Lethal action of bacteriophage lambda S gene.

(1)

0022-538X/82/120886-07$02.00/0

Lethal Action of Bacteriophage

X S

Gene

JINNIE M.

GARRETT

AND RY YOUNG*

Departmentof Medical Biochemistry, Texas A&M University,CollegeStation, Texas 77843

Received4June1982/Accepted5 August 1982

Thefunctionsofthe bacteriophage X lysis genes S, R, and Rz were investigated.

Different combinations of wild-type and inactive alleles ofall three lysis genes

wereclonedinto the plasmid pBH20 and were expressed under the control of a lac operator-promoter. Theinvolvement ofthe Rz gene inlysiswasproposedin our

previouswork andwasconfirmedby theMg2+-dependent lysisdefect of clones in

which part of the Rz gene is deleted. Membrane vesicles preparedfrom induced S+ cells were showntohaveaseverelyreducedcapacityfor active transport of glucose; thisdefect was detectableatleast 20minbeforelysis.Cellviability was also shown to decrease verysoonafterinduction,long beforephysiologicaldeath

and lysis;this decrease inviability isabsolutely dependenton Sexpression and

independent of R and Rz. The nonviable fraction of cells at any time after

induction was demonstrated tobe equal to the fraction committed to eventual lysis. Induction ofan Sts clone showed that the S gene product is stable and

capableofinducing lysis long afterthecessation ofsynthesisof S geneproduct.A

modelfor Saction is proposed.

Theterminatingeventof the infective cycle of

many viruses is lysis of the host cell with

con-comitantrelease of the progeny virus particles.

Three genes have been directly implicated in lysis of thebacteriophage X: S, R, and Rz (16, 19). The X lysis gene region has been cloned undercontrol of the lac promoter in theplasmid

pBH20 (9). Inductionof thelysis operon by the

addition ofthegratuitouslac inducer

isopropyl-P-D-thiogalactopyranoside

(IPTG)results in cell

lysisin the S+R+Rz+ clone. It is mucheasierto

studythemechanismofaction ofthe individual

lysis proteins in the cloned system, since the

lysis genes are removed from themultiple con-trols ofbacteriophage development.

Two of the lysis proteins, those coded by genes R and Rz, have been implicated in the

degradation of therigid peptidoglycan cell wall

(2-4). Together these activities constitute the

bacteriolytic activity previously designated as

endolysin (18). The R gene product (pR) has

beenpurifiedandbeenshowntobea

transglyco-sylase which degrades the peptidoglycan by attacking the glycosidic linkages (3). The Rz gene product (pRz) has been tentatively identi-fied as an endopeptidase which is believed to play aminor role inpeptidoglycan degradation bydigesting oligopeptide cross-links within the peptidoglycan (19).

The role of the S gene product (pS) is un-known, but there is evidence that it isrequired

for the transport ofbacteriolytic activityto the periplasm. Ourprevious work showed that in-duction ofthe S-R+Rz+ clone results in

exten-siveaccumulationof intracellular endolysin (9);

however, nolysis or reduction in the rate of cell growth was observed, indicating that endolysin was not reaching the vicinity of the peptidogly-can.

In early work on the mechanism of A lysis,

Reader and Siminovitch(17) suggested thatpS attacks the cytoplasmic membrane to allow transport of endolysin to the periplasm. They observed that in the case of induced

S+R-lysogens the cytoplasmic membrane becomes

permeabletosucrose; thatis,thesecellscan no

longer be plasmolyzed by hypertonic sucrose. This suggests that the normal integrity of the membrane has beendisrupted. However,

elec-tronmicroscopicexamination of thin sections of

induced XS+R- lysogens beyond the normal

lysis time revealedno detectable change in the cytoplasmic membrane structure. There is no detectablephospholipase activityattributableto pS; in the absence ofSgenefunction,celllysis

by mechanical methods results in a small amount of lipid degradation indistinguishable

from that occurring in X S+-infected cells (14). Other workers have noteddeleterious effects of pS action on membrane-dependent oxidative metabolism (1, 6), whichalso suggestsaninner membrane lesion of some kind. However, no model forSaction has beenproposed.

