Effect of Host Cell Wall Material on the Adsorbability of Cofactor-requiring T4

Effect of Host Cell Wall Material

on

the

Adsorbability

of

Cofactor-requiring

T41

DENNIS T. BROWN2 AND THOMAS F. ANDERSON

InstituteforCancerResearch, Philadelphia, Pennsylvania 19111

Receivedforpublication 7March 1969

Theadsorbability of T4onhostcells wasdetermined asafunction of timeafter

their liberation from infectedcells. Freshlyliberated (nascent) particles are readily adsorbed but lose theiradsorbabilitywithahalf-time of about 2daysat5C,butonly about 20minat37 C.Theycanbe made adsorbableagain withana-amino acid co-factor likeL-tryptophan, and thisstateofadsorbabilitycanbe stabilizedbycellwall material from Escherichiacoli.Such stabilized particles lose theiradsorbabilityata ratesimilartothatatwhichnascentparticles lose theirs. Mostfreshlyliberated parti-clesareobservedbymeansof electronmicroscopytohave "debris" attachedtotheir

baseplatesandtohavemostof theirsix,long tail fibersfree,whereas "old"particles

that have lost theiradsorbability appear relatively "clean" with mostof their tail fiberswrappedaround their sheaths. Nascentparticleshave densities thatarelower

than those of old particles. The material responsiblefornascentadsorbabilityseems tobeafragmentofthe host's cellwall,fornascentadsorbabilityisdestroyed by

lyso-zyme.Furthermore, nascentT4particles liberated from hostcells withradioactively labeledwallscarrythe label indensity gradientsbut lose itastheyloseadsorbability. Inaddition, onlyasmallproportion of particlesliberated from infectedspheroplasts

arenascentlyadsorbable, whereas mostparticlesliberated from intact cellsare ad-sorbable.

Bacteriophage particles in stocks of certain

strains of T4 and T6 are incapableofadsorbing to host bacteria unless they have been activated

byanadsorption cofactor (2, 3, 4). L-Tryptophan

wasfoundtobe themosteffectivecofactor, being capable ofinducingnearlymaximaladsorbability at concentrations of 2 x 10-6 g/ml.The

activa-tion of T4bycofactor isreversible;after the co-factor is dilutedto nonactivating concentrations, 99 % of tryptophan-induced adsorbability is

lost in 3to 5min.

It has been reported (6, 12, 15, 16) that the effect of the cofactor is to interact with the T4 particle in such a way as to induce a

configura-tional change in its tail fiber apparatus. In the absence of cofactor, cofactor-requiring T4 particles are often observed by means of

elec-tron microscopy to have most of their six tail

fibers lying in close association with the sheath and collar regions of the phage structure. The

same T4 observed in the presence of activating concentrations of tryptophan have mostof their

IPresented inpartby Dennis T. Brown in partial fulfillment oftherequirements foraPh.D. degreeattheUniversityof

Pennsyl-vania.

2Present address: Department ofBiological Sciences,

Dart-mouthCollege, Hanover,N.H. 03775.

tail fibers inanextendedstate,free of thesheath and available for interaction with the host-cell

receptors. It is thought that these changes in tail fiber arrangement can account for the observed change in therateofsedimentationfrom

S2,,

=

1,100 for inactive particles that are streamlined

with their tail fibers wrapped around the tailto

S20,

= 850 for tryptophan-activated particles

whose extended tail fibersexertafrictional drag as they move through the medium. Cummings et al. (7) suggested thatalterationsin thesize of the phage head may also contribute to changes inthe rates ofsedimentation.

Gamow and Kozloff (12) postulated that the function of tryptophan in activating cofactor-requiring phage is to combine with a folic

acid-like residue located in the baseplate of the T4

particle. This interaction results in a

configura-tional change in a postulated hinge structure located in the baseplate, thus changing the

orientation of the phage tail fibers.

Even though cofactor-requiring T4 particles were found to be incapable of forming plaques

when plated directly with Escherichia coli strain B on cofactor-free agar, it was noted (3) that

host bacteria that had previously been infected

withcofactor-requiringT4could formplaqueson

94

on November 11, 2019 by guest

http://jvi.asm.org/

CELL WALL MATERIAL

ANDWADSORBABILITY

OF T4

bacteria growingon suchmedia. Since the

produc-tion ofa plaque involves many rounds of

infec-tion, reproducinfec-tion, and liberation of phage, it

was thought that the daughter particles must

somehow have been infectious. However, when the progeny phage produced in such plaques were

isolatedandtestedbyplating in the presence and

in the absence ofcofactor, they were found to requirecofactor for adsorption justas the parent

particlehad. Anderson (3) attributed thisrather

surprising resulttothe possibilitythatthe progeny

phage had obtained cofactor from the host cells growing on synthetic agar.

In 1952, Wollman and Stent (25) named this state of adsorbability possessed by the newly

released T4 "nascentactivity" andtermed phage

which possessed thisactivity "nascent" phageas

opposed to "quiescent" or cofactor-requiring

phage. Theydemonstrated that nascent

adsorba-bility was due to an active state which was lost

much more slowly on dilution than

tryptophan-inducedadsorbability andsuggestedthat nascent

activity might be due to a tryptophan-like

co-factor (perhaps tryptophan itself) binding to a

special nascent surface possessed by the

newly

released T4 particles. This nascent surface was

presumed tobind cofactor morefirmly than the

cofactorsitesonquiescent phage,thusaccounting

for the slow loss of nascent adsorbability. The

special surface was supposed to be lost more

slowly than the

tryptophan-like

cofactor and

was

presumed

not tobelost atall so

long

as

co-factorwasboundtoit.

In1956, Jerne(14) discoveredthatthe

adsorba-bility of

tryptophan-activated

T4 could be

stabilized

by

"A8 serum." This was serum that

had beencollected fromahorseonly8days after

the animal had received an

injection

of T4.

Although thisearlyserumhad notdeveloped an

appreciable

titer of

neutralizing

antibodies, it

possessedtheremarkablepropertyinthepresence

oftryptophan of

rendering

T4 adsorbable toits

host after the tryptophan had been removed.

Jerne's results

opened

up another

possibility:

nascentadsorbability might be dueto asubstance

acquired during its latent

period

that stabilizes

the

phage

in itsadsorbablestate.

The

results

reportedinthis

study

indicatethat

the latter

possibility

is correct in a sense.

Nas-cently adsorbableT4 has alowerdensityin CsCl

than do quiescent

particles,

which suggests that thesubstance hasadensitythatis lower than that

ofquiescent particles and a volumethat is

rela-tively large. When viewed

by

means ofelectron

microscopy, quiescent

particles

appear "clean"

with their

long

tail fibers

wrapped

around their

sheaths, whereas

nascently

adsorbable

particles

are observed to have

particles

of "debris"

at-tachedto their baseplateregionswith most of their

long tail fibers free. A number of observations

indicate that the material that is responsible for

nascent adsorbability is derived from fragments ofthe host'scellwall. (i) Nascentadsorbabilityis destroyed by lysozyme. (ii) The adsorbability of tryptophan-activated T4 can be partially stabilized by a cellwallextract from E.coli.(iii) T4 particles that have been freshly liberated from host cells whose cell walls have been labeled with 14C-labeled glucosamine containthis label but lose it as they spontaneously lose adsorbability. (iv) Whereas most of the particles liberated from

whole cellshave nascentadsorbability,very fewof

those liberated from spheroplasts whose cell walls have been largely destroyed by lysozyme

before lysis possess this trait.

