferred bycon-
jugation, owingtosomedefectintheresistance
factor. Theresistance factors transduced to
Salmonella strains arequitestablyinherited,
unlikethosereceived byconjugation, inwhich
statetheyfrequently undergo spontaneous
segregation (seebelow). Themechanism by
whichtransduced resistance factorsinsalmonellae
cellsreplicate autonomously butcannot be
transferred byconjugation isnotknown at
present. Itispossible thatthegeneticdetermi-
nantsinvolved inthetransmissible resistance
factors, andresponsible fortheirtransfer by
conjugation, cannotbepickeduptogether with
theresistance factorsbytransducing phage.On
theotherhand,itisalsopossiblethattheresist-
ancefactorssegregated bytransducing phage,
fromthegenetic determinants responsible for
theirautonomous replication andtransfer by
conjugation, areabletoreplicate autonomously,
beingsupported bythegenome ofthetrans-
ducingphage.Infact,asimilarsituation was
reported byLuria,Adams, andTing(65)for
Pldlelements (defective P1phage,whichin-
corporated thelactoseregionofthehostgenome).
Theexceptional typeofTc-resistant trans-
ductant cantransfer itsTcresistance factorby
conjugation asindicated above.TheTcresistance
factorofthesetransductants isalsospontane-
ouslysegregated. Thefactorwhichstabilizes the
autonomously replicating nontransmissible re-
sistancefactorsinthetransductants isnotknown.
Thus,thepreviously discussed hypothesis of
Watanabe andFukasawa (119,124),according
towhichthetransduced resistance factorsin
salmonellae maybeintegrated, mayrequirerein-
vestigation.
Itwasdifficult tounderstand whynoSu,Sm,
Cm-resistant typeoftransductant couldbe
foundinK-12withPlkcintheexperiments of
Watanabe andFukasawa (117,125).Subsequent
studiesoftransduction withE.coliK-12and
PlkcbyKondo,Harada, andMitsuhashi (54),
usingadifferent four-drug resistant donor,
enabled themtofindthistypeoftransductant,
although withlowfrequencies.
Thekinetics ofphenotypic expression of
transduced resistance factors wasstudiedby
Watanabe andFukasawa (117,125).Theyfound
thatSu,Cm,andTcresistance isphenotypically
expressed rapidly, asinthetransfer ofresistance
factorsbyconjugation, whereas thephenotypic-MODEL
SuSmCmA-
-RTFTC
-MODEL B-
SuSmGmRTFTC
FIG.2.LinearmodelsofanRfactor(Su,Sm,
Cm,Tc)proposed byWatanabe andFukasawa
(125).
expression ofSmresistance requires aratherlong
incubation period.
ConceptofResistance Transfer Factor(RTF)
andRFactors
Watanabe andFukasawa (125)presented two
linearmodelsoftheresistance factorscarriedby
RTFonthebasisoftheirtransductional studies
(Fig.2).Theyassumed thattheresistance factors
mayhaveoriginated fromachromosome ofsome
bacteria. Thisassumption necessitates thepossi-
bilityofattachment ofRTFtohostchromosome
whenRTFpicksupthechromosomal genes.In
transduction, RTFwithout anyresistance factors
maywellbeproduced, butthereisnowayat
presenttoselectforsuchtransductants.
Beforethegeneticstructure andepisomality
ofthemultiple drugresistance factorsandother
infective resistance factorsweredisclosed, Iseki
proposed theterm"R-factors" fortheresistance
factorswhichareeasilytransferred fromcellto
cellbyconjugation. Thisisasimpleandcon-
venient termandhasbeenusedbyMitsuhashi
(72)andSuginoandHirota(109).Nakaya,
Nakamura, andMurata(84)proposed theterm
"resistance transfer agents"(Rta)forthesame
factors.ThetermRfactorshasalreadybeen
employed byseveralgroupsofworkers, andit
seemsunwisetoconfusetheterminology further
byusingthetermRta.Furthermore, Rtacanbe
confused withRTF.Iwill,therefore, usethe
termRfactorsinthepresentreview. Iwishto
emphasize, however, thatthetermRfactors
shouldbeusedinastricter sensethanheretofore.
Asindicated subsequently, Watanabe andLyang
(134,135)recentlyisolated twounusualstrainsof
multipledrugresistance fromacommon typeof
four-drug resistant parentofS.typhimurium
LT-2.Although theresistance factorsofthese94 WATANABE

VOL.27,1963INFECTIVE HEREDITY OFBACTERIAL DRUG RESISTANCE
unusualstrainsareinanautonomous state,they
cannotbetransferred byconjugation. Itisas-
sumedthatRTFinthesestrainshasundergone a
"defective" mutation. Ibelievethattheterm
Rfactorsshouldbeusedtodenoteagentswhich
arecomposed ofRTFandresistance factors,
irrespective oftheirtransmissibility byconju-
gation.
Watanabe andFukasawa (115,116,123,129)
testedShigellaandE.colistrainswithRfactors
fortheproduction ofcolicins. Somestrainswere
foundtobecolicinogenic, buttheRfactorsand
colicinogenic factors alwayssegregated from
eachotherintheirtransfer byconjugation, in-
dicating thattherearetwoindependent factors.
StrainsofE.coliK-12whichreceived Rfactors
weretestedforcolicinogeny, usingmorethan
150strainsofE.coliandshigellae asindicators.
Theseincluded E.coliK-12and4andShigella
sonneiE-90,whichareknowntobesensitive to
manycolicins. Colicinogeny wasnotdetected.
Watanabe andFukasawa (115,116,123,129)
alsostudiedthepossibility thattheRfactors
maymakeF-strainsofE.coliK-12fertile,but
failedtoprovefertility, usingvariouschromo-
somalgenesasselective markers. Suginoand
Hirota(109),however, recently reported that
RfactorsendowF-strainsofK-12withfertility;
thiswillbediscussed later.
ItwasshownbyOchiaietal.(89),Akibaetal.
(5,6),Mitsuhashi, Harada, andHashimoto (74),
andHaradaetal.(30)thatcell-free filtrates of
donorcultures areineffective intransferring
multiple drugresistance. Itis,therefore, evident
thatphagesarenotinvolved intheusualRTF
systems, exceptintheratherexceptional in-
stanceswheretransduction occurs.Therefore,
itiscertainthatRTFisnotaphage.
Rfactorsarethefirstclear-cut examples of
episome-mediated transfer ofdrugresistance.
Sager(101)reported onnon-Mendelian inher-
itanceofSmresistance inChlamydowmnas rein-
hardi,butthechromosomal attachment ofthe
responsible factorhasnotbeendemonstrated.
ConceptofEpisomes
Thetermepisome wasfirstusedbyThompson
(110),andonlylaterbyJacobandWollman (50),
inadifferent sense.Episomes cannotberegarded
asessential constituents ofcells,forcellswithout
themhavenodifficulty insurviving undernormal
conditions. Whenpresentincells,episomes may
existintwoalternative states,autonomous orintegrated. Intheautonomous state,theyrepli-
cateindependently ofthehostchromosome, and
usuallyatafasterpace.Theintegrated state
referstothestateofattachment ofepisomes to
thehostchromosome; inthisstate,theepisomes
replicate atthesamepaceasthehostchromo-
some.Ithasgenerally beenobserved thatau-
tonomous replication ofepisomes isinhibited by
thepresence oftheintegrated stateofthesame
episomes. Jacob,Schaeffer, andWollman (47)
included temperate phages,sexfactorsofE.coli,
andcolicinogenic factorsamongepisomes. The
"sporogenic factor"ofBacillus specieswasalso
considered tobeapossible episome. Thesein-
vestigators pointedoutacloserelationship among
theepisomes, andassumed thateachmaybe
abletomutatetoothertypesofepisomes. Luria,
Adams,andTing(65)suggested thatepisomes
otherthantemperate phagesmaybeproduced
astheresultofincorporation offragments ofhost
genomebytemperate phages.
Campbell (17)proposed toinclude among
episomes otherhereditary cytoplasmic factors,
whoseintegrated statehasnotyetbeenproved.
Forexample, "lac-episome" (13,14,15,22),
"sporulation factor" inBacillus species(47),
mycelial factorinAspergillus (100),and"con-
trollingelements" inmaize(69)wouldbere-
gardedasepisomes, inaddition totemperate
phages, Fagents,colicinogenic factors,andre-
sistance transfer factor.According toCampbell,
theseepisomes couldbedividedintotwoclasses,
namely, temperate phages(or,morestrictly,the
geneticmaterial thereof), andtransfer factors.
Transfer factorsaretheepisomes whichcanpass
fromcelltocellduringconjugation, independently
ofthebulkofthehostgenome. Someofthetrans-
ferfactorsplayacausative roleintheconjuga-
tionprocessitself.
Subsequent tothesuggestion ofWatanabe and
Fukasawa (115,120,125)thatRTFisakindof
episome, muchworkhasbeendevoted toproving
theepisomal natureofRTFandRfactorsand
studying itsrelationship tootherknownepi-
somes.Theautonomous stateofRfactorswas
proved(seeabove),andtheintegrated stateof
RTFwasdemonstrated (seebelow). Wecan,
therefore, classifybothRTFandRfactorsas
episomes.
Spontaneous Segregation ofResistance Factors
Manyinvestigators working onRfactorshave
notedoccasionally thespontaneous lossofpart95

BACTERIOL. REV.
oralloftheresistance factors. Itisourimpres-
sionthatthespontaneous segregation ofresist-
ancefactorsoccursmorereadilyinagedcultures
keptatroomtemperature thaninfreshcultures.
Thereasonforthisisunknown. Watanabe and
Fukasawa (118,124)studied thespontaneous
segregation ofresistance factorswithapenicillin
screening method, modified fromtheoriginal
method ofDavis(18).Thecellswithmultiple
drugresistance wereincubated inbrothcon-
tainingpenicillin andeitherCmorTc.Thespon-
taneous segregants, lackingresistance factorsto
thelatterdrugs,cannotdivideinbrothcontain-
ingthesedrugs,andselectively survivethelethal
actionofpenicillin. Thespontaneous segregants
werefoundtoberesistant eithertoSu,Sm,and
Cm,ortoTc.Complete lossoftheresistance
factorswasalsonoted.Thefrequency ofspon-
taneous segregation ofresistance factors was
highestinsalmonellae andlowestinE.coli.With
shigellae, intermediate frequencies ofspontaneous
segregation wereobserved.
Itshouldbeemphasized thatallthespon-
taneous segregants whichpossess anyofthe
resistance factorscantransfer theseresistance
factorsbyconjugation (Watanabe andFukasawa,
115,118,124).Yamanaka (141)assumed that
someoftheresistance factorsaremechanically
cutoffandcastoff.Thissuggests thateachre-
sistance factorcontrols itstransferbyconjuga-
tion.Suchanexplanation isunlikely inviewof
transduction experiments (Watanabe andFuka-
sawa,117,125),whichshowedthatanepisome
RTFoccupies aparticular position onRfactors.
Thereisotherevidence againstthehypothesis
thateachresistance factorcontrols itstransfer
byconjugation. Forexample, Watanabe and
Fukasawa (128)foundthatalltypesofRfactors,
including thespontaneous segregant types,sup-
presstheacceptance ofRfactorsandalsoinhibit
thefunctions ofFfactorofE.coliK-12(de-
scribed inSectionV-I).Thesefunctions ofR
factors werefoundnottobespecific forany
particular resistance factor,andmaybeascribed
toRTF,whichiscommon toallRfactors.
Watanabe andFukasawa (115,125)and
Watanabe andLyang(134)suggested thatspon-
taneous segregation ofresistance factors is
brought aboutbygeneticexchange between
RTF-carried resistance factorsandhostgenome
(Fig.3).Theyaccounted fordifferences inthe
frequencies ofspontaneous segregation ofre-
sistance factorsamongvariousbacterial speciesHOSTCHROMOSOME-
RTFSuSnCan Tc
RTFSuSRnCm
HOSTCHROMOSOME-
A '@1 RTFSuSMnCm Tc
Ic RTFTc
FIG.3.Mechanism ofspontaneous segregation
ofresistance factors.Hypothesis ofgeneticexchange
between Rfactorandhostgenome.
bythedegreesofgenetichomology between the
Rfactorsandthehostgenome.
Watanabe andLyang(134)foundthatthe
frequency ofspontaneous segregation ofresist-
ancefactorsissohighinS.typhimurium that
clonalanalysis ofsegregation ispossible with
thereplicaplatingtechnique (Lederberg and
Lederberg, 60).Whenapproximately 500cells
ofS.typhimurium withanRfactor(Su,Sm,
Cm,Tc)wereinoculated into5mlofbrothand
incubated at37Cfor24hr,theresulting culture
consisted ofabout75%four-drug resistant cells
and25%cellslackingsomeoralloftheresist-
ancefactors.Thecombination ofSu,Sm,Cm
resistance factorswaslostmostfrequently. It
wasassumed thatthegeneticexchange takes
placewhentheRfactorsareintheintegrated
state,thuspreventing theirautonomous replica-
tionincytoplasm. Otherwise, geneticexchange
shouldnoteasilygiverisetosegregant clones
because oftheautonomously replicating Rfac-
tors.Theseinvestigators didnotsucceed iniso-
latingdrug-resistant cloneswithnoRTF,which
wouldbedirectevidence forgeneticexchange
between RTF-carried resistance factorsandhost
genome. TheydidisolateanSu,Sm,Cmresist-
antcloneandafour-drug resistant clone,which
wereunabletotransfer theirresistance factors
byconjugation. Thesecloneswerefoundtohave
"defective" RTF,aswillbediscussed later.
Inaddition tothepossibility ofgenetic ex-96 WATANABE

