DNA Synthesis and Gene Expression in Bacillus subtilis Infected with Wild-Type and Hypermodification-Defective Bacteriophage SP10

(1)

1982, p. 636-648 Vol.

0022-538X/82/050636-13$02.000/

DNA

Synthesis

and Gene

Expression

in Bacillus

subtilis

Infected with

Wild-Type and Hypermodification-Defective

Bacteriophage

SP10

HEMANWITMER*ANDMICHAELFRANKS

Departmentof Biological Sciences, University ofIllinois atChicago Circle, Chicago,Illinois60680 Received25September 1981/Accepted24 December1981

A hypermodified base (Y-Thy) replaces 20%o of the thymine (Thy) in mature

DNAof Bacillus subtilis phage SP10. Two noncomplementing hypermodification-defective(hmd)mutants aredescribed. At 30°C, hmd phage carriedout a normal program,butattemperaturesof.37TC,the infectionprocess wasnonproductive. When cellswere infectedat37°C with hmd phage, DNA synthesis started atits usual time(12 mmin), proceededat about half the normalratefor 6to8min, and

then stopped ordeclined manyfold. All, ornearly all, of the DNA made under

hmd conditions consisted offully hypermodifiedparental DNA strands H-bonded

to unhypermodifled nascent strands. The reduced levels of DNA synthesis

observed under hmd conditions were accompanied by weak expression oflate genes. A sucrosegradientanalysis ofSP10hmd+ replicatingDNA intermediates wasmade. Twointermediates, calledVFandF,wereidentified. VF consisted of condensed DNAcomplexedtoprotein;VF also containednegatively supercoiled

domains covalentlyjoinedtorelaxedregions.Fwascomposedof linear

concaten-ates from which mature DNA was cleaved. None of those intermediates was evident in cells infected at 37°C with hmd phage. Shiftup experiments were performed wherein cells infected with hmdphageat30°Cwereshiftedto37°Cata time when replication was well underway. DNA synthesis stoppedor declined manyfold10 minaftershiftup.ThehmdDNAmade aftershiftupwasconservedas

aformsedimentationally equivalenttotheFintermediate,but littlematureDNA

was evident. It is proposed that Y-Thy is required for replication and DNA maturation because certain key proteins involved with theseprocesses interact preferentially withhypermodifiedDNA.

Bacillus subtilis phage SP10 belongs to a

group of viruses in which thymine (Thy) in mature DNA is partially replaced by a hypermodified base (10,37,39). The hypermodi-fied base in SP10DNA(designated Y-Thy) has not beencompletely characterized with respect to chemical structure (37). Other members of this group include B. subtilis phage SP15 and Pseudomonas acidovorans phage 0W-14, in which Thy is replaced by 5-(4',5'-dihydroxypen-tyl)uracil (DHP-uracil) and a-putrescinylthy-mine, respectively (4, 21). In the case of SP15,

DHP-uracil and Thy are assembled as free

de-oxynucleoside triphosphates that are subse-quently inserted intothe DNA polymer (36). In thecaseof SP10and 0W-14, however, neither Thy nor the respective hypermodified base is assembledas afreemononucleotide but,rather, each originates by post-replicational modifica-tion ofacommonprecursor base that is

proba-bly 5-hydroxymethyl(HM)uracil(10, 27, 31,37, 39). SP10 and0W-14aretheonly known

biolog-icalsystemsin whichThyin DNA does not arise via the deoxythymidylate synthetase reaction

(37).

Little is knownabout thebiologicalrelevance ofhypermodified basesin DNA. Recent studies suggestedthatDHP-uracilplaysnomajor role in

differentialtranscription of SP15 DNA (16) but

thatnormallevelsofthehypermodifiedbase are required for proper DNA maturation, phage assembly, orboth (6, 36). These earlier studies were confounded by the fact that no situation couldbe definedinwhichSP15DNAdevoid of

DHP-uracilwas made (6, 36).

This paperconcernstheisolation and charac-terization of hypermodification-defective (hmd)

mutants of phage SP10. The results of these experimentssuggest that Y-Thyis indispensable to SP10 development. Specifically, Y-Thy

636

on November 10, 2019 by guest

http://jvi.asm.org/

(2)

ROLE OF Y-Thy IN SP10 DEVELOPMENT 637 seems to be required for proper replication of

thephagegenome aswellascleavageof concat-enatestounit-size DNA.

MATERIALSANDMETHODS

Phageand bacteria.SP10hmdl andSP10hmd2are

heat-sensitivemutantsdefective in hypermodification (see below and reference39). SP10 hmd+is the same

clear-plaque variant used in allprevious studies (29, 30).B.subtilis W23wasthe hostin all experiments.

Mutagenesis.Cells infectedwithSP10 hmd+at30°C

weremutagenizedwith

N-methyl-N'-nitro-N-nitroso-guanidineasdescribedby Kahan (20).

Screening forhmdmutants.Survivorsof

mutagene-sis were plated at 30°C. Plaques were picked with

steriletoothpicks and scoredfortheirabilitytoplateat

42°C.Lysatesofheat-sensitivemutantswereprepared at30°C and usedtoinfect cellsat42°Cin thepresence

of6-(p-hydroxyphenylazo)uracil,aninhibitor of

bacte-rialDNA polymeraseIII (5, 32) that hasnoremarkable

effect on SP10development(29, 30). Thecells were

labeledwith[6-3H]uracil(10 uCi/ml)for 15min. Cellu-lar DNA was isolated(6) and acid hydrolyzed to a

mixtureofpurine basesandpyrimidine deoxynucleo-sides(36). The digestswerechromatographedinone

direction on unmodified cellulose thin layers (10),

usingdeoxythymidine (dThd) andY-dThdas optical

markers. Spotscorrespondingto the optical markers

wereeluted (10, 29),and theamountof label present

was determined. In those caseswhere no labelwas

recoveredin theY-dThdspot,asampleof [3H]DNA

wasenzymatically degradedtomononucleotides (36, 39) thatwerefractionated by two-dimensional chroma-tography onunmodified cellulose thin layers (7). Of

the 184 single-plaque isolates tested to date, only 2 generatedDNAat42°C whichwastotally devoid of

Y-dThd. The basesreplacingY-Thy inSP10 hmdl and SPO hmnd2 DNAs, undernonpermissive conditions,

are 5-(hydroxymethyl-O-pyrophosphoryl)uracil and

HM-uracil,respectively (39).

Sucrosegradient centrifugational analysisof

replicat-ingDNA.Cultures were infectedasdescribed previ-ously (29, 30).Thelabeledprecursor,typically32p, (10 ,Ci/ml),wasintroduced12minafterinfection. Atthe desired times,5-ml sampleswereremovedand mixed with 20 ml of ice-cold medium that contained 10mM KCN. Thecellswereharvestedbycentrifugationand suspendedinto 2 mlof 10mMTris-hydrochloride (pH 7.9),10mM disodium EDTA(pH 7.9),1 mMKCN, and 1mgofeggwhitelysozymeperml. Thecellswere

incubatedat37°Cfor 20min,and 0.1 ml of11%(wt/

vol)sodiumlauroylsarcosinewasadded tocomplete lysis.

