Occurrence of the bacteriophage lambda receptor in some enterobacteriaceae.

(1)

Copyright01975 American Society for Microbiology Printed in U.S.A.

Occurrence of the

Bacteriophage Lambda Receptor

in

Some

Enterobacteriaceae

MAXIME SCHWARTZ* AND

LtON

LE MINOR

Unite deGenetique Moleculaire and Service des Enterobacteries, Institut Pasteur, Paris, France Received forpublication 25 November 1974

In

Escherichia coli K-12, the

receptorfor

phage lambda is

anouter

membrane

protein which inactivates the phage

in vitro.

Lambda

receptoractivitywasfound

in extracts from

all

wild strains of E. coli tested, although mostofthem failto

support

growth

of the

phage.

Insome cases

this

failure is

due

toamasking of the receptorin

vivo, the bacteria being unable

to

adsorb the

phage orto reactwith

antireceptor antibodies.

In

other

cases,

adsorption

doesoccur,

and

thenatureof

the block in

phage growth

wasnot

investigated.

Most Mal+ strains of Shigella

have

lambda

receptor,whereasmost

Mal

-strains donothave it.Synthesis of the

lambda

receptor in

Shigella

is thus

presumably controlled by

the positive

regulator

gene

of the maltose regulon

asis

the

caseinE.

coli

K-12.

Phage

lambda

adsorbs on many Mal+ strains of

Shigella and

even

yields plaques

on some of

them, althoughatalowfrequency. No lambdareceptoractivity could be found in extractsof

several strains

of

Salmonella and

Levinea.

Bacteriophage

lambda

was

originally isolated

from

a

culture of Escherichia coli K-12, which

happened

to

be lysogenic

for

this

phage

(8).

Since

its

discovery,

phage lambda

was

tusually

propagated

on

nonlysogenic derivative of

E. coli

K-12

ormore

rarely

in

E. coli C

(2),

instrains of

Shigella

flexneri (5),

and in

hybrids

between E.

coli and

Shigella (5,

11)

or

Salmonella strains

(1).

There

is

little information

in

the literature

on

the host range

of

this

bacteriophage.

We

found that

plaques of

phage lambda could be

obtained

on

only five

out

of

70

randomly

se-lected strains of E. coli. The

inability

of most

strains of

E. coli

to

support

growth

of

the

phage

could

be due

to one or moreof

several reasons,

the

most

obvious

being

that those strains do

not

adsorb the phage,

perhaps because

they

lack the

specific receptor.

However,

possession

of

the

receptor may

be

suspected

to

confer

a

selective

advantage since this receptor has

been shown

to

be

a

protein of

the

outer

membrane

(15)

in-volved in

the active transport

of

maltose

(S.

Szmelcman and M.

Hofnung,

manuscript

in

preparation)

and in chemotaxis towards this

sugar

(G. Hazelbauer,

J.

Bacteriol.,

in

press).

We show in this

paper that lambda receptor

activity is

present

in extracts of most wild strainsof E.

coli,

aswellasinextractsofseveral

Shigella

strains. In some of those strains the

receptor

is

apparently

not

accessible

to the

phage,

so

that the lack

of

phage

growth

can

still

be

attributed

to a lack of

adsorption.

In other

strains

adsorption

ofthe

phage

doesoccur, and

the

block

must

therefore lie at a further step in

the infection

process.

MATERIALS AND METHODS

Bacterial strains. All the strains used are listed in Table 1, except for E. coli K-12 CR63 used to plate phage XVh (15). HfrG6 is our standard wild-type, lambda-sensitive strain of E. coli K-12, while pop 1730 is a spontaneous lambda-resistant mutant of

HfrG6,

carryinga deletion (malBA17)encompassing

lamB, the structural gene for the lambda receptor protein (7, 15). The other strains, except for E. coli ML30 (3), were chosenatrandom inthe collection of theService des Enterobacteries de l'Institut Pasteur. The E. coli 0 strains are reference strains for 0

serotypes. The other strains have been isolated from various sources,mainly from hospital patients. Char-acterizationofthe strainswasdoneaccording tothe classicaltechniques(4).

Phage strains. Phage XV isavirulent mutant of

lambda. Although this is probably not relevant for this work, it carries a b2 deletion. Phage XVh is a spontaneoushost rangemutantofXV isolated in this

laboratoryby plating XVonE. coli K-12CR63 which islambda resistant(15). XCI57is a b2+clearmutant ofX.

Media.

