Analysis of coliphage lambda mutations that affect Q gene activity: puq, byp, and nin5.

(1)

Vol.30,No. 1

JouRNAL OFVIROLOGY, Apr.1979, p. 1-13 0022-538X/79/04-0001/13$02.00/0

Analysis of

Coliphage

Lambda Mutations That Affect Q Gene

Activity: puq, byp, and nin5

NAT STERNBERGt* D LYNNENQUISTt

Laboratory ofMolecular Genetics, National Institute of Child Health and Human Development, National

Institutesof Health,Bethesda, Maryland 20014 Receivedfor publication 13 September 1978

We describe in this paper the isolation and characterization of a class of

mutations, designatedpuq, thatallow phage X to grow better under conditions

that limit the synthesis of the phage Q gene product. These mutations were

locatedbetween phage genes P andQ, a region of the X chromosome containing twogeneN-independent mutations, nin5and

byp,

that we also show tobe puq mutations. Whereas the puq-3 and puq-16 mutations probably map under the nin5 deletion, the byp mutation maps between this deletion and the QX-Q<80

crossoverpoint.These mutationslikelyactbyincreasing thesynthesis of the Q

gene product. We demonstrate that the clear-plaque phenotype and reduced

lysogenization frequency of byp mutants depend on increased Q gene activity. The significance of these results in understanding how transcription proceeds

throughtheP-Q region of the A genome is discussed.

When coliphage A infects Escherichia coli,

oneoftwomodesofphagedevelopmentensues. In thelysogenic mode, the phage coexists with

thehost withoutkilling it.Inthelytic mode, the phage rapidly reproduces and, in the process,

kills the host. The lysogenic response is the result ofrepressionof mostphage functions by

the phage cI gene product. The faithful

segre-gationofrepressed phageDNA(prophage)

dur-ing cell division is ensured by integration of

prophage DNA into the hostgenome, whereit isreplicated passively bythe host. In contrast,

during the lytic response, the phage DNA is actively replicatedand transcribed, phage par-ticlesareassembled,andcell lysisoccurs.Inthis

report, weanalyze the action of several phage functions involvedinlytic growth.

Regulation of transcription during the lytic growth cycle dependsonthe actionof the prod-uctsofphagegenes Nand

Q (16).

Inthe absence of theN gene

product,

transcription

that

initi-atesatpromoterstothe left

(pL)

and

right

(pR)

of the immunity region

(Fig.

1) terminates at

sites tLand tR,,respectively (20, 21, 27,

37).

In the presence ofthe N geneproduct,

transcription

continues beyond these termination sites. By mutation,newpromoters(e.g.,c17or

ri5b)

have been isolated that initiate transcription to the

rightoftR1 (13, 23,25,26). This abnormal

tran-tPresent

addres:

CancerBiology Program, NCI Frederick CancerResearchCenter, Frederick,MD21701.

tPresentaddress:Laboratoryof MolecularVirology, Na-tional Cancer nsitute, National Institutes ofHealth,

Be-thesda,MD20014.

scription terminates beforereaching gene Q if theN geneproductisabsent, but continues into

geneQif the N geneproduct ispresent(9).This

resultsuggested the existence ofasecond right-ward ternination site in addition to

tR,.

This

site, designated tR2,waslocated between genes P and Q by analysis of two novel mutations, nin5(a5.4%deletion) and

byp

(5,8, 14, 19,28). Phage carrying eitheroneofthesemutations no

longer requirethe N geneproductfor transcrip-tion through tR2, although such read through wasenhanced in the presence of this product. Phagecarryingthenin5 deletionneed noother mutations to grow in the absenceof theNgene

product, whereas the c17 promoter mutation must be included with the

byp

mutation to achieve N independence. Because the N gene

productmustalsoactattRl for normal

synthesis

ofphagereplication proteins,it issurprisingthat

nin5phagegrowinthe absenceofthisproduct.

Thenin5 mutationmighteliminatethis

poten-tialreplication defect byeitherpermittingmore efficient

utilization

of the available

replication

proteinsorincreasingtheirsynthesis.

The transcription of latephagegenes (genes

S, R, and A through J [Fig. 1]) is initiated

primarily atp'R

andrequirestheQgeneprotein

(18). The modeofaction of theQgeneproduct

isunclearalthoughit has beensuggestedthat it

acts like the Ngene product in allowing tran-scription to continue beyond termination sites downstreamfromP'R (28).

Inthispaperwe describe theisolation, map-ping, and characterization ofmutationsthat

on November 10, 2019 by guest

http://jvi.asm.org/

(2)

ENQUIST

an tL PL

A J lnt rMed redp Wm N ci O P a S R

PR tmi c617 dCSb tR2 PR

L

L.hiimnht43 J I

lMwuMq21

..le nl

I

'l _a opue-3 16

-I

I

K

3.2 90.6

[image:2.501.73.397.66.231.2]

Pam3Paau/ byp OAO wm21QmM57

FIG. 1. Geneticmap ofphage A. Geneticdistances are not drawn to scale, although the relative positions

ofthe genes andsitesarecorrect. The structural genes are placed between the two DNA strands, and the

genetic sitesareplacedabove and below the DNA strands.pL, PR,andP'Rrefer to the phagepromotersites; tL,tRi,andtR2refertothephage terninator sites;cl7and ric5b referto newpromotermutations;and att refers tothephage attachmentsite. Thetwoarrowsindicatethedirection ofphagetranscriptionduringlyticgrowth.

The region betweenphagegenesP andQisexpandedwith the A map coordinates of the left and right ends

ofthe nin5deletionandtheQA-Q80crossoverpointshown(11). low thephagetogrowbetter whenthesynthesis

of the Q gene product is limiting. Included

among these mutations are puq mutations, whose isolationsaredescribedinthispaper,and

the nin5andbypmutations.

MATERIALS AND METHODS

Bacterial andphage strains.The bacterial and

phagestrains used in this paper are listed in Table 1.

Media. Tryptone broth contains 10 g of tryptone

(Difco),5gofNaCl,and 5 ml of 1 MTrisperliterof water.TBMMB1istryptone brothsupplementedwith 0.4%maltose, 10mMMgSO4,and1

,g

of vitaminB1

perml. TB top and bottomagarsaretryptonebroth

supplementedwith 7 or 12 g of agar(Difco)perliter, respectively. Eosin methylene blue-agar (EMBO) plates were prepared as previously described (30). TMGbuffer contains 10 mMTris-hydrochloride, pH

7.4, 10 mMMgSO4,and 0.1%gelatin.

Standard phageprocedures. Phagestocks were

prepared from purified plaques by the plate lysate method (29).Single-step growth experiments(33) and

phage crosses (32) wereperformed aspreviously de-scribed. Whenpotentialrecombinantplaquesfor per-tinent mutationswerescreened, thefollowing pheno-types were used:(i)c1ts857,clearplaquesat

420C;

(ii) by, clear plaques at32°C and the ability to form plaques on strains NS377 and NS577 (33); (iii) nin5, the ability to form plaques on strains NS377 and NS577 and a decreased sensitivity of the phage to EDTA inactivation (24,34); and (iv)puq, theability

to allow a Qam phage to form a plaque on strain CA169(seebelow). Phageusedtoprepare extracts for thelysozymeassay werepurifiedfromplate lysates by

banding in aCsCl equilibriumdensity gradient (34).