A100-1l portionofthe extractwasmixed with100 p.1 ofasolution containing marker [14C]DNA. Each

markerwaspresentat0.5 pLg.The entire mixturewas

layeredontoa5.8-mllinearsucrosegradient (5to25% sucrose in10 mMTris-hydrochloride, pH7.9-1 mM

disodiumEDTA,pH7.9-1 MNaCI-0.01% sarcosine). Allgradientswere spuninan SW50.1 rotor at150C.

Theexactconditionsofcentrifugationaregivenin the

appropriate figurecaptions. Fractions (100 pLI) were

collectedfrom thetopwithanISCOdensitygradient fractionator,usingFluorinert FC-48asthechase

solu-tion.The fractionswereindividuallyprecipitated (6), andthe amount of labelinDNAwasdetermined.

Isolation of RNA.Unlabeled RNA was isolated from

2-litercultures by the method of Bolle et al. (2) as modified by Legault-Demare and Chambliss (22).

[3H]RNA was isolated from 50-ml cultures labeled with[5-3H]uracil(5 uCi/ml).

Isolation of SP10 DNA strands. Strands of SP10 DNA were separated in CsCl on the basis of their differential binding of polyuridylate-polyguanidylate

(16). TheCsCl-purified strands were self-annealed and passed through hydroxyapatite columns to separate single strands from residual duplexes (19).

Other methods. DNA-RNA hybridization and hy-bridization-competition were carried out as described byGillespieandSpiegelman (14). Enzyme assays and definition of activity units are given elsewhere (29, 39). Serum blocking power was determined by the method of DeMars (8). Protein concentration was measured by the method of Lowry et al. (26). The incorporation of radioactive precursors into DNA was measured as described by Lembach and Buchanan (23).

Materials. Radioactive materials were purchased from New England Nuclear Corp. (Boston, Mass.), except [6-3H]HM-deoxyuridine, which was synthe-sized by the method of Alegria (1).Nitrosoguanidine was purchased from Aldrich Chemical Co. (Milwau-kee,Wis.). Enzymes were obtained from Boehringer-Mannheim (Indianapolis, Ind.). Sucrose and Fluorin-ert FC-48 were purchased from Schwarz/Mann (Orangeburg, N.Y.) and Instrumentation Specialties Co. (Lincoln, Neb.), respectively. All nonradioactive materials werepurchased from Sigma Chemical Co. (St. Louis, Mo.).

RESULTS

Thermolability of hmd phage. Wild-type (hmd+) phage produced high bursts at all tem-peratures but, by this criterion, the infection processwasmostefficient between 30 and37°C

(Table 1). At temperatures of < 30°C, mutant (hmd) phagegave bursts that werewithin60% of normal, butathigher temperaturesthe burstsize declinedsharply(Table1).Forthispaper,hmd+

and hmd phagewerecomparedat37°Csincethis is the temperature at which most biochemical studies with SP10 were performed (10, 29, 30,

39,40).

TABLE 1. Burstsize ofSPIOhmd+andSP10hmd phagesatvarioustemperaturesa

Burstsize

Phage

200C 300C

330C

37°C

420C

SPIOhmd+ 31 87 89 83 63

SPno

hmdl 20 54 12

sl

x 10-4

sl

x 10-4

SP10hmd2 21 58 12 sl x i-0sl x

i0-4

a

Cells

weregrown at37°Cto adensityofca. 2 x

108/ml.Thecellswereharvestedbycentrifugationand

resuspendedinto0.95 volume of fresh medium

equili-bratedtotheindicated temperature. Ten minuteslater, the cells were infected by adding 0.05 volume of medium thatcontainedthephage(inputmultiplicityof infection=0.5PFU/cell).Theremainder of the

proce-durewasidenticaltotheoneusedbyBottand Strauss (3). Allplatingsweredoneat300C.

VOL.42, 1982

http://jvi.asm.org/

(3)

638 WITMER AND FRANKS

E L 0 _ 2 * 04zz 0 ; X X M p

w 200 hmd50o 20

Iytice ( S ( 0 (S0m 0

0 3 6 912 15 1821 012 1824 3036 42 4854 012182430 36 4248 54 Minutespost infection

FIG. 1. SynthesisofSP10-codedproteinsunderhmd' andhind conditions. Cultures(300 ml)wereinfected

with SP10 hmd' or SP10hmdl at 370C. Atthe times specified inthe graph, 25-mi samples were removed.

Extractswere preparedand assayedfordeoxynucleoside triphosphatase (A), serumblocking power(B), and

lyticenzymes(C). Symbols: (0)SPIOhmd';(0)SP10hmdl.

Complementation of hmd mutations. When

cells weredoublyinfected withhmdlandhmd2, the burst sizewas nogreater than that obtained

when cells were infected with only one of the

mutants (data not shown). These data suggest thathmdl and hmd2do notcomplement, which, inturn, suggests that both mutations arein the

samegene.

Synthesis of phage-coded proteins. The

condi-tionally lethalnature of hmd mutations

suggest-ed thatsomeimportantaspectsofSP10

develop-ment were aberrant in the absence of

hypermodification. Toseeif thedefect involved early or late functions, synthesis of SP10

pro-teinswascompared under hmd+ and hmd

condi-tions. In this and all other experiments, only dataobtainedwithhmdl phage will be presented unless statedotherwise.

A typical early enzyme, deoxythymidine

tri-phosphate nucleotidohydrolase (29), was

syn-thesized normally under hmd conditions (Fig. 1A), whereastwolatefunctions, structural

pro-teinsresponsible forserumblockingpower(Fig.

1B) and lytic enzymes (Fig. 1C), were

synthe-sized poorly. Therefore, under the conditions

defined, theabsence of hypermodification limit-edlate geneexpression.

Synthesis of phage-coded RNA. Since

hyper-modification is an event at the DNA polymer

level (39), it seemed reasonable to assumethat

thelow levels of lateprotein synthesis observed

under hmd conditions were due to poor

tran-scription oflate genes. To confirm this,

DNA-RNA hybridization competition experiments

weredone (29). For these experiments, parallel

cultures of B. subtilis W23were infected under

hmd+ conditions. One culture waslabeled with

[5-3H]uracil from 4to12min postinfection(early

[3H]RNA), and the other was labeled from 12 to 20min postinfection (late [3H]RNA) (29). Unla-beledcompetitor RNAs were isolated 12 and 20 min postinfection by hmd+ and hmdphage.

Unlabeled 12-min RNAs isolated from cells infected underhmd+ and hmd conditions were equivalent with respect to theirability to com-pete against early [3H]RNA (Fig. 2A). By the same token, hmd+ and hmd 12-min [3H]RNAs each competed against sequences that account-ed for about 60% of the label in late [3H]RNA (Fig. 2A). Thesedata indicate that 12-min RNAs present in hmd+- and hmd-infected cells were equivalent. This conclusion is supported bythe observation that unlabeled hmd+ and hmd 12-minRNAscompeted equallywellagainsthmd 4-to12-min [3H]RNAs (datanotshown).

On the other hand, hmd+ and hmd 20-min RNAs displayed marked differences (Fig. 2B). Specifically,sequencesamountingto30% ofthe label in late [3H]RNA were either absent from 20-min hmd RNA or presentin low concentra-tions. This difference is better illustrated by a "mixed competitor" experiment (2) in which late [3H]RNA was incubated simultaneously with asaturating concentration of hmd+ 12-min RNA and various concentrations of hmd+ or hmd 20-min RNA(Fig.2C). Overall,these data areconsistentwiththeinterpretationthat genes transcribeduniquelyattimeslaterthan 12minof anormal SP10program were weakly expressed underhmd conditions.