Inthis work wehaveuseda complete setof cloned S+R+Rz+

lysis

gene allelesto study the action of theSgene.Wedocument theeffect of induction of these cloned lysis genes on cell viability, membrane permeability and active

886

on November 10, 2019 by guest

http://jvi.asm.org/

(2)

LETHAL S

transport,andlysis kinetics. A general model for

pS action is proposed.

MATERIALSANDMETHODS

Bacterial strains, phages, and plasmids. The host

bacterium formost experiments wastheEscherichia

coli strain WFK, which is W3101 pro phe/F-lacIq

lacZ::Tn5 (9). The methylation-deficient E. coli K-12

strain GM119 (F- dem dam-3 metBI gaIK2 gaIT22

lacY) tsx-18 supE44) (obtained from J. A. Arraj)was

usedfor the preparation of unmethylated phage and

plasmid DNA. Thebacteriophages and plasmids used

arelisted in Table1.

Chemicalsandenzymes.Ethidiumbromide, EDTA,

IPTG, phosphoenolpyruvate (PEP),

glucose-6-phos-phate dehydrogenase, lysozyme, and the restriction

enzymeEcoRI werepurchased from Sigma Chemical

Co., St. Louis, Mo. T4 polynucleotide ligase was

obtained from NewEngland Biolabs, Beverly, Mass.

Allother restrictionenzymes wereobtainedfrom

Be-thesda Research Laboratories, Gaithersburg, Md.

[14C]sucrose (477 mCi/mmol) and [14C]glucose (500

mCi/mmol)werepurchased from New England

Nucle-arCorp.

Media andgrowth conditions. Cultures for all

experi-ments except membrane vesicle preparations were growninminimal media (7) supplemented withamino

acids and thiamine. Antibiotics usedwere ampicillin

(50 ,ug/ml) andkanamycin (30 ,ug/ml).

For lysis kinetics and cell viability experiments,

culturesweregrowntoadensity of10'cellsperml and

were induced by the addition ofl0' M IPTG. Cell

viability was followed by plating dilutions of the

culture on broth plates (1% tryptone, 0.5% yeast

extract, 0.5% NaCl) atgiven times before and after induction.

Cultures for membrane vesicle preparations were

grownin rich media(3.2% tryptone, 2.0%o yeast

ex-tract, 0.5% NaCI) with vigorous aeration and were

harvestedataconcentration ofapproximately5 x 109

cellsperml.

Plasmid construction. The S+R-Rz+ (pJG1) and

S-R-Rz+ (pJG3) plasmids were constructed as

de-scribedpreviously (9).TheS+R+Rz- plasmids (pJG7,

pJG8, and pJG9) were prepared by digestion of the

unmethylatedparentplasmid DNA (prepared from the

dam- strain GM119) with BclI and BamHI and

religa-tion. TheStsR+Rz- plasmid(pJG13)wasconstructed

by digestion of unmethylated phage DNA withEcoRI

and BclI and by ligation of the isolated lysis gene

fragment into pBH20 DNA cleaved with EcoRI and BamHI.

Transformation was performed with cryogenically

stored competent cells by the method of Morrison

(13).Transformants containing the lysis plasmidswere

identifiedaspreviouslydescribed (9).

Membrane veside experiments. Membrane vesicles

werepreparedfrom bacterial cultures by a

modifica-tion of the lysozyme-EDTA methodof Kaback (11).

Cultureswereharvestedby centrifugationat16,000x

gandweresuspendedatroomtemperature in 30 mM

Tris-hydrochloride, pH 8.0, (1g[wet weight] ofcellsper

80ml) containing 20%osucrose. The cells were

con-verted to spheroplasts by adding potassium EDTA,

pH 7.0,andlysozymetofinalconcentrations of 10 mM

and 50,ug/ml, respectively, and incubating for 1 h. The

spheroplastswerecollectedby centrifugationat16,000

x g for 20 min and were suspended with a glass

homogenizerinthe smallestpossible volume of0.1M

potassium phosphate buffer, pH 6.6, containing20%

sucrose, 20 mM magnesium sulfate, and 5 ,ug of

DNaseperml. The suspension was poured into 300

volumes of 50 mM potassium phosphate buffer, pH

6.6,at37'C andwasincubated withvigorous swirling

for 45 min. After 15 min of incubation, potassium

EDTA, pH 7.0, wasaddedtoafinalconcentration of

10 mM, and after a further 15 min of incubation,

magnesium sulfatewasaddedto15 mMfinal

concen-tration.The suspensionwasthencentrifugedat45,000

x g for 1 h to collect the membrane vesicles. The

vesiclesweresuspended byhomogenization in0.1 M

potassium phosphate, pH 6.6,ataconcentration of 7

mgofproteinperml, quick-frozen in ethanol-dry ice,

and stored at -80°C. Vesicles were prepared from

inducedlysis clonesatvarious timesafter the addition

ofIPTG.