Shigella strains,

which are

auxotrophic,

were grown in complete ML medium (1% tryptone [Difco], 0.5% yeast extract, 0.5% NaCl, pH 7.2, supplemented with 0.4% maltose). The other strains were grown in minimalmediumM63 (7, 15) supple-mented with 0.4% maltose, as wellas0.01% L-histi-dinein thecase ofHfrG6.To growpop 1730, which cannotutilizemaltose,0.4%glycerolwasaddedtothe above medium. The complete medium usedfor plat-679

on November 10, 2019 by guest

http://jvi.asm.org/

(2)

TABLE 1. Lambdareceptoractivity incholate EDTA extractsofvarious enterobacteria

Ability to Protein concn in Receptoractiv- Receptoractivityasmeasured

Abil

nlity

to cholate EDTA itymeasured by

inactivation

of

XVc

Bacterial strains utilize extracts byinactivation I rsneo nasneo

maltosea (mg/ml)b of

XVhC

Inpresenceof Inabsenceof

1 ~~~~20%,ethanol ethanol

Escherichia coliK-12 Hfr G6 + 0.80 +-+

E.coliK-12pop1730 _ 0.85 _

E.coliML30 + 1.20 +

E. coli13.73 + 0.50 +

E. coli14.72 + 0.30 +

E. coli18.72 + 0.40 + +

E. coli63.71 + 0.65 + + +

E. coli66.71 + +

E. coli67.71 + +

E. coli 16.72 + + + +

E.coli19.72 + +

E. coli02 + 0.40 + +

E.coli04 + 1.00 t

E.coli 06 + 0.35 + + +

E. coli 07 + 0.50 +

E. coli 010 + 0.20 + +

E.coli 018 + 0.30 + +

E. coli 028 + 0.15 +

E. coli075 + 0.55 + +

E.coli081 + 0.65 +

E.coli 083 + 0.85 +

Alkalescensdispar4.74 + 0.30 + + +

A.dispar25.73 + 0.30 + +

Shigella sonnei22.70 + 0.25 + +

S. sonnei 20.73 + 0.15 +

S. sonnei 25.73 + 0.20 +

S.sonnei3.74 + 0.10 +

S.sonnei 28.73 + 0.15 + +

S.sonnei 27.70 + 0.10 + +

S.sonnei28.70 + 0.10 + +

S.sonnei29.70 + 0.15 + +

S. sonnei30.70 + 0.20 + +

S.sonnei 46.70 + 0.21 +

Shigella dysenteriae R.16.72 + 0.45 +

S.d'ysenteriae8.39.74 + 0.35 +

S.dysenteriae1.30.74 0.50 _

S.dysenteriae8.34.74 - 0.35 _

Shigella flexneri1.29.73 0.20 _

S.flexneri2.5.74 _

S.flexneri1.8.68 0.45 _

S.flexneri2.8.70 - 0.35 +

S.flexneri6.15.66 - 0.10 _

S.flexneri3.16.70 + 0.25 +

S.flexneri4.28.74 + |+

S.flexneri1.15.73 _ 0.10 _

S.flexneri1.15.73 Mal+d + 0.25 +

S.flexneri2.21.73

S.flexneri2.21.73Mal+d + +

S.flexneri2.26.74

S.flexneri2.26.74 Mal+d + +

S.flexneri1.29.74

S.flexneri 1.29.74 Mal+d + +

S.flexneri3.27.74

on November 10, 2019 by guest

http://jvi.asm.org/

(3)

TABLE 1-Continued

..Proteinconcnin Receptoractiv- Receptor

activity

asmeasured

[image:3.507.54.452.67.338.2]

Ability to

_tiliz

_e Prti.oc_cholate

_EDTA

nRcpo_ity_measuredci by inactivation of XVc Bacterial strains utlze extracts

by

inactivation

|maltosea

maltosea(mg/ml)b(mg/ml)5 of XVhcof

XVh~ 20/C

In presence ofethanol

|Inabsence

ethanolof

S.flexneri3.27.74Mal+d

Shigella boydii 1.37.74 - 0.15

-S.boydii9.36.74 - 0.25

-S.boydii14.37.74 - 0.20

-S.boydii11.38.74 _ _

S.boydii11.38.74Mal+d +

Salmonella ohio + 0.15

-SalmonellatyphimuriumLT2 + 0.25

-Salmonella minnesota S 1114 + 0.25

-S. minnesota R60 + 0.60

-Salmonelladerby + 0.70

-Salmonellavirginia + 0.30

-Salmonella strasbourg + 0.15 - _

Salmonella anatum + 0.15 - _

Levineamalonatica (7 strains) + 0.10to0.72

-Levinea amalonatica (3 strains) + 0.10to0.25

-Klebsiellapneumoniae (1 strain) + - _

Enterobacter cloacae (1 strain) + 0.70 _ _

aAsdeterminedbystreakingoneosine methylene blue maltose agar plates (9). ,Proteinconcentration wasdetermined according toLowryet al. (10).

c+, Extract contains at least

10'%

asmuch activity as does an extract of E. coli K-12 Hfr G6.

dThisstrain isaspontaneous Mal+ mutant isolated from the Mal- strain listedon the preceding line. By growthin amaltose-containing medium suchMal+ mutants canbe obtainedfrom many of the Mal- Shigella strains.

ingphage had the following composition: 1% tryptone (Difco),0.25%NaCl, and 1% agar (Difco).

