T-Even Bacteriophage-Tolerant Mutants of Escherichia coli B II. Nucleic Acid Metabolism

JOURNAL OFVIROLOGY, Sept. 1971, p. 275-285 Vol. 8, No. 3 Copyright © 1971 AmericanSociety for Microbiology Printed in U.S.A.

T-Even

Bacteriophage-Tolerant Mutants

of

Escherichia coli B

II.

Nucleic Acid

Metabolism

CHRISTOPHER K. MATHEWS AND MARTINEZ J. HEWLETT

Departmentof Biochemistry, University of Arizona, College of Medicine, Tucson, Arizona 85724

Received for publication 1 June 1971

T-evenbacteriophage-tolerant mutants are strains of Escherichia coli which can

adsorbT-evenphagesbutcannot support thegrowth of infective virus. Under some

conditions,theinfectedcellsarenotkilled.Mutantcells infected by phage T6 are able

tocarryoutseveralmetabolic functions associated with normal virus

development,

including arrest of bacterial nucleic acid and protein synthesis, incorporation of isotopic precursors intoviralnucleic acids and proteins, synthesis of early enzymes of deoxyribonucleic acid (DNA) metabolism, formation of rapidly sedimenting DNA intermediates, and formation of normallevels of early and late messenger ribonucleic

acid species.Phageareunabletomutate toforms capable of growth on these mutants.

Thenatureof the biochemicalalterationleading to tolerance is still unknown.

As anapproachtounderstanding the

contribu-tion of host cell metabolic functions to virus development, ourlaboratory hasisolated T-even

bacteriophage-tolerant (tet) mutants of

Escheri-chia coliB (12). These strains canadsorb T-even phages butare unableto forminfective progeny. Undersome conditions, the cells are not killed. Preliminary investigation showed that someviral genes are expressed after infection of these mutants by phage T6. Early enzymes of

de-oxyribonuwleic acid (DNA) metabolism are

synthesized, albeit in limited amounts (12), and

somephageDNA replication occurs.

The objectives of the work described in the

present paper were, first, to devise

experimental

conditions suitable for extensive analysis of

metabolic events in abortive infections of tet

mutants,and, second,toexploreeventsinnucleic

acid metabolismwhichmight providecluestothe

primary biochemical alteration in tet mutants.

We find that even though tet mutants are

ab-solutely defective in the ability to propagate

infective virus they are able to support a wide range of virus-directed metabolic processes.

MATERIALS AND METHODS

All of the present work was carried out with T6 phage and E. coli B, along with its derivatives E. colitet 1 and tet 2, both described previously (12).

T6 was used because of its rapid adsorption to all threebacterial strains. Culture and assaytechniques and culture media were also as described in earlier

communications from this laboratory, except that

glycerol-Casamino Acids (GCA) medium contains 5 mgofCasamino Acidsperml rather than 15. Except whereotherwise indicated, all experiments were car-ried out in GCA medium.

Because of the tendency of tet mutants to revert to wild type, stocks were reisolated from single col-oniesatfrequent intervals. Moreover, in each experi-mentreported herein, the let phenotypewaschecked as follows. At 6 min after infection, samples were removed to dilution tubes either containing chloro-form for determination of unadsorbed phage or to tubescontaining no chloroform for measurement of

total infectivecenters.At90min, anothersamplewas removed to adilution tube containing chloroform for

determinationoftotal progeny. Suitabledilutions of the abovewere plated, withE. coli Bas theplating

host. The following were taken ascriteria for

satis-factory expression of the tet phenotype: infective

center titer < unadsorbed phage titer and infective

progeny titer < unadsorbed phage titer. Thus, in each experiment reported, there was no detectable formation of infective phage in infection of tet mu-tants.

For studies on

03-galactosidase

induction, 10-ml cultures were induced as described in the legend to

Fig. 4 and then chilled and centrifuged. Thepellets

wereresuspendedin5ml each of0.02Msodium phos-phate buffer (pH 7.5) and sonically treated for 30

sec. i3-Galactosidase activity was measured in these

sonically treatedextracts asdescribedbyDuckworth and Bessman(4).

Incorporation of 14C-uracil into DNA and ribo-nucleic acid (RNA) and incorporation of 14C-leucine into protein were followed as described previously

(10, 11). Techniques for isolation of RNA and for DNA-RNA hybridizations were also as described 275

on November 11, 2019 by guest

http://jvi.asm.org/

MATHEWS AND HEWLETT

earlier (11), except that in isolating RNA diethyl pyrocarbonate (at 1() was present in the extracting buffer for inhibition of nucleases rather than poly-vinylsulfate.

In studies on DNAmetabolism, the methods used for DNA labeling, cell lysis, sucrose gradient centrif-ugation, and radioisotope countingwere as described by Murray and Mathews (14, 15). The technique used for DNA-DNA hybridization was the membrane filter method of Denhardt (3), as modified in this laboratory by M. Sisson (M. Sisson and C.K. Math-ews, inpreparation).Incorporation of parental phage DNA into"replicating DNA complexes" was meas-ured by the M-band technique of Earhart and his colleagues (5, 6).

Further experimental details are provided in indi-vidual figure legends.