Membrane vesiclesweresuspendedata

concentra-tion of 35 mg/ml in 40 ,ul of 0.1 M potassium

phos-phate, pH 6.6, containing4mMunlabeledsucroseand

10,uCi of [14C]sucrose andwereincubated overnight

at4°C. At timezero,thelabeledvesicleswerediluted

TABLE 1. Phagesandplasmids

Designation Genotype Sourceor Reference

Phages

X363R- XbSl9b5l5cI857nin5RTn903.979 R. Young

X363Sam7R- XbSl9bSl5cI857nin5Sam7RTn9O3.979 R. Young

ASts XcI8575ts9B D.B. Wilson

Plasmids

pBH20 AprTcr DerivedfrompBR322 (10)

pRF26 AprS+R+ Rz+ Derived frompBH20 (9)

pJH2 Apr Sam7 R+Rz+ Derived frompBH20 (9)

pJG1 Apr S+R-Rz+ Derived frompBH20 (this paper)

pJG3 AprSam7R-Rz+ Derived frompBH20(thispaper)

pJG7 Apr S+R'Rz- DerivedfrompBH20(this paper)

pJG8 AprS+R-Rz- Derived frompBH20(thispaper)

pJG9 AprSam7R-Rz- DerivedfrompBH20 (thispaper)

pJG13 AprSts9BR+Rz- DerivedfrompBH20(thispaper)

pKY2 AprTCs Derived frompBH20(thispaper)

VOL.44,

http://jvi.asm.org/

[image:2.491.46.446.517.677.2]

(3)

200-fold into 0.1 Mpotassiumphosphate(pH6.6)40

mM MgCl2, and atvarious times 2-ml samples were removed, filtered through Millipore filters (pore size, 0.45 pLm),andwashed under suction with 5 x 1 ml of

0.1 MLiCl. The filters were dried and counted in a

TracorAnalytic Delta 300 liquid scintillation counter.

Glucose-6-phosphatewasassayed by the method of

Olive and Levy (15), and the protein concentration

wasdeterminedby theBradford dye-binding assay (5).

RESULTS

Construction of hybrid lysis operons. In previ-ouswork(9), the Alysisgenes S(bothS+and S-alleles) and R were cloned into the plasmid pBH20(Fig. 1). Tocomplete the set of cloned genes, the S+R- and S-R- alleliccombinations werecloned by identical methods; the resultant

plasmids carryingS+R- and S-R- hybridlysis

operons were designated pJG1 and pJG3, re-spectively. Whereas induction of the S+R+ clone with 10-3 M IPTG results in a sharply

definedlysis after40min (9;Fig.2),induction of

acultureoftheS+R- cloneresults in an eventu-al heventu-alt in mass accumulation (Fig. 2), and the cells appear tolose motility undermicroscopic

observation (data not shown). Predictably,

in-duction of the S-R- clone has no detectable

effect on culturegrowth.

Verification that no other genes

promoter-distaltoRz (19;Fig.1) areinvolved in cellkilling

orlysiswasobtainedbyconstructinganother set

of plasmids in which the region of A DNA

beyondthe BclI site within the Rz gene (Fig. 1)

wasdeleted by digestion withBclI and BamHI and by subsequent religation. This series of

clones,which alllacktheintact Rz gene,

result-ed in lysis kinetics identical to those of the

parentalclones,showingthat there are no other

X genes involved in lysis (data not shown). However, the S+R+Rz- clone did show

aber-rantlysis in the presenceof10 mM Mg2+ (Fig.

3).This aberrantlysispatternisassociatedwith

theformationofmechanicallyfragilespheroidal

cells and constitutes the Rz- phenotype de-scribed forbacteriophages carryingthe kanamy-cin resistance transposonTn9O3 inserted in the Rzgene (19).