VOL.27,1963INFECTIVE HEREDITY OFBACTERIAL DRUGRESISTANCE
changebetween RTF-carried resistance factors
andhostgenome, thefollowing mechanism may
beinvolved also.Ifreplication oftheRfactor
startsatoneendandforsomeunknown reason
stopssomewhere between thetwoends,segregant
typesofRfactorsshouldbeproduced. Circular
modelsofRfactors(seebelow)maybemore
usefulinexplaining thedevelopment ofvarious
combinations ofresistance factorsinRfactors.
Epidemiological studies ofdrug-resistant
shigellae clearlyindicate thatthefour-drug re-
sistantstrainsdeveloped early.Thereisnoevi-
dencethattheoriginalmultiple-drug resistance
wasproduced innaturebythesuccessive ac-
quisition ofsingleresistance factors.Ontheother
hand,Ochiaietal.(89),Akibaetal.(5,6),and
Mitsuhashi etal.(76,77,81)foundvarioustypes
ofRfactorswithvariouscombinations ofthe
resistance factors.Furthermore, itwasfoundthat
Rfactorswithmultiple resistance factorscanbe
"synthesized" byaddingthemissingresistance
factorsbyconjugational transfer ofRfactors
(seebelow). Itispossible thatmultiple drug
resistance develops innatureviathismechanism,
between thespontaneous segregants ofmultiple-
drug-resistant strains.RaretypesofRfactors
suchas(Su,Sm,Tc),(Cm,Tc),(Su,Sm),(Sm),
and(Cm)couldnotbefoundasspontaneous
segregants of(Su,Sm,Cm,Tc)byWatanabe
andFukasawa (115,118,124)andWatanabe
andLyang(134),butmightbeexpected tode-
veloponthebasisofthetwohypotheses pro-
posed.
Recombination ofRFactors
Mitsuhashi, Harada, andHashimoto (74),
Nakaya, Nakamura, andMurata (84),Yama-
naka(140),andOchiai,Yamanaka, andKimura
(90)foundthatadditional Rfactorscouldbe
transferred torecipient strainspossessing other
typesofRfactor.However, according toMitsu-
hashi(72),transfer ofRfactorstorecipients
possessing othertypesofRfactorisonlyabout
0.01timesasfrequent aswithsensitive recipients.
Watanabe andFukasawa (115,120,131)in-
dependently carriedoutsimilarexperiments with
anRfactor(Su,Sm,Cm,Tc)anditsspontaneous
segregants, (Su,Sm,Cm)and(Tc).Theyalso
observed thatthefrequencies oftransfer ofR
factorsarereduced toabout10-2bythepresence
ofRfactorsinrecipients, andtheyconcluded
thatthissuppression ofacceptance ofRfactors
isdetermined byRTFandnotbyresistancefactors.Itwasassumed thatthisphenomenon
mayberelated tothe"immunity" or"para-
immunity" oftemperate phages (Lwoff, 66;
Zinder, 148).Watanabe andLyang(135)and
Mitsuhashi, Harada, andHashimoto (75)found
thatsomeofthefour-drug resistant strainsthus
obtained arequiteunstable genetically, segre-
gatingcloneswitheitheroftheoriginal two
typesofRfactorswithhighfrequencies. These
findings indicate twopossible explanations: (i)
thatRfactors arefrequently integrated and
suppress theautonomous replication ofanytypes
ofRfactors;and(ii)thatthenumberofparticles
ofRfactorsinacellissolimitedthatsegregation
takesplaceoccasionally asthecellsdivide.We
havenoevidence atpresentfordiscriminating
between thesetwopossibilities. Neither isun-
reasonable. Integration ofRTFhasbeenshown
(seebelow).Also,thenumber ofFfactorsina
cell(andtheseresemble RTFinmanyrespects)
isassumed tobeonlyaboutthree(Jacoband
Monod,46).Furthermore, FfactorofF+strains
isassumed tobelostspontaneously withrather
highfrequencies (Skaar,Richter, andLeder-
berg,105).CellswhichhavelostFfactorare
apparently easilyreinfected withFfactorby
othercellsintheculturethatcarryFfactor.In
contrast, cellswhichhavelostRfactorsareap-
parently notsoeasilyreinfected, judging from
thelowfrequencies oftransfer ofRfactors.
Someoftheclonesreceiving twotypesofR
factordevelop quitestablefour-drug resistant
clones(Watanabe andLyang,135).Thefour-
drugresistance thusestablished behaves inthe
samewayastheoriginalfour-drug resistance and
seldomsegregates spontaneously. Also,intrans-
ductionwithPlkcthepatterns oflinkageofre-
sistance factorsandRTFresemble theresults
obtained whentheoriginalfour-drug resistant
strainisusedasadonor(Watanabe andLyang,
135).Inotherwords,thefour-drug resistance
factorsaretransduced together inmostinstances.
Watanabe andLyang(135)assumed thatastable
Rfactor(Su,Sm,Cm,Tc)isformedbyrecom-
bination ofthetwoRfactors(Su,Sm,Cm)and
(Tc)(Fig.4).
Circular ModelsofRFactors
Afterformulation ofacircular chromosome
modelforavirulent phagebyStreisinger (per-
sonalcommunication toCampbell, 17),Campbell
(17)proposed asimilarcircularmodelforphage
X,Ffactor,colicinogenic factor,andF'.Hesug-97

BACTERIOL. REV.
' RTFSuSmCm RT
Tc RTF
SuSm Cm Tc
FIG.4.Models ofrecombination ofRfactors.
InmodelA,crossovers anywhere between Cmand
Tcshould giverisetoanRfactor(Su,Sm,Cm,Tc).
-MODEL A- -MODEL B-
FIG.5.Circular models ofanRfactor(Su,
Sm,Cm,Tc).
gestedthatthiscircularmodel maybeapplicable
toepisomes ingeneral. Ifsuchacircularmodel
wereapplied toRfactors,thetwolinearmodels
indicated inFig.2wouldbemodified asinFig.5.
Alloftheavailable data canbeunderstood
without theuseofcircularmodels.Ontheother
hand,circularmodels arenotincompatible with
thoseresults.
Using acircular model,Campbell (17)was
abletoexplainthecuriousdifference inlinkage
ofchromosomal markers ofphageXinthevegeta-
tivestate ascompared withtheprophage state,
reported byCalefandLicciardello (16).The
formation ofXdgintheabsenceofgenetichomol-
ogybetween thegenome ofphageXandgal
genes,andthemechanism ofintegration and
transition fromintegrated toautonomous state
ofphageX,couldbeclearlyunderstood interms
ofthecircular model.Themechanism ofspon-
taneous segregation ofresistance factorsalso
maybeclarifiedbyusingcircularmodels. Itmay
bethatwhenRfactors areintegrated, thecir-cularstructures arestretched tolinearforms,
andwhentheyreverttotheautonomous state
circularstructures areagainformed.Circlesmay
beformedwithonlyaportionoftheoriginal
structures ofRfactors,leavingtheremainder
onthehostchromosome. Thus,thesegregant
Rfactors,containing onlypartoftheresistance
factors,mayformsmallercirclesthantheoriginal
ones.Alternatively, thesegregant Rfactorsmay
pickupapartofthehostgenome asphageX
does.Withthishypothesis, theconceptofgenetic
homology between theRTF-carried resistance
factorsandhostgenomeisnotessential. Watan-
abeandTakano (136)recently succeeded in
isolating strainsofE.coliK-12withstablyin-
tegrated Rfactors. Thevalidity ofcircular
modelsforRfactorscannowbeinvestigated
experimentally, bycomparing thelinkage of
resistance factorsintransduction ofintegrated
andautonomous Rfactors.
Interactions betweenRTFandFFactorin
Escherichia coliK-12
Uniqueinteractions werefoundbetween R
factorsandFfactorinE.coliK-12,especially by
Watanabe andFukasawa (121,122,127,131).
Themostcharacteristic interactions arethesup-
pressions oftheFfactorfunctions byRfactors.
Aswillbediscussed later,theinhibitory effects
ofRfactorsuponFfactorwereascribed toRTF.
Veryrecently, aclassofRfactorswasfound
whichdidnotsuppress thefunctions ofFfactor
whentheyareinmalestrainsofE.coliK-12
(Watanabe andTakano, 136;Watanabe, Fu-
kasawa, andTakano, 133).Theseauthors pro-
posedthatthetypeofRTFwhichsuppresses the
functions ofFfactorbedesignated asRTFtype
1;andthetypewhichdoesnot,RTFtype2.
Among theprevalent strainsofShigella and
E.coliwithRfactors,thosewithRTFtype1
areapparently predominant. Discussion inthis
sectionwillbelimitedtoRTFtype1.RTFtype
2willbediscussed inthesectiononBiological
Functions andMutations ofRTF.
Frequency oftransferofRfactorstoF-andF+
strainsofE.coliK-12.Watanabe andFukasawa
(122,131)foundthatthepresence ofFfactorin
recipients slightlysuppresses theacceptance of
Rfactors(reduced toonlyaboutone-half the
frequency oftransfer ofRfactorstoF-recip-
ients).SinceanRfactor(Su,Sm,Cm,Tc)and98 WATANABE

VOL.27,1963 INFECTIVE HEREDITY OFBACTERIAL DRUGRESISTANCE
itssegregants (Su,Sm,Cm)and(Tc)allex-
hibitedthesameeffect,itwasconcluded that
theacceptance ofRTFissuppressed byFfactor
intherecipients.
Frequency oftransferofFfactor,F',andhost
chromosome torecipients withRfactors.Watanabe
andFukasawa (127,131)foundthatthepres-
enceofRfactorsinrecipients suppresses slightly
theacceptance ofFfactor,F',andhostchro-
mosome. Thetransfer ofFfactortothere-
cipientswithRfactorscannotbeeasilydetected,
because, aswillbeindicated below,thepresence
ofRTFinF+cellsdeprives themoftheirability
totransfer Ffactorandhostchromosome to
othercells.Incontrast, thetransfer ofF'and
hostchromosome canbeaccurately determined
byutilizing donorcellswithpropermarkers,
carriedbyFfactorandhostchromosome, and
recipient strainswithpropermarkers.
EffectsofthepresenceofRfactors onFWand
F'+strains.Watanabe andFukasawa (121,127,
128,131)foundthattheF+andF'+strainswith
anRfactor(Su,Sm,Cm,Tc)oritsspontaneous
segregants (Su,Sm,Cm)and(Tc)areunableto
transfertheirFfactor,F',andhostchromosome.
Itwasconcluded thattheinhibitory effectsof
Rfactorsaremediated byRTF.Thefrequency
oftransfer ofhostchromosome byF+andF'+
strainswithRfactors waslessthan10"per
donorcell.ThisfactsuggeststhatF+strainswith
Rfactorsalsocannottransfer Ffactor.(The
frequency oftransfer ofFfactorcannotbede-
termined accurately fortechnical reasons;i.e.,
thefertilityofeachrecipient colonyarisingfrom
themixedcultureofF+strainswithRfactors
andF-recipients mustbedetermined.) The
suppression oftransfer ofFfactor,F',andhost
chromosome isnotduetotheelimination or
irreversible inactivation ofFfactorbyRTF.
Thisisindicated bythefactthatspontaneous
segregants without Rfactors,obtained bythe
penicillin screening method(seeabove), allbe-
havedasF+andF'+strains.
TheresultsofWatanabe andFukasawa (121,
128,131)onthesuppression ofrecombination by
RfactorsinF+strainswereconfirmed byYo-
shikawa andAkiba(146),usinganRfactor(Su,
Sm,Cm,Tc)ofdifferentorigin.Theyreported,
however, thatsmallnumbers ofrecombinants
weredeveloped inthecrossesbetween F+strains
withanRfactorandF-recipients, suggestingthatthesuppression ofrecombination byRfac-
torspresentinF+donorsisincomplete. Weprefer
toassumethatthefewrecombinants theyfound
mayhavebeenproduced bythesegregants ofF+
cells,whichlosttheRfactorspontaneously.
Alternatively, theymighthavebeenproduced by
Rmating(SuginoandHirota,109),orthestrains
mayhavecontained RfactorswithRTFtype2
(seebelow).
Watanabe andFukasawa (127,131)excluded
thepossibility thatFfactormightbecomeclosely
associated withRfactorsbyconducting anex-
periment inwhichF'+strainswithRfactorswere
treatedwithacridine orange.ThiscausedF'to
beeliminated withhighfrequencies, andR
factorswithmuchlowerfrequencies. Theirfind-
ingthatF'+strainswithRfactorstransfer the
RfactorsbutnotF'isalsoevidence thatthetwo
factorsarenotcloselyassociated.
Effectsofthepresence ofRfactorsonHfr
strainsofE.coliK-12.Nakaya, Nakamura, and
Murata(84)foundthatanHfrstrainofE.coli
K-12withanRfactor(Su,Sm,Cm,Tc)canre-
combine withan 'recipient onlywithreduced
frequencies (about102ofthefrequencies ofre-
combination oftheoriginal Hfrstrainwithno
Rfactor). Theirfinding wasextended by
Watanabe andFukasawa (121,131)withthree
Hfrstrainsshowing different ordersofchromo-
somaltransfer (JacobandWollman, 48,49).The
frequencies ofrecombination byalltheHfr
strainswithanRfactor(Su,Sm,Cm,Tc)were
reduced toabout102.Similardegreesofsup-
pression ofrecombination werebrought about
alsobysegregant typesofRfactors(Su,Sm,Cm)
and(Tc)inHfrstrains.Theseresultsagainsug-
gestthatthesuppression ofrecombination of
HfrstrainsisduetoRTF.Theintegrated F
factorofHfrstrainswasnoteliminated orirre-
versibly inactivated bytheinfection ofthese
strainswithRfactors, asevidenced bythefact
thatthedrug-sensitive segregants withnoR
factorsallbehaved liketheoriginalHfrstrains.
Itwasfurthershownthattheoriginsandse-
quences ofchromosomal transferarenotaltered
byRfactorspresentinthesestrains.
Watanabe andFukasawa (128,131)found
thatHfrstrainswithRfactorscantransferthe
Rfactorswithfrequencies aboutequaltothose
oftheF-,F+,andF'+strainswithRfactors.By
useofareplicaplatingtechnique, therecipient99