MATERIALS AND METHODS

Bacteria and bacteriophage. E. coliB has been

de-scribed (2). The bacteriophage T4 used throughout this study was obtained by isolating a single clear (r-) plaque from Anderson's stock labeled T4 5/10/45 (2).

General methods. The methods used were mainly

those described by Adams (1). F medium contained per liter of distilled water: NH4Cl, 1 g; KH2PO4,

1.5 g; Na2HPO4, 3.5 g; sodium lactate, 10.0 ml of a 60% syrup (Sigma Chemical Co., St. Louis,Mo.);and MgSO4, 0.01 g (added separately). N broth contained perliter ofdistilled water:Bactotryptone, 10 g; yeast extract,5g;NaCl, 10 g; and glucose, 1 g. The pH was

adjustedto 7.0 to 7.2with 1 N NaOH. Agar was added totheseliquid media at final concentrations of 1.0 and 0.6%, respectively,for plates and overlay. To prepare

hightiter stocks, 3 XD medium was usedasdescribed by Fraser (10).

Assaysweremade on N-agartodetermine thetotal titerofasuspensionofcofactor-requiringT4.Assays

onF-agar wereusedto count theparticlesthat were

capableof adsorbingtobacteria in the absence of

co-factor. Thefraction(R)of thephagepopulation capa-ble ofadsorbingin theabsenceof cofactorisexpressed astheratio of the number (F) of plaques obtainedon

F-agar to the number (N) of plaques obtained on

N-agar.

Preparation of E. coli extracts. Crude E. coli

ex-tractswerepreparedfrom 64Fhard-agarplatesthat

had been seeded with E. coli Bandincubated over-night at 37C. The resulting bacterial lawn was har-vestedinto 160 mlofFmedium. The cellsuspension wasthenpelletedinaclinicalcentrifugeatabout5,000

X g for 15 min andresuspendedina10-ml volumeof

Fmedium. This very thicksuspensionwasthentreated threeorfour times inaFrenchpressurecell(American

Instrument Co., Inc., Silver Spring, Md.) at 25,000

psi,being carefultokeepthetemperaturesof both the cell and the treated material below4C. Theresulting

pastewasthentreated withadropletof chloroformto

kill any survivingbacteria. These chloroform-treated cellwallfragmentshad littleabilitytokillT4particles

added later. Theywerequicklyfrozen at -78Cand storedat -20 Cfor later

use-VOL. 4,1969

95

on November 11, 2019 by guest

http://jvi.asm.org/

Electronmicroscopy. Themethods used were gener-ally those of Anderson (5). Preparations were nega-tively stained with sodiumsilicotungstate atpH 7 to increase contrast andresolution. All preparations were studiedbyusingaSiemensElmiskopIelectron micro-scope.

RESULTS

Preparation of nascently adsorbable T4. Woll-man and Stent (25) demonstrated that their phage T4-38 exhibited a nascent state of in-fectivity byassayingthe yieldofphageproduced in a one-step growth experiment and comparing the titers obtained on F and N agar. We have

repeated this experiment with results that agree

with those of Wollman and Stent. During the first 25 min of the latent period before the in-fected cellsbegin to lyse, the assaysonFand N

agar are identical, for each infected bacterium

forms a plaque with equal efficiency on the two

media. After 25 min, however, when the cells

begin to lyse both titers increase, but the titers

on Fagarbegintofall below thoseonNby

ever-increasing amounts; thefreshly liberated particles

slowly lose theabilitytoformplaqueson Fagar

until after 2 hr only 1 particle in 1,000 forms a

plaque on F. If suchlysatesareallowedtostand

at room temperature for a day, the value of R (= Assayon Fagar/AssayonNagar) fallstoas

lowas5 X 10-6.

To study effectively the problem of nascent

adsorbability, itwasnecessary tohavereadily at

hand a stock suspension ofphage known to be

nascently active. Thiswasaccomplished by

cool-ing to0Casuspension ofinfected bacteria inF

medium after only 30 min of aeration.

Chloro-form was then added to induce premature

lysis.

A trace of deoxyribonuclease was also added to

decompose the liberated deoxyribonucleic acid

(DNA) and soreduce the viscosity. After

stand-ingat0Cfor 8to 10min,the titerson Fand N

gaveR values between 0.6 and 1.0. Itwasfound thatthenascent

adsorbability

of thephagein such lysates couldbe preserved indefinitely by adding glycerolto afinalconcentration of 5%and quick-freezing small samples in an acetone-dry-ice bath at -78 C. After storage at -40C for months, these lysates showed no measurable

de-creasein R,although someloss in total titerwas observed. Therefore,these preparations provided

areadilyavailable sourceofnascentlyadsorbable

phagefor the studies that followed.

Factors affecting the rate of loss of nascent adsorbability. The initial observation that the nascent stateofadsorbability is lost more slowly

thantryptophan-induced adsorbability prompted

astudyofconditions whichmightbe expected to

change this observed rate. A number of factors

hadnoobservableeffect. The rate wasessentially constant between pH 6 and 9 and in NaCl concentrations between 4 M and 0 (distilled water);nordid thepresence orabsenceofoxygen

orofvisible light affect the rate.

The experiment of Fig. 1 shows that tryptophan (and the other compounds and known cofactors such as phenylalanine and tyrosine in N) have no effect on the loss of nascent adsorbability; nor does the reactivation by tryptophan of a phage preparation which has lost part of its nascent adsorbability affect the rate of loss of adsorba-bility. The reactivated phage return to a state of

adsorbability equal to that of the control and

quickly lose their tryptophan-induced adsorba-bility when the tryptophan is removed. These

0.1

R

0.01

2 TIME (hours)

FIG. 1. Effect oftryptophan andNmediumon the

rate of loss of nascent adsorbability of T4. Samples (0.1 ml) ofafreshly thawednascentphagepreparation

wereaddedtoeachoffour tubescontaining:tube 1, 10 mlofFmedium (a); tube 2, 10mlofFmedium

con-taining 100 ,ug ofL tryptophan/ml (A); tube 3, 10

ml ofNbroth (O); and tube 4, 10 ml ofFmedium

(0). Allsamples wereincubatedat 37 C. At various

times, samples oftubes 1, 2, and 3 were dilutedand

plated with B on F and N. The three samples lost

nascent adsorbability at essentially identical rates,

in--dicating that neither tryptophan alone nor theother

components in N medium affect the rate of loss of

nascenttadsorbability.Thesampleintube 4wasallowed partiallytodeactivateforIhr, at which time tryptophan

wasaddedto makethefinalconcentration 100,ug/ml.

After3 min, thesamplewasdiluted to anonactivating

concentration of tryptophan (0.01 ,ug/ml). After an

additional 5 mintoallowthephage tolose its

trypto-phan-induced activation, the dilutedpreparation was

assayedonFandN.Hereagain,theRvaluesare

iden-ticaltothoseinthe other three tubes.

on November 11, 2019 by guest

http://jvi.asm.org/

[image:3.491.271.456.248.440.2]

WALL MATERIAL AND ADSORBABILITY

resultsdisagree with those of Woilman and Stent

(25) and willbe discussed later.

Temperature did affect the rate of loss of nascentadsorbability as shown in the experiment ofFig. 2 in which the ratio R is plotted against the time ofincubation at different temperatures. These results are in good agreement with the

data reported by Wollman and Stent (25) and

approachfirst-order kinetics for the initial period asthoughthereaction

T4(nascently absorbable)

-T4 (nonabsorbable)

were the principal one occurring in this system.