Kinetics of DNA synthesis. In many phage systems, expression of late genes is, in some measure,dependenton DNAsynthesis (15, 18). Although the screening procedure showed that some DNA synthesis took place under hmd conditions, the possibility still existed that the

http://jvi.asm.org/

[image:3.505.57.442.56.246.2]

(4)

ROLE OF Y-Thy IN SP10 DEVELOPMENT 639

100

z

cc

I

CcO

E D

D >E

E

--E

0

50I

0

.I

B._-A.

-l

I

-

I-I

a

0\

_ 0

0 0.5 1.0 0 0.5 1.0 0 0.10.20.3 0.

Competitor RNA (mg/ml)

FIG. 2. DNA-RNAhybridizationcompetition between early and lateSP10RNAs made under hmd' andhmd conditions. Early[3H]RNAwaslabeled from 4 until 12min postinfectionbyhmd+ phage; late [3H]RNAwas

labeled12 to 20minpostinfection by hmd+ phage (29). Unlabeled RNAs were isolated 12 or 20minpostinfection

byhmd+ or hmdl phage.Hybridizationmixtures contained 1 ,ug of[3H]RNAandfilters that contained 20p.gof

heat-denaturedphage DNA. (A)Hybridization competition of early[3H]RNA.Thespecific activity of[3H]RNA

wasca. 32,000cpm/4gand, intheabsence of unlabeledcompetitionRNA,ca.11% ofthe input was retained. Unlabeled RNA was extracted 12 min postinfection by hmd+ (0) or hmdl (0) phage. (B) Hybridization

competitionof late [3H]RNA. The specific activity of

[3H]RNA

was 23,500 cpm/p.gand, in the absence of

competitor, 33%of the input was retained by theifiter.Theunlabeledcompetitorswere(0) hmd+12-minRNA, (0)hmd+20-min RNA, and (0) hmdl 20-min RNA. (C) Mixed competition of late [3H]RNA. Late [3H]RNA was simultaneouslycompeted by 1 mg of hmd+12-minRNA and theindicatedamountsofeither(0) unlabeledhmd+

20-minRNA or(0)unlabeled hmdl 20-minRNA.

I D

apparently weak expression ofSP10late genes observed under hmd conditions could reflect someupsetin themechanism ofDNAsynthesis. Kineticsof DNA synthesiswasmonitored by the incorporation of [6-3H]uracil into

acid-pre-cipitable, alkali-labile material (23). By this

cri-terion, replication of the viral genome, under

hmd+ conditions, was firstevident 9 to12 min

postinfection and continued for the duration of

the latent period (Fig. 3). During an hmd pro-gram,DNAsynthesisstartedatthe normaltime,

proceededatabout half the normalrateuntil18

to21 min, and thenstopped ordeclined

many-fold (Fig. 3). The limited capacity for DNA

synthesis under hmd conditions could not be

attributed toadearth of substratesinasmuchas dATP,dGTP, dCTP,and HM-dUTPpools

actu-allyexpanded duringanhmdprogram(datanot

shown). It is worth noting that depletion of dTTPpools occurred normally underhmd con-ditions(datanotshown),which agreeswithour

observations that many early functions were

expressed normally underthese mutant

condi-tions(Fig. 1A and2A).

Intermediates inSP10replication. Inthe case

ofSPO0,severalkeyintermediates in the

replica-tioncyclehavebeenidentified

by

sucrose

gradi-entanalysis (24, 25). Initially, parental DNAis

converted to a form that sediments

through

neutral sucroseatca. 300S;this is the so-called

VF ("very fast") form. Approximately 10 min

later, theVFform disappearsand isreplacedby

100,000' '

50.000

0-,o'

/C

0

10,000_

5,000

L

0/

1,000_

500 0'

0@

100 *

[image:4.505.125.411.71.247.2]

0 6 12 18 24 30 36 42 48 54 Minutes postinfection

FIG. 3. Incorporation[6-3H]uracilintoSP10hmd+ andSPIOhmdl DNA.Cultures(12ml)wereinfected at37°C (29).Various timespostinfection,500-1l

sam-plesweretransferredtoafresh tube thatcontained20

p.

of

[63H]uracil

(200 ,uCi/ml). The samples were

labeled for 3 min and frozen inanacetone-solidCO2

bath. At the endof theexperiment,the

samples

were

thawed, and theamountof label in DNA was deter-mined(25).Symbols:

(0)

SP10hmd ;

(0)

SPIOhmdl.

VOL.42,1982

http://jvi.asm.org/

[image:4.505.297.449.441.590.2]

(5)

.2,000

4,000 T SP15

C. S1

2,000-splo

0 1 2 3 4 5

Distance from meniscus(ml)

FIG. 4. Sampleneutralsucrosegradientprofilesof intermediates in SP10 hmd+ DNAreplication. A cul-ture was infected with SP10 hmd+ at 370C. 32p; (10

tLCi/ml)

was added at 12 min. At the times given

below, [32P]DNA was extracted as described in the text.(A)[32P]DNAextracted 18 minpostinfectionthat wassedimented at 11,000 rpm for 4.5 h. (B)[32P]DNA

extracted 24 min afterinfection thatwassedimentedat 11,000 rpmfor 4.5 h. (C) [32P]DNAextracted 33 min after infection that was sedimented at 11,000 rpm for 15h.Arrowsmark the position of SP10 or SP15 virion

[14C]DNA(ca. 2,000 cpm) used as marker.

the more slowly sedimenting F form composed of linearconcatenates that initially contain up to 20 phage equivalents of DNA. Mature DNA (54S) appears roughly 5 min after the F form, and astheamount of mature DNAincreases,the average size of the F formdiminishes.

Asimilarsequence of events seems adequate to explain the replicative cycle ofSP10 hmd+. Figure4depicts sample gradient profiles. Figure 5presents a detailed summary of the amount of label inVF, F, and M (mature) forms ofSP10 hmd+ DNA. For these particular experiments, cells were labeled with3

Pi

beginning at the 12th

min postinfection. At various times, extracts

were prepared andsedimented through sucrose

at areducedspeedtodemonstratethe VF and F formsor at ahigher speedtoresolve the F and M forms. The internal markers were SP10 hmd+

virion [14C]DNA (80 megadaltons; 50S) and SP15 virion [14C]DNA (250 megadaltons; about lOOS).

Until 18 min, all [3H]DNA sedimented through sucrose at a rate 2.5 to 3 times faster than did the SP15 [14C]DNA marker; this evi-dently represented the VF form of SP10 hmd+ replicating DNA (see below). Unlike the situa-tion withSPO (24),the VF intermediate ofSP10 replication usually remained evidentuntil latein the program. Between 18 and 21 min, a form sedimenting in the vicinity of SP15 [14C]DNA appeared in SP10 hmd+-infected cells which presumably correspondedtothe F intermediate. TheFintermediatereached its maximal levelat

24 min and was maintained at approximately

that level for theduration ofthe infection pro-cess. At 24 min, the F intermediate had a

nominalsedimentation coefficientof 105to120S whichcorresponded to linear duplexesofmass 350 to 500 megadaltons (9). Experiments with alkaline sucrose gradients (not shown) suggest-ed, by comparisonwith othersystems (11, 25),

60,000

45,000~

30,000[

15,000[

0 6 12 18 24 30 36 42 48

Minutespostinfection

FIG. 5. Abundance of replicative intermediates various times postinfection by SP10hmd+. Sucrose gradient analyseswerecarriedoutasdescribedin the text.Duplicate100,-ulsamples of extracted [32P]DNA

werecentrifugedateither 11,000rpmfor4.5h(VFand

Fforms)or11,000rpmfor15 h (Fand M forms). The

totalamountof radioactivityrecovered inaparticular

formisrecorded. Since theFform is definedin both regimens used, the reported amount of label in this intermediate is theaverage oftwovalues. Symbols: (0) VF; (0)F;(O) M.