0.8-

0.4-

0.2-c

b

a

1 2

[image:3.491.277.446.61.209.2]

THMEAFTERFTGADDITKON (a)

FIG. 2. Induction of lysis gene clones. Cultures

were grown inglycerol-minimalmedia (7) at 37°C to a

concentrationof108cells per ml. Thelysisgenes were

induced by adding 1 mM IPTG and culture mass

accumulation followed. (a) S+R+Rz+; (b) S+R-Rz+;

(c)Sam7R+Rz+; A550, absorbance at 550 nm.

The StsR+Rz- clone was constructed by

di-gestion of

XSts

DNA with identical restriction

digestions andligation;the clonescontaining the

lysis genes were screenedfor inducible loss of

viability at the permissive temperature (30°C).

Inductionofthetemperature-sensitive Sgeneat

therestrictive temperature (42°C)hadno effect

on culture growth (Fig. 4). Subsequent shift of

the culture to 30°C resulted in immediate

lysis

(Fig. 4).

The functional stability of the

temperature-sensitive S protein was investigated. Cells that

had been induced for 1 h at the restrictive temperature were diluted 25-fold into

nonin-ducer medium(Fig. 5).Thisdilution(5 x

10-4

M

to 2 x 10-5 M) was shown to decrease

1B-galactosidasesynthesis50-foldfromtheinduced

level and to suppress gene S killing activity

(Table 2) and should therefore turnoffpS

syn-thesisatthe timeof dilution. Temperatureshift

ofthe cultureto30°Cafterincubationindiluted

mediumfor1 hresulted inimmediatelysis(Fig.

5). This indicates that pSts is stable within the

cellattherestrictivetemperature.

0.4-"' 0.2-H/ndE

Tn905 PR' TR

. S R z

4.5. l 46 465 47 T 47.5 48 485(KIo-bp)

EwRI Bc/I

FIG. 1. Genetic mapof thelysisregion of phageX.

Thepositions of the lysis genes and relevant

restric-tion enzyme sites are shown. The stem and loop

structurerepresents the transposonTn9O3 inserted at

48.5kilobasepairs.

1 2

TIME AFTERPTGADDMTON(w)

FIG. 3. The effect of magnesium on lysis in the

S+R+Rz- clone. Cultures of strain WFK(pJG7) were

grown and induced as described in thelegend to Fig. 2

inglycerol-minimalmedium without (a) and with (b) 10

mMmagnesium chloride.

b

a

http://jvi.asm.org/

[image:3.491.282.426.524.633.2] [image:3.491.53.247.568.675.2]

(4)

t

0.5-oc

0.3 b

1 2 3

TIME AFTERIPTG ADDITION (hrs)

FIG. 4. Induction of the temperature-sensitive S

protein. A cultureof strain WFK(pJG13)(StsR+Rz-)

was grown at42°C in glycerol-minimal medium and

wasinduced by adding 1 mM IPTG. Portions of the

culturewereshiftedto30°Cafter incubation in

induc-ing conditions for1 h (a)or2 h(b).

Effect of gene S on membrane permeability.

Since gene S is implicated in the transport of bacteriolytic activity totheperiplasmand in the permeability of the inner membrane to sucrose

(9, 17), it would be reasonable to expect that membranes isolated from induced S+ lysis cloneswould have altered permeability

proper-ties. To characterize the action of pS on the membrane, we prepared membrane vesiclesby

the method of Kaback (11) from the induced S+R- and S-R- lysis clones as well as from

cells containing no cloned lysis genes. Mem-brane vesicles prepared by this method have

0.10-

08-

0.6-

0.4-

02-

0.1-06.

0.04-

0.02-I

PI

3TC

1 2 3 4 5

TME Ow.)

FIG. 5. Stability of the temperature-sensitive S

protein. A culture ofWFK(pJG13) (StsR+Rz-) was

grownandinducedasdescribed in the legendtoFig. 4. After1 hofincubation, aportion of theculture was

[image:4.491.51.207.55.199.2]

diluted25-fold intononinducermedium; 2hlater, the culturewasshiftedto30°C.