Plotting log

ki

against 1/T, onecanestimate the

activationenergy(Ea)for the reactiontobe11,000

cal.

The kinetic observations indicate a low

tem-perature

dependence

for the conversion of the majorityofthe nascent phageto cofactor-requir-ing phage. On the other hand, a lack of

homo-geneity inthepopulationcouldgiveanapparent

value ofEa that islowerthanthatforanyof the

particles. Infact, at all the temperatures studied

therateof deactivationdecreasesasR

approaches

itsminimal value of5 X 10-6to5 X 10-A.Some,

but not all, of the particles that lose activation

very slowly prove to be mutants that have

re-duced requirements for cofactor

(4).

Itis important toemphasizethatnascentphage

deactivate much more slowly than

tryptophan-activated phage at each of the temperatures

studied.

Electron microscopy of nascently adsorbable

R

TIME

FIG. 2. Tlte deactivation of nascently adsorbable

T4r- at different temperatures. Samples (I ml) ofa

freshly thawednascentphage suspensionwereaddedto

9-ml volumes ofF medium at the indicated

tempera-turesandwereassayed from time to timeonFandN

agartodeterminethe R valueofthepopulation, which

is shownplotted against time.

phage. The electron microscope was used to

determine whether nascently adsorbable T4

possess any grossmorphological differencesfrom

cofactor-requiring phage. In the following

ex-periment, nascentphagewereobtained by

allow-ing infected bacteria to lyse on a carbon-coated

electron microscope grid. E. coli B growing in

the log phase in F medium were infected with

tryptophan-activated T4r [multiplicity of

in-fection(MOI) =5].After5minhadbeen allowed

for adsorption, the infected bacteria were

pel-leted. The supernatant fluid containing

unad-sorbed phage plus thecofactorwasdiscarded and thecells wereresuspendedinfreshFmedium. The

resuspended cells were incubated for 20 min at

37C, at which time a small droplet of the

suspension wastransferred to aprepared copper

grid. The grid, held with forceps, wasincubated

foran additional 10 minat 37C, afterwhich it

was washed with F medium and then negatively

stained and placed inthe microscope todry.

Microscopic fields produced by this method

were found to contain many unlysed bacteria

aswell as bacterial debris; however, some areas

suchasthatshown inFig. 3 wererelativelyclean.

In this instance one can see the end of a

bac-terium which has broken open to release its

contained phage onto the grid surface. One is

struckby thefactthatmostof the phagehave all

six of their tail fibers in the extended state. On

two ofthe particles thesix tail fibersare pointed

out with arrows, but many of the other phage

mayalsohave allsixfibers extended. The release

ofallsix tail fibers isgenerallynot seeninpurified

cofactor-requiring phage evenin the presence of

highconcentrationsoftryptophan (6, 15).

Comparison of the adsorption ratesof nascently adsorbable T4 and oftryptophan-activated T4. If

the rate ofadsorption ofa population ofphage

is limited by therate at whichthetail fibers can be made available for interaction with the host receptors, the nascently active phage, having all

tail fibersextended, might be expectedto exhibit

a more rapid adsorption rate than

cofactor-acti-vatedphage.

The rate of adsorption of cofactor-requiring

phage activated by N broth, where its rate is a

maximum(24), wascompared to thatofnascently

adsorbable phage in F medium at 15 and 37C.

Contrary to expectation, the results (Fig. 4), indicate thatnascentT4 inFadsorbsto B

some-what more slowly than does cofactor-activated

T4inN medium.

Data from five such adsorption experiments

were averagedtogivetherate constantsrecorded

in Table 1.Adsorptionrates wereappreciablefor

bothforms ofT4atboth37 and15C. Thisresult

97

on November 11, 2019 by guest

http://jvi.asm.org/

[image:4.491.36.229.425.572.2]

FIG. 3. Electronmicrogr-aph showinignascentphageproduced by

allowinig

bacteriainfectedwith

cofactor-requir-ing T4 tolyseonanelectronmicroscope grid maintained at 37 C. Tailfibers

oni

twoparticlesare

poinited

out with

arrows.Stainedwithsilico-tungstate.

differs from the observations of Wollman and Stent (25), who reported that nascent phage

adsorbedveryslowlywitharateonly0.015thatof

the same phage in broth at 15C. It should be

noted that the nascently adsorbable phage in F

medium appear to adsorb at only a slightly slower rate than do

cofactor-requiring

phage in broth. On the one hand, this difference in rates

may be due to some effect other than available

cofactor; for instance, the different nutrients

may have different effects on the bacterial sur-faces. On the other hand, the lower apparent

adsorptionratefor thenascentphage may bedue

toitsprogressive lossof

adsorbability

duringthe

experiment,whereasthecofactor-requiringphage

in broth have sufficient availablecofactor to

on November 11, 2019 by guest

http://jvi.asm.org/

[image:5.491.81.446.69.505.2]

T4

1.0o 0.8

0.6

0.4

0.2

0o

0

}-

W--'*

I

37°C

;////

}=-°

150C

/

I

5 10

[image:6.491.35.226.45.193.2]

TIME (MIN)

FIG. 4. Comparison of the rates ofadsorption of

nascent T4r- andof N broth-activated

cofactor-requir-ing T4r- at 15and 37 C. Cofactor-requiring T4r- was dilutedintoNbroth. Nascent T4r-fromafrozen

prep-aration wasdilutedinto Fmedium.At timezero,a

vol-umeofE. coliB wasaddedtoeach tube tomake their

finalconcentration 2 X 108 cells/ml and that of the phage S X 107 particles/ml. The reaction mixtures

wereincubated at either 37or15 C. At various times, samples of thereactionmixture werediluted1/100into

medium containing anti-T4 serum at a concentration

sufficient to inactivate99.9% oftheunadsorbedphage in 2min. Theinfectedbacteria were thenplatedonN agar. Solid line indicates T4 activated with nutrient

broth;broken line indicatesnascentlyactiveT4.

TABLE 1. Velocity constantsfor the adsorption of

nascent T4andcofactor-activatedT4 at 15and

37 Ca

Temp Cofactor-requiring Nascently-activephage phage in broth inFmedium

15C (2.3 7 0.5) X 10-10 (1.9 i 0.9) X 10-10

37C (12.5 4 0.3) X 10-10 (8.2 4 0.5) X 10-10

aValues areexpressed as milliliters perminute.

sure thatthey will all be adsorbable throughout the experiment. It was concluded from these

experiments that nascently adsorbable phage possess an ability to adsorb to bacteria that is

nearly equivalent to that of cofactor-activated phage.

Retention of nascent adsorbability by intra-cellular phage. The experiments described above

have been confined to free phage that had been released from the hostcellatthe end of thenormal

30-min latent period. The experiment of Fig. 5 wascarriedout todetermine whether thenascent

state of adsorbability can be retained by the

phage inside the host cell beyond the normal

latent period. If the nascent adsorbability were dueto some special nascent cofactor which was

incapableofdiffusingout of the cell, one might

expectthephagetoretain theirstate of

adsorba-bility throughout an extended latent period. If,

[image:6.491.239.434.66.171.2]

TIME AFTER INFECTION(hrs.)