Measurementoflysozymesynthesis.A culture of strain W3350 growinginexponential phaseatlol celsper ml was concentratedbycentrifugation (SS34

Sorvallrotorfor10minat6,000 rpm), and thepellet

was suspended in 1/5 volume of a 10 mM MgSO4

solution. A 2-mlaliquot of this cell suspension was infected withphageat amultiplicity of infection (MOI) of fourphagepercell. After5minat370Cforphage adsorption,the infectedcellsweredilutedwith 8 ml of tryptonebroth, andthe culturewasgrownwith

vig-orousaeration at37°C.Samples (1 ml)wereremoved atdesignated intervals, chilled to4°C, andsonically

treated with the fine probe of a Branson Sonifier (setting 3) for two 20-s intervals. The debris was removedbycentrifugation (15min at6,000 rpm in an SS34Sorvall rotor),and thesupernatantwasused to assaylysozymeaspreviouslydescribed(33).One unit oflysozymeis that amountofenzyme whichproduces

0.001decrease in cell absorbanceat 600 nm(A6s)in1

min. The extract was adjusted so that the enzyme level used in each assay reduced cell absorbance 0.1 to 0.2

Awos/inin.

Thebackgroundlysisof sensitized cells alone was 0.0005 Awo/min, and the activity is

ex-pressedasunitsper 100

pl

of infected cells.

Measurement ofIy8ogenization frequency. A culture of strain YMC or 594growinginexponential phase at 10'cells per ml inTBMMB1 at 320Cwas concentrated by centrifugation, and the pellet was

suspendedin 1/10 volume ofa10mMMgSO4solution. Thecellswerestarved for30min at320C,and then 0.1-mlsamples(10W cells) wereinfectedat anMOI of either 5 or 0.2phageperinfected cell. After10min for phage adsorption at320C,dilutionsof these infected

cells were spread on EMBO plates on which 108 Ah80(attA-int)Ac selector phage had previously been

spread. These plates were incubated overnight at

320C,and colonieswerescoredasAlysogens.As the

infecting A phage in these experiments carried the cIts857mutation,wecouldindependently confirmthat thesecolonieswerelysogens bytheirinabilitytogrow at420C.Thelysogenizationfrequencyis the number oflysogensper number of infected cells, the latter

beingthecelltiteratthemomentof infection in the MOI=5experimentand 20% of this titer in theMOI

=0.2experiment.Wealso measured thelysogenization

frequency byspreadingtheinfectedcellson EMBO

I

J. VIROL.

http://jvi.asm.org/

(3)

PHAGE A MUTATIONS AFFECTING Q GENE ACTIVITY

TABLE 1. Bacteria and phagestrains

Strain Pertinent genotype Comments/source/reference

supF

YMC(A) YMC(Ximm434) YMC(Ximm2lSam7)

sup+8trr sup+

supE

supC

sup+nusA-1rif'-2

NS377(.o80psu3+)

N215(sup+)

OW

recA

XcIts857Pam80+(Qam501-attR)'

AcIts857Pam80+(Qam501-attR)"

X(attL-Pam80)AQam21+

A(attL-Pam80)AQam21+

X(attL-Pam80)AQam21+ X(attL-Pam80)Anin5Qam21+ X(attL-Oam29)APam3+nin5Qam21+

Phage

AcIts857Qam2l XcIts857Qam57 XcIts857Qam5Ol

AcI+Qam21

Aimm2lcItsSam7

AcIts857byp

Ximm434byp

XcIts8570am29byp

AcIts857b3pQam21

XcIts857nin5 AcIts857PamSOnin5 XcIts857nin5Qam2l

AcIts857nin5byp

AcIts857nin5bypQam21 AcIts857Pam80nin5byp AcIts857Pam3nin5Qam2l

XcIts857puq-3 (orpuq-16)Qam2l XcIts857puq-3 (or puq-16)

AcI+puq-3(orpuq-16)Qam2l Ah8Oatt8Oimm434(QSR)4080 Ah8O(attAint)ac

12;usedto assayAammutants

6 7 3 4

33;used to assayAbypand Anin5 phage

33; used to assay Abyp am or

Anin5am mutants 35

Phage strains were either

ob-tained from the National In-stitutes of Health collection

orwereconstructed by

recom-bination betweenappropriate A mutants, using selection procedures described in the text.

plates without selector phageandindividually testing theresulting coloniesfortemperaturesensitivity.The results obtainedbythe two methodswere identical.

RESULTS

Isolation ofpuqmutants.Weisolated

mu-tationsthatspecificallyaffect either the synthe-sis of theQgeneproductoritsactivityasfollows.

AQam mutants fail to form plaques on ahost

containing a supC ochre suppressor (strain CA169). Since these same phage form normal

plaquesonahostcontaininganamber suppres-sor (YMCsupF) that inserts the same amino

acid (tyrosine) at the amber site as the supC

suppressor,thegrowth defect of XQammutants

in strain CA169 is probably a consequence of theirinabilitytomakesufficientQgeneproduct. The levels of amber suppression in strains CA169and YMCare,respectively, about15and 50%(15).

WhenXQam2l isassayedonCA169, the

effi-ciency of plaque formation is about 6 x

1O-'

relativetothatonYMC. Of theserareplaques,

three classescanbe distinguished onthe basis

ofplaquesize onCA169. The first class, repre-senting about 10% of the plaque forners,

con-tainsphagethat formnormal-sizeplaqueswith

thesameefficiencyonstrainsCA169, YMC,and

594

(sup').

Thesephagearepresumablyamber'

revertants. Asecondclass ofphage,also

repre-Bacterial YMC

NS62 NS617 NS61

594 W33S0

C600 CA169 NS377 NS577 NS460

N205

SA268 SA431 SA613 SA741 SA307

DC401

DC402

VOL. 30,1979 3

http://jvi.asm.org/

[image:3.501.61.464.64.524.2]

(4)

sentingabout 10%of theplaques, forms pinpoint

plaques on strain CA169 and 594, but normal-size plaqueson strain YMC. The efficiency of plaqueformationis about the same inall cases.

Thesemutants areprobablythe same as the "Q-independent(qin)"phage studied by Herskowitz and Signer (17) that exhibit an increased level of Q-independent transcription of late phage genes. These were not further characterized. The thirdclassofphage mutants, representing the remaining 80% of the plaque formers on strain CA169, made normal-size plaques on YMC, small plaques with unit efficiency on

CA169, andplaques with lessthan

10'5

efficiency on 594. Our hypothesis was that these phage

acquired a second-sitepuq mutation, enabling themtogrow better when eitherthe synthesis oractivity ofthe Q gene product waslimiting. SinceApuqQam2l phagedonotformplaqueson

the

&itp+

host, we suggest that at least some Q gene product is essential for improved

phage growth. Three representative class 3

Qam2l pseudorevertants of independent

origin

(Qpuq3Qam21,

Apuq4Qam21,

and

Apuq16Qam21)werechosen forfurther analysis.