I

0

II

*-

I

1,

I X

http://jvi.asm.org/

[image:5.505.71.217.57.403.2] [image:5.505.274.423.368.573.2]

(6)

ROLE OF Y-Thy IN SP10 DEVELOPMENT 641 that the F intermediate was a linear

multige-nomic structure containing, at 24 min, between four andsix SP10 phage DNA equivalents. The MformofSP10hmd+ DNA appeared between 24 and 27 min postinfection and, thereafter, increased in level until it accounted for 60 to 70% of the [3H]DNA present at 51 min (i.e., 4 minbefore lysis).

Apulse-chase experiment was performed to determine the precursor relationship between VF,F,andMformsofSP10DNA.Label, as

[6-3H]HM-deoxyuridine,wasintroducedat 12 nun,

andincorporation was terminated at 17.5 min by the addition of dThd (25). Initially, all

radio-activity was recovered in the VF intermediate

(datanotshown). Between 18 and 24 min, label appeared in the Fintermediateatapproximately thesame rate atwhich itwaslostfrom the VF

intermediate.Labeldidnot appearin theMform

until 24to 27minpostinfection. These dataare consistent with the

progression

VF-3 F-- M.

Under hmdconditions,replicationof the SP10 genome waslimitedto anintervalof time when VF was the only observable intermediate in normally infected cells (Fig. 3 and 4). This

impliesthat, under hmd conditions,formationof

the VF intermediate should nottake place.

In-deed,all[3H]DNAmadebetween12and 18 min

ofanhmdprogram sedimented throughsucrose

at60S, i.e., 20%6faster than matureDNA (data

notshown). The fate of this intermediate hasnot been studied.

Characterization of the VF and F intermedi-ates. The VF form was stable in sarcosine,

which indicates that binding to the cell mem-braneplays little, ifany,roleinmaintainingthe structure of this intermediate (25). Similarly,

treatmentof VFwithRNaseAhad noobvious effect on the sedimentation properties of this intermediate(Fig. 6A).On the otherhand,

treat-mentwithpronaseconvertedVF toaformthat sedimented at a rate characteristic of the F intermediate (Fig. 6B); material produced by pronase treatment of VF will be calledF*. These dataindicate thatVF represented acondensed intermediate caused by

complexing

of DNAto

protein.

Novobiocin,aninhibitorofDNAgyrase(13),

immediatelyarrestedSP10DNA

synthesis

atall

timespostinfection(datanotshown).

Evidently,

negatively supercoiled

(underwound) regions

must be maintained for continuous replication

(13). Sedimentation coefficient of

negatively

su-percoiled DNA varies with ethidium bromide

concentration in a characteristic fashion (35).

Lowdye concentrationsunwind the

supercoils,

which results in a reduction in sedimentation

coefficient;higher

dye

concentrations introduce

positive supercoils with a

corresponding

in-crease in sedimentation coefficient. Forms of

DNA thatcannotbenegatively supercoiled (i.e., linear duplexes and open circles) also show characteristic changes in sedimentation coeffi-cient;inthiscase,the .sedimentationcoefficient decreases approximately linearly with increas-ing dye concentration.

Toshow thatour system would give reliable

results, we compared the effect of ethidium bromide concentration on the Escherichia coli nucleoid (41) in its native, negatively super-coiled form and its DNase I-relaxed form (Fig. 7A). For the native form, the characteristic biphasic curve, indicative of negatively super-coiledDNA, was readily obtained, whereas the sedimentation coefficient of the relaxed form decreased linearly across therange ofdye con-centrationstested.

The F and F* forms of SP10 hmd+ DNA responded to increasing dye concentrations as

relaxedforms(Fig. 7B). VF, onthe other hand,

responded in a peculiar fashion (Fig. 7B). For dyeconcentrations ofc2,ug/ml,the sedimenta-tion coefficient of VF decreased linearly in a manner similar to that obtained for F and F*. Higher dye concentrations, however,produced

vI ,

4,000 A. SP15

2.000-0 vi

4,000*

B. 'P

2,000

-ft

SP15

0

0 1 2 3 4 5

Distance from meniscus (ml)

FIG. 6. Effect of pronase and RNase A on sedi-mentation coefficient of the VF intermediate. Cells infectedat370C with SPIOhmd+received32P,(10 F.Ci/ ml)at12min.DNAwasextractedat18 min.A100- l

portionoftheextractwastreatedwith either RNaseA

(10pg/ml)orpronase (10,ug/ml)at370C for1 hand

thencentrifugedat11,000rpmfor 4.5 h.(A)RNaseA; (B)pronase.Symbols:(0)incubatedwithoutenzyme; (0)incubated with enzyme. Arrowslocatethe posi-tion of marker[14C]DNA.

VOL. 42,1982

http://jvi.asm.org/

[image:6.505.299.443.365.585.2]

(7)

'~~~~~~~~~~Al

0.8o

0_7 ,_,1_J.

01 23 4 5 6 7 8 0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8

[image:7.505.100.392.70.274.2]

Ethidium bromide(pg/ml)

FIG. 7. Effectof ethidium bromideonsedimentation behavior ofreplicativeintermediates. Extracts of

32P-labeled cells received the indicated concentration ofdyeandweresedimented ingradientscontaininganidentical concentration ofdye. (A)E. colinucleoid(41) inits native,negatively supercoiledform(0)orits DNase I-relaxed form(0).(B) The F(A),F*(A),and VF(0)forms ofSP10hmd+replicatingDNA. F* isgeneratedby

pronase treatment of VF (see text). (C) DNase I-relaxed VF. Extracts were first treated with increasing

concentrations of DNase I inanice-water bath for 10 min under conditions that limit nucleolytic damageto

single-strandbreaks(41).Symbols:(0)noDNaseI;(0)0.5 ,ugof DNaseI/ml;(m)1 1Lgof DNaseI/ml; (O)2 ,g of DNaseI/ml.

nofurtherreduction inthesedimentation coeffi-cientof VF. That istosay,atcohcentrationsof dye where positive supercoiling of restrained DNA would be expected, VF underwent no furtherreduction in sedimentation coefficient.