BACTERIOL. REV.
cellswhichreceived RfactorsfromHfrstrains
withRfactors werefoundtobelargely nonre-
combinants. Ontheotherhand, someofthere-
combinants fromtheHfrstrainswithRfactors
hadRfactors. Although thesefindings willbe
discussed againinthesection Evidence for
Chromosomal Attachment ofRTF,itshouldbe
emphasized herethatthehostchromosome and
Rfactors canbetransferred together insome
instances.
Suppression byRfactorsofthefphage-sensitive
mechanism inE.coliK-12conferred byFfactor.
Agroupofphages(fphages), whichspecifically
attackthemalestrainsofE.coliK-12, wereiso-
latedbyLoeb(63)andLoebandZinder(64).
Someofthese wereshowntocontain ribonucleic
acidandnodeoxyribonucleic acid.Sincethese
phages werenotadsorbed by F-strains, the
authors assumed thatFfactorinduces thesyn-
thesisofareceptor substance forthephages.
Watanabe, Fukasawa, andTakano (132,133)
observed thatinfection ofF+,F'+,andHfr
strainswithvarious typesofRfactorsinduces
resistance tofphages. Inaddition, theyreported
thatstrainswithbothFfactorandRfactorsdo
notadsorbfphages. Because thesensitivity of
thesestrainstothephagesoftheTseriesisnot
changed byRfactors, thedevelopment ofresist-
ancetofphagesisconsidered tobespecific.
Watanabe andFukasawa (127,131)haveshown
thattheFfactorisnoteliminated orirreversibly
inactivated byRTF.Therefore, theinability of
theabovestrains withRfactors toadsorb f
phages cannotbeattributed totheabsence of
Ffactor.ItseemsthatRTFsomehow inhibits
thesynthesis ofthefphage-specific receptor
substance. Itisalsopossible thatsomeadditional
grouping isaddedtotheendofthereceptor sub-
stance. Asimilarsituation wasfoundbyUetake
andHagiwara (111)insomatic antigens of
salmonellae. PhageE34induces thesynthesis of
antigen 034onlyinthecellswithantigen 15.Itis
believed thatphageE34modifies antigen 15by
adding glucose tothefreeendofantigen 15
(Robbins andUchida, 99).
F-strainswithRfactors donotadsorb f
phages,indicating thatRTFdoesnotinducethe
synthesis ofthefphage-specific receptor sub-
stance.
Sneath andLederberg (106)foundthattreat-
mentofFtandHfrcellsofE.coliK-12with
sodium periodate induces temporary infertility,andpostulated thepresence ofa"mating sub-
stance"ofpolysaccharide natureonthecellsur-
face,thesynthesis ofwhichisinduced byF
factor.Theyfurtherpostulated thatthissub-
stanceisessential formating. Recently, Iijima
(40)foundthatcellswithRfactorsalsolose
theirabilitytotransfer Rfactorsaftertreatment
withsodiumperiodate, suggesting thepresence
ofa"mating substance" whichissynthesized in
thepresence ofRfactors.Thematingsubstance
synthesized bycellswithRfactorsmustbedif-
ferentfromthatsynthesized bycellswithF
factor,becauseF+andF'+strainswithRfactors
cantransfer Rfactors,butnotFfactor,F',and
hostchromosome. Ontheotherhand,Qrskov
andQrskov (91)discovered anewsomatic
antigen(f+)inthemalestrainsofE.coliK-12.
Although itisnotknownatpresentwhether the
receptor substance forfphages,thematingsub-
stance,andantigenf+areidentical, wecansafely
assumethattheyareatleastcloselyrelatedand
possibly identical, forthesynthesis ofthesesub-
stancesisinducedbyasingleagent,Ffactor.
Onthebasisoftheabovefindings, Watanabe
andFukasawa (131)andWatanabe, Fukasawa,
andTakano(132,133)suggested thatthegenetic
determinant ofRTF,controlling itstransfer, is
epistatic tothecomparable geneticdeterminant
ofFfactor.According tothishypothesis, inthe
presence ofbothFfactorandRTF,thepoly-
saccharide specificforFfactorisnotsynthesized
butapolysaccharide specificforRTFissynthe-
sized.Wearenowattempting todetermine
whether RTFinducesthesynthesis ofsomenew
somaticantigen. Therelationofepistasis between
RTFandFfactorsuggests aclosesimilarity
between thesetwofactors.
Interactions betweenRTFType1andColicinogenic
FactorsinEscherichia coliK-12
Watanabe andFukasawa (121,131)foundthat
thefrequency oftransfer ofcolicinogenic factor
Elwaslessthan10-perdonorcellwhenRTF
waspresentinthesamecells,whereas thefre-
quencyoftransferbyFWstrainswithout RTF
wasabout40%.Thiscolicinogenic factorre-
quiresthepresence ofFfactorinthesamecells
foritstransfer (Fredericq, 25).Iijima(40)re-
portedthatF-(col.E,+)cellswithRfactorscan
transfer thecolicinogenic factorEl,butonly
withlowfrequencies (about10Operdonorcell).
Watanabe andFukasawa (131)alsostudied100 WATANABE

VOL.27,1963INFECTIVE HEREDITY OFBACTERIAL DRUGRESISTANCE
theeffectofRTFonthetransferofcolicinogenic
factorI,whichdoesnotrequirethepresence of
Ffactorforitstransfer (95,96).Itwasfound
thatRTFdoesnotsuppress thetransfer ofthis
colicinogenic factor.Hence,thesuppression of
transferofcolicinogenic factorElbyRTFmay
wellbemediated bytheepistasis ofRTFto
Ffactor.
Evidence forChromosomal Attachment ofRTF
Although Campbell (17)classified asepisomes
severalfactorsforwhichchromosomal attach-
menthasnotyetbeenproved,itisessential to
demonstrate anintegrated statefortheautono-
mouslyreplicating elements topermittheirin-
clusioninthecategory ofepisomes asoriginally
defined (Jacob andWollman, 50;Jacob,
Schaeffer, andWollman, 47).Alfoldietal.(10)
havealready demonstrated theintegration of
colicinogenic factorEl,byemploying several
different typesofHfrstrainscarryingthisfactor
andbystudying thekinetics ofitstransfer. In
anattempt toestablish thechromosomal attach-
mentofRfactors,Watanabe andFukasawa (130)
carriedoutkineticstudiesonthetransferofhost
chromosomal markers andRfactorsbyHfr
strainswithRfactors.Withselection fortransfer
ofRfactors,theresultsobtained werenotclear-
cut,probably becausetheautonomously replicat-
ingRfactorsaretransferred muchmorefre-
quently thantheintegrated Rfactors. With
selection fortransfer ofhostchromosomal
markers, itwasfoundthatonlytherecombinants
whichreceived thesegment ofhostchromosome
between theB1andmannitol locihaveRfactors.
Itwasconcluded, therefore, thatRfactorsare
integrated ataspecificsitebetween B1and
mannitol loci,andaretransferred torecipient
cellsalongwiththehostchromosome. Thefact
thatsomeoftherecombinants whichreceived
thischromosomal segment donothaveRfactors
wasinterpreted asfollows.Eithertheserecombin-
antswereproduced bysensitive segregants with
noRfactors,orRfactorscouldalsobeintegrated
atsomeothersite(s)beyondthemannitol locus.
Thesetwopossibilities weresubsequently con-
firmedexperimentally. Watanabe andTakano
(136)succeeded inisolating cloneswithstably
integrated RfactorsfromHfrstrainswithR
factorsbyemploying areplicaplatingtechnique
forselecting cloneswithhighfrequencies ofre-
combination. Asindicated above,HfrstrainswithRfactorsrecombine onlywithreduced fre-
quencies. Theclonesthusisolated wereoftwo
types,withandwithout Rfactors.Integrated
Rfactorswerefoundinthefirsttype.Thesiteof
integration hasnotyetbeendetermined accu-
ratelyintheseclones,butitisbelieved tobenear
thesiteofchromosomal attachment ofFfactor
inoneclonederived fromHfrstrainW-2252.
Thissuggests thatintegration ofRfactorsneed
notbelimitedtothesitebetween B1andmanni-
tolloci.Theintegrated stateofRfactorsap-
pearedtobequiteunstable insomeoftheclones
butfairlystableinothers,judging fromthefre-
quencies ofsegregation ofcloneswhichrecombine
onlywithlowfrequencies.
Segregant typesofRfactors,(Su,Sm,Cm)
and(Tc),aswellas(Su,Sm,Cm,Tc),were
foundtobefrequently integrated between B.
andmannitol loci,suggesting that,inE.coli
K-12,integration iscausedbyRTFandnotby
resistance factors. Studies ontransduction of
RfactorsinE.coliK-12withphagePlkcalso
support thefollowing assumption. Resistant
transductants withnoRTFseemtodevelopthe
stateofabortive transduction inK-12,becauseof
thelowgenetichomology between thehost
genomeandtheRTF-carried resistance factors.
HighFrequency Resistance Transfer System
(HFRT)
OzekiandStocker(95)andOzeki,Howarth,
andClowes(96)foundthatcolicinogenic factor
Icanbetransferred withveryhighfrequency by
thecellsofS.typhimurium thathavejustre-
ceivedit,whereas cellscarrying thiscolicino-
genicfactorarenormally onlypoordonors.They
succeeded inobtaining apopulation whichcon-
tainedahighratioofcompetent donorcells,by
growing asmallnumber ofcolicinogenic cells
withalargenumber ofsensitive cellsovernight
withoutshaking, dilutingthemixedculture1:10
infreshbroth,andthenincubating for2hrwith-
outshaking. Asimilarprocedure wasemployed
byWatanabe andFukasawa (129)withRfac-
tors,andthefinalpopulations thusobtained
werefoundtotransfer Rfactorswithextremely
highfrequencies. Thissystemwasreferred toas
HFRT,becauseofitsanalogy totheHFCTsys-
temofOzekiandStocker(95).Thefrequency of
transferofRfactorswasashighas1.5to7.6per
resistant cell,whenthemixture ofrecipient and
HFRTpopulation wasincubated at37Cfor101

BACTERIOL. REV.
1hr.Thesesurprisingly highvaluesindicatethat
theexconjugants oftherecipient, whichreceived
Rfactorsduringthisincubation period,probably
serveascompetent donorsforfurthertransferof
Rfactors.Itisalsopossiblethateachdonorcell
intheHFRTpopulation conjugates withand
transfers Rfactorstomorethanonerecipient
cell.Inaddition, recipient cellsthatreceived
Rfactorsrequireincubation at37Cforabout
8minforthephenotypic expression ofthese
factors(Watanabe andFukasawa, 126).Thus,
thephenotypically drug-resistant cellsinthe
population ofHFRTdonotnecessarily reflect
theactualnumbers ofcellswithRfactors.Some
ofthecellswithRfactorsmaynotbedetected
asresistant cells,owingtoincomplete phenotypic
expression ofthedrugresistance.
Ozeki(personal communication) proposed the
following explanation forthemechanism of
HFCT.Colicinogenic factorIiseasilyintegrated,
andtheintegrated factorisnotnormally trans-
ferred.Itenterstheautonomous staterarely,
andisseldomtransferred totherecipient cellsby
conjugation. Thecompetent donorcells,thusde-
veloped, losetheircompetence, ifthecolicino-
genicfactorisintegrated. Thisexplanation is
predicated ontheassumption thatthecolicino-
genicfactorIisinanautonomous stateinmany
cellsoftheHFCT.Ozeki'sinterpretation may
notbeabletoaccount forthemechanism of
HFRT,becauseclonescontaining Rfactorsbut
unabletotransferthem,orabletotransferthem
onlywithlowfrequencies becauseoftheirinte-
gration,areratheruncommon. Itispossiblethat
theintegrated stateofRTFisusuallyunstable.
Iprefertoassumethattheautonomous state
ofRfactorsinthecompetent donorcellsof
HFRTisratherunusualinthatthesefactorsare
transferred moreeasilybythesecellsthanby
cellswithautonomously replicating Rfactors
("stable autonomous state"). The"unstable
autonomous state"ofRfactorsinthecompetent
donorcellsofHFRTmayresultfromtheun-
usuallyvigorous replication ofRfactors.
MatingofE.coliK-12wasnotdetected with
HFRT,although OzekiandStocker (95)and
Ozeki,Howarth, andClowes(96)foundmating
ofS.typhimurium withHFCT.
Biological Functions andMutations ofRTF
Watanabe andFukasawa (115,116,123,129)
wereunabletodetectthespecificproduction ofcolicinsbyRfactors,asindicated above.They
concluded, therefore, thatRTFmaynotbea
colicinogenic factor.
Inanefforttodetermine whether Rfactors
arerelatedtotemperate phages,Ochiaietal.(89),
Akibaetal.(5,6),Mitsuhashi (71),andMitsu-
hashi,Harada, andHashimoto (74)studiedthe
effectsofcell-free filtrates ofstrainswithR
factorsbutcouldnotdemonstrate theabilityof
thefiltrates totransfer Rfactors. Theusual
transfer ofRTF,therefore, doesnotinvolve
transduction. However, asmentioned previously,
Rfactorshavebeentransduced inseveralsys-
temsofbacteria andphages. Ibelieve, how-
ever,thattransduction ofRfactorsisratherex-
ceptional undernatural conditions, andthat,
invivo,conjugation isprobably themostcommon
modeoftransfer.
Inthetransduction experiments ofWatanabe
andFukasawa (117,125)withS.typhimurium
LT-2andphageP-22,RTFwasshowntoassume
thecausative roleinconjugation andtransfer of
Rfactors. ThefactthatFfactorandother
transfer factorsarenotrequired forthetransfer
ofRfactors(80,116)alsoindicates thatRTF
itselfeffectstheconjugation andthetransfer of
Rfactors.Theformation ofaconjugation bridge
(orcanal),suchasthatrevealed inthemating
ofE.coliK-12(11,58),hasnotyetbeendem-
onstrated duringtransfer ofRfactors. How-
ever,thereisevidence tosupport theassump-
tionthataconjugation bridgedevelops. First,
thetransfer ofRfactors canbeinterrupted
bysubjecting themixture ofdonorandre-
cipienttotheactionofablender (Watanabe
andFukasawa, 119,123),aftertheprocedure
employed todemonstrate conjugation bridgesin
thematingofE.coliK-12(Wollman, Jacob,
andHayes,139).Furthermore, transfer of
Rfactorsisalsointerrupted byaddition ofphage,
towhichthedonorcellsaresensitive andthe
recipient cellsareresistant (Watanabe and
Fukasawa, 126),anexperiment similartothat
performed byHayes(35),whointerrupted the
conjugation causedbyFfactorwithphageT6.
Finally, theobservation ofIijima(40)that
colicinogenic factorElcanbetransferred byF
(col.E1+)cellswithRfactors,butnotbyF
(col.El+)cellswithout Rfactors, alsosuggests
theformation, inthepresence ofRfactors,ofa
conjugation bridgewhichcannotbeformedby
colicinogenic factorElalone.102 WATANABE