RESULTS

Effects of host cell survival. The basis for isolation of tet mutants is theirabilityto grow on

agarplatesin the presenceofalargeexcessoftwo

different T-even phages. Under these conditions, infection is not lethal to the hostcell. However, in our early studies wefound that considerable

cell killing occurs in infection of tet mutants in

liquid media, even though infection is always abortive. It is now clear that the ability of a tet

cell tosurvive infection dependsuponits environ-ment; cell survivalishigher in solid media than

in liquid media, and in liquid survival is higher

in nutritionally rich media than in minimal or

other defined media. Figure 1 presents data on

the

ability

oftetcellstoformcolonieswhen

plated

in the presence of various amounts of T6. The mutants did lose the ability to form colonies

ultimately but only at an input of phage some

100-fold greater than that necessary to abolish

the colony-forming

ability

of E. coli B plated

under the same conditions. Similarly, in liquid

media tet cells could lose the ability to form colonies when plated after infection at high

multiplicities, but the difference between E. coli

B and the mutants was considerably less under

these conditions. Figures 2 and 3 show bacterial

survival curves at various multiplicities of

infec-tionby T6in nutrient broth and GCAmedium,

respectively.

Note that the slopes of the tet

survival curves are considerably steeper inGCA

medium than in nutrient broth. Similar results

are seenininfection by T2 (data notshown). In all cases, the survival curves are approximately linear, indicating that killing of the tet mutants is a one-hit process; infection by a single phage

particle is a lethal event, although this occurs

with lower probability in infection of tet mutants than ininfectionofwild-type bacteria.

Because the environment on an agar plate is considerably more anaerobic than that of a

104

105

1O6

107

108

T6 Particles per plate

FIG. 1. Expression ofthe tet phenotype on agar

plates. Foreachstrain, a constant number of

colony-formers was plated on each of a series ofnutrientagar

plates in thepresence of the indicatednumbers of T6 particles.

MOI

FIG. 2. Survival curves bor bacteria injected in

nutrient broth. Bacterial cultures at about 5 X 108

per ml were infected with T6 at the indicated

multi-plicities. At 6 minl, samples were removed, diluted, and platedfbr determination ofcolony-forming

abil-ity.MOI, multiplicity of infection.

2)7

6 J. VIROL.

on November 11, 2019 by guest

http://jvi.asm.org/

[image:2.480.255.445.60.295.2] [image:2.480.254.445.359.578.2]

T-EVEN-TOLERANT MUTANTS OF E. COLI. 11

.i_

L

Ul)

. c

4-f,

CID

ic

IA

AB

N 0 tet1

0 * tet2

0

.-0

0.

A~~~

6.

0 6 12

M01

FIG. 3. Survivalcurvesforbacteria infectedin

glyc-erol-Casamino Acids (GCA) medium. Except for the

substitution of GCA medium for broth, the protocol

wasidenticalto thatdescribedinthelegendtoFig. 1. MOI,multiplicity of infection.

vigorously bubbled liquid culture, it seemed

possible that there is a relationship between

anaerobiosis and cellkilling:themore anaerobic the culture, the lower the probability that a

particularinfection will be lethal. This wastested

by comparing cell survival in GCA medium

either aeratedorbubbledwithnitrogen (datanot

shown). Cell survival was no greater in the anaerobic thanintheaerobic culture and indeed was somewhat less, indicating the absence of a direct relationship between anaerobiosis and cell survival.

Inability of phage to overcome tolerance. We wished to know whether T-even phages could

undergo mutations which would allow them to overcome tolerance, i.e., to plate oneither tet 1

or tet2. Themappositionsof such "host-range" mutations, at least in the genetically

well-char-acterizedT4, might providecluestothebacterial functions altered intheoriginaltetmutations. In

ourpreviousstudy,wefoundthatT2, T4,and T6 wereallunabletoplateoneithertetmutantwhen

as many as 106 viable phage were added to a

single plate. Wehavenowextended thiswork by subjectingboth T4andT6tochemical mutagene-sis by nitrous acid, hydroxylamine,

2-amino-purine, 5-bromodeoxyuridine, and N-methyl-N'-nitro-N-nitrosoguanidine. Even when asmanyas

108 plaque-forming units from mutagenized

stocks were plated on tet 1 or tet 2, no plaques were seen. Thus, phage seem quite unable to

circumvent the effects of the tet mutations by

extending their own host range. In this context, it isof interest that phage T7 plates on tet 1 with an

efficiency equal to that seen on E. coli B (W. C. Summers, personal communication).

By analogy with temperate phages, we consi-dered whether tolerance in the tet systems is at all akin toprophage immunity. In other words, could the tet character result, not from mutation, but from a stable association of phage DNA with the bacterial genome either as an episode or a

plasmid. Although this seemed improbable, we

tested it by attempted hybridization between bacterial DNA, fixed on membrane filters, and

"4C-labeled T6 DNA. Fewer than

0.2%

of the

input counts were bound to bacterial DNAfrom either E. coli

B,

tet 1 or tet 2, and no differences were seen among the three strains. Weconclude, therefore, that the tet genomes do not contain

significantamounts ofphage-specific DNA.