Such data could be interpreted to mean that VF DNA contains underwound and restrained regions that are covalently joined to relaxed regions. High concentrations ofdye introduced

positivesupercoiling intothe restrainedregions

whichcompensatedforthecontinuedunwinding of the relaxed regions. If so, introduction of sufficient single-strand breaks into VF DNA should relax therestrained regions, thereby re-sulting ina structure whosesedimentation coef-ficientdeclines linearlyatdyeconcentrationsof >2 ,g/ml. Controlled incubation of VF with DNase I did, indeed, generate a form that re-spondedtoethidiumbromide concentrationas a fullyrelaxed structure (Fig. 7C).Theamountof DNase I treatment required to give a fully re-laxed formofVF DNA did not affect

the

initial sedimentation coefficient (data not presented). Therefore, ifwe are correct inassumingthat VF DNAis acondensed, protein-complexed inter-mediate, thensupercoiling makes only a

nminor

contributiontothe rapidly sedimentingnature of this DNA. Similar conclusions have been reached for thecondensedform of T4 DNA (17). CharacterizationofDNAmade under hmd

con-ditions. The smallamount of DNA made under hmdconditions seems to beinaform thatdoes not occurin hmd+ infections (unpublished data described above). Inasmuchashmdl andhmd2 DNAsreplaceY-Thy with different bases(39),it followed that DNAmade under hmd conditions would have abuoyantdensity in CsCl different from that of hmd+ DNA and that this density

differentialcould beused asaprobe toexplore the relationship between parental and newly replicated strands generated under hmd condi-tions.

For these experiments, cellswere infectedat 37°C with hmd phage thatcontained [32P]DNA. At 12 min, [6-3H]uracil was added. DNA was extracted 20 min postinfection and banded in CsClgradients.In the caseofhmdl,about75% of parental [32P]DNA and all of the nascent [3H]DNA were denser than the SP10 hmd+ marker [14C]DNA (Fig. 8A). In the case of hmd2, 65% of the parental [32P]DNAand all of thenascent[3H]DNAwererecoveredin a band that was less dense than hmd+ marker DNA (Fig. 8B). These data are consistent with the observation thattheuniquebase inhmdl DNA is more acidicthan Y-Thy, whereas the one in hmd2 DNAisneutral (39).

When newly replicated hmdDNA was dena-tured andrebandedin neutralCsCl,theparental [32P]DNAwasrecoveredatthe samedensityas

http://jvi.asm.org/

(8)

3.000

2.000[

1,0001

3,000

2.000

IL0

1.000

01 2 3 4 5 6

W23 SP1O SPOl B. '

A_**w0~~~~~~w*-J

j|*

*~~~~~

1,500 Shiftupexperiments. Strictly speaking, the

re-sults shown above indicate that hypermodifica-tion is required for the first round of DNA synthesis. However, is hypermodification

con-tinuously required for DNA synthesis? To

ob-,i

tainsomeideaonthis point, shiftupexperiments

<, were

performed.

Cells wereinfected with hmd

phageat

30°C.

Various timeslater,asamplewas S00-2,000 removed andheat shockedat50°Cfor 2minto

,

rapidly

destroy

thethermolabile function. After

_1.000 < the heat shock, the samplewas placed at 37°C

and

32P,

was added. Control experiments

dem-O °0 onstrated that heatshock hadnoeffecton

hmd+

infections (datanotshown).

Only the results ofa typical experiment are

shown here (Fig. 10). In thiscase, shiftupcame

1,500 50

min

postinfection at

30°C,

which is roughly

equivalentto36minpostinfectionat37°C.When

an hmd+ culture was shifted up in the manner

1,000o described, incorporation of 32p; continued for

about 21 min,atwhich timethe culture lysed. In

the hmd culture, however,the apparent rate of

500 2,000 9

0

1,000 0

O1 2 3 4 5 6 Distancefrommeniscus (ml)

FIG. 8. Isopycniccentrifugation of DNA extracted fromcells infected with SP10 hmdl or SP10 hmd2.

Bacteria wereinfected at370C withhmdphage that contained[32P]DNA. At 12 min, [6-3H]uracil (10 ,uCi/ ml)wasadded. DNAwasextracted 18min

postinfec-tion and banded in neutral CsCl density gradients (6). (A) SPIO hmdl. (B)SP10 hmd2. Symbols: (0) parental

[32P]DNA;(0) newly replicated[3H]DNA; (0)

mark-er[(4C]DNAs runinaparalleltube.

denatured hmd+ DNA, whereas the nascent

[3H]DNAwasrecovered inaseparateband(Fig.

9). [3H]DNA annealed tobothH and L strands

ofSP10 hmd+ DNA (Table 2), suggesting that

newly replicated hmdDNA contained both

com-plementary strands. Thus, itappears that repli-cation under hmd conditions generated hybrid DNAmoleculescomposedofafully

hypermodi-fiedparental strand H-bonded to anascent

un-hypermodified strand. Inotherwords, fully

un-hypermodified duplexes couldnotbegenerated

in vivo.

In ethidium bromide, material made under

hmd conditions behaved as if it were fully

re-laxed,and treatmentwith pronase hadnoeffect

on its sedimentation coefficient (data not

shown).Therefore,itisnotimmediatelyobvious

whymaterial madeunderhmdconsistently sedi-mented more rapidly than matureDNA. Many possibilitiesexist, butnonehasbeentested.

2.000

1.000

1.500

1.000 L, x

0

500-2:000

2

10 JO

FIG. 9. Isopycnic centrifugation of denatured DNAextracted from cells infected withSP10 hmdl andSPIO hmd2. This is the companion experiment to the one reported in Fig. 8. Material with adensity

different from hmd+ native DNAwas isolatedfrom pooled gradient fractions and denatured by dialysis againstdistilledwaterfollowedby heatingfor 2minin

a boiling water bath (39). The denatured DNA was rebanded inneutral CsClgradients. (A) SP10 hmdl; (B) SP10hmd2.Symbols: (0) parental [32P]DNA; (0) newly replicated [3H]DNA; (0) denatured marker [14C]DNAsruninaparalleltube.

.A. W23 S;1o 2 ISPOll

LA

I

I.q$o %1.0~.I

I I

I 3

B ~ IQ

ILA

VOL. 42,1982

NL

9 0

.00 r.

3.000

2

a.

u

(L

11

http://jvi.asm.org/

[image:8.505.71.257.61.369.2]

(9)

TABLE 2. Annealingof DNA made in vivotothe L andH strands ofSP10hmd+ DNA'

Si nuclease-resistantcpm Labeled DNA_Labeled_DNA Unlabeled_DNA

Input

_cpm_b _Mock _bHbiie

hybridized" yrdzd

32P-labeledparental None 2,463 13 82

L strand 2,876 71 686

H strand 2,626 66 649

3H-labelednascent None 1,251 18 47

L strand 1,338 41 305

H strand 1,119 27 381

32P-labeledLstrandc None 2,476 41 93

L strand 2,509 73 176

H strand 2,818 81 1,572

32P-labeled H strandc None 2,011 45 87

Lstrand 2,256 53 1,703

H strand 2,256 58 201

aCellswereinfected,at

37°C,

withSP10hmdl thatcontained

[32P]DNA

and labeled with

[6-3H]uracil

(10

,uCi/

ml)from 12 until 18 minpostinfection.DNAwasextracted and banded in neutralCsCl density gradients(Fig. 8A). All material thatbanded more heavily thanmarker SP10hmd+ DNAwascollected and denatured by dialysisagainstdistilledwaterfollowedbyheatinginaboilingwaterbath for 2 min(36, 39).The denatured DNAs were rebanded in neutral CsCl; material corresponding to parental [32P]DNA and nascent [3H]DNA was reisolated(Fig. 9). Hybridization was carriedoutin 200 mM NaCl-10mMTris-hydrochloride (pH7.9).The reaction mixtures contained'0.1,ugof labeled DNA and 25jigof unlabeled DNA. The final volumewas200 ,u. The reaction mixtureswereincubatedat67°Cfor 4 h. This isnotenoughtimetogivemaximalhybridizationto unlabeledDNA,butself-annealingof the labeled DNA isnegligible duringthis interval. The reaction mixtures weredilutedtwofold with 60 ,uMZnSO4and 250 ,ug of denatured calfthymusDNA perml. The Si nuclease of