VOL.27,1963INFECTIVE HEREDITY OFBACTERIAL DRUGRESISTANCE
ThefindingofWatanabe andFukasawa (121,
127,131)andWatanabe, Fukasawa, andTa-
kano(132,133)thatthefunctions ofFfactor
aresuppressed byRTFwasascribedbythemto
theepistasis ofRTFtoFfactorinestablishing
theirtransfer systems. Forelucidating the
mechanism ofthisepistasis, thereportof
Watanabe, Fukasawa, andTakano (132,133)
thatthereceptor forfphagesissynthesized in
thepresence ofFfactor,butnotinthepresence
ofbothFfactorandRfactors, seemstobeof
fundamental importance. Aswehavealready
pointedout,itisnotknownwhetherthereceptor
substance forfphages,f+antigen(Orskov and
Qrskov,91),andmatingsubstance (Sneathand
Lederberg, 106)areidentical. Ifmatingsub-
stanceisidentical tothereceptor substance for
fphages,failureofthecellstosynthesize this
receptor substance shouldresultinlossofthe
abilitytotransfer Ffactorandhostchromosome.
Although Watanabe andFukasawa (115,116)
couldnotobtainevidence thatRfactorsmake
F-strainsofE.coliK-12fertile,Suginoand
Hirota(109)subsequently reported suchan
effect.Theyreferred tothematingcausedby
Rfactors asRmating. Thefrequencies of
Rmatingdiffered forvarious Rfactors, and
rangedfrom10-6to10-8perdonorcell.Rmating
occurred withhighfrequencies (10-4to10-6per
donorcell)inanF-derivative ofW-3876ofan
Hfrstrain(Richter, 98).Theorderofchromo-
somaltransferbyW-3876withRfactorswas
foundtobeidentical tothatwithW-3876and
Ffactor.Thehighfrequencies ofRmatinginan
F-derivative ofanHfrstrainmayindicate the
difference between theoriginal F-strainsand
theF-strainsderivedfromHfrstrains.Jacob
andAdelberg (45)andAdelberg andBurns(1)
discovered whattheycalled"souvenir" (or
memory) asaresultofgeneticexchange between
Ffactorandhostgenome. RTFmighthavesome
genetichomology toFfactor,promoting the
integration ofRTFtothesiteofmemory which
Ffactorleftonthehostchromosome. Watanabe
andFukasawa (131)laterfoundthatRfactors,
whichdonotsuppress thefunctions ofFfactor,
givefertility totheF-strainsofK-12.IfR
matingwereageneralphenomenon forRfactors,
RTFshouldbeconsidered tobeakindofF
factor. Thediscrepancy intheresults of
Watanabe andFukasawa (115,116)andSugino
andHirota(109)maybeattributable tovaria-tionsintheconditions ofmatingemployed.
Watanabe andFukasawa (115,116,131)carried
outtheirexperiments inbroth;Sugino and
Hirota(109)onselective plates.
Egawa (personal communication) recently
foundamutantofRfactor(Su,Sm,Cm,Tc)
whichdoesnotsuppress therecombination of
Hfrstrains.Watanabe, Fukasawa, andTakano
(133)andWatanabe andTakano(136)alsofound
severalsuchRfactors. Inaddition, theyob-
servedthattheseRfactorsdonotsuppress the
recombination ofF+andF'+strainsandthe
transferofFfactorandF'byF+andF'+strains.
Another important distinguishing characteristic
oftheseRfactorsisthattheydonotinhibitthe
synthesis byFfactorofthereceptor substance
forfphages.Hence,F+,F'+,andHfrstrains
withtheseRfactorsaresusceptible tofphages.
ThetypeofRTFassociated withtheseRfactors
isreferred toasRTFtype2;andthemore
common RTF,asRTFlype1.Egawa(personal
communication) foundIFiattypes1and2can
mutatetoeachother.Acorrelation between the
inhibition ofthesynthesis ofthereceptor sub-
stanceforfphagesandtheinhibition ofmating
inmalebacteria withRfactorsisnowestab-
lished.Thesestudiesprovide furtherevidence
thatasubstance essential formatingisidentical
tothereceptor substance forfphages.
Although HfrstrainswithRTFtype1cannot
synthesize thereceptor substance forfphages,
theyarestillcapableoftransferring theirhost
chromosomes, butwithreducedfrequencies. This
factindicates thatthe"transfer system" ofhost
chromosome isswitched overtothatofRfactors
inthesestrains.Itisconsidered thatRfactorsin
theautonomous stateinhibitthetransferofhost
chromosome; therefore, chromosomal transferby
HfrstrainswithRfactorsmaybeeffectedby
cellswithintegrated Rfactors.
Attempts toisolatepossible mutants ofRTF
havebeencontinued buthavenotbeenverysuc-
cessful.Thisisprobably duetothetechnical
difficulty ofselecting cellswithmutantRTF.
Thisdifficulty iseasilyunderstood, iftheprob-
lemsencountered withFfactorandothertrans-
ferfactorswherenomutants havebeenisolated
todatearerecalled. Forexample, theF2variant
oftheFfactorisconsidered toariseasaresultof
geneticexchange andnotofmutation (Adelberg
andBurns,1).Incontrast tothetransferfactors,
mutant typesoftemperate phages arerather103

BACTERIOL. REV.
easilyisolated. Perhaps thenatureofthetransfer
processmakestheisolation ofmutant typesof
transfer factorsdifficult. Thus,transfer factors
cannotbetransferred intheirintegrated state
unlesshostchromosome istransferred. Onthe
otherhand,theycanbetransferred whenthey
areintheautonomous state.Inthisstate,the
numbers ofparticles oftransfer factorsarecon-
sideredtobeusuallymultiple and,evenifone
ofthemmutates, itcannotbedetected easily
unlessitissegregated fromthewild-type transfer
factors.
AdefinitemutantofRTF,differing fromRTF
type2,wasdiscovered bychancebyWatanabe
andLyang(134,135),whentheywerestudying
thespontaneous segregation ofresistance factors
inS.typhimurium LT-2carrying anRfactor
(Su,Sm,Cm,Tc).Twounusual cloneswere
found;onehadSu,Sm,Cmresistance, andthe
otherwasresistant tothefourdrugs.However,
neitherclonewasabletotransfer itsresistance
factorsbymixedcultivation withsensitive re-
cipients. Thesestrainswerefoundtosegregate
theirresistance factorsspontaneously withfre-
quencies ashighasthestrainswithnormal
Rfactors(Su,Sm,Cm,Tc)and(Su,Sm,Cm),
indicating thattheirRfactorsmightbeinan
autonomous state.Furthermore, thetransfer of
asegregant typeofRfactor(Tc)tothestrain
with<<Su,Sm,Cm>>(mutant Rfactorwhich
cannotbetransferred byconjugation) gavefour-
drugresistant clones,someofwhichwerefound
tobegenetically stableandabletotransfer all
fourresistance factorsbyconjugation asaunit.
Otherswerequiteunstable andtransferred only
(Tc)byconjugation. Thedevelopment ofthe
stablefour-drug resistant strainswasassumed to
beduetorecombination ofthetwoRfactors.
TheRTFoftheRfactors,whichcannotbe
transferred byconjugation, isassumed tobea
mutant whichiscomparable tothedefective
mutants oftemperate phages(Jacob,44).Itis
notknownyetwhichsteporstepsofthetransfer
ofRfactorsareblockedbythismutation.
Ifvariousstrainsofshigellae andE.coliwith
Rfactors,whichhavebeenisolatedfromhuman
specimens arestudied, someofthemarefound
totransfertheirRfactorsbyconjugation very
easily,otherswithdifficulty, andstillothersnot
atall.Thereasonsforthedifficulty orfailureto
transfer resistance factors aremultiple. Some
donorsandrecipients producecolicinsandphageswhichkilltheirpartnercells.Somedonorstrains
areonlyabletotransfer resistance factorswith
lowfrequencies ornotatall,because theycontain
defective mutants ofRTF,ormanifest stable
integration ofRfactors,orpossessonlychromo-
somalresistance factors andnoRfactors.
Watanabe andFukasawa (unpublished data)
foundthattheRfactorsofthedonorstrainswith
lowfrequencies oftransfer sometimes acquire
hightransferability whentheyaretransferred to
otherstrains.Thelowtransferability ofRfactors
oftheoriginalstrainsmust,therefore, beascribed
tosomespecificcharacter ofthehostcells.Al-
mostnothingisknownabouttheroleofthehost
cellsindetermining thefrequencies oftransferof
Rfactorsbyconjugation.
Watanabe andFukasawa (116,123)foundthat
Salmonella strainsarepoorrecipients forR
factors.Baron,Carey,andSpilman (13,14)also
reported thatsalmonellae arenotgoodrecipients
inthematingofHfrstrainsofE.coliK-12.They
wereabletoisolategoodrecipient clonesfrom
theoriginalSalmonella strainsandassumed that
thecellshadmutated fromastateF0tosome
otherstateinwhichtheyfunctioned asgood
recipients. Watanabe andFukasawa (123)origi-
nallypostulated thatthecellsofsalmonellae
mightcontain afactorcloselyrelatedtoRTF
(possibly RTFitself),whichconfersimmunity
against infection byRfactors.Theyisolated
spontaneous segregants withnoresistance factors
(andpossiblywithnoRTF)fromS.typhimuriumr
LT-2withanRfactor(Su,Sm,Cm,Tc),usinga
penicillin screening method. Thesesegregants,
however, werefoundtoreceiveRfactorswith
frequencies aslowastheoriginalsensitive strain
LT-2.Thereasonforlowfrequencies oftransfer
ofRfactors tosalmonellae isnotknown at
present.
Comparison ofRTFwithOtherEpisomes
Thecomparison ofRTFwithotherepisomic
elements issummarized inTable2.Theterm
episomeappliesnotonlytothegeneticelement
responsible forepisomal behavior, butalsotoall
associated material. However, whenvariousepi-
somesarecompared, itismorerationalandcon-
venienttoexcludethenonessential anddefinitely
accessory material carriedbytheoriginaltypes
ofepisomes, asCampbell (17)suggested. Other-
wise,eachepisome mustbeaccompanied by
numerous relatives. UsingFfactorasanexam-104 WATANABE

VOL.27,1963 INFECTIVE HEREDITY OFBACTERIAL DRUG RESISTANCE
TABLE2.Comparison ofresistance transfer factorwithotherrepresentative episomes
Auton- Transfer Pick-up Killinof Productionomous Inte- Cellconju- ofhostofhostcellsbyColicin ofextra-Kindofepisome repli- gration gation chromo- crm autono pro- cellular cation somalm~~~~~~~orousduction infectiouscation ~~~~~~some gns replication particles
Temperate phages + + – – + + – +
Temperate phager + ++(?)+(?)-(?) + – +
Ffactors + + + + +
Colicinogenic factors + +-or+–(?) – +
Colicinogenic factor1 ++ + +-(?) – +
Resistance transfer factor + + ++(?)+ – – –
ple,wewouldhavetoplaceF'-lac,F'-galand
manyotherkindsofF'wouldbeplacedonthe
samelevelasFfactor,colicinogenic factor,and
temperate phage.RTFwithout anyresistance
factorshasnotbeenobserved asyet,buta
varietyofevidence suggests thattheresistance
factorsarenotessential fortheepisomality of
theRfactors.
Thisproblem mayberelatedtothatofphage
r(tau).Thisisauniquetemperate phage,which
wasisolatedbyHakura andHirota(27)and
claimed togivefertility toF-strainsofE.coli
K-12inwhichitbecomes aprophage. Itinfects
alsoF+cellsbutcannoteitherlyseorlysogenize
them.
OriginofResistance FactorsCarriedbyRTF
TheoriginofRTF,likethatofotherepisomes,
isobscure. Watanabe andFukasawa (117,125)
presented ahypothesis thatthemultiple drug
resistance factorscarriedonRTFwereacquired
byRTFinasinglestepfromthehostchromo-
someofsomeunknown bacteria. Theypostu-
latedthatRfactorsarecomparable toF'(Jacob
andAdelberg, 45;HirotaandSneath,38),and
thatthetransferofRfactorsbyconjugation isa
phenomenon comparable toF-duction (sex-
duction;Jacob,Schaeffer, andWoilman, 47).It
isdifficult tounderstand whyonlytheresistance
factorsandnootherdetectable markers should
belocatedincloselinkage onthehostchromo-
some.Itisquitepossiblethatothermarkers are
presentamongtheresistance factors,buthave
notbeendetected. Furtherstudies areawaited.
Watanabe andFukasawa (125)alsosuggested
thatRTFmaypickupotherchromosomal genes
inaddition toresistance factors.However, only
theRTFwithresistance factorswouldhavea
selectiveadvantage inareassuchasJapanwhere
chemotherapy hasbeenadministered extensively.Iftheresultsofepidemiological studies of
drug-resistant shigellae inJapan(Table1)are
examined, itisevidentthatthefour-drug resist-
antstrainsappeared veryearly.IfRTFhad
acquired resistance factorsseparately, strains
resistant toone,two,orthreedrugsshouldhave
beenisolated morefrequently andpriortothe
development ofthefour-drug resistant strains.
Japanese workers inthefieldofpublichealth
haveextensively studied, onanation-wide scale,
thedrugresistance oftheisolated Shigella
strains;ifanysingly,dually,ortriplyresistant
strainsweredeveloped, theyshouldhavebeen
detected.
Asmentioned earlier,multiple drugresistant
Rfactorcanbe"synthesized" experimentally by
combining variousRfactors,anditislikelythat
thesamemechanism operates inthehuman
intestinal tract.Thisshouldnotbetaken,how-
ever,asevidence forthesuccessive acquisition of
resistance factorsinthedevelopment ofthefirst
multiple drugresistant strainsinthenatural
environment.
Watanabe andFukasawa (116,117,123,131)
demonstrated thatthelevelsofdrugresistance
expressed byRTF-carried resistance factorscan
differconsiderably fromhosttohost.Forexam-
ple,Shigella strains,whichhaveacquired the
Smresistance factorcarriedbyRTF,withstand
Smconcentrations ofmorethan1,000Agperml,
whereas theSmresistance ofE.coliwiththis
factorisaslowas10ugperml.Thelevelof
Tcresistance ofShigella andE.colistrains
havingRTF-carried Tcresistance factoris
higherthan100,ugpermlbutincorresponding
Salmonella strainsitisaslowas10,ugperml.
Eveninthevarioussubstrains ofE.coliK-12,
thephenotypic expression oftheCmresistance
factormaydifferconsiderably. Ifthesefactsare105