TABLE 2. Effect of IPTG concentrationonlossof

cellviabilitya

% Survivalattime after

IPTG concn (M) induction:

20min 60min

1 x lo-, 27 3.6

5 x 10-4 35.6 4.4

1 x 10-4 75.7 7.6

2 x10-5 128.0 285.0

a Cultures of strain WFK(pRF26) (S+R+Rz+)were

induced with various concentrations of IPTG. Cell

viability was followed by plating dilutions of the

cultureon noninducer rich medium atvarious times

afterIPTG addition. Cell viability is expressedasthe

percentageofsurvivingcellsatthattime,with cellsat

timezero as100o.

been shown tobe whole cell vesicles, withthe samepermeabilitybarriersasthe intact cell(11). These vesicles were used in passive sucrose diffusion experiments. Vesicles from the S+ clonewereapproximatelyfourfold more perme-able to sucrose thanwerevesiclespreparedfrom control bacteria(Table3).Surprisingly,the

vesi-clesaffected by Sam7 (S-) were alsofound to

have increased permeability to sucrose (Table

3).Webelievethis is an indication that the Sam7 nonsense fragment retains some function; in

fact, induction of the Sam7 clone results in

accumulation oflargequantitiesof thenonsense fragmentpolypeptide (55 amino acidresidues;8; D. L.Daniels, Ph.D. thesis,Universityof Wis-consin, Madison, 1981) in the membrane (E. A. Altman, J.Garrett,R.Grimaila,R.Schultz, and R. Young, manuscriptinpreparation).

Effect ofgene S on PTS-mediated active trans-port.In E.coli membranevesicles,theuptakeof severalsugars (e.g.,glucose,fructose)occursby

TABLE 3. Passive diffusion of ["4C]sucrose from

membrane vesiclesa

Initial Strainsupplyingvesicles t1/2(min)b rateof

effluxc

WFK(pBH20) 2 7.5

WFK(pJG)(S+R-) 0.46 32.6

WFK(pJG3) (Sam7R-) 0.43 34.9

a Membrane vesicles were passively loaded with

[14C]sucrose overnight, washedquickly, diluted into

label-free medium, and assayed for the retention of

[14C]sucroseatvarious times(14). Maximumcapacity

of the vesicle samples was estimated as 30 pmol of

[14C]sucrose per mgofprotein.

btj/2, Time taken for one half of the maximum

capacity of vesicle samples to diffuse out of the

vesicles.

cCalculated from [14C]sucroselost in the first30 s

ofincubation;expressed aspicomolesofsucroseper

minute permilligramofprotein.

VOL.44,1982

II I

http://jvi.asm.org/

[image:4.491.86.210.430.624.2]

(5)

phosphorylation via the

PEP-phosphotransfer-ase system (PTS). We used the S+-induced vesicles to determine the effect of pS PEP-driven [14C]glucose uptake in these membrane vesicles. S+-affected vesicles were completely

deficient in PTS activity, in contrast to

S--affected and control vesicles (Fig. 6). This result was verified by assaying the

glucose-6-phosphate produced by this system. Each

ex-periment was allowed to continue for 2 h to

allow all theglucose-6-phosphate todiffuseout

of the vesicles before starting the assay. In vesicles madeafter only 15 min of S+ induction, the membraneloses the ability to support PTS-mediated phosphorylation of glucose (Table 4). No effect is observed on glucose-6-phosphate

production in the control and S--affected vesi-cles. Thus,theloss of theabilitytosupport PTS-mediated activetransportof glucoseoccurswell

before normal lysis time in S+-affected

mem-branes. Furthermore, the Sam7-affected

mem-braneswereessentially normal in their abilityto

support PTS-mediated active transport, despite

theirincreased permeabilityto sucrosein vitro.

Kinetics of the loss of cell viabilityoninduction

of the Sprotein. The observed effect of thegene

Sprotein on the PTS before normal lysis time

led us toinvestigate the kinetics of the loss of cellviability in bacteria exposedtotheaction of the gene S protein. The sharply defined lysis curveobservedat40minfortheS+R+ clone is accompaniedby the loss ofcolony-forming units beginning much earlier, about 8minafter induc-tion(Fig. 7). Identical resultsareobtained with

x

70- 60-50

40

30

20

10

lb io 3'0 io 5o 60

TIME(mine)