Effects on bacterial polymer synthesis. One of the most powerful biochemical research tech-niques is the use of radioisotopes for labeling

metabolic intermediates, whereas one of the most attractive features of T-even phages as objects for metabolic studies is the fact that infection com-pletely arrests the synthesis of macromolecules coded by the bacterial genome. Thus, one can label T-even phage-infected cells with isotopic protein or nucleic acid precursors secure in the knowledge that only virus-specific polymers will belabeled. In ourearlier work intetmutants,we

had no such security; since bacterial survival was high, the level ofbacterial nucleic acid and protein synthesis almost certainly was also high, and a number of informative biochemical experiments could notbe performed. However, once we had

learned that infection at high multiplicity in

chemically defined media leads to extensive cell

killing,

itseemedthat theseapproaches mightbe available after all. First, of course, we had to

establish that host cell gene expression is

ar-rested in the tet mutants in a fashion similar to

that seen inwild-type cells. As a prerequisite to

theseexperiments, weneededtoknow thatphage adsorb to tetcellsasrapidlyas toE.coli B. This

wasindeedthe case,atleast underourconditions of infection(5 X 108to8 X101cells/ml,in GCA

medium,andmultiplicities of6to10). With each

bacterialstrain, morethan 60% of added phage

was adsorbed by 2 min after infection andover

85%7,

by5min.

As afirstapproachtohost cell geneexpression,

we examinedthe ability of infected cells to

syn-thesize 3-galactosidase after induction by

iso-VOL.8, 1971 277

on November 11, 2019 by guest

http://jvi.asm.org/

[image:3.480.39.231.61.281.2]

MATHEWS AND HEWLETT

propylthiogalactoside (IPTG). The specific

ques-tionaskedisthefollowing:does everycell which

is killed (i.e., loses colony-forming ability) also

lose the capacity to synthesize

#3-galactosidase?

This was asked by infectingcultures at different

multiplicities. Six minutes after infection, samples

wereremoved for determination of viablecount. IPTG was added tothe remainder of eachculture

for a 12-min period, after which the cells were

harvested and assayed for

3-galactosidase

activity. Now, if every cell which is killed also loses the

capacityfor induced enzymesynthesis,we expect

a plot ofresidual enzyme induction versus

bac-terial survivaltobeastraight linewitha

450

slope.

However,ifsomekilled cells do notloseenzyme inducibility with the same kinetics, then we expect

significant upward deviation from a

450

line. As

seen inFig.4,thedatamostclosely approximated

450

linesforallthreestrains.Inotherexperiments

(Fig. 5), nodifferenceswere observedamong the

three strains in enzyme inducibility when IPTG

wasadded earlier ininfection, i.e.,at 3 min.Thus,

it appears that the kinetics of loss of induced

enzyme

synthesis

arethesamein themutants as

in the wild-type strain and that every killed cell

losesthecapacityfor bacterialenzymeinduction.

We have investigated the kinetics ofarrest of

host-specific RNA synthesis

indirectly by

fol-lowing the rate ofphage-specificRNA

synthesis

early in infection. RNA was

pulse-labeled

with

uracil-5-3H, either in uninfected cells or at

early

times after T6 infection. After

purification,

the

RNA

samples

werehybridized toT6DNA. The

RNA toDNA

input

ratio was

adjusted

1:20, a

condition

favorable

for

complete hybridization

of

all

RNA

species

which are

homologous

to

phage DNA (9). Thus, the percentage of

input

counts bound to phage DNA represents the

percentage of the total labeled RNA which is

phage-specific.Ourresults for T6-infectedE.coli

B(Fig. 6)aresimialrto those reportedbyKennell

(9) for T4-infected E. coli; a relatively

large

fraction of theRNAbeing synthesized aslateas

5 min after infection is not phage-specific and

hence is host-specific. The important point from Fig. 6 isthat the data for E.coil

B,

tet 1, and tet 2 areidentical. In otherwords,the rate of

replace-ment of host transcription by virus-directed

transcriptionis the same inall threestrains.

The kinetics of arrest of host cell DNA

syn-thesis were studied in similar fashion (Fig. 7).

Thymidine-methyl-3Hwas used tolabelDNA for

2-min periods either before or at the indicated

periodsafterinfection. Figure7A showsthe total

radioactivity incorporated into acid-insoluble

material during each labeling period. The

dif-ferences among the three strains in counts

in-corporated before infection may well represent

not truedifferences in rates of DNAsynthesisbut

rather differences in thymine nucleotide pool

sizesor someotherfactor. At any rate,thedata

on total counts incorporated during each

post-infection labeling period are quite similar,

sug-gesting that the rates ofhost DNA shutoff are

similar, ifnotidentical, inthe threestrains. The

upturn in rate oftotalthymidine incorporation

seen with each strain between 5 and 7 min

representsthe initiationof phage DNAsynthesis,

since DNAlabeled during thisperiodhybridizes

efficiently withT6 DNA(data not shown).