BACTERIOL. REV.
takenintoconsideration, failuretoisolatefrom
thehumanintestinal tractstrainswithhighre-
sistance tothefourdrugsunrelated toRTF
doesnotnecessarily excludethepossibility that
theresistance factorscamefrombacterial chro-
mosomal genes.Inthehumanintestinal tract
Pseudomonas specieswithratherhighresistance
toSm,Cm,andTc,andratherlowresistance to
Su,arefairlyprevalent. Attempts totransfer
theseresistance factorstoE.colibymixedculti-
vationhavebeenunsuccessful sofar(unpub-
lisheddata).However, itshouldnotbetoo
surprising iffutureexperiments revealthatthe
resistance factorsofRfactorscamefromPseudo-
monasorsomeothergenuswhichdoesnotbelong
toKauffmann's Enterobacteriaceae. Forexample,
Rfactorscanbetransferred toS.marcescens
(Falkow etal.,22)andV.comma(Baronand
Falkow, 15).Eventhoughtheoriginofthe
resistance factorsofRfactorsisstillobscure, the
hypothesis thattheyderived fromsomebac-
terialchromosome helpsexplainmanyexperi-
mentalfindings.
Ontheotherhand,itisalsopossiblethatthe
Rfactorsnowprevalent inJapanareallde-
scendants oftheRfactor(Su,Sm.Cm,Tc),
whichwasfoundinaShigellastrainisolatedby
Kitamoto etal.(52)fromapatientwhoreturned
fromHongKongin1955.Thisassumption may
seemfruitless, becauseitagainleadstoaques-
tion.HowwasthisoriginalRfactorproduced?
However, inasmuch asitisnotknownwhether
the"genepick-up" (Campbell, 17)byRTFi;
nowtakingplaceinJapan,thispossibility cannoiG
beneglected.
Molecular BasisofRFactors
SinceRfactorsreplicate autonomously, itis
mostlikelythattheyarecomposed ofnucleic
acids.Asmentioned above,theresistance factors
carriedbyRTF,inallprobability, originated
fromabacterial chromosome, andtherefore may
becomposed ofDNA.Theepisome RTFisalso
possiblyDNA,judgingfromitsaffinityforhost
chromosome.
Theeffectsofultraviolet lightandmitomycin
ConRfactorswerestudiedbyWatanabe and
Fukasawa (118,124)andWatanabe andTakano
(136).Itwasfoundthatsomeofthedonorcells
inactivated byultraviolet lightormitomycin
Cwerestillabletotransfer Rfactorsbyconju-
gation,thusincreasing thefrequency oftransferpersurviving donorcell.WhenHfrstrainswith
stablyintegrated Rfactorsweretreated with
ultraviolet lightormitomycin C,notonlythe
frequencies oftransferofRfactorspersurviving
donorcellbutalsotheabsolute frequencies of
transfer weresignificantly increased (Watanabe
andTakano, 136).Thissituation isprobably
comparable totheinduction ofprophages by
ultraviolet light(Lwoff,66)andbymitomycin
C(Otsujietal.,93).
Nakaya, Nakamura, andMurata (85)ir-
radiated phagePlkcgrownonE.coliK-12
containing anRfactor(Su,Sm,Cm,Tc)with
ultraviolet lightandstudiedthedose-response
curvesofplaque-forming abilityandtransduci-
bilityofhostchromosomal markersandRfactors.
Theirresultsareessentially thesameasthose
ofArber(12),whocarriedoutsimilarexperi-
mentswithFfactor.Theplaque-forming ability
andtransducibility ofRfactorsweremuchmore
sensitive toultraviolet lightthanthetrans-
ducibility ofthehostchromosomal markers.
(Asmentioned earlierPikcusuallytransduces
theentireRfactor.)Theyselectedtheresistant
transductants onmediacontaining individual
drugsoramixture ofthedrugs.Thedose-
response curvesoftransducibility ofRfactors
wereessentially thesamewiththesemedia.These
resultsindicate theinactivation ofRTFby
ultraviolet lightratherthaninactivation ofthe
individual resistance factors.
Watanabe andTakano (unpublished data)
studiedtheeffectofdecayofincorporated P32
onRfactorsinE.coliK-12withthetechnique
usedbyLavall6andJacob(56)inF'-lac,F'-gal,
andcolicinogenic factorE1.Theirresultsin-
dicatethatRTFiscomposed ofDNAwitha
molecular weightclosetothoseofF'andphage
X(Stent,Fuerst,andJacob,107).Theyfound
alsothatinactivation ofRTFresultsininactiva-
tionoftheentireRfactor,eveniftheresistance
factorsareintact.
S.marcescens, whichacceptsRfactors(Falkow
etal.,22),hasDNAofabasecomposition
different fromthatofentericbacteria. Fractiona-
tionofRfactorsbythedensitygradient cen-
trifugation technique (Meselson andStahl,70),
whichwassuccessful infractionating F'-lac
(Marmur etal.,67),willundoubtedly provide
important information onthechemical nature,
andperhapstheorigin,ofRfactors.106 WATANABE

VOL.27,1963INFECTIVE HEREDITY OFBACTERIAL DRUGRESISTANCE
BIOCHEMICAL MECHANISMS OFMULTIPLE
DRUGRESISTANCE
Because multiple drugresistance factorscan
betransferred asaunitbyconjugation, some
workers assumed thatmultiple drugresistance
maybecontrolled byasinglegeneticdetermi-
nant.Hence,thebiochemical mechanism of
multiple drugresistance wasascribed tonon-
specificreduction ofcellpermeability (Mizuno,
83).However, ithasnowbeenestablished that
theRfactorsinvolvemultiple geneticdetermi-
nants.Asimpleunitary explanation forthe
biochemical mechanism ofmultiple drugre-
sistanceis,therefore, unlikely.
Miyamura (82)reported thatShigella and
E.colistrainswithCmresistance factorenzy-
matically decompose Cm.Elaboration ofthe
enzyme, chloramphenicolase, bythecellswas
therefore assumed tobeamechanism ofCm
resistance. However, Yokota andAkiba(144)
laterfoundthattheactivity ofthisenzymeis
aboutequalinCm-sensitive andCm-resistant
strains. TheCm-resistant strainsdecompose
moreCmonlybecausetheycangrowinthe
presence ofCm.Thus,theenzyme whichde-
composes Cmisnotessential forthemechanism
ofCmresistance.
Yokota andAkiba(143)andAkibaand
Yokota(8)studied theaccumulation offolic
acidprecursors bytwostrainsofE.coliresistant
toSualone(thisSuresistance factorisnot
carriedbyRTF)andalsobyastrainofE.coli
withanRfactor(Su,Sm,Cm,Tc).Oneofthe
strainswithSuresistance alonewasisolated
fromanaturalsource,andtheotherwasobtained
asamultistep mutantinvitrofromasensitive
parent.Priortothesestudies,SevagandYokota
(103,104)reported thatSu-resistant mutants of
E.coliisolated invitrosynthesize increased
amounts oftheintermediates betweenpteridines
andtetrahydrofolic acids,whichcounteract
theactionofSu.(Suinhibitsfolicacidformation
byblocking thecoupling ofpteridine and
p-aminobenzoic acid.)YokotaandAkiba(143)
alsofoundthattheintracellular levelsofpteri-
dines,folicacids,andtetrahydrofolic acids
werethreetofourtimeshigherinthestrains
withsimpleSuresistance thaninthesensitive
strain.Themultiple drugresistant E.colidid
notshowthismetabolic change.
UsingSu-resistant andmultiple drugresistantstrainsofE.coli,AkibaandYokota(8)studied
theeffectofSuontheenzyme,folicase, which
catalyzes thecombination ofpteridine with
p-aminobenzoic acid.Itwasfoundthatthis
enzymewasnotinhibited bySuinintactcellsof
Su-resistant orofmultiple drugresistantstrains,
whereas itwasinhibited inSu-sensitive cells.
Incell-freehomogenates, however, thefolicase
ofthemultiple drugresistant strainwasin-
hibitedtoapproximately thesamedegreeas
theenzymeofthesensitive strain.Thefolicase
inacell-free homogenate oftheSu-resistant
strainswasnotinhibited bySu.Whether this
resulted fromalteredsensitivity oftheenzyme
toSuorfromincreased accumulation ofthe
pterines whichantagonize Suisnotcertain.
TheyfoundalsothatS35-sulfathiazole readily
enterssensitive andSu-resistant strainsbutnot
themultipledrugresistant strain.Thesefindings
indicate thattheSuresistance involved in
multiple drugresistance isduetoreduced
permeability ofthecellstoSu.
Yokota andAkiba(144)thenstudiedCm
resistance inastrainofE.coli,whichhadacquired
Cmresistance invitroasaresultofmultistep
mutations fromasensitive parent,andalsoina
strainofE.coliwithanRfactor(Su,Sm,Cm,
Tc).Theydetermined theeffectofCmonthe
synthesis offl-galactosidase byosmotic lysates
ofpenicillin-spheroplasts andlysozyme-proto-
plastsinthesystemofReiner(97).Thesynthesis
ofthisenzyme wasfoundtobeinhibited by
Cminthelysatesofbothresistant strains,
whereasitssynthesis, aswellasthenetsynthesis
ofprotein,byintactcells,penicillin spheroplasts,
andlysozyme protoplasts wasnotinhibited by
Cm.Accordingly, itwasconcluded thatCm
resistance resultsfromreduced permeability of
cellstoCminbothtypesofresistant strains.
Thefactthatbothspheroplasts andprotoplasts
manifest Cmresistance suggests thattheper-
meability barriertoCmliesinthecytoplasmic
membrane ratherthaninthecellwall.Similar
studies wereundertaken byOkamoto and
Mizuno(92)ontheeffectofCmontheincorpora-
tionoflabeledaminoacidsintoproteins, using
sonically disrupted cellfractions. Thesame
conclusion wasreached.
AkibaandYokota(9)studiedtheeffectsof
CmandTcontheincorporation oflabeledamino
acidsintoproteins byisolatedribosome prepa-
rationsfromsensitive andmultiple drugre-107