FIG. 5. Effect ofthe extended latent periodon the

nascent stateofadsorbability.E.coliB,rapidlygrowing in F medium, were infected with cofactor-activated T4r-, centrifuged, and resuspendedas previously de-scribed. After incubationat37Cfor26min,KCNwas

addedto themixture toafinalconcentration of6 X

10-' mtoprevent spontaneouslysis. The suspensionof infectedcells wasstoredat 37C. At various time

in-tervals, samples of this suspension were lysed with chloroformat0C andwereassayedonFandNagar.

The assay dilution tubes were then returned to 37 C andreplated from timetotimetofollow thecourse of thephageactivityafter thedelayedlysis. The topline shows theR valueof thephageat thetimeof induced

lysis(indicated by arrows).The descending lines

repre-sentthedecreasein Rwith timeat37 Cafter the

in-ducedlysis.

on the other hand, nascent adsorbability were due to a diffusable cofactor or to some time-labile configurational difference at the cofactor sites, one would expect the phage to lose its

nascent adsorbability and be released from the

cellinareduced stateofadsorbability.

The results of such an experiment are shown inFig. 5, in whichRisplotted against time after infection. It is shown that phage released from thehost cell upto 12 hrafter the normal latent

period possess nascent adsorbability and that this adsorbability is lost only after the phage is

released from the host cell. The fact that phage retained thenascent stateofadsorbabilitysolong as they remained inside the host cell was inter-preted to support the notion that nascent ad-sorbability is due to some special "nascent

co-factor" that cannot diffuse from the cell.

Nascent cofactor from bacterial cells. An

at-tempt to find thenascentcofactor in extracts of

uninfected E. coliBthathad beengrownin F is describedin Fig. 6. Evidently, something similar

tothenascentcofactorcanbe obtained from the

uninfected E. coli cell. The activating ability of

this nascent cofactor is considered to be rather

low, for6 hrwasrequiredtoreactivate thephage from R = 10-a toR = 10-3, whereas the same

phage had been released withR = 1.0atthe end

ofa normal 30-min latentperiod.

A situation similarto thiswas found by Jerne a

w

cs:

m

0

Io

z

0

Li.

99

on November 11, 2019 by guest

http://jvi.asm.org/

[image:6.491.36.228.349.432.2]

1.0r

o0-1

10-2

,o--0 -° -=

-I\ o tc

1o a

10-3['

I ~~~~~~~~~~~~~b

10-4 ta

015l & I I

2 3 4 5 6

TIME (hours)

FIG. 6. Reactivation of quiescentT4r- by F-grown

E. coli extract. T4r- (R = 5 X 10-5) wasaddedto

1-ml samples of E. coli homogenatethat had been pre-paredasdescribed in MaterialsandMethods. The re-sulting mixture was incubated at 37C, and samples

weredilutedfrom timeto timeandplatedonFand N

with the result shown incurvea. Onlya slight

activa-tion isproduced bytheextractina6-hrperiod.Oneof

thedilution tubes takenfromthisreactionmixturewas

allowed tostandat37Candsampleswerereplatedat

various intervals onF and N. The inducedactivity is

lostatarate (curve b)that is much slower thanwould

have been expected for tryptophan-induced activity.

Curve c shows the results ofa similar experiment

in which activation by E. coli extract occurred in

thepresenceof50 jAg of L-tryptophan permland in

which samples were diluted 1/1,000at various times

intocofactor-free F medium. They wereassayedonF

and Nafterincubationfor5 min hadbeenallowedat

37 C to eliminate any tryptophan-inducedactivation. The1/1,000dilution tube taken at2hrwasincubated

furtherat37C andassayed fromtimetotimeonFand

Nagarstogivethe resultsplottedincurved. L-Trypto-phan greatly increases therateofactivationbythe E.

coli extractandtheactivatedphagelosesits

adsorba-bilityatthe slowratethat is characteristicof nascently adsorbable phage.

(14) in his studies ofa phage-specific antiserum

produced during the early period of immuniza-tion of a horse against T4. This antibody was found to activate T4 permanently, but at a rate

that was very slow unless the phage had been activated by tryptophan. We have found (Fig. 6, curve c) that tryptophan similarly enhances

therateof activation of T4byE.coliextractsand

that, when the tryptophan is diluted out, this

adsorbability is lost slowly at a rate equivalent tothat ofnormally producednascentphage (Fig. 6, curve d). It is also shown that the activated

phageseemto retain thisactivitysolongasthey remain in the E. coli paste but the maximal observed activation represents only about 5%

of the total phage population. This could well

reflect competition between the activation

reac-tion and the spontaneous loss ofrenascent ad-sorbability. No appreciable loss of phage titer was observed during this experiment, which is

possibly a result ofpretreatingthe extracts with

chloroform.

Attempts weremadeto increase theactivating properties ofthe E. coli extractbyuse of various purification procedures. The extract was

centri-fugedat5C for 10 min at about 6,000 X gina

clinical centrifuge. When the supernatant fluid and theresuspended pelletweretested for

activat-ingabilityin the presenceof50 ,gof tryptophan

perml,the supernatantfluidwasfoundtopossess no detectable activating material, whereas the

pellet was found to induce adsorbability with

about the sameefficiency asdid theoriginal

ma-terial.Attempts to remove theactivating material

from the whole paste by exhaustive dialysis at

5Cagainst F medium failed.

The above results indicate that the nascent

cofactor is associated with the larger

faster-sedimenting portions of the bacterial paste,

possibly fragments of thecell

envelope.

Since E. coli cell walls possess substrate sites

for theenzymelysozyme, thisenzymewastested

for its effect on the

activating

properties of the

E. coli extract. The resuspended pellet ofa 10 min, 6,000 rev/min centrifugation of crude E.

coli extract was treated with lysozyme at 3 x

10-5g/mland allowedtoincubateat37Cfor 10

min. Cofactor-requiring T4r- were then treated

withthismaterial for 50 minat37C in the

pres-ence of 50 ,tg of tryptophan per ml. A control

containing pelleted material had undergone a

similar incubation procedure, but it was not

treatedwithlysozyme.

The results (Table 2) indicate that lysozyme

destroys the activating ability of this material.

Other enzymesweretested for theireffecton the

activating properties ofthispellet fraction, with

uniformly

negative results.

Effect of enzymes on the deactivation rate of nascently adsorbable T4.The

experiments

already

described indicate that nascent adsorbability is

due to a special nascent cofactor. To further

characterizethe chemicalnature of thiscofactor,

the deactivation kinetics ofnascent phage were

studied in the presence of various enzymes. A suspension of freshly thawed nascent T4r- was treated with the enzyme to be studied and was

incubated at37C. Samples of this reaction

mix-ture and a controlwithout enzyme were diluted

at intervals andplated on F and N media. The enzymes studied included Pronase (5 x 10-4 g/ml), trypsin (2.5 X 10-3 g/ml), deoxyribo-nuclease (1.0 X 10-3g/ml), ribonuclease (1.0 x

on November 11, 2019 by guest

http://jvi.asm.org/

[image:7.491.61.250.48.225.2]

CELL WALL MATERIAL AND ADSORBABILITY OF

10-3 g/ml), phospholipase C (8.0 x 10-4 g/ml),

and lysozyme (at 2.1 X 10-4 and 8.8 x 10-3

g/ml).