Hybridization

of DNA isolated from each

labeled cultureagainst unlabeled E. coilB DNA

100/1 10 10C

1/o Bacterial Survival

FIG. 4. Comparisonofcellkillingwithf-galactosidase inducibiliiy.Log-phase bacterial cultures at about 8 X 108 per mlinglycerol-CasaminoAcidsmedium were subdivided into 10-ml amounts, each of which was infected with

T6 over arangeof multiplicities between 0 and 5. At6min, samples were removed for determination of viable

count,andisopropylthiogalactosidewasadded to the remainder of each culture to a level of S X 10-4M. Twelve

minutesafter addition of inducer, each culture was chilledandcentrifuged. The pellets were resuspended, and

fl-galactosidase

wasassayedasdescribedinMaterials and Methods. Thefigure presents combined datafrom three separateexperimentsoneach bacterialstrain.

278 J. VIROL.

on November 11, 2019 by guest

http://jvi.asm.org/

[image:4.480.112.397.445.560.2]

T-EVEN-TOLERANT MUTANTS OF E. COLI. II

0

x

-I._

oracy act---, 0h- 0-0-.--O

t0-0

O.-U'.

-0 15 30 0 15 300 15 30

Minutes After Infect ion

FIG. 5. fi-Galactosidase induction after T6 infection. Log-phase bacterial cultures inglycerol-Casamino Acids medium were infected with about 6 T6particles per bacterium. Isopropyllhiogalactoside (IPTG) was added to 5 X 1O4 Xat3min,andsampleswereremovedatthe indicated times andassayedfor

f3-galactosidase

asdescribed inthelegendtoFig.4.Results are compared with control values from uninfected cultures.

B

tet1

tet

2

80

2

60-

r 0

0

40- F--0

0

20-0

2 4

0

2 4

0

2 4 6

[image:5.480.94.389.71.318.2]

Minutes Af

ter

Infection

FIG. 6. T6-specificRNAsynthesis. Bacterial cultures (100-ml)weregrown in glycerol-Casamino Acids (GCA) mediumplus20figofuracilper ml. At aturbidity corresponding to about8X 108cells/ml, each culture was cen

-trifuged, washed in GCA medium, and resuspended in 100 ml offresh GCA medium containing nouracil.

Pulse-labelingwascarriedoutfor 1.5-minperiods with50MCieach ofuracil-5-3H onuninfected cellsand T6-infected cultures (multiplicity of infection = 6),beginning at 2.5min and at 4 min after infection. RNA was isolatedas describedin MaterialsandMethods.Labeled RNA samples containing 2.5Mgeach were hybridizedagainst 50lsg eachofT6 DNA immobilizedonmembranefilters.Eachpointinthefigure correspondstothemidpointofalabeling

period. The ordinate denotes the percentage of labeled RNA which is phage-specific in each preparation.

279 VOL. 8, 1971

on November 11, 2019 by guest

http://jvi.asm.org/

[image:5.480.96.387.383.555.2]

MATHEWS AND HEWLETT J. VIROL.

[image:6.480.99.387.70.251.2]

-1 1 3 5 7 -1 1 3 5 7

FIG. 7.Incorporation ofthymidine-methyl-H intoE. coliDNA. Bacterial cultures(20 ml) wereinfectedwith

T6atamultiplicity of about6. At various times, 1-ml sampleswereremovedtosmallincubation tubescontaining

10 MACieachof 3H-thymidiiie. After 2minof labeling,metabolism washaltedbyadditionof

cyanide-ethylenedi-aminetetraaceticacid-lysozyme lysing fluid (14). Lysiswascarriedoutbyahigh-temperature procedure (14).Each

lysate wasextracted twice in succession with water-saturatedphenol, and the solutions weredialyzed overnight

against 0.02Vsodiumphosphate buffer, pH6.8. A 0.5-mlamountofeachpreparation wasmixed with0.5 mlof

I NNaOH, and allpreparationswereheatedinaboiling-waterbathfor7min. Eachpreparation wasneutralized

andmadeupto3.0 mlil 3X SSC (SSC = 0.15 mNaCl plus0.015-fsodiumcitrate). Duplicate0.5-mnlsamples

ofeachwerehybridizedagainstmembranefilters containing20jugeachofE. co/i B DNA.Hybridizedcountswere

corrected forthe small numberofcountsboundtoblankfilters. (A) Totalcountsincorporatedintoacid-insoluble

material. Eachpointon thegraph represents the midpoint ofapulse-labelingperiod. (B) Hybridizable counts.

TheseareplottedaspercentofE.coliDNA-hybridizablecountsrelativetothelabeled unionfected cultures.Again,

each pointrepresentsthemidpoint ofalabeling period.

providesmorespecificinformationontherelative

ratesof hostDNA shutoff (Fig. 7B).These data

make it appear that host DNA synthesis is

actually turned off faster in infection of the tet

mutantsthaninnormal infection. Thedifferences

are more apparent than real, since the E. coli

DNA-hybridizablecountsineachlabeling period

areplotted relative to the counts incorporatedin the uninfected cultures. Thus, the apparent

differences derive in part from the differencesin

total counts per minute incorporated amongthe three uninfected cultures. However, it does ap-pearthat host cell DNAsynthesis is arrested at

least as fast in the tet mutants as in the parent strain,ifnotfaster. Ofcourse,quantitativestudies

of DNA metabolism mustawait the isolation of thymine auxotrophsof the tetmutants, andthese

are currently being preparedinourlaboratory.