BACTERIOL. REV.
sistantstrainsofE.coli.Atconcentrations as
lowas5mugperml,Cmcompletely inhibited the
incorporation ofaminoacidsintoproteins in
bothpreparations. Ontheotherhand,evenat
251Agperml,Tcproduced onlypartialinhibition
inpreparations frombothstrains.SinceTcacts
asachelating agentandcombines withMgion,
whichisessential fortheintegrity ofribosomes,
theresultsarenotclear-cut. However, theysug-
gestthattheTcresistance involved inmultiple
drugresistance mayalsobecausedbyreduced
permeability ofcellstoTc.
Watanabe andTakano (unpublished data)
transferred anRfactor(Su,Sm,Cm,Tc)to
Sm-dependent mutants ofE.coli.Therecipients
whichreceivedtheRfactorwerestilldependent
onSm.Sincethebiochemical mechanism of
Smdependence isnotyetknown,itisdifficult
todrawsignificant conclusions fromthisresult.
ExceptforSmresistance, aboutwhichvery
littleisknown,theresultsdiscussed abovein-
dicatethatmultiple drugresistance mustbe
ascribed toreducedpermeability ofcellstothe
drugs.Itshouldbeemphasized againthatthe
reduced permeability isspecifically controlled
byeachresistance factor.
SUMMARY
Iwillnowsummarize brieflythisoutlineof
thestudiesoftransmissible (or,moreaccurately,
infective) drugresistance whichhavebeen
carried outalmostexclusively byJapanese
investigators. Several groupsofworkers are
continuing workonthisproblem. Although I
havetriedtoincludetheresultsofmostofthe
pertinent publications, anumber ofnewfacts
willhaveemerged bythetimethepresent re-
viewappearsinprint.
Theproblem oftheinfective heredity ofdrug
resistance aroused interestinitially because of
itsmedicalimportance. Onlylaterdiditdevelop
intoanareaofconsiderable geneticinterest. A
greatdealofdatahavealreadybeenaccumulated
bymanyworkers onthesubjectofbacterial
drugresistance buttheinfective drugresistance
described inthisreviewisthefirstcaseofepi-
some-mediated transfer ofdrugresistance. Itis
interesting tonotethatthediscovery oftransfer
ofmultipledrugresistance resultedfromcareful
epidemiological observations ofmultiple drug
resistantshigellae, andthatthisledsubsequentlytothediscovery ofanewepisome, theresistance
transferfactor(RTF).
Themedical importance ofinfective drug
resistance, especially multiple drugresistance, is
apparently limitedtoJapanatpresent, butit
couldbecomeaseriousandworld-wide problem
inthefuture.ThefactthatRfactorscanbe
transferred toeverygenusofEnterobacteriaceae
andtoothergenera,including V.comma, by
nonpathogenic bacteria suchasE.coliconstitutes
aseriouspublichealthproblem. Inindividuals
whoingestE.coliwhichcontainRfactors,the
E.coliandotherbacterial strainsalreadypresent
intheintestinal tractcanbeconverted todrug
resistance. Suchanindividual willbecome a
carrierofdrug-resistant intestinal flora.If
infection withpathogenic bacteria suchas
salmonellae, shigellae, andV.commaoccurs,
thesepathogenic organisms couldeasilyacquire
Rfactorsfromtheintestinal flora.Although the
transferofRfactorsnormally occurslessreadily
intheintestinal tractthaninvitro,because of
thepresence offattyacidsandotherinhibitory
agents,itmaybeaccelerated intheintestinal
tractundertheselective pressure ofchemothera-
peutics.
Thehistoryofthediscovery ofRfactors,and
subsequently RTF,isquiteunusualinthatRfac-
tors,whichpresumably resulted fromgene
pick-up, werediscovered beforeRTFwas
recognized. Examples ofepisome-mediated trans-
ferhavebeenreported inotherepisomes, butin
mostinstances theoriginalepisomes weredis-
covered firstbecause oftheiressential, discrete
functions. Lac-episome (Falkow etal.,22;
BaronandFalkow, 15)isapossibleexception.
Lac-episome behaves asanepisome, although its
integrated statehasnotbeenshown.Itisassumed
thatthelacgenesoflac-episome derived from
thechromosome ofE.coli.Baron,Carey,and
Spilman (13),however, failedtodifferentiate
thelacgenesandanepisome whichpresum-
ablycarriesthemintheexperiment ofinter-
ruptedconjugation. Theepisome RTFwas
recognized onlybecause oftheresistance fac-
torsitcarried.Itisassumed thatRTFwithout
anyresistance factorsshouldexist,butwould
notbeidentified becauseofitsinability tomani-
festadiscretefunction, suchaslysisofbacteria
byphages,conversion tofertilitybyFfactors,
orproduction ofcolicinsbycolicinogenic factors.
TheonlymarkerofRTF(RTFtype1)which108 WATANABE

VOL.27,1963 INFECTIVE HEREDITY OFBACTERIAL DRUGRESISTANCE
hasbeenfoundsuitable forselecting cellswith
RTFissuppression ofFfactorfunctions in
E.coliK-12.Eventhisisnotagoodmarkerfor
selection.
RTFwasaddedasanewmember tothecate-
goryofepisomes, anditwasanticipated that
comparative studiesofRTFwithotherknown
episomes willincrease ourunderstanding ofepi-
somesingeneral.Temperate phageshaveproved
tobeveryvaluablesubjectsforgeneticanalysis,
andhavebeenstudiedingreatdetail.Incontrast,
geneticstudiesontheotherepisomal "transfer
factors" areextremely difficult,primarily because
thesefactorscannotbetransferred extracellularly.
Wemaysucceedinisolating cell-freeparticles ofa
transferfactorbutatpresentthereisnowayto
testfortheirbiological activities. Theobserva-
tionthatRTFinteracts withFfactorandthe
chanceisolation ofmutanttypesofRTFoffer
somehopeforthestudyofthegenetics ofRTF
andothertransferfactors.
Themostessential features ofepisomes are
autonomous replication andintegration. Many
beautiful resultshavebeenobtained onthese
pointswithtemperate phages(Jacob,44;Camp-
bell,17).Another important characteristic of
certainepisomes, thetransferfactors,istheir
transferbyconjugation, causedbytheepisomes
themselves. Thesynthesis ofthe"mating sub-
stance," whichwasfirstimplicated bySneath
andLederberg (106)asessential formating, was
foundtobeinhibited byRTFtype1,which
suppresses fertilityduetoFfactor,butnotto
beinhibited byRTFtype2,whichdoesnot
affectfertility duetoFfactor.Inhibition by
RTFtype1ofsynthesis ofmatingsubstance is
considered tobethemechanism forsuppression
ofthefunctions ofFfactor.Untilthisobserva-
tion,geneticstudiesonRfactorsandRTFhad
tobeconducted alonglinessuccessful withother
episomes, especially temperate phages and
Ffactors. Thediscovery oftheinteractions
between RTFandFfactorhasopenedanew
fieldofresearch. Itisexpected thatfuture
studiesofRfactorsandRTFwillmakefunda-
mentalcontributions toourunderstanding of
episomes ingeneral.
ACKNOWLEDGMENTS
Theauthorisindebted toD.Ushiba, De-
partment ofBacteriology, KeioUniversity
SchoolofMedicine, B.D.Davis,Department ofBacteriology andImmunology, Harvard Medical
School,andH.W.Scherp,National Institute of
DentalResearch, fortheirencouragement, which
madethisreviewpossible. Heismostthankful
toB.D.DavisandZ.Miller,whokindlyread
andcorrected themanuscript. Hisdeepthanks
arealsoduetoD.Ushibaforhelpfulcriticisms.
HealsoowesmuchtoT.Fukasawa forhis
cooperation inmanyoftheauthor's experiments
citedinthepresentreviewandalsoforfruitful
discussions.
Someoftheauthor's investigations cited
hereweresupported byresearch grantsfrom
theMinistry ofEducation, Japan,andfrom
theWaksman Foundation ofJapan.Publication
ofthisreviewwassupported byaresearch grant
(E4740) fromtheNational Institute ofAllergy
andInfectious Diseases, U.S.PublicHealth
Service.
LITERATURE CITED
1.ADELBERG, E.A.,ANDS.N.BURNS. 1960.
Genetic variation inthesexfactorof
Escherichia coli.J.Bacteriol. 79:321-330.
2.AKIBA,T.1959.Mechanism ofdevelopment
ofresistance inshigella. Medicine ofJapan
in1959.[InJapanese] Proc.15thGen.
Meeting oftheJapanMed.Assoc.5:299-
305.
3.AKIBA,T.,ANDS.IWAHARA. 1961.Studies on
themechanism oftransfer ofdrug-re-
sistance inbacteria. IX.Influence oftem-
perature andpHonresistance-transfer.
[InJapanese] Med.Biol.(Tokyo) 60:42-44.
4.AKIBA,T.,ANDS.KIMURA. 1959.Anepidemic
ofdysentery causedbyshigella highly
resistant totetracyclines. [InJapanese]
NipponIjiShimpo 1837:21-24.
5.AKIBA, T.,K.KOYAMA, Y.ISHIKI,S.KIMURA,
ANDT.FUKUSHIMA. 1960.Onthemecha-
nismofthedevelopment ofmultiple drug-
resistant clones ofShigella. Japan. J.
Microbiol. 4:219-227.
6.AKIBA, T.,K.KOYAMA, Y.ISHIKI,S.KIMURA,
ANDT.FUKUSHIMA. 1960.Studies onthe
mechanism ofdevelopment ofmultiple
drug-resistant Shigella strains. [InJapa-
nese]Nippon IjiShimpo 1866:46-50.
7.AKIBA, T.,K.KOYAMA, S.KIMURA, AND
T.FUKUSHIMA. 1961.Studies onthemecha-
nismoftransfer ofdrug-resistance in
bacteria. VIII.Experiments onthetransfer
ofdrug-resistance invivo.[InJapanese]
Med.Biol.(Tokyo) 59:185-188.
8.AKIBA,T.,ANDT.YOKOTA. 1961.Studies109

BACTERIOL. REV.
onthemechanism oftransfer ofdrug-
resistance inbacteria. V.Enzyme for
conjugation ofPABandpteridine, isolated
fromthesulfonamide-resistant, drug-
sensitive andmultiple-resistant strains
ofE.coli.[InJapanese] Med.Biol.(Tokyo)
58:161-165.
9.AKIBA,T.,ANDT.YOKOTA. 1961.Biochemi-
calmechanisms ofresistance ofmultiple
drug-resistant enteric bacteria. 4.Effects
ofdrugsonthebiosynthesis of3-galactosi-
daseandincorporation ofC'4-amino acids
byribosomes ofdrug-sensitive andmultiple
drug-resistant E.coli.[InJapanese]
Proc.15thGen.Meeting KantoBranch
JapanBacteriol. Soc.,p.25.
10.ALFOLDI, L.,F.JACOB, E.L.WOLLMAN,
ANDR.MAZP,.1958.Surledeterminisme
g6netique delacolicinogenie. Compt.
Rend.246:3531-3533.
11.ANDERSON, T.F.,E.L.WOLLMAN, AND
F.JACOB.1957.Surlesprocessus deconju-
gaisonetderecombinaison chezEscherichia
coli.III.Aspects morphologiques en
microscopie 6lectronique. Ann.Inst.
Pasteur 93:450-455.
12.ARBER,W.1960.Transduction ofchromosomal
genesandepisomes inEscherichia coli.
Virology 11:273-288.
13.BARON, L.S.,W.F.CAREY,ANDW.M.SPIL-
MAN.1959.Genetic recombination between
Escherichia coliandSalmonella typhi-
murium. Proc.Natl.Acad.Sci.U.S.45:976-
984.
14.BARON, L.S.,W.F.CAREY, ANDW.M.
SPILMAN. 1959.Characteristics ofahigh
frequency ofrecombination (Hfr)strain
ofSalmonella typhosa compatible with
Salmonella, Shigella andEscherichia
species. Proc.Natl.Acad.Sci.U.S.45:1752-
1757.
15.BARON. L.S.,ANDS.FALKOW. 1961.Genetic
transfer ofepisomes fromSalmonella
typhosa toVibriocholerae. Genetics 46:849.
16.CALEF, E.,ANDG.LICCIARDELLO. 1960.
Recombination experiments onprophage
hostrelationships. Virology 12:81-103.
17.CAMPBELL, A.1962.Episomes. Advan.Genet.,
inpress.
18.DAVIS,B.D.1950.Studies onnutritionally
deficient bacterial mutants isolated by
meansofpenicillin. Experientia 6:41-50.
19.EGAWA, R.,H.HASHIMOTO, T.HONDA, S.
MITSUHASHI, K.HARADA, M.KAMEDA,
ANDM.SuzuKI. 1961.Studies ondrug
resistance ofentericbacteria withspecial
reference totheinhibition oftransfer ofdrugresistance. [InJapanese] Japan. J.
Bacteriol. 16:703-704.
20.EPHRUSSI, B.,H.HOTTINGUER, ANDA.M.
CHIMENENES. 1949.Actiondel'acriflavine
surleslevures. I.Lamutation "petite
colonie." Ann.Inst.Pasteur 76:351-367.
21.FALKOW, S.,ANDL.S.BARON. 1961.An
episomic element inastrainofSalmonella
typhosa. Bacteriol. Proc.,p.96.
22.FALKOW, S.,J.MARMUR, W.F.CAREY,
W.M.SPILMAN, ANDL.S.BARON. 1961.
Episomic transfer between Salmonella
typhosa andSerratia marcescens. Genetics
46:703-706.
23.FISHER, K.W.1957.TheroleoftheKrebs
cycleinconjugation inEscherichia coli
K-12.J.Gen.Microbiol. 16:120-135.
24.FISHER, K.W.1957.Thenature ofthe
endergonic processes inconjugation in
Escherichia coliK-12.J.Gen.Microbiol.
16:136-145.
25.FREDERICQ, P.1954.Intervention dufacteur
depolarit6 sexuelle Fdanslatransduction
desproprilt~s colicinogenes chezEscher-
ichiacoli.Compt. Rend.Soc.Biol.148:746-
748.
26.HAGIWARA, A.1958.Recombination ability
ofEscherichia coliK-12protoplast. Nature
182:456-457.
27.HAKURA, A.,ANDY.HIROTA. 1961.Sexde-
termining temperate phage r(tau) in
Escherichia coliK-12.[InJapanese] Abstr.
33rdMeeting Genet.Soc.Japan., p.4.
28.HARADA, K.1959.Studies ondirected varia-
tionofSalmonella 0-34bybacteriophage.
1.Antigenic transformation bybacterial
autolysate andantiserum. Japan.J.Micro-
biol.3:53-G0.
29.HARADA, K.1959.Studies ondirected varia-
tionofSalmonella 0-34bybacteriophage.
2.Antigen-transforming agentinbacteria
ofgroupE3.Japan.J.Microbiol. 3:61-69.
30.HARADA, K.,M.KAMEDA, M.SUZUKI, AND
S.MITSUHASHI. 1961.Onthedrug-resist-
anceofentericbacteria. 5.Mechanisms of
thetransmission ofthedrug-resistance
between Shigella andE.coli.[InJapanese]
Japan. J.Bacteriol. 16:6-16.
31.HARADA, K.,M.KAMEDA, M.SUZUKI, AND
S.MITSUHASHI. 1962.Drug-resistance of
entericbacteria. IX.Acquisition oftrans-
missibility ofnon-infective R-factor trans-
ducedbyphageepsilonfollowing F-factor
infection. [InJapanese] Med.Biol.(Tokyo)
62:21-24.
32.HARADA, K.,M.SUZUKI, M.KAMEDA, H.
HASHIMOTO, M.MORIMURA, S.TANAKA,110 WATANABE