FIG. 6. The effect ofpSonPTS-active transport in membrane vesicles. Membrane vesiclespreparedfrom

the S+R- and S-R- clones were suspended at a

concentration of 20mgofprotein perml in 50 mM

potassium phosphate buffer(pH 6.6)-10mM

magne-siumsulfate-0.3 MLiCI-0.1 M PEP(adjusted topH

6.5 to7.0 with sodiumcarbonate). Thesamples were

incubatedat40°C. Atvarious timepoints,40-,ul

sam-pleswereremoved, dilutedinto 2 ml of0.5 MLiCl,

rapidly filtered through a Millipore filter (pore size,

0.45 Lm)undersuction,and washedwithanother 2 ml of 0.5 MLiCl. The filtersweredried andcounted.

TABLE 4. Effect of gene S expression on

PTS-mediated active transporta

Time Glucose PTS

Relevant ~~of transport

Relevant Strain

incuba-genotype tion Amtb %of

(min) control

Host WFK 16.2 81

Control WFK(pKY2) 20.9 100

S-R- WFK(pJG3) 0 14.9 74

15 14.9 74

30 14.3 71

45 16.7 83

S+R- WFK(pJG1) 0 10.4 52

15 0.12 0.6

30 0.09 0.4

45 0.08 0.4

a Transportof glucose by the

PEP-phosphotransfer-asesystem was measuredbyassayingthe

accumula-tionofglucose 6-phosphate. Purified membrane

vesi-cles were preincubated with PEP and then were

suspended in a reaction buffer containingglucose by

the methodof Kaback (12). Glucose 6-phosphate was

assayed in a coupled system by using

glucose-6-phosphate dehydrogenase and NADP+. Strain WFK is the lacIq host; plasmid pKY2 is a

tetracycline-sensitive derivative of pBH20, and pJG1 andpJG3are

lysis gene clones (9).

bExpressed as nanomoles of glucose 6-phosphate

permilligram of protein.

the S+R- clone (Fig. 7), so actual lysis is not

required for loss of viability. The loss of viability isabsolutely dependent on a functional S allele (Fig. 7); in fact,S-R+ clones can be subcultured indefinitely in the presence ofinducer without detectablekilling (9).It isclear fromcomparison of the results in Fig. 7 that the loss of viability function of the gene Sprotein takes effect much

sooner than the observable effect on the

accu-mulation of culture mass.

Loss ofviability and commitment to lysis. To pursue the relationship between loss of cell viability and lysis, induced cells were diluted into noninducer medium at various times after induction(Fig. 8). The dilutionwas asdescribed for the Sts dilution experiment (Fig. 5). Com-parison ofFig.7 and 8shows that thepercentage of cells no longer viable at a given time after induction will lyse at the normal lysis time. These results suggest that the function of the

gene S protein consists oftwodistinct phases: initial attack on the cytoplasmic membrane, which is required for the destruction of PTS-mediated activetransport andcausesloss of cell viability, and subsequentlysisingeneR+ condi-tionsorphysiological death inR- conditions.

DISCUSSION

Wehave shownhereandinpreviouswork(9)

that theexpressionof thebacteriophage A S and

http://jvi.asm.org/

[image:5.491.90.211.440.567.2]

(6)

X 891

U,

< 0.1- -100

0.06- 50

0.04--20

-10 AB

-5

-2

1 2 34

TIME(hrs)

FIG. 7. The effect of pSoncellviability

ofS+R+Rz+ (a,*), S+R-Rz+ (b,A), and

(c,(D) were grown and induced as descrit

legendtoFig. 2. Portions of the culturewere

diluted, andplatedonrichmediatoassayce

at times before and after dilution. Cell v

plotted as percentage of survival with th

induction as 100% for S+R+Rz+ (A,),

(B,O), and S-R-Rz+ (C,).

R genes will result in a scheduled ly:

normalgrowth conditions. The expres

thirdgene,Rz,isrequiredinX-infected orininducedlysisgeneclones(Fig. 3)

containingahigh

Mg2'

concentration.

phage proteins are requiredforlysis c

timing of lysis;theexpression ofthely is necessary and sufficient for the phenomenon.