DNA metabolism. Our earliest work onDNA

metabolism in T6 infection of tet mutants

in-volved studies on the fate of labeled parental

phage DNA. We find (Table 1) that parental DNA doesbecome attached tothe "M band," a

complexdescribed by Earhartand his colleagues (5, 6), which apparently contains replicating

DNA and the membrane material present atthe

TABLE 1. Incorporation of pareiiial T6 DNA ilto

Mbandsa

Bacteria Totalcounts/min Percentoftotal Bacteria Tot~nl

zin

Mband parentalcounts/min

inMband E.coli B... 1.48 X 104 25.1

E. colitet 1.... 1.78 X 104 30.2 E. coli tet 2.... 1.47 X 104 24.8

aBacterialcultures(5.0ml) growing in tryptone

broth mediumwereinfectedatamultiplicityclose to 5.0with T6 containing 14C-labeled DNA. At 7 min after infection, the cells wereharvested and M bands were prepared and visualized on

mag-nesium-sarkosylgradientsasdescribed by Earhart (5). EachM bandwas removed from itsgradient

centrifuge tubebylowering the collection tube of a Buchler gradient collector to the level of the band andpumpingoutthewhite, viscous material

by suction.

replicating site. This, of course, isnotunexpected, since wehad earlier found (12) that tetmutants

support T6DNAreplication.

Inourprevious study, wefound that parental

phage DNA is not degraded to acid-soluble

280

Iq₀

x

E

0

on November 11, 2019 by guest

http://jvi.asm.org/

[image:6.480.258.447.386.476.2]

T-EVEN-TOLERANT MUTANTS OF E. COLI. II

material in infection of tet cells. In the present 4 Minutes

Alkaline

Gradients study, we asked whether any breakdown is

de-tectable. Again, parental DNA was labeled by tet 2

growth of phage in the presence of 14C-uracil (10). 10 Cells were harvested at 4minand lysed by a

low-shear, low-temperature

process

(7, 15);

the

lysates

were subjected to sucrose gradient analysis in the 5

presence of

32P-labeled

mature T6DNA present as a sedimentation marker. At 4

min,

no break- 0 down is evident, as shown by the fact that the t t

sedimentation profiles of marker and parental et DNA are virtually identical in neutral gradients 10

(Fig. 8). The same results are seen when portions Nc

of the same lysates are sedimented through alka- °

line sucrose gradients (Fig. 9); parental and x 5. I

marker profiles are identical. Thus, within the E o **-.

first 4

min

of infection, parental phage DNA is a 0 -0

neither broken nor subjected to significant B

endonucleolytic cleavage. 10

4 Minutes NeutraI Gradients

tet 2

10- 0

-.__.0

b0to0o,

'Q

10 20 3C

Fraction Number

[image:7.480.242.433.60.319.2]

5 /

FIG. 9. Alkaline sucrose gradientcentrifugation of Hi\_ *

parental

T6 DNA at 4 min

after

infection.

Samples

0 ofthe lysates preparedfor the experiment described

in Fig. 8 were adjusted to0.3Xy with NaOH, and 0.2-tet 1 mlportions were layered onto 5.0-ml alkaline sucrose

10 gradients (14), which were centrifuged at 40,000 rev/

0/ min for 60 min. Symbols: Q, 32p marker colts per

a minute; *, 14C parental counts per minute.

x5 / \

E / \t Analysis of the fate of parental DNA later in

Q

i

nfection

reveals

differences between

the wild

and

B mutant strains. Asshown by Frankel (7), much of

the parental DNAduringperiods ofactive DNA

10

replication

ininfection

by

T4 ispresentinaform

which sediments more

rapidly

than does mature

DNA and which contains individual DNA

i \ strands longer than those of mature DNA. As

J

v*

**- ;* shown in Fig. 10, much of the parentalDNA at 0>&e~e...*g t 0 13 min afterT6 infection ofE. coli B sediments

10 20 30 through neutral gradients heterogeneously but

Fraction Number more rapidly than mature

T6

DNA. Infected tet FIG. 8. Neutral sucrose gradient centrifugation of cells also form rapidly sedimenting intermediates parental T6 DNA at 4 miii after infection. Bacterial containingparentalDNA. However, thismaterial

cultures (about 8 X 108 per ml) were infected with does not sediment as rapidly as that seen in

in-T6 phage containing '4C-labeled DNA (15). At 4 fection ofE.

coli

B,

and a smaller proportion of

min,

cells were lysedby a low-temperature procedure the total parental DNA sediments

rapidly.

The

(7) in the presence of 32P-labeled T6 DNA added as a

sedimentation marker. Portions (0.2 ml) were layered

rue

anitue

ofthe ifferenebetwendtheswil

onto 5.0-ml neutral 5 to20% sucrose gradients, which strain and the mutants is probably

underesti-were then centrifuged at 32,000 rev/min for 35 min. mated in this figure. Recovery of input parental Symbols: Q, 32P marker counts per minute;

*,

14C countsin the tet 1 and tet2 gradients was 87 and

parental counts per minute.