Aspergillusoryzae(250U)wasadded. The mixturewasincubatedat50°Cfor4h,and theamountof precipitable labelwasdetermined.

b Threehybridizationmixtureswereconstructed. Onewasprecipitatedimmediatelyand usedtodetermine the

input label. The secondwas mock hybridized byincubation at 2°Cfor4h andwasthen incubated with S1

nuclease at 50°C. The third reaction mixturewashybridizedat67°Cfor 4 h anddigestedwith S1 nucleaseat50°C.

cPrepared from SP10hmd+ [32P]DNA.

DNA synthesis decreased rapidly and stopped within 10 to 12 min of the shiftup. The hmd culturelysed 30min after the shiftup.

In a subsequent experiment, the shiftup was

performedasdescribed above.After shiftup, the

cellswerelabeled with32p,for 10min.TheDNA was extracted and centrifuged through neutral sucrose gradients. Approximately 50% of the

hmd+ [32P]DNAwas recoveredas theMform,

whereas the remainder sedimentedasthe Fform

(Fig. 11A). In the case of hmd [32P]DNA, no

morethan10% of the material sedimentedasthe

Mform(Fig. llB). When denaturedand banded

in neutral CsCl gradients,thebuoyant densityof

hmd[32P]DNA made after the shiftup was

be-tween that of fully hypermodified strands and

unhypermodified strands (Fig. 12), implying

that, under these conditions, the

unhypermodi-fied DNA was covalently linked to

hypermodi-fied DNA made before shiftup. All in all, these

data are consistent with the interpretation that concatenates containingunhypermodified DNA

areinefficientlycleavedtothe Mform.

Shiftdown experiments. Shiftdown experi-mentswereperformedtoseeifrestorationof the

thermolabile hmd functions permitted

comple-tion ofthe infection process. For the sake of

comparison, dataarealso shown forSP10repd, a heat-sensitive mutant defective in an as yet undefined aspectof replication.

Inthe caseofrepd, shiftdown resulted inthe reestablishment of DNA synthesis (Fig. 13) and production of viable phage (not shown); this mutantcould be incubatedat 37°C forup to42 minwithoutaffecting the subsequent production of DNA andphage at30°Cinanymajorway. In the case of hmd mutants, normal amounts of DNA andprogeny were generated at 30°C pro-vided the shiftdown took place <12 min post-infection at the nonpermissive temperature; thereafter, the amount of DNA and progeny generatedat30°C decreasedsharply (Fig. 13 and unpublished data). Under hmd conditions, a shiftdown laterthan18min postinfectionyielded virtuallynoDNA synthesis or progeny produc-tionat 30°C (Fig. 13 andunpublished data).

DISCUSSION

Eventhough itis a minorcomponentofSP10

DNA (37, 39), Y-Thy is clearly essential for intracellular development. At nonpermissive temperatures, SPIO hmdl andSP10hmd2 gener-ateDNAs in whichdifferentintermediates

http://jvi.asm.org/

[image:9.505.48.447.87.238.2]

(10)

80,000

60,000

a. 40,000[

20,000

0

0 10 20

Minutesafter shift-up"

FIG. 10. Effectofa shiftuponDNA synthesis in cellsinfected with SP10hmd+andSP10hmdl.

Bacte-ria were infected at 30°C, which is a permissive temperaturefor hmdl. At 50 min, the culture (5 ml)

was transferred to a waterbath (50°C) and swirled gently for 2min. The culturewasthenplacedat37°C. Periodically, 500-p,lportions werepulse-labeled with

[6-3H]uracilasdescribed in Fig.3.Symbols:(0)SP10 hmd';(0)SPlO hmdl; (A)SPnorepd,aheat-sensitive mutantthatisdefective in DNA synthesisat37°C(see text).

mulateinplace of Y-Thy (39),yetboth mutants showbasicallythesameproperties withrespect tointracellular development (thispaperand

un-published data). Therefore, intermediatesin hy-permodification cannotsubstituteforY-Thy.

Under hmd conditions, replication of SP10 DNAstopped ordeclined manyfoldafter afew

minutes (Fig. 3). The principal, if not only, product of SP10 DNA replication under hmd conditions consisted ofduplexes composed of parental (hypermodified) strands H-bonded to nascent(hmd)strands(Fig.8 and9).DNA-DNA annealingexperiments (not shown)indicatethat hmd duplexes can be formed quite readily in

vitro, sothereis no reasontobelieve thatany

intrinsic properties ofhmd strands per se pre-ventformationof suchduplexesinvivo. Conse-quently, it seems reasonable to propose, as a

working hypothesis, that the replicational

ma-chinery of SP10 cannot use an hmd strand as

template,soreplicationstopsat thepointwhere

these strands wouldhave toserve as templates

tosustain DNAsynthesis.

Shiftdown experiments indicated that the

thermolabile hmdfunction could notbe rescued

later than 18 min postinfection (Fig. 13).

Al-though it could be thatinactivation of the

ther-molabile hmd function becomes irreversible at

18min, another explanation seems equally val-id. The time at which hmd functions become unrescuable by shiftdown corresponds to the time at which DNA synthesis stops at the non-permissive temperature (Fig. 3), i.e., the time at which hmd+-hmdhybrid duplexes are complet-ed (Fig. 8 and 9). Such hybrid duplexes are evidently inert as templates for DNA synthesis

(Fig. 3 and 10). At 37°C, hypermodification of

nascentSP10 hmd+ DNA occurs no more than 8 to 10s afterpolymerization (39). It couldwell be that hypermodification can occur only within a short time interval of DNA synthesis and that,

beyond this interval, hypermodification cannot

occureven if thehmdfunctions are restored. In thiscontext, it is worth noting that hmd+-hmd

hybrid duplexes cannot be rescued by fection with hmd+ phage even though

superin-fectingphagedevelops normally (H. Witmer and

M. Franks, unpublished data).

Insofarasreplication under hmd+ conditions

was concerned (Fig. 4 and 5), SP10 showed

basically the same intermediates as SPOl (25).

10,000 _ SP1O SP15

A. * 4

5,000

10,000 SPl

B.

5,000-0 1 2 3 4 5

Distance from meniscus(ml)

FIG. 11. Neutralsucrosegradientcentrifugationof DNA made after a shiftup. Cultures (5 ml) were

infectedat30°CwitheitherSP10hmd+orSP10hmdl. At 50 min, the cultures were shifted up to 37°C as

describedinFig.12.32P,(10FCi/ml)wasadded 1 min

aftertheculturewasplacedat37°C.DNAwas extract-ed 10 minlater and sedimented through neutral

su-crosegradientsat11,000rpmfor 15 h.Onlythe F and M forms of DNA (see text) are resolved by this

procedure. (A) SPIO

hmd';

(B) SPIO hmdl. Arrows locate thepositionof marker

[14C]DNAs.