Specificity ofpuqmutations.The puq

mu-tationsmight allow AQam2l to grow better in

strainCA169because thesemutationsalterthe

Q-gene protein so thatits activity is increased orbecause these mutations increase thelevelof

Q gene expression. If the first explanation is correct, wemightexpect thatsuppression of the

Qam growth defect byaparticularpuqmutation

would bespecificforthemutationschosen;that

is, the protein alteration produced bythe puq mutationmight onlyincrease the specific

activ-ityof the protein whentyrosine is inserted by

an amber-suppressingtRNA at a particular

lo-cationin theprotein. If the secondexplanation

is correct, puq suppression of the Qam defect

should be independent of the particular Qam

mutation used.

To test these two possibilities, we constructed phage by recombination, carrying each of three puqmutations with three differentQam

muta-tions. We firstselectedQam+revertants ofeach

ApuqQam2l phage by selecting

normal-size

plaquesonstrain 594(sup+). The Qam+puq

mu-tants were then crossed with

XQam501

or

XQam57

in strainYMC(supF).Theprogeny of

thecrosses wereplated on CA169. About 5% of the plaques were small (characteristic of Qam

pseudorevertants)and, of these, about 20% were indistinguishable in their properties from

ApuqQam(see above). The Qam mutation could

alwaysbe recoveredfrom theseApuqQamphage

in backcrosses with wild-type X by selecting

phagethatfailed to formplaquesonCA169 and 594butplated nornallyonYMC. We obtained the same resultswith all three puq

mutations,

soninecombinationsofApuq

Qam phage

(three

puqmutationsand three

Qam mutations)

have the same properties. Wesuggest that the puq

mutationincreases thesynthesisof the Qgene product andnotthespecific

activity

of the pro-teinitself. Mappingdata described in thenext sectionplacethepuqmutationsat a

position

to

theleftoftheQgene(Fig. 1),moredefinitively ruling out a direct alteration of the Q gene product by puq.

Mapping puq mutations. (i) Phage

crosses between

AimmAcIpuqQam21

and Aimm21cItsSam7. We identified Aimm-21cItsSam+ recombinantsamongthe progeny of

thiscross asclear-plaque fornersonstrainC600 at420C (Table 2A).Atotal of75 to90%ofthese

recombinantscarried theQam21 mutation and,

amongthese,60 to70% carriedthepuqmutation

(recombinant classa). We conclude thatthepuq

mutations tested (puq3, puq4, andpuql6) lie betweentherightend of theimm2l substitution and theQam21mutations. Because most recom-binants thatcarried theQammutation also car-ried the puqmutation,puqmustbe closetoQ. (ii) CrossesbetweenXimmAcI+puqQam21 andXcIts857

(Q-attR)A prophage

deletions.

Wetestedabatteryofprophage deletions that

enterthe Xprophagefromtherightandremove

the Qgene, yet endwithin the region between genes 0 and Q (Table 2B). We infected these

lysogens withtheAimmAcI+puqQam21mutants

and selectedphagethatrescuedthecIts857 im-munitymarker from theprophage. Such phage form clear plaques on YMC at 42°C. If the

prophage deletion contains thewild-typeallele of puq, then someof the cIts857 recombinants

shouldhavelost the puqphenotype. Sincenone ofthe lysogenscarrythe prophageQgene, the

infecting phagecannot lose theQam mutation.

None of the recombinants tested (at least 100 from each cross) lost the puq mutation. As a

control, we used

XcI+Pam80

to infect either

SA268 orSA431 andfound that about 10% of the AcIts857 recombinantprogenywere Pam'.

We suggest that the wild-type allele(s) of the puq mutations is removedbytheprophage dele-tions. These results and those shown in Table

2Aplace thepuqmutations between the left end oftheSA268 and SA431 deletions and the wild-type allele of the Qam21 mutation. We cannot

rule out,however, that the puq mutations map outside and just to the left of the SA268 and

SA431deletions.

(iii) Crosses between XcIts857puqQam21

http://jvi.asm.org/

(5)

PHAGE A MUTATIONS AFFECTING Q GENE ACTIVITY TABLE 2. Mapping ofpuqmutatl

immn2l marker, the Q gene, andpi

A

imm,Acr

I

imm2lclt-pug mutation puq recombinant frc

14

16

B Infecting

pfropa

0.7 0.6 0.7 0.6 0.7

clt

-

c1tsa857-Prophage

Aclts8570am29+(Pam3.attR)A=SA439

Iclts857Pam3 +(Pam8OatR)A=SA444 Acils8S7PamO+(Qam2t-1tfR)6=SA268

Acits8S7Pam8O+(Qam2t1atIR)6=SA431

a

m(A)

Anm21cItsSam'

recombi asclear-plaque-formingphageonj

(the Sam7 mutation isnotsuppr suppressor in strain C600). Eighl

recombinantscarrytheQam2lm

unable to form plaques on str

ionsrelative to the nants would be a measure of the relative dis-rophage deletions0 tances between imm434 and either puq or byp. Incontrast,ifpuqwerebetween byp andQam2l pvq

am21

+ (Fig. 2, order 2), we would expect all of the

-J

Sam?

Ximm434Qam

recombinants to carry thepuq

mutation since the formation of puq+

re-combinants would require a triple recombi-tion[(a)/(a)+(b) nation event. For

puq3

and

puql6,

respec-tively, 12% (16/132) and 21% (22/106) of the

'o

Aimm434byp+Qam21

recombinants were puq+.

so Therefore, both puq mutations are to the leftof ,,

Ibyp,

puq3

atafractional distance 0.88

andpuql6

57 atafractional distance 0.79 of the distance

be-tweenthe

right

end of the

imm434

substitution '° andthe byp mutation. Ifwe assumethatthese geneticdistances aredirectlyproportionaltothe puq

Gam21

physical

distancesbetweenthese

markers,

then the puq mutationscanbelocated on the A

phys-,__,,,,,,,,,

ical map.Sincethe A mapcoordinate of theright

end oftheimm434 substitution is79.1 (11) and

the right-most coordinate of the byp mutation q+phagt

anong

he

kldt,57

mustbe90.6

(see

below),

wecanlocate

thepuq3

Q#m2t

recombinants and puql6 mutations on the A map between

puq-3 puq-16 coordinates 88 and 89. This would

put

thesepuq

mutationsin the DNAsequences deletedbythe 0/120 0/123 nin5 deletion (DNA deleted between 83.6 and o/laO 0/110 89.2 on the X map) or just to the right of that

0/211 0/143 deletion

(Fig.

1).

0/140 0/126 (iv) Crosses with

APam3nin5Qam21

and

ApuqQam67.

To determine whether the puq

nantswere selected mutationsareindeed removed

by

thenin5

dele-strainC600 at42°C tion, we crossed

APam3ninNQam2l

with ressed by the supE

ApuqQam57

(Table 3). The Qam57 mutation is ty percent of these located to the right of

Qam2l

(18), and phage

utation as

they

are with this mutation

will

form plaques on am

594(sup').