Neither the mechanism oflysis nor

is understood. The results presented I

gest that there is damage to the cyl

membrane mediatedby thegeneSpro

lesion is manifested as an increased

permeabilitytosucroseandadestructi

PTS-mediated active transport of gluc Sam7 nonsense fragment polypeptid

has a length of 55 amino acid resi(

accumulates in the membrane as does man et al., manuscript in preparati increases the passive sucrose permeal

does not destroy the PTS active trai

glucoseand hasno effectoncell viabif

Surprisingly,themembranelesionsa ent invesicles preparedvery soon aft

tionofthebacteriophage lysisgenes ra

just in vesicles prepared from the normal lysis

timeorlater. Thisresult ledus toinvestigatethe

kinetics of cell viability after induction of the

lysis genes since a membrane lesionleadingtoa

collapse ofactivetransportmightbeexpectedto

be alethal event. The results (Fig.7) show that

cell viability, measured as colony-forming

po-tential, begins to decrease dramatically very

soon after induction of the lysis genes, much

_1% earlierthan the onsetlysisinS+R+organismsor

cessation of mass accumulation in S+R-. The

>_ simplestinterpretation ofthese resultsis that the early membrane lesion detected in the vesicle

6 transport

experiments

and the

early

loss in

via-bility are manifestations ofanearly membrane

ul action ofpS.Once asufficientquantityofpS has

w

o beensynthesizedorhas exerted its effectonthe

membrane, the cell is committed to lysis at a

latertime(Fig. 8), eventhoughprotein synthesis

(and thus respiration) continue atexponentially

increasing rates for at least 30 min after the

lethal event.

What is theeventuallysistrigger, then, ifthe cell iscommittedtolysisatsuchanearlier time?

.Cultures Ourresults with the Sts9B allele suggest that pS

S-R-Rz+ is stable;furthermore, once acell has

accumu-bed

in the latedenough pS nofurthersynthesis isrequired

,,,°removed,

for the eventual lysis to occur (Fig. 8). In

1viabiityiYprevious studies, we have shown that the

phe-e

time

of nomenonof

premature

lysis

is

duplicated

in the

S+R-Rz+

cloned

lysis

gene system

(9).

Premature

lysis

was first observed in cells infected with the

bacteriophageT4,which haslysisgenestand e,

apparently analogoustothephageXgenes S and

sis under 0.3

sion of a cells(19) inmedia

Noother b

rforthe

0.2-sis genes

complete I

0.1-itstiming

here sug-

<0008

toplasmic

tein. The 0.06

I passive a

ionofthe

ose. The 0.04

e, which

dues and ; 2 3

pS

(Alt-on), also TIME(hra)

biity but FIG. 8. Loss of cell viability results in commitment

bsYty

bof

to

lysis.

Acultureof strainWFK(pRF26)

(S+R+Rz+)

nsport wasgrown and induced asdescribed in the legend to

lity. Fig. 2 (a). After 10 min (b) and 20 min (c) ofinduction,

ireappar- portions of the culture were removed and diluted into

Lerinduc- noninducermedium. TheA550was normalized to the

tther than undiluted culture in (b) and (c).

VOL.44, 1982

http://jvi.asm.org/

[image:6.491.59.219.60.316.2] [image:6.491.282.421.425.614.2]

(7)

R;severallaboratorieshavedocumented prema-ture lysis in X infections. The phenomenon is simple and dramatic; ifcyanide orany

respira-torypoison is addedtocells in which the Sgene hasbeen sufficientlyexpressed (but long before thenormallysis time), lysis occursimmediately (6). No lysis is observed ifcyanide is addedto

S--affected cells(6). Thus, poisoning respiration

apparently acts to trigger lysis prematurely by

subvertingthe normalscheduling mechanism.

Takentogether with the results in this work,

theprematurelysis phenomenonsuggestsa gen-eralmodelforS-mediated scheduled lysis. Once

acertainquantityofpShas accumulated in the

membrane, themembrane is thensusceptibleto

alysis-triggering event. Thistriggering eventis

likelytobecollapse of the chemiosmotic

gradi-ent across the cytoplasmic membrane since it can be emulated by respiratory poisons. Since nofurtherpS accumulationisnecessaryfor the

triggerto occur, one possibilityis that the

trig-geringeventoccursduringtheinception ofcell

divisionor septumformation,which is theonly

knownsingularityinmembranebiogenesis.That is, the onset of septation in the presence of sufficient pScausesa catastrophiclesion in the

membrane, resulting in the immediate collapse

of the membrane potential. Alternatively, the accumulation ofpSmayresult inadeterioration of theabilityof the celltomaintain the chemios-moticgradient.Thecellcanresistcollapseof the

gradient bycompensatoryhydrolysisof ATPin

the cellularpools or by increased utilization of the oxidative metabolic pathways. The resist-ancewould beovercomebyfurtherpS

synthesis

orbyany substance whichpoisonsrespiration.