VOL.27,1963INFECTIVE HEREDITY OFBACTERIAL DRUG RESISTANCE
K.ATSUMI, F.TABEI, S.MITSUHASHI, Y.
AKITA, ANDM.TAKAHASHI. 1961.Onthe
drug-resistance ofenteric bacteria. 7.
Transmissibility ofthedrug-resistance
amongfamilyEnterobacteriaceae. [InJapa-
nese]Japan.J.Bacteriol. 16:189-194.
33.HARADA, K.,M.SUZUKI, M.KAMEDA, AND
S.MITSUHASHI. 1960.Onthedrug-resist-
anceofentericbacteria. 2.Transmission of
thedrug-resistance among Enterobac-
teriaceae. Japan.J.Exptl.Med.30:289-299.
34.HARADA, K.,M.SUZUKI, M.KAMEDA, AND
S.MITSUHASHI. 1962.Drugresistance of
enteric bacteria. VIII.Transduction of
transmissible drug-resistance (R)factors
bye15andeuphage,andproperties oftheir
transductants. [InJapanese] Med.Biol.
(Tokyo) 62:13-16.
35.HAYES, W.1957.Thekinetics ofthemating
process inEscherichia coli.J.Gen.Micro-
biol.16:97-119.
36.HIROTA, T.1960.Theeffectofacridine dyes
onmating typeinEscherichia coli.Proc.
Natl.Acad.Sci.U.S.46:57-64.
37.HIROTA, Y.,ANDT.IIJIMA. 1957.Acriflavine
asaneffective agentforeliminating F
factorinEscherichia coli.Nature 180:655.
38.HIROTA, Y.,ANDP.H.A.SNEATH. 1961.F'
andFmediated transduction inEscherichia
coliK-12.Japan.J.Genet.36:307-318.
39.IIJIMA, T.1959.Transfer ofacolicinogenic
character byprotoplast ofEscherichia coli
K-12.Japan.J.Genet.34:213-219.
40.IIJIMA, T.1961.Transfer ofR-factor and
colicinogenic factorinEscherichia coli
K-12.Abstr. 33rdMeeting Genet. Soc.
Japan., p.7.
41.ISEKI,S.,ANDT.SAKAI. 1953.Artificial
transformation of0antigens inSalmonella
Egroup. I.Transformation byantiserum
andbacterial autolysate. Proc.Japan
Acad.29:121-126.
42.ISEKI,S.,ANDT.SAKAI. 1953.Artificial
transformation of0antigens inSalmonella
Egroup. II.Antigen-transforming factor
inbacilli ofsubgroup E2.Proc.Japan
Acad.29:127-131.
43.IWAHARA, S.,ANDT.AKIBA.1961.Studies on
themechanism oftransfer ofdrug-resist-
anceinbacteria. X.Influence ofpreincuba-
tiontimeofdonorandrecipient strains on
resistance transfer. [InJapanese] Med.
Biol.60:45-48.
44.JACOB,F.1961.Genetic controlofviralfunc-
tions.Harvey Lectures Ser.1959-1960,
p.1-39.
45.JACOB,F.,ANDE.A.ADELBERG. 1959.Trans-fertdecaractbres gendtiques parincorpora-
tionaufacteur sexueld'Escherichia coli.
Compt. Rend.249:189-191.
46.JACOB, F.,ANDJ.MONOD. 1961.Genetic
regulatory mechanisms inthesynthesis of
proteins. J.Mol.Biol.3:318-356.
47.JACOB,F.,P.SCHAEFFER, ANDE.L.WOLL-
MAN.1960.Episomic elements inbacteria.
Symp.Soc.Gen.Microbiol. 10:67-91.
48.JACOB,F.,ANDE.L.WOLLMAN. 1957.Ana-
lyzedesgroupes deliaisongenetique de
differentes souchesdonatrices d'Escherichia
coliK-12.Compt.Rend.245:1840-1843.
49.JACOB,F.,ANDE.L.WOLLMAN. 1958.Genetic
andphysical determinations ofchromo-
somalsegments inEscherichia coli.Symp.
Soc.Exptl.Biol.12:75-92.
50.JACOB,F.,ANDE.L.WOLLMAN. 1958.Les
episomes, elements genetiques ajoutes.
Compt. Rend.247:154-156.
51.KAGIWADA, S.,S.KATO,M.ROKUGO, J.
HOSHINO, ANDK.NISHIYAMA. 1960.Studies
onthemechanism ofdevelopment ofre-
sistantShigella inintestinal tractswith
specialreference tothepossibility ofcon-
verting sensitive Shigella toresistance by
multiple-drug-resistant E.coliinintestinal
tracts.[InJapanese] Nippon IjiShimpo
1886:5-9.
52.KITAMOTO, O.,N.KASAI, K.FUKAYA, AND
A.KAWASHIMA. 1956.Drug-sensitivity of
theShigella strains isolated in1955.[In
Japanese] J.Japan. Assoc. Infectious
Diseases 30:403-404
53.KOBARI, K.,ANDI.TAJIRI. 1959.Recent
trends ofantibiotic-resistant Shigella.
[InJapanese] Modern Media5:205-209.
54.KONDO, E.,K.HARADA, ANDS.MITSUHASHI.
1962.Transduction ofthetransmissible
drug-resistance factors bybacteriophage
Plkc.Japan.J.Exptl.Med.32:139-147.
55.KOYAMA, K.,ANDT.AKIBA.1961.Studies on
themechanism oftransfer ofdrug-resist-
anceinbacteria. XI.Inhibitory effecton
theresistance transfer bySS-medium and
itsingredients. [InJapanese] Med.Biol.
(Tokyo) 60:159-162.
56.LAVALLP, R.,ANDF.JACOB. 1961.Surla
sensibilite desepisomes sexueletcolicino-
gened'Escherichia coliK12aladesint6gra-
tionduradiophosphore. Compt. Rend.
252:1678-1680.
57.LEDERBERG, J.1951.Streptomycin resist-
ance:agenetically recessive mutation. J.
Bacteriol. 61:549-550.
58.LEDERBERG, J.1956.Conjugal pairing in
Escherichia coli.J.Bacteriol. 71:497-498.illl

BACTERIOL. REV.
59.LEDERBERG, J.1956.Linear inheritance in
transductional clones.Genetics 41:845-871.
60.LEDERBERG, J.,ANDE.M.LEDERBERG. 1952.
Replica plating andindirect selection of
bacterial mutants. J.Bacteriol. 63:399-406.
61.LEDERBERG, J.,ANDE.M.LEDERBERG. 1956.
Infection andheredity. Symp.Soc.Growth
Develop. 14:101-124.
62.LEDERBERG, J.,ANDJ.ST.CLAIR. 1958.
Protoplasts andL-type growth ofEsch-
erichiacoli.J.Bacteriol. 75:143-160.
63.LOEB,T.1960.Isolation ofabacteriophage
specific fortheF+andHfrmating typesof
Escherichia coliK-12.Science 131:932-933.
64.LOEB,T.,ANDN.D.ZINDER. 1961.Abacterio-
phagecontaining RNA.Proc.Natl.Acad.
Sci.U.S.47:282-289.
65.LURIA, S.E.,J.N.ADAMS, ANDR.C.TING.
1960.Transduction oflactose-utilizing
ability among strains ofE.coliandS.
dysenteriae andtheproperties ofthetrans-
ducingphageparticles. Virology 12:348-
390.
66.LWOFF, A.1953.Lysogeny. Bacteriol. Rev.
17:269-337.
67.MARMUR, J.,R.ROWND, S.FALKOW, L.S.
BARON, C.SCHILDKRAUT, ANDP.DOTY.
1961.Thenatureofintergeneric episomal
infection. Proc.Natl.Acad.Sci.U.S.47:
972-979.
68.MATSUYAMA, T.,K.HARADA, M.SUZUKI,
M.KAMEDA, ANDS.KOMIYAMA. Anepi-
demicofdysentery caused byShigella
flexneri 3ahighlyresistant tocommon
antibiotics andaconsideration ofthe
mechanism ofdevelopment ofdrugresist-
ance.[InJapanese] Nippon IjiShimpo
1795:22-25.
69.MCCLINTOCK, B.1956.Controlling elements
andthegene.ColdSpringHarbor Symp.
Quant.Biol.21:197-216.
70.MESELSON, M.,ANDF.W.STAHL.1958.The
replication ofDNA.ColdSpringHarbor
Symp.Quant.Biol.23:9-12.
71.MITSUHASHI, S.1960.Multidrug-resistance
ofentericbacteria. I.Induction ofdrug-
resistance following mixedcultivation of
drug-resistant anddrug-sensitive bacteria.
[InJapanese] Medicine andBiology 55:
49-52.
72.MITSUHASHI, S.1960.Drugresistance andits
heredity inentericbacteria. [InJapanese]
Kagaku (Tokyo) 30:628-633.
73MITSUHASHI, S.,K.HARADA, ANDH.HASHI-
MOTO.1960.Drug-resistance ofenteric
bacteria. IV.Generality oftheappearance
ofmultidrug-resistant enteric bacteria and
experimental approach toobtaining multi-drug-resistant E.coli.[InJapanese] Med.
Biol.(Tokyo) 55:157-160.
74.MITSUHASHI, S.,K.HARADA, ANDH.HASHI-
MOTO.1960.Multiple resistance ofenteric
bacteria andtransmission ofdrug-resist-
ancetootherstrainbymixedcultivation.
Japan.J.Exptl.Med.30:179-184.
75.MITSUHASHI, S.,K.HARADA, ANDH.HASHI-
MOTO. 1961.Drug-resistance ofenteric
bacteria. VI.Combination oftwotypesof
R-factor inhostbacterium. [InJapanese]
Med.Biol.(Tokyo) 61:145-148.
76.MITSUHASHI, S.,K.HARADA, H.HASHIMOTO,
ANDR.EGAWA. 1961.Onthedrug-resist-
anceofenteric bacteria. 4.Drug-resistance
ofshigella prevalent inJapan. Japan. J.
Exptl.Med.31:47-52.
77.NIMTSUHASHI, S.,K.HARADA, H.HASHIMOTO,
ANDR.EGAWA. 1961.Drug-resistance of
enteric bacteria. 5.Drug-resistance of
Escherichia coliisolated fromhumanbeing.
Japan.J.Exptl.Med.31:53-60.
78.MITSUHASHI, S.,K.HARADA, ANDM.
KAMEDA. 1961.Elimination oftransmissi-
bledrug-resistance bytreatment with
acriflavine. Nature189:947.
79.MITSUHASHI, S.,K.HARADA, ANDM.
KAMEDA. 1961.Onthedrug-resistance of
enteric bacteria. 6.Spontaneous andarti-
ficialelimination oftransmissible drug-
resistance factors. Japan.J.Exptl.Med.
31:119-123.
80.MITSUHASHI, S.,K.HARADA, M.KAMEDA,
M.SUZUKI, ANDR.EGAWA. 1960.Onthe
drug-resistance ofenteric bacteria. 3.
Transmission ofthedrug-resistance from
Shigella toF-orHfrstrainsofE.coliK-12.
Japan.J.Exptl.Med.30:301-306.
81.MITSUHASHI, S.,H.HASHIMOTO, K.HARADA,
R.EGAWA, ANDT.MATSUYAMA. 1961.
Studies onthedrug-resistance ofenteric
bacteria. 7.Typeofthetransmissible drug-
resistance factor. Gumma J.Med.Sci.
10:59-66.
82.MIYAMURA, S.1961.Chloramphenicolase in
dysentery bacilli,withspecial reference to
chloramphenicol resistance. [InJapanese]
Japan. J.Bacteriol. 16:115-119.
83.MIZUNO, D.1961.Mechanisms ofresistance
ofdrug-resistant bacteria. [InJapanese]
IgakuNoAyumi39:309-312.
84.NAKAYA, R.,A.NAKAMURA, ANDY.MURATA.
1960.Resistance transfer agentsinShigella.
Biochem. Biophys. Res.Commun. 3:654-
659.
85.NAKAYA, R.,A.NAKAMURA, ANDY.MURATA.
1961.Transduction ofresistance transfer
agentsandchromosomal genesbyUV-112 WATANABE