We

YMC(supF)

butnot onC600(supE).Phagewith

screeneca tnes AunmMLcLUswamZI recomUmants ror

the puq mutation based on their ability to form

plaquesonstrainCA169.(B)We selectedphagethat

had rescued the clts857allele from theprophageas

clear-plaque-forming phage onYMCat42°C. These werescreened forthe puq mutationonstrain CA169.

andXimm434bypQ+. Topositionthe puq mu-tations with respect to the

byp

mutation,

we selectedXimm434Qam21recombinantphageas turbid plaque formers on

YMC(supF)

at

42°C

(phage carrying the byp mutation form clear

plaques [5]) (Fig. 2). The absence of the byp mutation among these recombinants was

con-firned by their inability to form plaques on

strainNS377(see above and reference 33). Since the byp mutation is located between phage

genes P and Q (seenextsection), thepuq

mu-tation might be located either to the right or

totheleft ofbyp. If puqwerebetweenimm434

and byp

(Fig.

2, order 1), the fraction of

Ximm434puqQam21 phage amongthe

recombi-orderI

order 2

imm434 + byp +

cItsg5T puq + I m21

imm434 byp + +

[image:5.501.62.254.74.386.2]

cJfts57 + P 0 1

FIG. 2. Turbid-plaque-forming recombinant phages(recombinants between theimm4andbyp

markers) were selected on strain YMC at 42C. Ninety-eight percent of these phage caried the

Qam21mutationasthey were unabletoform plaques

onstrain594(sup+).Thepresenceofthepuqmutation amongtheseXimm434byp'Qam2lrecombinantswas

assessed bytheir ability toformplaquesonstrain CA169.

VoL 30,1979 ₅

3

PU4

http://jvi.asm.org/

[image:5.501.266.460.452.656.2]

(6)

6 STERNBERG AND ENQUIST

the Pam3 mutation will not form plaques on

C600(supE) at420C. Thus, onlyQam57+

Pam'

recombinant phage will form plaques on C600 (supE) at420C. These recombinantphagewere

purifiedandtested fornin5,puq, andQam2las

described in footnote a of Table3andabove. An

ambiguityarises in thisanalysisbecause it isnot possible to distinguish between Xnin5Qam2l

andXnin5puqQam2l because nin5isapuq

mu-tation (see next section). In addition, itis not

possible to distinguish between Xam+puq and

Xam+puq+.In any case, we wereunabletodetect anyQam21puq+nin+ phage,suggestingthat the

nin5 deletionremoves the wild-type alleles of puq3 andpuql6. Wecannotruleoutthe possi-bilitythat puq maps very closetobutjust out-sideofthe nindeletion.

byp

andnin5arepuq mutations. The puq,

nin5,andbypmutationsarelocatedinthesame regionoftheX genome. Thisproximity encour-agedusto testnin5andbypfor the puq

pheno-type.To dothis,theQam21 mutationwas intro-duced into Xnin5 and Xbyp. Xnin5 and Xbyp phage were crossed with XQam21, and phage

[image:6.501.54.241.366.539.2]

thatformed smallplaquesonCA169 (about2% of the yield from the cross) were selected for

TABLE 3. Mappingofpuq mutations relative to the

nin5deletiona

PsWnm3A Oem21 +

I(c) (b '(a)

+ ..puq-3orpuq-16 + 0am57

puqmuaation No. of recombinants in class

(a) (b) (c)

Aam+ (ApuqQam2lI (Auin Qam2lor

AinSpuqQam2l1

puq-3 10 20 35

puq-16 3 24 31

0The

level ofwild-type (am+) recombinants was measured as plaque formers on strain 594(sup+).

AQam21recombinants were assayed on strain C600 at

420C

(neither AQam57 norXPam3 form plaques on C600 atthis temperature) and represent 10% of the totalphage burst. We determined whether the nin5 mutation was present intheseAQam2lrecombinants by measuring either their ability to form plaques on strain NS577ortheir EDTAsensitivity. Among the

AQam21recombinants without the nin5 mutation, the presence or absence of the puq mutation could be assessedbytheability of these phage toformplaques onstrainCA169(supC).

further analysis. The small-plaque phenotype itself strongly suggested that nin5 and bypwere

puqmutations. We demonstrated that these Q pseudorevertants were either XbypQam21 or Xnin5Qam2l bythefollowingcriteria: like both the nin5 and byp parents, the recombinants formedplaquesonNS577 (33); thepresumptive

nin5 recombinant isasresistanttoEDTA treat-ment as is thenin5parent(wild-typeX is EDTA sensitive; deletions are EDTA resistant); and both phage do not form plaques on 594(sup+) but do plaque on YMC(supF). Furthermore, they both fail to complement XQam2l in 594(sup+). Because these phage are able to grow

on CA169, both byp and nin5 must be puq mutations.XbypQam2lfonnslargerplaques and Xnin5Qam2l forms smaller plaques than does ApuqQam2l on CA169. The kinetics of phage production by these Qam2l phage inCA169(see

below) confirm thisincreasing ability to express thepuqphenotype (byp>puq >nin5).

Mapping nin5 and byp mutations. Al-though both nin5 and byp mutations have been located in the interval between phage genes P and Q, their location relative to each other is unclear. (Are they at the same site? Is bypto the left orright ofnin5?) The following three

crosses locate more precisely these two muta-tions.

(i)

Crosses between either Pam8Onin5

or XOam29byp andXcIts857 (Q-attR)A

pro-phage deletions. We infected each prophage deletion with XPam8Onin5 or

AOam29byp

and

isolated

amber'

recombinantsfrom the progeny

byplatingthe phage yield onstrain

594(sup')

(Fig. 3A). These

amber'

plaques were then

screenedfor loss of nin5orbyp.Wedetectedno nin+ or byp+ plaques among several hundred

amber'recombinants. We conclude that the two

prophage deletions contained instrains SA268

and SA431 carryneitherthewild-typeallele of the byp mutation nor DNA to the right of the nin5 deletion.

(ii) Crosses between Xnin5Qam2l or

XbypQam2l and (attL-P)A prophage dele-tions. We used another set ofprophage dele-tions that enter the P-Q region from the left

(Fig. 3B). These strains were infected with

Xnin5Qam2l orXbypQam21. Amber'

recombi-nantswere selectedbyplatingthe phage yield

on594(sup'),andphage in the resulting plaques werethen tested for the loss ofthe nin5 or byp mutations.Among several hundred recombinant plaques obtained from the three crosses, none contained nin+ phage. Similarly, byp+ phage

were not found in crosses involving SA613 or

SA741.However,20% of theamber'phage from

http://jvi.asm.org/

(7)

Cross Prophag Infecting

Strain Phage A) I SA268or

SA431

APamSOnin5

SA2S8or

SA431

10am29byp

8) 1 SA613or

SA741

AbypQam2l

SA307

Abyp0am2l

lo-P-Qregion

0+_P+________

Pam8O 0 +

o+P+ ,,aS,...