Lysozymealonewasfoundtoincrease therate ofdeactivation withaconcentrational dependence

that is shown in Fig. 7. Lysozyme's activity is

probably not due to its high net-positive charge

(isoelectric point = pH 11) causing it to react

nonspecifically with the phage particles, for trypsin with essentially thesameisoelectric point

atpH 10.8 (18) hadnoeffect. On the other hand,

the rate of deactivation is not proportional to lysozyme concentration and was much slower

than might have been expected for lysozyme act-ingon a readily available substrate. Itmay well

be that the lysozyme-sensitive site is associated with some other type of macromolecule on the

phage particle that limits its accessibility to the

TABLE 2. Effect of lysozyme on the activating

properties of E. coli extract

Material tested pRvdlued4

T4r- + 50S g of tryptophan per

(no extract)... T4r- + 50SAg of tryptophan per + extract...

T4r- + 50 jAg of tryptophan per + lysozyme-treated extract (3 105goflysozymeperml,50min 37C)...

aMeasured after dilution to

tryptophanconcentration.

1.0

0.1

R

0.01

0.001

b

I

I

jC I

20 40

[image:8.491.246.438.207.323.2]

TIME

(min)

FIG. 7. Effect oflysozyme on the

nascent adsorbability. Freshly thawe nascentphage wasallowedto incuba

(a) andinthepresenceof2.1 X

10-perml(b)and 8.8X 10-3goflysozyn

various times, the threepreparations

F and N agar and the R value ofi

against incubationt time.

ml

*-- 5 X 10-5

- 1

enzyme. At any rate, it would appear that the nascent cofactor contains material that is

sus-ceptible to lysozyme.

Buoyant density of nascently adsorbableT4.To determinehow muchof the relatively low-density

cellwallmaterial is boundtonascentT4,wehave

compared its density in CsCl with that of T4

from the same preparation that had been

incu-batedat37 Cto acquire the needfora cofactor. The cofactor-requiring particles (R = 10-4)

band in peaks at p = 1.49 and p = 1.51 g/ml

(Fig. 8). In contrast, the nascent particles (R =

1.60 L

1.50

*T4r-R 10-4 oT4-R .8

--o."~!t 1

so ~:~o---f-Q

0.0.

.AI 0

6 12 18

FRACTION NUMBER

24

LO

.8

.6 0m260

.4

.2

30 36

ml FIG. 8. Densities in CsCl of nascently adsorbable

3 X

10j

T4r- and of cofactor-requiring T4r- from the same

ml preparation. Nascent phage were prepared by the

X technique ofone-stepgrowth;30ml ofE.coliB,rapidly

at growing in3 XDmediumataconcentrationofSX 108

* X 10- cells/ml, were infected (MOI = 5) with cofactor-re-nonactivating quiringT4r-inthepresenceof 10-4gofadded trypto-phanperml.After5 minat37 C had beenallowed for adsorption, the infected bacteria werepelletedand

re-suspendedin 10 ml of fresh 3 XD. The infectedcells

wereincubatedwith aerationforan additional 26 min

and then were quickly cooled in an ice-salt bath. Chloroform was added to lyse the cells and a trace

of deoxyribonuclease was added to reduce the

vis-cosity of the mixture. After 10 min at 0 C had

been allowedtoobtain maximallysis, themixturewas

divided intotwo5-mlportions.One samplewasallowed

to remainat 0C; the otherwasincubatedat37 Cfor

5hr. The R valueofthe material storedat0Cwasthen foundtobe close to 1.0,whereas thatofthe37C

ma-terialwas10-4. CsCIwasdissolved in thelysates to a

concentration of0.715 g/ml. The samples were

cen-trifugedinlusteroid tubesat39,000to 40,000rev/min

for18 hrat3to4 C(SpincomodelL-4,SW50 rotor).

60 80 Fractions were collecteddropwise andused to

deter-mine OD260 (optical densityat 260nm) profiles,

via-bility profiles, andindices of refraction. If viability

? rate ofloss of assays were tobedone, theentirecollection, dilution,

d preparation of and plating procedure (toadditionof agar) wascarried

ite at 37 C alone out at5Cto reduce the lossofnascentadsorbability.

4 goflysozyme The secondsamplewasgivenalow-speed centrifugation neperml(c). At to remove some ofthefast-sedimentingdebris. It was

were assayed onz allowedto standfor24 hrat37Cto losenascent ad-each wasplotted sorbability. Figure10b shows the sedimentationprofile

ofa10-p.liter portion ofthissample.

101

on November 11, 2019 by guest

http://jvi.asm.org/

[image:8.491.39.234.252.568.2]

0.8) evidently have lower densities, for they distribute themselves overa broadregion whose

densityrangesfrom 1.46to 1.50g/ml. Assays of the phage in various fractionsindicated that the R value is about the same at all densities; ap-parently the activatingnascent cofactor material comes inavarietyof sizes.

Electron microscopy of nascent T4. If the

nascentcofactor iscapable of shifting the density of phage particles as much as we observe, it should be possible to see the attached material byelectron microscopy. Freshlyprepared lysates

ofnascentphagewerefoundtocontaintoomuch

bacterial debris to give unequivocal results. Therefore, freshlyprepared lysates werepurified by several rounds of alternate high- and low-speed centrifugation at low temperature. Ap-proximately 40to 60% of the phage population

lost nascent adsorbability during this procedure.

Figure 9 shows phageparticles from several such preparations. Material can be seen attached to thebaseplate regions ofmanyparticles. This ma-terial isseentovaryinamountfrom rather large pieces on someparticles to small, hardly visible fragmentsonthe tail pins of others. This observa-tion agrees with the density-gradient studies in indicating that thematerial bound to the phage has a wide range of sizes. Thearrow in Fig. 9a

points out a piece of free material identical in

size andappearancetothat attachedtothe nearby phage particle. This fragment maybe a pieceof

material which had been liberated from the tail of a phage particle which has lost its nascent

adsorbability.

Binding of cell wallfragmentsto nascentphage.

Thatnascentadsorbabilityis relatedtothe

inter-action of phage with host cell wallswas further

demonstrated by the experiments presented in Fig. 10 and 11.

Since it isknown thatN-acetylglucosamine is aprimary constituent of theE.coli "rigid layer" (19, 20), E. coli B cells were grown in 14C N-acetyl glucosamine to make their rigid layers radioactive. Nascent phage produced from in-fection of these hot cells were sedimented on

sucrose gradients (Fig. 10) and banded on cell

density gradients (Fig. 11) before and after losingnascentadsorbability.

As isshown in Fig. 10a, phage havingnascent adsorbability are able to affect the sedimenta-tion profile of 'IC-labeled fragments of the

bac-terial wall. Cofactor-requiring phage on the

r)ther hand do not (Fig. 10b).

When nascent phage prepared from

"4C-labeled cellsare banded on CsCl (Fig. 11), the

phage having nascent adsorbability are seen to bandpolydispersely in the gradient (see also Fig. 8). The 14C isseentoband in regions

correspond-ing tothe infectivity peaks (Fig. Ila). It is also

found that the ratio of counts 'IC to infectivity

is higher for those phage taken from the less

dense regionofthegradient, indicatingthatphage

bandingintheseregionshavelarger fragments of

cell wall material boundto them. The

cofactor-requiring phage(R = 1 X 10-s) inFig. lib are

found tobandin a morerestricted region of the

gradient and have less of the label associated withthem than do the nascentphageinFig. 1 la. Adsorbability of phage produced from

sphero-plasts.The

preceding

studies have indicated that

nascent adsorbability is the result of the

inter-action ofT4 phage with host cell wall material

atsometimeduringthelatentperiod.Ifthis is so,

it should be possible to produce

tryptophan-requiring phage directly from spheroplasts,

cells whichhavehadmostof theirwalls removed

enzymatically (11).