VOL.27,1963INFECTIVE HEREDITY OFBACTERIAL DRUG RESISTANCE
irradiated phagePlkc.[InJapanese]
Abstr.9thMeeting Virol.Japan,p.9.
86.OCHIAI, K.1959.Distribution andclinicof
dysentery caused byantibiotic-resistant
strains. Medicine ofJapanin1959.[In
Japanese] Proc.15thGen.Meeting Japan
Med.Assoc.5:306-316.
87.OCHIAI, K.1962.Studies onthetransfer of
drugresistance. [InJapanese] Chemo-
therapy (Tokyo) 10:47-50.
88.OCHIAI, K.,H.UWATOKO, S.NAITO, J.
KAWABATA, ANDW.KAWABE. 1959.Amass
incidence ofbacillary dysentery causedby
oxytetracycline resistant strainonanedu-
cational tripamongboysofAsahigaoka
HighSchoolinNagoya. [InJapanese] J.
Japan.Assoc.Infectious Diseases 32:721-
729.
89.OCHIAI, K.,T.YAMANAKA, K.KIMURA, AND
0.SAWADA. 1959.Studies oninheritance of
drugresistance between Shigella strains
andEscherichia colistrains. [InJapanese]
Nippon IjiShimpo 1861:34-46.
90.OCHIAI, K.,T.YAMANAKA, ANDK.KIMURA.
1961.Combination ofdrugresistance by
mixedcultivation. [InJapanese] Nippon
IjiShimpo 1835:38-42.
91.0RSKOV, I.,ANDF.QRSKOV. 1960.Anantigen
termed f+occurring inF+E.colistrains.
ActaPathol. Microbiol. Scand. 48:37-46.
92.OKAMOTO, S.,ANDD.MIZUNO. 1961.Unpub-
lisheddatacitedinMizuno, D.1961.
Mechanisms ofresistance ofdrug-resistant
bacteria. [InJapanese] IgakuNoAyumi
39:309-312.
93.OTSUJI, N.,M.SEKIGUCHI, T.IIJIMA, AND
Y.TAKAGI. 1959.Induction ofphagefor-
mation inthelysogenic Escherichia coli
K-12bymitomycin C.Nature 184:1079-
1080.
94.OZEKI, H.1956.Abortive transduction in
purine-requiring mutants ofSalmonella
typhimurium. Genetic studies withbac-
teria.Carnegie Inst.Wash.Publ.No.612,
p.97-106.
95.OZEKI,H.,ANDB.A.D.STOCKER. 1960.Un-
published datacitedinintroduction:
Microorganisms ingenetics. Symp.Soc.
Gen.Microbiol. 10:1-11.
96.OZEKI,H.,S.HOWARTH, ANDR.C.CLOWES.
1961.Colicine factorsasfertility factorsin
bacteria. Nature 190:986-989.
97.REINER, J.M.1960.Induced enzyme syn-
thesisincell-free preparations ofEsch-
erichiacoli.J.Bacteriol. 79:157-165.
98.RICHTER, A.1957.Complementary determi-
nantsofanHfrphenotype inE.coliK-12.
Genetics 42:391.99.ROBBINS, P.W.,ANDT.UCHIDA. 1961.Identi-
fication ofglucose asaphage-determined
component ofSalmonella polysaccharides.
Federation Proc.20:86.
100.ROPER, J.A.1958.Nucleo-cytoplasmic inter-
actions inAspergillus nidulans. Cold
Spring Harbor Symp.Quant. Biol.23:
141-154.
101.SAGER, R.1954.Mendelian andnon-Men-
delianinheritance ofstreptomycin resist-
anceinChlamydomonas reinhardi. Proc.
Natl.Acad.Sci.U.S.40:356-363.
102.SEKIGUCHI, M.,ANDY.TAKAGI. 1960.Effect
ofmitomycin Conthesynthesis ofbac-
terialandviraldeoxyribonucleic acid.
Biochim. Biophys. Acta41:434-443.
103.SEVAG,M.G.,ANDT.YOKOTA. 1960.Evolu-
tionofthephenomenon ofdrug-resistance.
II.Studies oftheextracellular andintra-
cellular derivatives offolicacidinrelation
toresistance tosulphathiazole andgrowth
factor requirements. Proc.Roy.Soc.
(London) Ser.B.153:205-219.
104.SEVAG,M.G.,ANDT.YOKOTA. 1960.Evolu-
tionofthephenomenon ofdrug-resistance.
III.Theoriginofthedevelopment ofthe
resistance tosulphathiazole inEscherichia
coliandStaphylococcus aureus. Proc.Roy.
Soc.(London) Ser.B.153:220-233.
105.SKAAR, P.D.,A.RICHTER, ANDJ.LEDER-
BERG.1957.Correlated selection formo-
tilityandsexincompatibility inEscherichia
coliK-12.Proc.Natl.Acad.Sci.U.S.
43:329-333.
106.SNEATH, P.H.A.,ANDJ.LEDERBERG. 1961.
Inhibition byperiodate ofmating in
Escherichia coli.Proc.Natl.Acad.Sci.
U.S.47:86-90.
107.STENT,G.S.,C.R.FUERST, ANDF.JACOB.
1957.Inactivation d'unprophage parla
desintegration duradiophosphore. Compt.
Rend.244:1840-1842.
108.STOCKER, B.A.D.1956.Abortive transduc-
tionofmotility inSalmonella; anon-repli-
catedgenetransmitted through many
generations toasingledescendant. J.Gen.
Microbiol. 15:575-598.
109.SUGINO, Y.,ANDY.HIROTA. 1961.Mediation
ofsexuality ofE.coliK-12bythedrugre-
sistance factors R.Abstr.33rdMeeting
Genet.Soc.Japan,p.22.
110.THOMPSON, D.H.1931.Theside-chain theory
ofthestructure ofthegene.Genetics 16:
267-290.
111.UETAKE, H.,ANDS.HAGIWARA. 1961.Genetic
cooperation between unrelated phages.
Virology 13:500-506.
112.UETAKE, H.,S.E.LURIA, ANDJ.W.113

BACTERIOL. REV.
BURROUS. 1958.Conversion ofsomatic
antigens inSalmonella byphageinfection
leading tolysisorlysogeny. Virology
5:68-91.
113.UETAKE, H.,T.NAKAGAWA, ANDT.AKIBA.
1955.Therelationship ofbacteriophage to
antigenic changes ingroupEsalmonellas.
J.Bacteriol. 69:571-579.
114.WATANABE, T.1954.Genetic studies onthe
mechanism ofacquired streptomycin-
resistance inmicroorganisms. KeioJ.Med.
3:193-213.
115.WATANABE, T.,ANDT.FUKASAWA. 1960.
"Resistance transfer factor", anepisome
inEnterobacteriaceae. Biochem. Biophys.
Res.Commun. 3:660-665.
116.WATANABE, T.,ANDT.FUKASAWA. 1960.
Episomic resistance factors inEntero-
bacteriaceae. I.Transfer ofresistance
factors byconjugation amongEnterobac-
teriaceae. [InJapanese] Med.Biol.(Tokyo)
56:56-59.
117.WATANABE, T.,ANDT.FUKASAWA. 1960.
Episomic resistance factors inEnterobac-
teriaceae. II.Transduction ofresistance
factors inEscherichia coliandSalmonella
typhimurium. [InJapanese] Med.Biol.
(Tokyo) 56:64-67.
118.WATANABE, T.,ANDT.FUKASAWA. 1960.
Episomic resistance factors inEnterobac-
teriaceae. III.Elimination ofresistance
factorsbytreatment withacridine dyes.
[InJapanese] Med.Biol.(Tokyo) 56:71-74.
119.WATANABE, T.,ANDT.FUKASAWA. 1960.
Episomic resistance factors inEnterobac-
teriaceae. IV.Kinetics oftransfer ofresist-
ancefactors. [InJapanese] Med.Biol.
(Tokyo) 56:98-100.
120.WATANABE, T.,ANDT.FUKASAWA. 1960.
Episomic resistance factors inEnterobac-
teriaceae. V.Transfer ofresistance factors
mediated byanewepisome. [InJapanese]
Med.Biol.(Tokyo) 56:201-206.
121.WATANABE, T.,ANDT.FUKASAWA. 1960.
Episomic resistance factors inEnterobac-
teriaceae. VI.Suppression ofsexfactorof
Escherichia colistrainK-12byresistance
transfer factor.[InJapanese] Med.Biol.
(Tokyo) 57:30-33.
122.WATANABE, T.,ANDT.FUKASAWA. 1960.
Episomic resistance factors inEnterobac-
teriaceae. VII.Suppression oftransfer of
resistance transfer factorfromresistant
Shigella toEscherichia colistrainK-12by
thepresence ofF-factor inrecipients. [In
Japanese] Med.Biol.(Tokyo) 57:169-172.
123.WATANABE, T.,ANDT.FUKASAWA. 1961.
Episome-mediated transfer ofdrugresist-anceinEnterobacteriaceae. I.Transfer of
resistance factors byconjugation. J.
Bacteriol. 81:669-678.
124.WATANABE, T.,ANDT.FUKASAWA. 1961.
Episome-mediated transfer ofdrugresist-
anceinEnterobacteriaceae. II.Elimination
ofresistance factors withacridine dyes.
J.Bacteriol. 81:679-683.
125.WATANABE, T.,ANDT.FUKASAWA. 1961.
Episome-mediated transfer ofdrugresist-
anceinEnterobacteriaceae. III.Transduc-
tionofresistance factors. J.Bacteriol.
82:202-209.
126.WATANABE, T.,ANDT.FUKASAWA. 1961.
Episomic resistance factors inEnterobac-
teriaceae. VIII.Kinetics oftransfer ofre-
sistance factors studied byinterrupted
conjugation withphageT6.[InJapanese]
Med.Biol.(Tokyo) 58:202-206.
127.WATANABE, T.,ANDT.FUKASAWA. 1961.
Episomic resistance factors inEnterobac-
teriaceae. IX.Interaction between resist-
ancefactorsandF'inEscherichia coliK-12.
[InJapanese] Med.Biol.(Tokyo) 59:42-47.
128.WATANABE, T.,ANDT.FUKASAWA. 1961.
Episomic resistance factors inEntero-
bacteriaceae. X.Relationship between re-
sistance factors andresistance transfer
factor.[InJapanese] Med.Biol.(Tokyo)
59:115-118.
129.WATANABE, T.,ANDT.FUKASAWA. 1961.
Episomic resistance factors inEntero-
bacteriaceae. XI.Highfrequency resistance
transfer system (HFRT). [InJapanese]
Med.Biol.(Tokyo) 59:153-157.
130.WATANABE, T.,ANDT.FUKASAWA. 1961.
Episomic resistance factors inEntero-
bacteriaceae. XII.Chromosomal attach-
mentofresistance transfer factor in
Escherichia colistrainK-12.[InJapanese]
Med.Biol.(Tokyo) 59:180-184.
131.WATANABE, T.,ANDT.FUKASAWA. 1962.
Episome-mediated transfer ofdrugresist-
anceinEnterobacteriaceae. IV.Interactions
between resistance transfer factorand
F-factor inEscherichia coliK-12.J.Bac-
teriol.83:727-735.
132.WATANABE, T.,T.FUKASAWA, ANDT.
TAKANO. 1961.Episomic resistance factors
inEnterobacteriaceae. XV.Suppression by
RTFofthephagef-sensitive mechanism
conferred byF-factor onEscherichia coli
strainK-12.[InJapanese] Med.Biol.
(Tokyo) 61:92-96.
133.WATANABE, T.,T.FUKASAWA, ANDT.
TAKANO. 1962.Conversion ofmalebacteria
ofEscherichia coliK-12toresistance to
fphagesbyinfection withtheepisome114 WATANABE

VOL.27,1963INFECTIVE HEREDITY OFBACTERIAL DRUG RESISTANCE
"resistance transfer factor." Virology
17:218-219.
134.WATANABE, T.,ANDK.W.LYANG. 1961.
Episomic resistance factors inEnterobac-
teriaceae. XIII.Spontaneous segregation of
resistance factors inSalmonella typhi-
murium. [InJapanese] Med.Biol.(Tokyo)
61:18-23.
135.WATANABE, T.,ANDK.W.LYANG. 1961.
Episomic resistance factors inEnterobac-
teriaceae. XIV.Genetic recombination of
resistance factors carried byRTF.[In
Japanese] Med.Biol.(Tokyo) 61:42-48.
136.WATANABE, T.,ANDT.TAKANO. 1962.
Episomic resistance factors inEnterobac-
teriaceae. XVI.RTFwhichdoesnotsup-
pressthemating ofHfrstrains ofEsch-
erichia coliK-12.[InJapanese] Med.
Biol.(Tokyo) 62:37-41.
137.WATANABE, T.,ANDM.WATANABE. 1959.
Transduction ofstreptomycin resistance
inSalmonella typhimurium. J.Gen.Micro-
biol.21:16-29.
138.WATANABE, T.,ANDM.WATANABE. 1959.
Transduction ofstreptomycin sensitivity
intoresistant mutants ofSalmonella
typhimurium. J.Gen.Microbiol. 21:30-39.
139.WOLLMAN, E.L.,F.JACOB, ANDW.HAYES.
1956.Conjugation andgenetic recombina-
tioninEscherichia coliK-12.ColdSpring
Harbor Symp.Quant. Biol.21:141-162.
140.YAMANAKA, T.1961.Studies ontheheredity
ofantibiotic-resistance ofdysentery ba-
cilli.II.Thecombination ofantibiotic-
resistances bymeans ofmixedculture.
[InJapanese] J.Japan. Assoc. Infect.
Diseases 34:1026-1033.
141.YAMANAKA, T.1961.Studies ontheheredity
ofantibiotic-resistance ofdysentery ba-
cilli.III.Deprivation-sensitization of
antibiotic-resistances andtheirpartialtransference. [InJapanese] J.Japan.Assoc.
Infect.Diseases 34:1034-1042.
142.YOKOTA, T.,ANDT.AKIBA.1961.Studies on
themechanism oftransfer ofdrug-resist-
anceinbacteria. III.Transfer oftheresist-
ancefactorinpenicillin-spheroplasts of
Enterobacteriaceae. [InJapanese] Med.
Biol.(Tokyo) 58:141-144.
143.YOKOTA, T.,ANDT.AKIBA. 1961.Studies on
themechanism oftransfer ofdrug-resist-
anceinbacteria. IV.Biochemical aspects
onthemechanism ofresistance tosulfon-
amideinthesulfonamide-resistant and
multiple-drug-resistant strainsofE.coli.
[InJapanese] Med.Biol.(Tokyo) 58:
151-154.
144.YOKOTA, T.,ANDT.AKIBA.1961.Studies on
themechanism oftransfer ofdrug-resist-
anceinbacteria. VI.Biochemical aspects
onamechanism ofresistance tochloram-
phenicol inthechloramphenicol-resistant
andinthemultiple-drug-resistant strains
ofE.coli.[InJapanese] Med.Biol.(Tokyo)
58:172-175.
145.YOKOTA, T.,ANDT.AKIBA.1961.Studies on
themechanism oftransfer ofdrug-resist-
anceinbacteria. VII.Inhibition oftransfer
oftheresistance-factor fromE.colito
Shigellaflexneri bysomechemicals. [In
Japanese] Med.Biol.(Tokyo) 58:188-191.
146.YOSHIKAWA, M.,ANDT.AKIBA.1961.Studies
onthemechanism oftransfer ofdrug-
resistance inbacteria. I.Relationship be-
tweenthetransferable resistance-factor
andF-factor. [InJapanese] Med.Biol.
(Tokyo) 58:126-129.
147.ZINDER, N.D.1953.Infective heredity in
bacteria. ColdSpringHarborSymp.Quant.
Biol.18:261-269.
148.ZINDER, N.D.1958.Lysogenization and
superinfection immunity inSalmonella.
Virology 5:291-326.115