Oam29 byp Q +

byp 0am21

_____________bp _0+

byp ORM21

3 SA307

AninS5Om21

byp + a+ I

FIG. 3. Recombinantphage in these crosses between AOam, APam, and AQam phage and deleted A

prophages were selected asplaque formers on strain 594(sup+). Recombinants lacking thenin5 deletion

(crosses Al and B3)weredetectedby theirinabilitytoform plaquesonstrainNS377, and those lackingthe bypmutation(crosses A2,Bl,andB2)weredetectedby their turbid plaquephenotypeand confirmed by their inabilitytoform plaquesonstrain NS377.

the

AbypQam2l

xSA307cross were

byp'

(cross

2 recombinant classa). SA307, therefore,

con-tains thewild-type allele of

byp

butnotDNAto

the leftonthe nin5deletion. We conclude that the

byp

mutation liesbetween the left endpoint

of the nin5 deletion andgeneQ.

(iii) Cro88e8 with Aclts857byp and

(attl-Oam29)A or (attL-Pam8O)A cryptic

pro-phage deletions derivedfromaAnin5

pro-phage.Todeterminewhether the nin5 deletion

removesthewild-type allele of

byp,

wecrossed

XcIts57byp with prophage deletions derived froma?nin5 prophage. (Fig. 4). In bothcases, byp+ recombinants were detected. Two of four

byp+ recombinantsfrom the DC401 cross and three of eight byp+ recombinants from the

DC402cross weremoresensitivetoEDTA treat-ment than werenin5phage and failedtoform

plaquesonstrainNS577,bothexpected

charac-teristics ofbyp+ nin+ phage (recombinantclass

b). If thebyp+allelewerelocatedinaregionof

DNA removed by the nin5 deletion, then all

byp+ recombinants in these crosses should be nin5. As this is notthe case, the byp+ allele is

notcoveredby the nin5 deletion. Basedonthese data and those presented above, we conclude that the

byp

mutation is located between the

right end of the nin5 deletion and the Qgene. IsolationofXnin5bypdouble mutants. If

nin5andbyp areatdifferentsites,it should be

possible toisolate the

nin5byp

double mutant. We did this by treating the progeny of the crossesshown inFig.4 with EDTAtoinactivate

nin+ phage. The surviving phage were then

platedon

594(sup').

Ten to twenty percentof theresulting plaques wereclear and thuswere

potentialAnin5byp phage.Toverifytheir

geno-type, weisolated both nin5 and bypsingle

mu-tantsfromthe double mutant. Theexperiments were asfollows. We first crossed either Pam8O

or Qam2l mutations into the presumptive

Xnin5byp

phage and then crossed the

XPam80nin5byp phage with XQam2l and the

Anin5bypQam2l phagewith XPam8O (Table 4).

In both cases amber+ recombinants were

se-lected and clear

(byp)

orturbid

(byp')

plaques were tested for the nin5 mutation. We found

that: (i) in cross 1, 30% of the turbid plaques

VOL. 30, 1979 ₇

http://jvi.asm.org/

[image:7.501.115.417.60.350.2]

(8)

8

cross Prophage Infecting strain phage

I DC401

2 DC402

Aimm-OP-Q region

nin5 byp+ °+

=

_

I )

J5b)

ArIts857byp c1tS857 byp

Pam3~+ WOl byp~ Q+

Acls857byp cUts857 byp

FIG. 4. Recombinantphageswith thebyp+allele were selected as turbid plaqueformers on strain 594 at

42°C. The EDTA sensitivity ofthephage contained in these plaques was tested under conditions in which

nin5phage are completely resistant and nin+phage are inactivated to a level of10-2 survival (EDTA treatmentfor10min at32°C).

containednin5byp' phage, and in cross 2, 35% ofthe clearplaques contained nin+ byp phage (in both cases, recombination occurred in seg-mentb;therefore, theoriginal phagemusthave carriedthenin5and thebypmutations); (ii)the location of byp between nin5 and Qam21 is

confirmed because all of the clear-plaque

for-mersin cross1werenin5,and all of the turbid-plaque

forners

incross 2 werenin+

(recombi-nation in both crosses is in segment c);(iii) the relative recombination frequencies in intervals

a,b,andcare ameasureof the relativelengths of the intervals (based on the results given in Table 4, thevaluesfor these distancesexpressed

asafraction ofthe total recombination distance

betweenPam8O andQam2lare:segment a, 0.44

to0.49; segment b,0.21 to0.24; and segment c, 0.30to0.32).

Mapping

byprelativetotheA-480

recom-bination site in gene

Q.

A unique site for

recombination betweenphagesA and

480

is

lo-cated within the Q gene of these phages at position 90.6ontheA map (11, 14). Asthissite isimmediatelytotherightofthe nin5deletion,

we decidedto use a

X-+80

hybridphage with A DNAtotheleft of this recombination site and

480

DNAtotherightof thissitetolocalizemore

specifically the byp mutation. We infectedstrain

YMC with this hybrid phage and with

AbypQam2l (Fig. 5) and selected hAQam+

re-combinantphage from the phageyieldasplaque

forners on a +80-resistant derivitive ofstrain

N215. Three of 1,120 recombinantplaques were clearat

320C.

Presumably,they contained phage withthe byp mutation andwereformed by

re-combination event c. The

bypQ480

genotype of theserecombinantphagewasconfirmedby the

following tests: (i) they forned plaques on NS577,aproperty of byp mutants; and (ii) when

anN205(recA) lysogen containing the presump-tive AbypQ4)80 prophage was infected with

Xred3imm434Ram5, plaquesappeared at an

ef-ficiency of< 10-4 relative to that found after infection of an N205(recA) lysogen carrying a

AbypQA prophage. Plaque formation by heter-oimmune R- phage on lambda lysogens of

N205(recA) is anindication of the transactiva-tion of the prophage R gene by gpQ of the infecting phage (36). Thefailure to transactivate the R gene from thepresumptive bypQ480

pro-phageimpliesthat itindeed carried

480

DNA to therightof the A-480recombination site. Trans-activation of the )80R gene isnotstimulatedby

the Q gene product of phage A (10, 36). We conclude that

byp

is located between theright

end of thenin5 deletion (89.2onthe map) and the QX-Q480recombination site (90.6on the A map).

Single-step growth experiments. Up to

now,wehave discussedthepuqphenotype only

intermsofplaque sizeandplating efficiency.In thissection,wequantitatively analyzetheeffect

of various puq mutationson kinetics ofphage development andyield ofphage carrying Qam mutations. We used strains 594(sup+), CA169(supC), and YMC(supF) and infected

them with wild-type A, AQam21, AbypQam2l,

Anin5Qam2l, Apuq3Qam21, and

Anin5-bypQam2l (Fig. 6). In YMC(supF), all phages

grewwell,althoughAnin5Qam2lseemedtogive aslightlyreducedphageburst. The puq

pheno-type isexpressedin strainCA169(supC).Inthis host, AbypQam2l andAnin5bypQam21 grew

al-most aswellasAwild type.However, the growth of Apuq3Qam21, Anin5Qam2l, and XQam2l

phageswasincreasinglymore defective both in the kinetics ofappearance and in the yield of phage. In the

sup'

host, only the wild-type J. VIROL.

http://jvi.asm.org/

[image:8.501.62.391.58.217.2]

(9)