E. coli B cellswere

infected

with T4r- in the

presence oftryptophan (MOI = 10). A sample

ofthecellswasconvertedtospheroplastsin

16%

sucrose by a modification of the method of

Guthrie and Sinsheimer (13). After incubation

for 30 min to allow development of daughter

phage particles, the spheroplasts were diluted

1:100 at 0 Cin Fmediumtorupturethem and

releasethe daughter particles. Only 16% of the

phage produced by spheroplasts were able to

form plaqueson Fmedium,whereas 82% ofthe

particles produced by control cells formedplaques

on F medium (Table 3). The simplest

explana-tion is that theremoval of the cell wall

prior

to

the release ofprogeny T4preventsmost

daughter

particles from becoming

nascently

adsorbable.

Theequivalence in burst size in

experiment

and

control indicates that thepresenceof thecellwall

itself isprobably unimportant for phage

produc-tion.

DISCUSSION

Particles ofT4 are liberated fromhostcells in

a state of nascent adsorbability in which they

canreadily adsorbtohostcellswithout additional

cofactor (25). Once liberated into the medium, however, the particles spontaneously lose this nascentadsorbabilitywith ahalf-time ofonly 20 min at 37 C.Thus, particlesinolder stocksor in

preparationsthat havebeenpurifiedin the course

ofafewhourslose theiradsorbability. They

can-not adsorb to host cells unless they have been

reactivated by a cofactor like L-tryptophan.

Activation by tryptophan is reversible; it is so

rapid that half-times oftheforward and reverse

reactionsaremeasuredinseconds.

Ourresultsalso indicatethatnascent

adsorba-bilityis due to a substance we shall call"nascent

cofactor" that is attached to the phage particle.

on November 11, 2019 by guest

http://jvi.asm.org/

a

I.

b

.us1.C

'h:t.X

FIG. 9. Electron micrographs ofnascentT4. Phagewereprepared by lysing infected bacteriaat28 minafter

inzfectionbychloroformtreatment.Thesenascently adsorbable particles were purified by several rounds ofhigh- and

low-speedcentrifugation carriedout at 4 to5C. The arrowspointoutmaterial attachedtothebaseplate regions of the particles. InFig. 9a, the arrowpointsout afragment equalin sizeandshapetotheoneattachedtothtephage; this fragment may havebeenreleasedbyaparticle which has lost adsorbability.Barsrepresent100 nm.

on November 11, 2019 by guest

http://jvi.asm.org/

[image:10.491.44.450.67.570.2]

Thenascent cofactor is evidently solarge that it

cannot diffuse from infected cells, for if their

lysis is delayed in KCN,daughter phage particles

canremain trappedinside for days without losing

their nascent adsorbability. Furthermore, the

N~

2

Q-U

35 40

15-b

L

0

5c

5 10 15 20 25 30 35 40

[image:11.491.160.442.60.379.2]

FRACTION NUMBER

FIG. 10. Sedimentation profile ofnascent (a) and

cofactor-requiring phage (b) prepared fromcells which

have their "rigid layers" labeled with 14C. E. coli B

were growntoaconcentrationofS X 108cells/ml in

F medium andthenwerediluted into 5 mlofF medium

containing 0.05 mc of N-acetyl-'4C-D-glucosamine

(specificactivity, 10.4 mc/mM). The initial

concentra-tionofcells in the hot medium was 107/ml. After the

cells had been allowedto grow to S X 108/ml, they

were washed three times in coldmedium. These cells werenextinfected withcofactor-requiringT4atMOI=

10 in thepresenceof50 ,ug of tryptophanperml. The

infectedcells were centrifugedandresuspended in co-factor-free medium toafinalconcentration of2.5 X 101in2ml ofFmedium.Thismixturewasincubatedat

37 Cfor30 mininamicro-Fernbachflask.Atthe end

ofthe incubationperiod, themixture wascooledinan

icebath and treatedwith 0.2 mlof chloroform for 10

min. Twosampleswerecollected. One10-,lAitersample

containingnascentlyadsorbableT4waslayered quickly

ontoanice-cold5to30%sucrosegradient andspuI at

20,000rev/minfor10 min;equal-volumefractionswere collected, assayedfor infectivity, and countedfor14C content. The results ofthis experiment are slhown in

Fig. 1Oa.

15

-~lo

0

n

a.

-5 10

---10

L)

5'

0.

5 10 15 20 25 30 35 40

FRACTION NUMBER

FIG. 11. Densityin CsCl ofnascentT4 (a) andof

cofactor-requiring T4 (b) produced from cells which

have theirrigid layerslabeled with14C.Bacteriophage

werepreparedasdescribed inFig. 10. Onesamplewas

maintainedat0C toprevent loss ofadsorbability. A

secondwasplacedat37Cforaperiod of2 hr. CsCI

wasaddedtoboth whileat 0 Cto raise thedensity to

about 1.475g/cm3. Thesesampleswerecentrifugedin

theSW50-1rotorat35,000

revlmin

for18 hrat2C.

Titer andcounts/minareshownplotted against fraction

numberfor nascentphage (a) and cofactor-requiring

phage (b).

mean buoyant density of nascently adsorbable T4 particles in CsCl is considerably lower than that ofparticles that have been allowed to lose

their nascent adsorbability. This indicates not

onlythatthe nascent cofactor islarge enoughto

affect the overalldensityofaT4particlebut also

that the nascent cofactor has a density that is

lower than that ofparticles that have been freed

of nascent cofactor. The fact that nascent

parti-cles display a considerable range of densities indicates that the nascent cofactor can have a widerangeofsizes. It is evident that this cofactor is not a discrete molecule but a substance or

aggregation of substances. The electron

micro-scope confirms this view: whereas particles in

on November 11, 2019 by guest

http://jvi.asm.org/

[image:11.491.57.249.135.414.2]

CELL WALL MATERIAL

TABLE 3. Adsorbability of phage newly released from spheroplasts

Expt

~~Total

phage Rvalue

Expt produced

Controla 6 X 109 0.82

Experimentalb 4 X 109 0.16

aInfected cells suspended in F medium +

16%

sucrose.

6 Infected cells suspended in F medium + 16%

sucrose + 104 g oflysozyme/mi + 10-3M ethyl-enediaminetetraacetate.

"purified preparations" that have lost their

nas-cent adsorbability appear relatively "clean,"

nascently adsorbableT4particlesare seentohave

pieces of material of various sizes adhering to

their baseplate regions. This could well be the

nascent cofactor.

By means of electron microscopy, it is also

observedthat T4

particles

inpurified

preparations

often have their

long

tail fibers

wrapped

around

their tailsheaths, whereasnascentparticles have

all six long tail fibers freed from their sheaths,

withtheirtips freetocontactthesurfaces ofhost

cellsand initiatetheadsorption process (15, 22).

L-Tryptophan

also frees long tail fibers fromthe

sheathsofquiescentparticlestoactivatethem,but

oddly enough, even in a large excess of

trypto-phan, very few particles have all of their tail

fibers visibly free of their sheaths (6, 15). Thus,

it seems highly

probable

that thisfreeing ofthe tail fibers is the

significant

morphological

con-figurationthatenablesT4particlestobe

adsorba-ble,nascentlyorotherwise.