96%,

respectively, whereas that in the E. coli B

VOL.8, 1971

281

on November 11, 2019 by guest

http://jvi.asm.org/

[image:7.480.36.226.265.527.2]

MATHEWS AND HEWLETT

2. 0 15

)~~~~~'

LoDoooo0

oo@

03-0-0-0-0 0- 00-0Di0--0I0:~ ie ~0

0 10 20 30

[image:8.480.55.245.59.298.2]

Fr2ction Number

FiG. 10. Neutral sucrose gradient centrifugation of parental T6 DNA at 13 min after infection. Except for thetime ofinfection, theprotocol wasidenticalto

thatdescribedinthelegendtoFig. 8.

gradientwasonly78 %. It isprobableinthelatter case thatconsiderable parentalDNA sedimented

to the bottom of the tube and was lost in this

gradient.

When the above lysates, obtained at 13 min,

wereanalyzedinalkaline gradients,an even more

strikingdifferencewas seen.As shown inFig. 11,

parental DNA strands in infection of E. coli B sedimentheterogeneously,butmostof the strands sediment asthough they are longerthanmature

DNA strands. By contrast, virtually no

long-stranded DNA was seen in infection of the tet

mutants. This implies that parental DNA is

subjectto considerable endonucleolytic breakage

or nicking. Whether this is a direct or indirect consequence ofthe tet alteration remains to be

determined.

Oncewehadestablished thatarrestof host cell

gene expression occurs normally in T6 infection

of the tet strains, it became possible to design isotope incorporation experiments for measuring

relative rates of nucleic acid and protein syn-thesis. When cells are grown in the presence of

high levels of uracil, such that they arerepressed

in the pyrimidine biosynthetic pathway, it

be-comes possible to use 14C-uracil for following

accumulation ofboth DNA and RNA (10, 11).

When such an experiment was performed (Fig.

0 0 2C 30

[image:8.480.259.452.59.277.2]

Fraction Number

FIG. 11. Alkaline sucrose gradient centrifugation of parental T6 DNA at 13 mill after infection. Except forthe time ofinfection, the protocol was identical to

thatdescribedinthelegend to Fig. 9.

12), we found, somewhat to our surprise, that

the ratesofaccumulationofisotopeinboth DNA

andRNAwere

comparable

ininfectedtetcellsto

thoseseeninnormal infection. Similarly,ratesof

incorporation of

'4C-leucine

into protein reveal no gross defects inprotein synthesis ininfection

of the mutants (Fig. 13). Similar results for T4

infection of thesemutantshave been observed by

A. Kaji (personal communication). We have

relatively little information on the synthesis of

specificphageproteins. However, itis ofinterest

that infected tet cultures incubated for several

hoursultimately undergo partial

lysis.

Thus, even

though the synthesis ofphage

lysozyme

is

sub-normal(12), someexpressionof thislategenedoes

occur.Wehavenot yet

investigated

the

synthesis

ofotherspecific late phage

proteins.

RNA synthesis. When tet mutants were

origi-nally isolated, an attractivehypothesis toexplain

their

inability

tosupportphage

replication

was an

alteration in the enzymatic

machinery

for

tran-scription: perhaps an alteration in RNA

poly-merase which prevented the replacement of the

host cell sigma factor by comparable phage

factors (17,18).Ifthisweretrue,thenonewould

expectto seeafailureofinfectedtetcellstocarry

outthenormalprogram ofphagegene

transcrip-tion (forreview, seereference 13).Wehave used

competitive DNA-RNA hybridization to try to

detect this. To ask whether any abnormalities

282 J. VIROL.

on November 11, 2019 by guest

http://jvi.asm.org/

T-EVEN-TOLERANT MUTANTS OF E. COLL II2

0

6

0

x

LT)

4-,

.E_

z

4-,-C:

j

(13

z

0

x

cE

0.

E

z :3

E

U3

u u

0 20 40 60 0 20 40 60

[image:9.480.90.380.62.291.2]

Minutes After Infection

FIG. 12.Accumulation ofphageDNA and RNA. Bacterial cultures (20ml) inglycerol-CasaminoAcids medium were grown inthe presenceof20,ugofuracilperml, centrifuged, resuspendedinfresh medium,andinfectedinthe presenceof14C-uracil, asdescribedpreviously (10, 11). Removalof samples,determination oftheradioactivity incorporatedintoDNA andRNA,and calculationoftheradioactivityin DNAasphage-equivalentunitsojDNA permilliliterofculture were alsoas describedpreviously (11). Note thedifferentordinatesforthe twopanels.

exist in transcription of early phage genes, we

hybridized earlypulse-labeled (2to5minat37C)

RNAfromT6-infected E. coli B toT6 DNA in

the presence of unlabeled early RNA species prepared from cultures of E. coli B, tet 1, and tet2 harvested 5 min after infection by T6. As

showninFig. 14, both the initial slopesand the

final plateau values of all threecompetitioncurves are similar. Hence, we conclude that at 5 min infected tet cells contain the same messenger RNA (mRNA) species, and in thesamerelative abundance, asthose present 5min afterinfection

ofwild-type bacteria.