TABLE 4. Orderingof Pam80, nin5, byp, and Qam2l mutations byfour-factor crossesa

cross1 a b c

+

n1nS

byp Oam2l

PamSO + +

Recombinant Segmentin which amber+recombinant No. isolated frequency recombination event occured

bypnin+ 0 --

-bypnin 17 0.30 (c) byp +nin+ 28 0.49 (a) byp+nins 12 0.21 (b)

cross 2 a b c

+ + Oam2l

Recombinant Segmentin which amber + recombinant No. isolated frequency recombination eventoccured

bypnin+ 24 0.24 (b) bypnin

byp +ni + byp +ninS

44 0.44 32 0.32 0

(a) (c)

a(Cross1)Wild-type(am')recombinants were

as-sayed byplaqueformation on strain 594 and repre-sented2% ofthephageyield. Thirtypercentof these

plaqueswereclear(byp),andof 17 clearplaquestested

all contained phagewith the nin5 mutationbyvirtue of their increased EDTA resistance. Among the

re-maining 70% turbid plaques, 12 of 40 tested were

judgedtobe nin5 byvirtue ofboth their increased EDTAresistance and theirabilitytoform plaqueson strainNS377.(Cross2) In this cross 2.8%ofthephage yieldweream'recombinantsand of these, 32% were

turbid(byp')and 68% were clear(byp).The presence of the ninallele among these recombinants was as-sessedasincross 1.

phage grew normally. Anin5bypQam21 and

AbypQam2lproduced onlytwo- tofourfoldmore

phage than did otherAQam2l phageinthis host. Whereas AbypQam2l and Anin5Qam2l phage did not forn plaques on 594(sup+) [less than

10'

relative to plaque formation on YMC(supF)], Xnin5bypQam21 phage fonned pinpointplaqueson594(sup+) withalmost wild-typeefficiency.

Kineticsofendolysinproduction.The

ef-ficiencyof activation of latephagegenesbythe Q geneproductisgenerally assayedby

measure-0

a

I

a.

ment of phage endolysin (the phage R gene product) (9, 22, 33). We measured endolysin

after infection of strains YMC(supF),

CA169(supC), and 594(sup+) by XQam2l and

XQam+ phages (Fig. 7). In YMC(supF), all phages produced normal or near-normal amounts ofendolysin activity, ranging from 800 to1,000 unitsby 40 min after infection (Fig. 7a). After infection of CA169(supC) with XQam2l, endolysinproductionwasmuch reduced relative

h8O att8O Imm434' 080

(a) (b) (c)

hA attA cItS857 byp Qam2l

FIG. 5. AbypQp8O recombinant phage were de-tected as clear-plaque-forming phage on strain

NS460 at 32°C. Wecould confirm that these clear

plaquescontainedphage with the bypmutation as thesephageformedplaquesonstrain NS377.

TimeafterInfection(min)

FIG. 6. These single-step growth experiments were

performed aspreviously described (33). Thephage

yieldwassampledatdesignated times after infection

byblendingsamplesoftheinfectedcellculture ina Vortex mixerwith several dropsof chloroform. The bacterialstrainsusedwereYMC(a), CA169 (b), and 594 (c). The phage used were XcIts857Qam+ (0), AcIts857Qam21 (A), AcIts857nin5Qam21 (A),

Ac1ts857puq-3Qam21(0),AcIts857bypQam21(),and

AcIts857nin5bypQam21 (0).

9 VOL. 30,1979

http://jvi.asm.org/

[image:9.501.52.448.70.559.2] [image:9.501.265.446.302.567.2]

(10)

E

000

400- c)594

300-200

100

20 40 60 80 100

[image:10.501.60.220.62.357.2]

Time after Infection (min)

FIG. 7. Lysozyme assays were perforned as

de-scribedin the text. Thebacterial strainsusedwere

YMC (a), CA169 (b), and 594 (c). The symbols for phage usedwerethesame asthosegiven in the legend toFig.6.

to the YMC(supF) infection. The nin5, puq3, and byp derivatives ofAQam2l produced,

re-spectively, increasingamountsofendolysinafter

infectionofCA169(supC) (Fig. 7b).In contrast,

endolysin production by both XQam2l and

AbypQam2lin594(sup+)wasmuchreduced;the

byp mutation enhanced enzyme production by onlyafactor of 2(Fig. 7c). Comparingtheresults shown inFig.7b andc, weconcludethat thebyp mutation produces atwofoldincrease in

endo-lysin synthesis in the absence of the Q gene

product (XQam2linstrain594)anda morethan

sixfold increase inendolysin synthesiswhen syn-thesis of theQgeneproductislimiting(AQam2l instrain CA169).

Clear-plaque phenotype of Abyp phage. Phage that carry the byp mutation form clear

plaquesandareimpairedintheirabilitytoform lysogens. The frequency of lysogen formation

was generally three- to sixfold less than that found forbyp+ phage (Table5). Wehave shown

inthetwoprevioussectionsthatthe byp

muta-tion increasestheexpressionoftheQgeneand latephage functions. This increasedexpression

could channel a phage infection into a lytic

rather thanlysogenic response. We would expect that if this were so, XbypQam2l phage would

form lysogens with a higher frequency than )bypQ+ phage inYMC(supF) and with even a

higher frequency in 594(sup+). This is, in fact,

what weobserved(Table 5). InYMC(supF)and 594(sup+) hosts, the Qam21mutation increased the frequency of lysogen formation by Abyp phage, respectively, two- andfourfold. Indeed,

XbypQam2l

forms a more turbid plaque on

YMC(supF)than does

Xbyp.

DISCUSSION

Inthis paperwehavedescribedthe isolation and characterization of mutations (designated puq)inphageA thatincreasesynthesisof theQ

gene product. Normally, these mutations have no phenotype because the level of the Q gene productsynthesized by wild-typeAorby

XQam

phage grown in strainYMC(supF) ismorethan sufficient for phage production. However, if

XQam

phage are grown in the weak

am-sup-pressing strain CA169 (under these conditions, synthesis of the Qgene product islimiting for phagegrowth), thenthe puq mutation helpsto

alleviate

thephage growthdefect. Thus, whereas

XQam21

fails to plaque on strain CA169,

ApuqQam2l

plaqueswith unitefficiencyonthis

host.SincepuqQam phagedo not grow in strains

lacking either am or oc suppressors, at least somesynthesis of the

Q

geneproductis neces-saryfor puqtoexertanyaffecton

XQam

plaque

formation;

puqdoesnotcreateaQ-independent

phenotype,asdomutationspreviouslydescribed

by Herskowitz and Signer (17). Various puq

mutations have been mapped between phage genesP andQ,aregion ofthe genome knownto containtwomutations(nin5andbyp)that allow A to grow better in the absence of the phage

gene Nproduct.Bothbypand nin5 are also puq

mutations. Based on plaque size, phage burst, and phage endolysin synthesis, byp is a better

TABLE 5. LysogenizationfrequencyofAbypphagea

Lysogenization

Bacterial frequency at:

Infecting phage strainI

=5 =0.2

AcIts857 594(sup+) 0.4 0.5

XcIts857Qam21 594(sup+) 0.25 0.36

AcIts857byp 594(sup+) 0.08 0.12

XcIts857bypQam21 594(sup+) 0.30 0.40

AcIts857 YMC(supF) 0.5 0.25

AcIts857Qam21 YMC(supF) 0.25 0.30

AcIts857byp YMC(supF) 0.10 0.06

AcIts857bypQam21 YMC(supF) 0.17 0.12

aLysogenizationwasmeasuredasdescribed in the text.

http://jvi.asm.org/

[image:10.501.255.447.527.651.2]

(11)

puq mutationand nin5 isaworsepuqmutation than are the mutations (puq-3 and puq-16)

whoseisolationsaredescribedin this paper. A mapof theAgenome,withparticularfocus ontheregion between phage genes P and Q, is shown inFig. 1. The map coordinates of the nin deletion are well documented (11). Crosses

de-scribed in Results located the puq mutations

either underthenindeletion or veryclosetothe

leftorrightendofthisdeletion.Thebyp

muta-tion waslocalizedtotherightof thepuq muta-tions andthe nin deletion(rightmapcoordinate

89.2)andtotheleft of thecrossoverpointin the Q genes of phageAand phage80 (map

coordi-nate 90.6). We have isolated Anin5byp double

mutants,confirmingthe locationof byp.

Based on ouranalysis of the properties of nin,

byp,and puqmutations,we candraw the

follow-ing conclusions about theregulation ofQgene

expression andtheactivityof the Qgene

prod-uct.

(i) The byp mutation increases the level of phageendolysin synthesis byafactorofnomore than2when the Qgene productis not synthe-sized (Fig. 7). Transcription ofthe phage

endo-lysingene(geneR)intheabsenceof theQgene

product initiates exclusively at pL and reads

throughinto genes Qand R (9). Since R gene

transcription is enhanced twofold by the byp mutation, Q gene transcription should be

en-hanced to the sameextent.Yet, whenalimiting amount ofQ gene product issynthesized (Fig.

7b), theenhancement ofendolysinsynthesis by bypisinexcessof sixfold.Theseresultsindicate that theconcentrationdependenceof the activ-ity of the Q gene product is not linear. This might be the case if the product of gene Q is

requiredas adimeror atrimer. In sucha circum-stance,atwofold reductioninmonomer

synthe-sismightmeanafourfoldreductionor even an eightfold reductioninactivity.

(ii) Abyp phage form clearplaques and lyso-genize standard bacterial hosts withathree-to sixfold-reduced frequency relative to byp+ phage. In contrast,

XbypQam

phage lysogenize strain

594(sup')

at the same frequency as do

Xbyp+

phageand lysogenizestrain YMC(supF)

at a frequency internediate between

Xbyp

and

Xbyp+

phage (Table 5). These results suggest

that afterinfection thelysogenization defectof

Abyp phage reflects an increased expression of

the Q gene and late phage functions, with a

consequent channeling of the infection into a

lytic rather than the lysogenicresponse. If the

synthesisofQgeneproductisreducedor

elimi-nated, thenthelysogenizationfrequency ofbyp

phage is thesame asthat ofbyp' phages.

(iii)

How do nin5, byp, and puq mutations

exerttheir affectsongpQactivity?We can rule outthepossibilitythat thesemutationsdirectly alter the Q geneproteinbecausenin5 is a

dele-tionthat maps to theleft of the Q gene, and the puqphenotype exhibitedby puq-3,puq-16, and byp is independent of the Qam mutation used (see Results; data for byp is not shown). More likely, these mutations act by increasing Q gene

transcription. As we cannot detect endolysin production from Apuq-3, Apuq-16, Anin5, and

Xbyp lysogens(data notshown; also see reference

5), these mutations mustnot create a constitu-tive Q and R gene promoter like the qinlOl mutation (10). Since theregion between phage genes P and Q contains a transcription

termi-nation site

(tR)

that is sensitive to the N gene product,thepuq, nin5, and byp mutations might alteroreliminate this sitesothattranscription

beyond itbecomes Nindependent. If

transcrip-tion antitermination mediated by the N gene product wereless than 100% efficient attR2,then thesemutations would enhance transcription of downstream genes (Qand R) even in the pres-enceof theNgeneproduct. Indeed, transcription stimulated bythis productdecreasesasthe num-berofterminationsites traversed by the

polym-erase increases (1, 2), suggesting that N-me-diatedtranscriptionantitermination is not 100% efficientateach terminationsite.

Althoughthismodelfitswellwiththe nature of the nin mutation, a deletion which could

eliminate the tR2 termination site, it does not

readilyexplainsomeof the properties of thepuq andbypmutations. First,whereasnin5andbyp mutationsare N-independentmutations, puq-3 andpuq-16 are not

[XNam7Nam53cl7byp

and

XNam7Nam53nin5

plaque onstrain 594(sup+),

but ANam7Nam53cl7puq does not (data not

shown)].

Yet,puq-3andpuq-16 mutationsshow

a stronger puq phenotype than does the nin5

mutation(Fig.6and7). Thesediscrepanciesare

mosteasilyreconciledifthe puq mutation does

notaffectatermination site but rather creates

apromoter upstream from tR2. Thus, itsability

to stimulate Q gene transcription would still dependonNproductactivity.Second,sincebyp is not covered by the nin5 deletion, if nin5 deletestR2,then byp must either affect a second

transcription termination site in thisregion or must create a new promoter capable of tran-scribing theQ gene in theabsence of the N gene product.Assuchapromoter is not active in the prophage state (5), wesuggest that it requires upstream transcription (e.g., from pR) for its activation.Inthe absence of theNgeneproduct, the lowlevel oftranscription that might proceed through thetR2sitecouldactivate this new pro-moter, resulting in an increased Q gene

tran-VOL. 30,1979 11

http://jvi.asm.org/

(12)

scription.

In additionto the nin5, byp, and puq muta-tions, a fourth mutation which maps in thetR2

region of the A map is pasB (31). This mutation enables a Xred-gam- phage togrow better in strainscontainingapolA mutation (thefeb phe-notype [38]). Preliminary experiments indicate that nin5 and puq-3 are pasB mutations but that byp is not. As the pasB phenotype is not

well understood, thesignificance of these find-ings must await further experimentation.

An analysis of the transcription patterns in the P-Qregion of the A map produced by the various mutant phages shouldhelp to elucidate their mode ofaction.

ACKNOWLEDGMENTS

We thankR.Weisberg for his support and encouragement during thecourseof thiswork and L.Stockmanforcarefully typing thismanuscript.

LITERATURE CITED

1. Adhya, S.,and M.Gottesman. 1978.Control of tran-scription termination. Annu. Rev. Biochem. 47:967-996.

2. Adhya, S.,M.Gottesman,B. deCromburgghe, and D. Court. 1976. Transcription termination regulates gene expression,p. 719-730. InR. Losick and M. Cham-berlin(ed.), RNA polymerase Cold Spring Harbor Lab-oratory,ColdSpringHarbor, N.Y.