0

t

\

IE

O.I .h *...x.s

VOL. 42,1982

http://jvi.asm.org/

[image:10.505.91.237.70.290.2] [image:10.505.299.438.356.578.2]

(11)

a-0

ctJ

splo

spol

3 4

2,000

Ci:

1,000 c0

[image:11.505.97.394.64.221.2]

-Distance from meniscus (ml)

FIG. 12. Isopycnic centrifugationofdenatured DNA made afterashiftup.Material that sedimentedthrough

sucroselikethe Fintermediate(Fig. 11)wasreisolated,denatured,and banded in neutralCsCldensity gradients

(6, 39).Symbols: (0) [32P]DNAextracted fromSP10hmdl-infectedcells;(0)denatured marker

["4C]DNAs.

Therefore, atthepresenttime, it doesnot seem

necessary to assumethat replication of hyper-modified DNAs requires novel intermediates. However,onepointof interest should be noted.

Replication of the SP1O genome is novobiocin

sensitive, yet underwinding of SP1O DNA is

seemingly limitedtorelatively small domains in

the VF intermediate (Fig. 7B). Maintenance of theseunderwoundregions, whichweassume to be essential forreplication,requires continuous presenceofproteinsbecause F* DNA(Fig. 6B)

behaves as if it were fully relaxed (Fig. 7B).

Whencellswereinfected underhmd

conditions,

SP1ODNA assumednothing comparable to the

VFform (Witmer and Franks, ourunpublished

data), so it is possible that some proteins

re-quiredtogenerateand maintain theunderwound

regions requireafully hypermodifiedDNA. We arecurrently attempting toidentify SPlO-coded DNA-binding proteinsto seeifanybind

prefer-entiallytofullyhypermodifiedDNAs.

Consider-ing the complex and reactive nature of the substituents in hypermodified bases (37), they would be ideally suited as target molecules for DNA-protein interactions.

DHP-uracil, the hypermodified base in SP15 DNA(4), isrequired for proper DNA maturation or phage assembly (6, 36). Evidence for this conclusion stemsfrom the fact that, under con-ditions where only5% of the replicating pool has anormal DHP-uracil content, over50% of the DNA packaged into phage is normal with re-spect to base composition (6, 36). A similar conclusion seems to hold for Y-Thy in SP1O

DNAinasmuchas little, if any,hmdDNA made aftera shiftup matured (Fig. 11). Late protein synthesis isessential for DNA maturation, since

chloramphenicol, added 12minpostinfection by

hmd+ phage, allowsthe VF and Fintermediates toaccumulate but no mature DNA is generated (datanotgiven). Atpresent, it is not known if

SP1Omature DNA is cut fromtheconcatenates beforepackaging, as is the casewith SPO1 (24), or, asin thecase ofcoliphages 0X-174, T4, and

X (12,33, 38), cleavagefromtheconcatenate is the final step ofpackaging. Eitherway, one or

more late proteins must interact with the F

intermediate to generate mature DNA and,

when the F intermediate contained

unhyper-modified regions (Fig. 12), this interaction was

inefficient(Fig. 11).

60,000[

Shift-down(minutes post infection)

FIG. 13. Restoration of thermolabile functions: Shiftdown experiment. Parallel 12-ml cultures were

infectedat37°Cwith eitherSP10hmdlorSP10repd

(see text). Varioustimes postinfection, 1-ml samples were transferredto afresh tube and incubated, with

gentleagitation,at20°C for2min. The sampleswere

placedat30°C, and32Piwasaddedtoafinal

concentra-tionof10,Ci/ml.The sampleswereincubatedat30°C until all the SP10 repd-infected cultures lysed. The

amount of label in DNAwas determined (23). SP10 hmdl-infected culturesfailedtolyse orlysedpoorly if

shiftdown came later than 18 min after infection at

37°C. Symbols: (0) SP10repd; (0)SP10hmdl.

-jr--a-I, E

1;--111

\\

IIz

I *

2

(L

L)

http://jvi.asm.org/

[image:11.505.270.420.395.556.2]

(12)

Thus far, it is clear that Y-Thy is required for replication as

well

as cleavage of mature DNA

from concatenates. In turn, it is proposed that both defects arise through faulty interaction of certain key proteins with unhypermodified DNA. It is difficult to judge from the available data if Y-Thy plays a role in differential tran-scription of theSP10genome. Although it is true that SPIO late genes were weakly expressed under hmd conditions (Fig. 2), this could just as easily be attributed to the low levels of DNA synthesis that occur under those conditions (Fig. 3), since late transcription of numerous phage genomes is, to some extent, coupled to

replica-tion (15, 20). The importance of unusual bases in differential gene expression varies somewhat from system to system. In the case of coliphage T4, late genes can be transcribed only from an HM-cytosine-containing template (42). Indeed, the alc gene product of phage T4 seems to modify

cellular

RNA polymerase such that HM-cytosine-containing DNA is a preferredtemplate (34). On the other hand, Thy can substitute for, at least, 20% of the HM-uracil in 4e DNA without affecting phage functions (28).Similarly, Thy can replace up to 70% of theDHP-uracilin

SP15DNA without affecting gene expression (6, 36). Thus, there is no a priori reason to believe that Y-Thy would play a direct role in transcrip-tion of the SP10 genome.

ACKNOWLEDGMENT

This research was funded by grant PCM 79018-03 from the National Science Foundation.

LITERATURECITED

1. Alegria, A. H. 1967. Hydroxymethylation of pyrimidine nucleotides with formaldehyde. Biochim. Biophys. Acta 149:317-324.

2. Bolle, A., R. H. Epstein, W. Salser, and E. P.Gelduschek. 1968. Transcription during early bacteriophage T4 devel-opment: synthesisand relative stability of early and late RNA. J. Mol. Biol. 31:325-348.

3. Bott, K., and B.Strauss.1965.The carrier state ofBacillus subtilisinfected with the transducing bacteriophageSP1O. Virology 25:212-225.

4. Brandon,C., P. M.Gaflop,J.Marmur,H. Hayashi,and K. Nakanishi. 1972.Structure of a new pyrimidinefrom Bacillussubtilisphage SP-15. Nature (London) New Biol. 239:70-71.

5. Brown, N. C. 1970. 6-(p-Hydroxyphenylazo)uracil: a

se-lective inhibitor of host DNAreplicationinphageinfected Bacillussubtilis. Proc. Natl. Acad. Sci. U.S.A. 67:1454-1461.

6. Csella,E.,0.Markewych,M.Dosnr, and H.Witiner. 1978. Production and expression of dTMP-enriched DNA of bacteriophage SP15.J. Virol. 28:753-766.

7. Dawid, I. B., D. D. Brown, and R. H. Reeder. 1970. Composition and structureofchromosomal andamplified ribosomal DNA's of Xenopus laevus. J. Mol. Biol. 51:241-260.

8. DeMars, R.I. 1955. The production of phage related material when bacteriophage developmentisinterrupted by proflavine. Virology 1:83-99.

9. DoneUl, G., E. Dore, C.Frontali, andJ.Grandolfo. 1975.

Structure and physico-chemical properties of bacterio-phageGIll. A homogeneous DNA of molecular weight 5

x 16§.J. Mol. Biol.94:555-565.