The ability of infected tet cultures to form

mRNA transcripts normally synthesized late in

infection was tested by using unlabeled 20-min

RNA preparations as competitors against T6-B

RNAlabeled from 15to20min. Theleftpanel of Fig. 15 shows competition curves for unlabeled

5-and20-min RNA.Thismerely confirmsforT6

what has long beenknown for T2 andT4 (1, 8, 11), namely, that there are somemRNA species

whicharesynthesized late butnotearly in infec-tion. If the tetmutantswereunableto carryout part or all of the transitions involved between earlyand latepatternsoftranscription,onemight

expectcompetitioncurves withlate unlabeled tet

RNA to resemble those seen with normal early

RNA. However, as shown inFig. 15, the

com-petitioncurves for late infectedtet RNA species

are

virtually identical

tothecorresponding curve

for late T6-B RNA. Thus, within the limits of

resolution of this technique, we can see no

dif-ference intheability of infectedtet cellstocarry

out the normalprogram ofviral gene

transcrip-tion.Itis

possible

thatone ora small number of

critical lategenes are not transcribedin infected

tet

cells,

but no gross defects in transcription

patternsexist.

DISCUSSION

The genetic alteration in tet mutants which

prevents phage replication seems to affect a

relatively late viral functionin view of the large

number ofphage-directedprocesses which occur

ininfection oftetcells.Theseinclude adsorption,

cell killing,

arrest of host cell nucleic acid and

protein synthesis, early changes in phospholipid

metabolism (L. I. Pizer,personalcommunication),

initiation of DNA replication, formation of

rapidly sedimenting DNA intermediates,

ac-cumulation of newly synthesized nucleic acidand

protein, formation of specific early enzymes,

formation of

early

and late mRNA species in

normal amounts, and partial lysis of infected

cultures. Thusfar, thechiefdistinctive properties

of tet cells relative to wild-type include

slightly

slower growthrate in defined media(12),relative

283 VOL. 8,1971

on November 11, 2019 by guest

http://jvi.asm.org/

MATHEWS AND HEWLETT

, resistance to ultraviolet light (12), subnormal

o levels of both early and late enzymes

(12),

AB extensiveendonucleolytic cleavageof intracellular

otet1 0

DNA, high

survival to infection in solid

media,

3-

e*

tet 2 and, of course, failure of infective progeny

phage

to be formed under any conditions yet devised.

An additional difference relates to the ability to

adsorb T4. When the mutants were

originally

a o/ isolated, both tet 1 and tet 2 could adsorb T4,

///-

although

more

slowly

and to a lesser extentthan

2 the

adsorption

of either T2 orT6.

However,

with

increasing time of cultivation in the laboratory,

/./ both strains gradually lost theabilitytoefficiently

0 adsorbT4,although T2 and T6areboth adsorbed

0 at rates

comparable

tothose seen with

wild-type

X 0 bacteria. Whether this is a result of the

primary

E tet mutation or some

secondary mutation

we do

E1 // not know. We are presently tryingto

isolate

new

tet derivatives capable ofadsorbingT4.

Westill have verylittleidea of thebiochemical difference between

wild-type

and tet bacteria. A

* difference in membrane structure and function

seems an attractive

possibility,

first,

because so

many phage-directed processes, including tail

o 20 40 60 morphogenesis (16), occur at membrane

attach-Minutes After Infection ment sites and, secondly, because bacterial cells

FIG.

13. Incorporation of

14C-leucine

into protein with membrane

defects

can lose the ability to after T6 infection. Bacteria were grown in M9 medium

support

T-even

phage

growth

(2).

We

plan

to containing leucine, centrifuged, washed, resuspended

analyze

further intracellular

phage

DNA

inter-in fresh medium, inter-infected with T6, and labeled with mediates, since the extensive endonucleolytic

'4C-leucine as described previously (10). Determina- cleavageofintracellularparentalphage DNAmay

tion of the amount ofradioactivity incorporated into

protein was also as described previously (10). The po.int t

ards

a

efectininA

met

asnthe

higher rate of incorporation seen in tetI relative toB primary basis for explaining the let phenotype.

was noted in three separateexperiments. Finally, we planto analyze the process ofphage