3. Appleyard, R. K. 1954. Segregation ofnew lysogenic typesduring growthof adoubly lysogenicstrainderived from Escherichiacoli K12.Genetics39:440-454. 4. Brenner,S., and J. R. Beckwith.1965.Ochre mutants,

a newclassofsuppressiblenonsensemutants.J. Mol. Biol.13:629-637.

5. Butler, B., and H.Echols. 1970.Regulationof bacterio-phage Xdevelopmentby gene N:properties of a muta-tion thatbypassesNcontrol of lateprotein synthesis. Virology40:212-222.

6. Campbell,A. 1965. The steric effect inlysogenizationby bacteriophage A. II. Chromosomalattachment of a b2 mutant.Virology27:340-345.

7. Campbell,A., and E.Balbinder. 1958. Propertiesof transducingphage. CarnegieInst.WashingtonYearb. 57:386-389.

8. Court, D., and K. Sato.1969.Studies on novel transduc-ing variantsoflambda:dispensibilityof genes Nand Q. Virology39:231-252.

9. Couturier,M., C.Dambly,and R.Thomas.1973. Con-trol ofdevelopment in temperate bacteriophage. V. Sequential activation of the viral functions. Mol. Gen. Genet. 113:244-250.

10. Dambly, C.,M.Delstanche,and A. M.Gathoye.1979.

qinlOl:promoter mutation which allows the constitu-tive expression of the late genes. J. Virol. 30:14-20.

11.Davidson,N., and W. Szybalski. 1971. Physicaland chemicalcharacteristics of lambdaDNA, p.45-82. In A. D. Hershey (ed.), The bacteriophage lambda. Cold SpringHarborLaboratory,ColdSpringHarbor,N.Y. 12. Dennert, G.,and V.Henning.1968. Tyrosine-incorpo-ratingambersuppressors in Escherichia coli K12. J. Mol.Biol. 33:322-329.

13.Dove, W. F., H. Inokuchi, and W. Stevens. 1971.

Replicationcontrol inphagelambda,p.747-771.InA.

D. Hershey (ed.), The bacteriophage lambda. Cold Spring HarborLaboratory, Cold Spring Harbor, N.Y. 14. Fiandt,M., Z. Hradecna, H. A. Lozeron, and W. Szy-balski. 1971. Electronmicrographic mapping of dele-tions, insertions, inversions, and homologies in the DNAs of coliphages lambda and phi 80, p. 329-354. In A. D.Hershey (ed.), The bacteriophage lambda. Cold Spring HarborLaboratory, ColdSpringHarbor,N.Y. 15.Gorini, L 1970. Information suppression. Annu. Rev.

Genet.4:107-134.

16.Herskowitz, L. 1973. Control of gene expression in bac-teriophage lambda. Annu. Rev. Genet. 7:289-324. 17.Herskowitz, I., and E.Signer.1970. Control of

tran-scription from the r strand ofbacteriophage A. Cold Spring Harbor Symp. Quant.Biol. 35:315-323. 18.Herskowitz,I., and E.Signer.1970.A siteessential for

expression of all late genes inbacteriophageA. J.Mol. Biol. 47:845-856.

19.Hopkins,N. 1970.Bypassingapositiveregulator: isola-tion of aA mutant that does notrequire N product to grow.Virology 40:223-229.

20. Kourilsky, P.,L.Marcaud,M. M.Portier,M. H. Za-mansky, and F. Gros. 1968. Studies on the messenger RNA ofbacteriophage A. I. Various species synthesized early after induction of theprophage. Proc. Natl. Acad. Sci. U.S.A.61:1013-1020.

21. Kumar, S., K.Bovre,A.Guha, Z.Hradecna,V. M. Maher, and W.Szybalski.1969.Orientationand con-trol of transcription of E. coli phage A. Nature (London) 221:823-824.

22. McMacken,R.,M.Mantei,B.Butler,A.Joyner,and H.Echols. 1970. The effect of mutations in the clI and cIII genes ofbacteriophageA onmacromolecular syn-thesis in infected cells. J. Mol. Biol. 49:634-654. 23. Packman,S.,and W. S.Sly.1968.Constitutive A DNA

replication byA c17, a regulatorymutant related to virulence.Virology34:778-781.

24. Parkinson, J.,and R.Huskey.1971.Deletionmutants

ofbacteriophagelambda. I. Isolation and initial char-acterization. J. Mol. Biol. 56:369-384.

25. Pereira DaSilva, L.H., and F. Jacob. 1968. Etude genetique d'une mutation modificant la sensibilitea l'immunite chez le bacteriophage lambda. Ann. Inst. Pasteur Paris 115:145-158.

26. Roberts, J. W.1969.Promoter mutation in vitro. Nature (London)223:480-482.

27. Roberts,J.W. 1969. Termination factor for RNA syn-thesis. Nature(London)224:1168-1174.

28. Roberts,J. W. 1975.Transcriptiontermination andlate controlinphagelambda. Proc. Natl. Acad. Sci. U.S.A. 72:3300-3304.

29. Schrenk,W.J.,and R. A.Weisberg.1975. Asimple method formakingnewtransducinglines ofcoliphage A. Mol.Gen. Genet.137:101-107.

30. Shimada, K., R. Weisberg, and M. E. Gottesman. 1972.Prophagelambdaatunusual chromosomal

loca-tions.I.Locationofthesecondary attachmentsites and

theproperties of thelysogens.J. Mol.Biol. 63:483-503.

31. Shizuya,H.,D.Brown, and A.Campbell.1974.DNA polymeraseI-dependentmutantsofcoliphagelambda. J.Virol. 13:947-952.

32. Sternberg,N.1976. Ageneticanalysisofbacteriophage A headassembly.Virology71:568-582.

33. Sternberg, N. 1976. A class ofrifR RNA polymerase mutations that interferes with theexpression of coli-phage A late genes.Virology73:139-154.

34. Sternberg,N.,D.Tiemeier,andL.Enquist.1977.In vitropackaging ofaADamvectorcontainingEcoRI DNAfragments of Escherichia coli andphageP1. Gene 1:255-280.

35. Sternberg,N.,and R. A.Weisberg.1975.Packagingof

http://jvi.asm.org/

(13)

PHAGE A MUTATIONS AFFECTING Q GENE ACTIVITY prophage and host DNA by coliphageA.Nature

(Lon-don) 256:97-103.

36. Szpirer,J., and P. Brachet. 1970. Relation physiolo-giquesentreles phagestemperesAet N80.Mol.Gen.

Genet. 108:78-89.

37. Szybalski, W., K. Bovre, M. Fiandt, S. Hayes, Z. Hradecna, S. Kumar, H. A. Lozeron, H. J. J.

Nijk-amp,and W. F. Stevens.1970.Transcriptional units

and their controls in E. coliphage A: operons and

scrptons.Cold Spring Harbor Symp. Quant. Biol. 35: 341-353.

38. Zissler, J., E. Signer, and F. Schafer. 1971.Therole of recombination in growth of bacteriophage lambda.I. Thegammagene, p.455-468.InA.D.Hershey(ed.), The bacteriophage lambda. Cold Spring Harbor

Labo-ratory,Cold Spring Harbor, N.Y.

VOL. 30,1979 ₁₃

http://jvi.asm.org/