10. Dosmar, M., and H. Wltmer. 1979. An atypical deoxynu-cleoside triphosphate in Bacillus subtilis infected with bacteriophage SPlOc. Curr.Microbiol. 2:361-364. 11. Frankel,F. R.1968. Evidence for long DNA strands in the

replicatingpool after T4 infection. Proc. Natl.Acad. Sci. U.S.A.59:131-138.

12. Frankel, F., M. Batcheler, and C. Clark. 1971. The role of gene 49in DNA replication and head morphogenesis in bacteriophageT4. J. Mol. Biol. 62:439-463.

13. Gdlert,M., M. O'Dea, T.Itoh, and J.Tomizawa. 1976. Novobiocin and coumermycin inhibit DNA supercoiling by DNA gyrase. Proc. Natl. Acad. Sci. U.S.A. 73:4474-4478.

14. GIllespie, D., and S. Splgeman. 1965. A quantitative assayfor DNA-RNA hybrids with DNAimmobilizedon a membrane. J. Mol. Biol. 12:829-842.

15. Glamberg, J., R. A.Somiany, and C. R. Stewart. 1977. Selective screening procedure for the isolation of heat-andcold-sensitive, DNA replication-deficient mutantsof bacteriophage SPOl and preliminary characterization of the mutants isolated. J. Virol. 21:54-60.

16. Guha, A., and W. Szybalski. 1968. Fractionation of the complementary strands of coliphage T4DNA based on the asymmetric distribution of the PolyU andPoly U, G bindingsites. Virology 34:608-616.

17. Hamilton, S., and D. E. Pettljohn. 1976. Properties of condensed bacteriophage T4 DNAisolated from Esche-richia coli infected with bacteriophage T4. J. Virol. 19:1012-1027.

18. Hemphill, H. E., and H. R. Whiteley. 1975. Bacterio-phages of Bacillus subtilis.Bacteriol. Rev.39:257-315. 19. Jayaraman,R., and E. B. Goldberg.1969. Agenetic assay

for mRNA's of phage T4. Proc. Natl. Acad. Sci. U.S.A. 64:198-204.

20. Kahan, E. 1966. A genetic study oftemperature-sensitive mutants of thesubtilis phageSP82.Virology 30:650660. 21. Kropinskl, A. M. B., R. J. Bose, and R. A. J. Warren. 1973. 544-Aminobutylaminoethyl)uracil, an unusual py-rimidine fromthedeoxyribonucleicacid ofbacteriophage 0W-14. Biochemistry12:151-157.

22. Legault-Demare, L., and G. H.Chambliss. 1974. Natural messenger ribonucleicacid-directed,ceil-free protein syn-thesizing system of Bacillus subtilis. J. Bacteriol. 120:1300-1307.

23. Lembach, K., and J. M.Buchanan.1970.Therelationship of protein synthesis to early transcriptive events in bacte-riophageT4-infected Escherichia coli B. J. Biol. Chem. 245:1575-1587.

24. Levner, M. H. 1972. Replication of viral DNA in SP01-infected Bacillus subtilis. II. DNA maturation during abortiveinfection. Virology 48:417-429.

25. Levner, M. H., and N. R. Cozzarefl. 1972.Replicationof viral DNAinSPOl-infectedBacillus subtilis.I. Replica-tiveintermediates.Virology 48:402-416.

26. Lowry, 0. H., N. J.Rosebrough, A. L. Farr,and R. J. Randall. 1951. Protein measurement with Folin phenol reagent.J. Biol. Chem. 193:265-275.

27. Maltman, K. L., J. Neuhard, H. A.Lewis, and R. A. J. Warren. 1980. Synthesis of thymine and a-putrescinylthy-mine in bacteriophage 0W-14-infected Pseudomonas aci-dovorans. J. Virol. 34:354-359.

28. Marcus, M., and M. C. Newlon. 1971. Control of DNA synthesis inBacillus subtilisbyphageoe.Virology 44:83-93.

29. Markewych, O., A.Boghosan, M.Dosmar, D.Ende, and H. Witmer. 1977. SP-10 bacteriophage-specific nucleic acid and enzyme synthesis in Bacillus subtilis W23. J. Virol. 21:84-95.

30. Markewych,O.,E. Casella,M.Dosmar,and H. Wltmer. 1979. Deoxythymidine nucleotide metabolism inBacillus subtilis W23 infected withbacteriophage SPlOc: prelimi-naryevidence that dTMP inSPlOc DNA issynthesizedby VOL.42,1982

http://jvi.asm.org/

(13)

anovel,phage-specific mechanism. J. Virol. 29:61-68. 31. Neuhard, J., K. L.Maltman,and R. A.J. Warren.1980.

Bacteriophage 0W-14-infected Pseudomonas

acidovor-ans synthesizes hydroxymethyldeoxyuridine triphos-phate. J. Virol. 34:347-353.

32. Neville,M.M., andN. C. Brown.1972. Inhibition ofa

discrete bacterial DNA polymerase by 6-(p-hydroxyphen-ylaso)-uracil and 6-(p-hydroxyphenylazo)-isocytosine. Nature(London) NewBiol. 240:80-82.

33. Sinsheimer, R. L.,R. Knippers, and T. Komano. 1968. Stages in thereplication of bacteriophage 0W174 DNAin

vivo.Cold Spring Harbor Symp. Quant. Biol. 33:443-448. 34. Snyder, L., L. Gold, and E. Kutter. 1976. A gene of

bacteriophage T4 whose productprevents truelate

tran-scription on cytosine-containing T4 DNA. Proc. Natl. Acad.Sci. U.S.A. 73:2098-2102.

35. Vinograd,J., J.Lebowitz,R.Radloff,R. Watson, and P.

Laipis.1965. The twisted circular form of polyoma viral DNA. Proc.Natl. Acad. Sci. U.S.A. 53:1104-1111. 36. Walker, M. S.,andM.Mandel.1978. Biosynthesis of

5-(4',5'-dihydroxypentyl)uracilas anucleosidetriphosphate inbacteriophageSP15-infected Bacillus subtilis. J. Virol.

25:500-509.

37. Warren, R.A.J. 1980.Modified bases in bacteriophage DNA's. Annu.Rev.Microbiol. 34:137-158.

38. Weissbach, A., P. Barti, and L. Salzman. 1968. The structureofreplicative lambda DNA-electron microscope studies.Cold Spring Harbor Symp. Quant. Biol. 33:525-532.

39. Witmer, H. 1981. Deoxythymidylate and the unusual nucleotide inmatureSP10 phage DNAaresynthesized by postreplicational modification of 5-hydroxymethyldeoxy-uridulate. J. Virol. 39:536-547.

40. Witmner, H., and M. Dosmar. 1978. Synthesis of 5-hydroxymethyldeoxyuridine triphosphate in extracts of SPlOc-infected Bacillus subtilis W23. Curr. Microbiol. 1:289-292.

41. Worcel,A., andE. BurgL.1972. On thestructureof the folded chromosome of Escherichia coli. J. Mol. Biol. 71:127-147.

42. Wu,R., and E. P.Geiduschek. 1975.Therole of replica-tionproteins in theregulation of bacteriophage T4

tran-scription. I. Gene 45 and hydroxymethyl-C-containing

DNA.J. Mol. Biol. 96:513-538.

http://jvi.asm.org/