Competition vs. 2-5' Labeled T6-B RNA

10 'B teti tet2

d~o

0'\lo

0

*280

L

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

C

E40-

_C1 °

~~~~~~~0

0

a.

00

00 0

3'20 0

0 10 20 30 0 10 20 30 0 10 20 30 40

[image:10.480.60.251.63.323.2]

Unlabeled/Labeled RNA Ratio

FIG. 14. CompetitiveDNA-RNA hybridization: early RNA. Pulse-labeledRNA waspreparedasdescribedin

the legendtoFig.6, withlabelingcarriedoutfrom 2to5 min after infection ofE.coliBwith T6. UnlabeledRNA

samplesforcompetitioncurveswerepreparedfrom cells harvestedatS min afterinfection of E. coliB,tet1,ortet

2 with T6.Hybridizationmixtures contained5sAgeachofT6 DNAimmobilizedonfilters, either2.5or5,ugeach of labeledearly T6-BRNA,andamountsofunlabeled RNAspecies sufficienttogivethe indicatedratiosof

com-peting(unlabeled) to labeledRNA. Thefigurecontainscombineddata fromtwo separateexperiments.

284

J. VIROL.

on November 11, 2019 by guest

http://jvi.asm.org/

[image:10.480.114.401.422.574.2]

T-EVEN-TOLERANT MUTANTS OF E. COLI. IL

Competition vs. 15-20' Labeled T6-B RNA

10O B

D 80

>I

ox 95 RNA

a*40 O

-0 0

20 \ 2o' RNA

o .

0 20 40 60 0

\ 20' RNA

of~~

20 40 60 0

2'RA

20 40 60

Unlabeled/Labeled RNA Ratio

FIG. 15. Competitive DNA-RNA hybridizations: late RNA. The experimentalprotocol was similar to that

described in thelegendtoFig.14,exceptthatlabeledRNA wasobtainedinaperiodbetween 15and20min after

infectionat37 C,and unlabeled competitorRNAspecies wereisolatedfrom cellsharvestedat20min.Acontrol hybridization withunlabeled5-min T6-B RNA isalsopresentedinthe left panel. Thisfigurepresents combined datafromtwoseparateexperiments.

morphogenesis, since the high rates of phage nucleic acid and protein synthesis seen

through-out, plus the normal operation of the transcrip-tionprogram, suggest that the tet phenotype, at

least in liquid

media,

is expressed at a very late

stage. At present, we have no idea why tet cells

survive infection under some conditions but are

efficiently killed under others. Further investiga-tionof thispointwill bedeferreduntilwehavea

better understanding of why these cells are so

totally unabletoform infectivephageprogeny. ACKNOWLEDGMENTS

This investigation was supported by Public Health Service researchgrant AI-08230from theNational Institute of Allergy and Infectious Diseases and research grant 70 1013 from the American HeartAssociation.

We thank MildredCeizyk and Celta Lacambra for capable technical assistance.

LITERATURE CITED

1. Bolle, A., R. H. Epstein, W. Salser, and E.P.Geiduschek. 1968. Transcription during bacteriophage T4 development: synthesisandrelativestabilityofearlyand late RNA. J.

Mol.Biol.31:325-348.

2. Cronan,J.E., Jr., andP.R.Vagelos.1971.Abortive infection by phage T4 under conditions of defective host membrane lipid biosynthesis.Virology43:412-421.

3. Denhardt,D. T. 1966. Amembrane-filter technique for the detection ofcomplementary DNA. Biochem. Biophys. Res. Commun.23:641-646.

4. Duckworth,D.H., and M. J. Bessman. 1965. Assay for the killing properties of T2 bacteriophage and their "ghosts." J. Bacteriol. 90:724-728.

5. Earhart, C.F.1970.Theassociationof host andphageDNA

with the membrane ofEschertchiacoli. Virology42:429-436.

6. Earhart, C. F., G. Y. Tremblay, M. J. Daniels, and M.

Schaechter. 1968. DNA replication studied byanewmethod fortheisolation of cell membrane-DNAcomplexes. Cold Spring Harbor Symp. Quant. Biol.33:707-710.

7. Frankel, F. R. 1968. DNA replication after T4 infection. Cold Spring Harbor Symp. Quant.Biol. 33:485-493.

8. Hall, B. D.,A.P.Nygaard, and M. H. Green. 1964. Control ofT2-specificRNAsynthesis. J. Mol. Biol.9:143-153. 9. Kennell, D. 1968. Inhibition of host protein synthesis during

infection ofEscherichiacolt by bacteriophage T4.I.

Con-tinuedsynthesisof host RNA. J. Virol.2:1262-1271. 10. Mathews, C. K. 1966. DNA metabolism and virus-induced

enzyme synthesis in a thymine-requiring bacterium

in-fectedbyathymine-requiringbacteriophage. Biochemistry 5:2092-2100.

11. Mathews, C. K. 1968. Biochemistry ofDNA-defective amber

mutantsofbacteriophage T4.I.RNAmetabolism. J. Biol.

Chem. 243:5610-5615.

12. Mathews, C. K. 1970. T-Evenbacteriophage-tolerantmutants

ofEscherichta colt B. I. Isolation and preliminary char-acterization.J. Virol. 6:163-168.

13. Mathews, C. K. 1971. Bacteriophage biochemistry. American Chemical Society Monograph No. 166. Van Nostrand-ReinholdBooks, NewYork.

14. Murray, R. E., and C. K.Mathews. 1969. Addition of nucleo-tidestoparentalDNAearlyin infectionby bacteriophage T4.J. Mol. Biol. 44:233-248.

15. Murray, R. E.,and C. K. Mathews. 1969. Biochemistry of DNA-defective amber mutants ofbacteriophage T4. II.

Intracellular DNA formsininfectionbygene44mutants.

J. Mol.Biol. 44:249-262.

16. Simon, L. D. 1969. The infection of Escherichtacoltby T2 and T4bacteriophagesasseeninthe electronmicroscope.III.

Membrane-associated intracellular bacteriophages. Virology 38:285-296.

17. Travers, A.A. 1969.Bacteriophage sigma factor for RNA polymerase.Nature(London)223:1107-1110.

18. Travers, A. A. 1970. Positive controloftranscription bya

bacteriophage sigma factor. Nature (London)

225:1009-1016.

VOL. 8, 1971

285

tet1l tet 2.

on November 11, 2019 by guest

http://jvi.asm.org/

[image:11.480.92.381.67.245.2]