This modelpredictseitherdecreasingATPpool

levels as the time oflysis approaches or more

rapidoxygenconsumption.

Specific

predictions

of theseptationtriggerarelessclear,exceptthat

some septation delay mutants should show

se-verely retarded lysis scheduling. These

predic-tions are currently being tested in this labora-tory.

ACKNOWLEDGMENTS

Thisstudywassupported byPublic Health Service grant 1-R01-GM27099-03 from the National Institutes of Healthand by aBiomedical ResearchSupportgrant from Texas A&M UniversityCollegeof Medicine.

J.VIROL.

LITERATURECITED

1. Adhya, S., A. Sen, and S. Mitra. 1971. On the role of gene S, p. 743-746. In A. D.Hershey (ed.), The bacteriophage lambda. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.

2.Bienkowska-Szewczyk, K., B. Lipinski, and A. Taylor. 1981. The Rgene product of bacteriophage lambda is the mureintransglycosylase. Mol. Gen. Genet. 184:111-114. 3. Bienkowska-Szewczyk, K., and A. Taylor. 1980. Murein

transglycosylase from phage lambdalysate; purification andproperties. Biochim. Biophys. Acta615:489-4%. 4. Black, L. W., and D. S. Hogness. 1969. Thelysozymeof

bacteriophage X.I.Purification and molecular weight. J. Biol. Chem. 244:1968-1975.

5. Bradford, M. 1976. Arapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72:248-252.

6. Campbell, J. H., and B. G. Rolfe. 1975. Evidence for a dual control of the initiation of host-celllysiscaused by phage lambda. Mol. Gen. Genet. 139:1-8.

7. Clark, A. J., and0.Maaloe.1967.DNA replication and the divisioncycle in Escherichia coli. J. Mol. Biol. 23:99-112.

8. Daniels,D.L., and F. R. Blattner. 1982. Nucleotide se-quenceof the Q gene and the Q to S intergenic region of bacteriophage lambda. Virology 117:81-92.

9. Garrett, J. M., R. Fusselman, J. Hise, L. Chiou, D. Smith-Grillo, J. Schultz, and R. Young. 1981. Cell lysis by induction of cloned lambdalysis genes. Mol. Gen. Genet. 182:326-331.

10. Itakura, K., T. Hirose, R.Crea, A.Riggs, H. L. Heyneker, F.Bolivar, and H. W.Boyer. 1979. Expression in Esche-richia coli of a chemically synthesized gene for the hormonesomatostatin. Science 198:1056-1063. 11. Kaback, H. R. 1971. Bacterial membranes. Methods

En-zymol. 22:99-120.

12. Kaback, H. R. 1974. Transport in isolated bacterial mem-brane vesicles. Methods Enzymol.31:698-709. 13. Morrison, D. A. 1979.Transformation and preservation of

competent bacterial cells by freezing.Methods Enzymol. 68:326-331.

14. Mukherjee, P. K., and R. K. Mandal.1976. Role of Sgene ofbacteriophagelambda in hostlysis.Biochem. Biophys. Res.Commun. 70:302-309.

15. Olive, C., and H. R. Levy. 1967. The preparation and someproperties of crystalline glucose-6-phosphate dehy-drogenase from Leuconostoc meserteroides. Biochemis-try6:730-736.

16. Reader, R. W., and L.Siminovitch.1971. Lysisdefective mutantsofbacteriophage lambda: genetics and physiolo-gyof Scistronmutants.Virology43:607-622.

17. Reader, R. W.,and L.Siminovitch.1971. Lysisdefective mutants ofbacteriophage lambda: on the role ofthe S functioninlysis.Virology43:623-637.

18. Taylor,A.1971.Endopeptidaseactivity of phage-endoly-sin. Nature(London) New Biol.233:144-145.

19. Young, R., J. Way, J. Yin, and M. Syvanen. 1979. Transposition mutagenesis of phage lambda: a new gene affectingcelllysis.J. Mol.Biol. 132:307-322.

http://jvi.asm.org/