Therateof

adsorption

ofnascentT4particles

is nearlyequal tothat oftryptophan-reactivated

particles. One

might

suppose that the presence

ofwhat lookslikebacterial wall debrisassociated

with the

baseplates

of nascent particles would

interfere with the

complex

function of this

struc-ture (22), but thismaterial

might

readily merge

with compatible substances endogenous to the

bacterial surface when the baseplate region is

brought intocontactwiththe hostcellwall. If so,

this material would

quickly

be prevented from

interfering

with the function of the baseplate, and the rate of infection would scarcely be affected.

We have found that the cell wall fraction of

E. colican conferan active "renascent" state of

adsorbabilitytomany,butnottoall,T4particles

that havelosttheirnascent

adsorbability

on

stand-ing.Two observationssuggest that thiswall

ma-terialis very similar ifnotidenticaltothenascent

cofactor. First, both the nascent

adsorbability

of the phage and the

activating ability

of the

extracts are destroyed by lysozyme. Second, the renascent particles spontaneously lose their

adsorbabilityatmuch the same rate as the nascent

particles lose theirs.

Tryptophan greatly enhances the rate and ex-tent to which the cell wallfraction activates T4, as though tryptophan prepared the phage for reaction with the wall fraction by extending its

tail fibers. Thewallmaterial might thenlockthe

particleinthis adsorbable configuration inwhich

it canremain evenafterthe tryptophan has been removed. It seems probable that here tryptophan actsinmuch the same way asit does in promoting the permanentactivationof T4 by A8 serum(14). It would appear that cofactors like tryptophan reversibly activate T4 particles by extending their long tail fibers, whereas substances likethe

nas-centcofactorand A8 serum stabilize theparticles

inthis adsorbable configuration.

The susceptibility of nascent adsorbability of

T4 to egg-white lysozyme suggests that,6-1-4

bonds between N-acetyl-glucosamine residues

are essential to the activity of the nascent

co-factor. These bonds probably are not readily

accessible to the enzyme, however, for the rate

of lysozyme action is not

proportional

to its

concentration. Nascent phage produced from

cells that had been prelabeled with N-acetyl-l"C

glucosamine are found to exhibit a variety of

densities thatis paralleled by the density profile

of this radioactive material. These phage also

showalargeramountof label associatedwiththe

less-denseparticlesthanwith thephageofgreater

density.

Sincetheradioactive labelintheseexperiments

isincorporated intotherigid layerofthecell wall

andsince theloss ofnascent adsorbability

coin-cideswithlossof label fromnewlyreleasedphage,

it may be concluded that the nascent cofactor

contains (in part at least) some of the host cell

wallmucopolymer layer.

In light ofthe evidence presented here, it is

possible to present a reasonable model for the

nascent state ofadsorbability. Since the nascent

cofactor is an activity stabilizer rather than a

phage activator, one can assume that this

co-factorisbound tothe phage during thetime that

the tail fibersareintheextendedstate.

Electron microscopy indicates that the cell

wall material may be attached to the baseplate

and tail-pin region of the phage structure. The

baseplate of the phage may acquire cell envelope materialasit isformed within thecelland

some-time during thelatent periodeither prior to the

synthesis oftail fiberorbefore thetail fibers are

retracted. On the otherhand,thebaseplateofthe finishedparticle may associate witha portion of the cell wall of the infected bacteria. Since the

VOL. 4, 1969

105

on November 11, 2019 by guest

http://jvi.asm.org/

[image:12.491.45.239.69.166.2]

cell membrane is interposed between the cyto-plasm and the cell wall proper, one must either

include the membrane in thefragment boundto

the phage or assume that the cell membrane has

been broken downatthis point,making possible

directaccess tothecellwall.

The location of the nascent cofactor on the

underside of the baseplate may account for its

abilityto preventthe retraction of the tail fibers.

Location ofthehinge (12, 17) which allowsthe tailfibers tochangeorientation isnotknownbut

couldwell be in the area where thebaseplate is

attached to the tail fiber. The cell wall material bound to theunderside ofthebaseplate may act insuch a way as to preventthishinge from work-ing, thus creating arigid tail fiber-extended

con-figuration. This might be accomplished by

link-ingthetail fiber to thetail pin. Theloss of

nas-centadsorbability would betheresultofremoval

ofthe cell debrisfrom the baseplate, freeingthe

hinge and allowing the tail fibers to bindto the

sheath. Theforcesremoving thenascentcofactor

from the phage may

simply

be due to kinetic

forces acting on the rigid tail fibers, eventually

freeing the hinge. The essential aspects of this

model are depicted

diagrammatically

in Fig. 12.

ParticleA representsaT4particle in thenascent

state.Thenascentcofactoris showncross-linking the tail pins to the tail fibers, maintaining them in an extended state. T4 may lose nascent

ad-sorbabilityin two ways. (i) Thenascentcofactor

may be removed fromthebaseplate regionasin

particle C, allowing the retraction of the tail

fibers, or (ii) the tail fibers may free themselves

from the cofactor and retract, leaving the

co-factor attached to thebaseplate asin particle B.

MODEL FOR NASCENT ACT!VITY

0'

FIG. 12. Modelfornascentadsorbability. (See text

forexplanation.)

The reaction A -* B may be

reversible, allowing

the

phage

toreturntothenascent state ifthe tail fibersarereassociated with the cofactor after

dis-placement by

some phage activator. [This may

account for the return to the nascent state after

treatmentwithtryptophaninthe system of Woll-man and Stent

(25).]

Thereaction A -* Cwould

notbereversible,asthecofactor is gone from the

baseplate. It would also be possible for particles

in the B state to convert to the C state, losing theirability to revertbackto the A state.

Thus,

anoverall reaction

might

be

A

(adsorbable)

* B

(nonadsorbable

but

capable

of

reverting)

->C

(nonadsorbable)

The reaction A

'->

B -÷ C may accountfor the

shape ofthecurves obtained from the studies of

therateofloss ofadsorbabilityofnascent

phage.

Thedeactivationcurve (Fig. 2) ofT4r- seemsto

possessaninitialregionoflinearity,followedbya decrease in deactivation rate prior to obtaining the lowest observed R value. The low R value observed for T4r-wasfound to be R = 5 X 10-6. This

change

inratemaybethe resultof

kB

-c

being

less thankA

B-Thereactivation ofphage by the cell wall frac-tion of E. coliextractsis presumedto follow the modeljust described, with theexception that the tailfibersarefirstextended bytryptophan. If the

nascent cofactor is made up of sections of the cell wall consisting of protoplasmic membrane and complete multilayered wall, increase in the deactivationratebylysozyme may be by removal of some of the cell wall layers, leaving behind the membrane and other protein-containing

layers.Under thesecircumstances, the rate of loss

of nascent adsorbability may not reflect the

di-gestion by lysozyme, as there would still be

lysozyme-insensitive material to be removed. The spontaneous loss of nascentadsorbability by free phage could be dueto a simple dissocia-tion ofnascentcofactor from thephage. On the other hand, it is conceivable that spontaneous deactivation is duetothe action of the samelytic enzymes in T4 that hydrolyze cell wall material when T4 is adsorbed to host cells (19, 21, 23). In the latter case, to account for the intracellular retention of nascent activity, one would haveto

assume that these enzymes areeither blocked or cannotreach their substratesduring an extended intracellularsojourn of the virion.

Three possible models may be suggested to explain how phage obtain these cell wall frag-ments. (i) Phage particles maturate with their baseplates in association with the cell envelope. (ii) At some timeduring the latentperiod,phage attach by their baseplates to the cell envelope.

on November 11, 2019 by guest

http://jvi.asm.org/

[image:13.491.60.248.447.616.2]

CELL WALL MATERIAL AND ADSORBABILITY OF

(This attachment need not be permanent and at any given time only a few of the intracellular phage particles may actually be in contact with the wall structure.) (iii) Fragments of wall

ma-terial might be obtained during the physical

breakdown of the cell envelope (lysis).

Theplausibility of the third model is strength-ened by observations of T4-infected bacteria in

V.~~~~~~~~~~~

$,J

FIG. 13. Electron micrograph ofabacterium infectedwith cofactor-requiring T4 in the process oflysis. The

arrowpointsout amembrane-like structurewhich seems tosurroundthenewly released particles. Thismembranze seems tobeconitinuouswith the bacterial cell wall.

VOL. 4,1969

107

on November 11, 2019 by guest

http://jvi.asm.org/

[image:14.491.36.442.109.609.2]

the process of lysis. By briefly exposing infected

bacteria to chloroform vapor on an electron

microscope grid, cells can be observed in the

process of releasing their contained phage. In

Fig. 13, thewall structure ofa bacterium seems

to be broken down insomeareaswhereprogeny

phageare seenspilling out, although outside the

normal confines of the host cell some of the

progenyparticlesseemtobe retained by a

mem-brane-like structure which is continuous withthe

bacterial envelope. (This structure may consist

of the lipid layers of the envelope.) It may be

that the phage obtain the fragments of the rigid layer while retained inside this baglike structure duringa temporary state of partial lysis. It thus seems possible that, much as certain animal

viruses acquire their envelopes from nuclear (8)

or cytoplasmic (9) membranes, the T4 phage

particle acquires its nascent cofactor from cell

wall material at the time of its release from the cell.

ACKNOWLEDGMENTS

Weacknowledge the excellent technical assistance of John M. Mackenzie, Jr. Manfred Bayer contributedmanyhelpful sugges-tions in thepreparation ofthismanuscript.

This investigation was supported by Public Health Service traininggrant5-TOIGM-00694-07totheUniversityof Pennsyl-vania from the National Institute of General Medical Sciences, by Public Health Service grants FR-05539 and CA-06927 to the Institute for Cancer Research, bygrantGB-4640fromthe Na-tional ScienceFoundation,andbyanaward toDartmouth Col-lege from the Brown-Hazen Fund administered by the Research Corporation ofNew York. Anappropriationfromthe Common-wealth of Pennsylvania totheInstituteforCancer Researchis

gratefully acknowledged.

LITERATURE CITED

1. Adams,M.H. 1959.Bacteriophages. Interscience Publishers, Inc., New York.

2. Anderson,T. F. 1945. The role oftryptophaninthe adsorp-tion oftwobacterial virusesontheir host E.coli.J.Cell. Comp.Physiol.25:17-26.

3.Anderson, T. F. 1948. Theactivation of thebacterial virus T4by L-tryptophan.J.Bacteriol. 55:637-649.

4. Anderson, T. F. 1948. Theinheritance ofrequirements for adsorptioncofactors in the bacterial virus T4.J. Bacteriol. 55:651-658.

5.Anderson, T. F.1962.Negative stainingand itsusein thestudy of virusesand theirserological reactions,p.251-262.InR. J. C. Harris, (ed.), The interpretation of ultrastructure, vol. 1. Symp. Int. Soc. Cell Biol. Academic Press, Inc., New York.

6. Cummings, D. J. 1964. Sedimentation and biological proper-ties of T phages ofEscherichia coli.Virology23:408-418. 7. Cummings, D. J., V. A. Chapman, and S. S. DeLong. 1969.

The sedimentation andconformational varianceamong T-evenbacteriophages.Virology 37:94-108.

8. Darlington, R. W., and L. H.Moss,III.1968. Herpes virus

envelopment.J.Virol.2:48-55.

9.Davis, B. D., R. Dulbecco, H. N. Eisen, H. S. Ginsberg, and W. B. Wood, Jr. 1967. Microbiology. Harper and Row Publishers, New York.

10. Fraser, D., and E. A. Jerrel. 1953. The amino acidcomposition of T3bacteriophage. J. Biol.Chem.205:291-295. 11. Fraser,P., H. R.Mahler, A. L. Shug, and C. A. J. Thomas.

1957. The infection of sub-cellular E.coli strainBwitha DNA preparation from T2bacteriophage.Proc. Nat. Acad. Sci. U.S.A.43:939-947.

12. Gamow,R.I.,andL.M.Kozloff. 1968. Chemically induced cofactor requirement for bacteriophage T4D. J. Virol. 2:480-487.

13. Guthrie, G. D., and R. L. Sinsheimer. 1963. Observationson

the infection of bacterialprotoplasts with theDNA of

bac-teriophage -X 174. Biochim. Biophys. Acta 72:290-297. 14. Jeme, N. K. 1956. Thepresenceinnormalserumofspecific

antibody against bacteriophage T4 and its increase during the earlieststagesof immunization. J. Immunol. 76:209-216.

15. Kellenberger, E., A. Bolle, E. Boy de la Tour, R. H.Epstein,

N.C. Franklin, N. K. Jerne, A. Reale-Scafate, J. Sechaud, I.Bendet, D. Goldstein, andM. A.Lauffer. 1965. Functions and properties relatedto the tailfibers ofbacteriophage T4. Virology 26:419-440.

16. Konner, L., and L. Kozloff. 1964. The reaction of indole and T2bacteriophage. Biochemistry 3:215-223.

17. Kozloff, L. M. 1968. The large DNA bacteriophages, p. 435-496.In H. Fraenkel-Conrat (ed.),Molecular basis of

vir-ology. ReinholdBookCorp.,NewYork.

18. Long, C. 1961.Biochemistry handbook. D.vanNostrandCo., Inc., Princeton, N.J.

19. Martin, H. H. 1966. The biochemistry of bacterial cellwalls, p.457-484. In Paul D.Boyer (ed.) Annual reviewof

bio-chemistry, vol. 35.AnnualReviews, Inc.,PaloAlto, Calif. 20. Salton, M.R.J.1964. The bacterial cell walls. American

Else-vierPublishing Co.,NewYork.

21. Sechaud, J., E. Kellenberger, and G. Streisinger. 1967. Perme-ability of cellsinfected with T4r andr+ phages. Virology 33:398-404.

22. Simon, L. D., and T. F. Anderson. 1967. The infection of Escherichiacoli byT2andT4bacteriophagesas seeninthe

electronmicroscope.II.Structure andfunction of the

base-plate. Virology 32:298-305.

23. Weidel, W., and J. Primosigh. 1957. Die gemimsame Wurzel derLysevonEscherichia colidurch Penicillin oder durch

Phagen.Z.Naturforsch.12b:421-427.

24. Wollman, E. L., and G. S. Stent. 1950. Studiesonthe activa-tionof T4bacteriophage bycofactor. I.Thedegreeof ac-tivity.Biochim. Biophys. Acta 6:292-306.

25. Wollman, E. L., and G. S. Stent. 1952. Studies on the

activation ofT4 bacteriophage by cofactor. IV. Nascent

activity.Biochim.Biophys.Acta9:538-550.

on November 11, 2019 by guest

http://jvi.asm.org/