Extraction and assayof the phage lambda re-ceptor. Extraction ofthe receptor from whole cells with 1% sodium cholate and 2 x 10' M ethylenediaminetetraacetate (EDTA), as well as assay ofreceptor activity, weredescribed previously (15). The rate constantforphage XVh inactivation is proportionaltothe concentrationofreceptorpresent.

Preparation of antiserum directed against the lambda receptor. Partially purified receptor was preparedasdescribed (15) exceptforminor modifica-tionduetothe scalingupofthe procedure. Themain modification is the introduction of a precipitation with 66%ethanolbefore the chloroform-ethanol treat-ment.Protein (150 ug)fromthemostpurified fraction (representing about 100 jtg of receptor), emulsified with complete Freund adjuvant and 5

Mg

of methyl-ated bovine serumalbumin, was injected subcutane-ouslyinto arabbit. The injectionwasrepeatedonce a weekover 4 weeks. The fifthweek,300

tig

ofproteins was injected intravenously and the serum was col-lected 10days later. Exhaustion ofthe serum from antibodiesreacting with E. colisurfaceantigensother than thereceptor wasdoneby incubatingthe serum with 2 volumes ofa pop 1730 suspension (10l

bac-teria/ml) insaline (0.9% NaCl) for 1 h at 37 C and then 10h at4C,andby eliminationofthebacteriaby

centrifugation. Lambda receptoractivity isinhibited in extracts by this antiserum, but not by sera from

nonimmunized rabbits. Protection of the receptor from inactivation by the antiserum is provided by an extract fromCR63 (a lambda-resistant mutant with a point mutation in lamB) and not by an extract from pop1730(a strainwhose lamB gene is deleted). These results, which will be described in detail elsewhere, demonstratethe specificityofreceptor inactivation by the antiserum used inthiswork.

Immunofluorescentlabeling. The technique used consists in first reacting the bacteria with rabbit anti-lambda receptor antiserum prepared as de-scribed above and then adding fluorescein-labeled goatimmunoglobulins directed against rabbit

immu-noglobulins(13). Thistechniqueclearly differentiates between cells of HfrG6, which become labeled with fluorescent antibodies, and cells of pop 1730, which remain unlabeled. In control experiments, the anti-lambda receptor antiserumwasomitted andno label-ingwasobserved.

Electronmicroscopy. Exponentially growing cells werecentrifuged andresuspendedat a concentration of

10'

bacteria/ml in 10-2 M MgSO,. Purified XV phage was added at a multiplicity of 400 active particles per bacterium. After 15-min incubation at 37C, most oftheunadsorbedphagewasremoved by

centrifugation,and thecellswereresuspendedin102 M MgSO4 and fixed by adding 0.5% formaldehyde.

Thebacteriawerethendepositedoncarboned, Form-var-coatedgridswhich hadpreviously been submitted to a glow discharge. The preparations were stained

VOL. 681

on November 10, 2019 by guest

http://jvi.asm.org/

(4)

with 1%//. uranyl acetate and examined under a Sie-mensElmiskop 101 electron microscope.

RESULTS

Lambda receptor activity in extractsfrom various enterobacteria. Extracts of E. coli K-12 inactivate host range mutants of phage lambda, such as the XVh strain used in this

work (15). They also inactivate wild-type phage lambda if saturating amounts of chloroform

(15)or20%ethanol (M. Schwartz, unpublished

data) were present duringthe reaction. Inacti-vationofbothX and Xh isduetothepresencein

the extractsof lambdareceptor, aprotein from

theouter membrane. Asignificant activitywas

found in 21 randomly selected E. coli strains includingtwo strains from the

Alkalescens

dis-pargroup. An apparently less clearcutresult is

obtained with

Shigella,

since extracts from

some strains have lambda receptor activity

whereas extracts from other strains do not.

Although all testedE. coli strains are Mal+, as are most strainsofthisspecies, thesame isnot true of Shigella strains. An almost perfect

correlation is observed between the ability of

Shigella strains to utilize maltose and the

presence in their extracts of lambda receptor

activity. Hence, 19Mal+strains outof 20make

an active lambdareceptor, and 18Mal- strains

outof 19 do not make it.

Furthermore

someof

the Mal+strainsofShigella listed in Table1are in fact Mal+ variants of strains which were

originally Mal- anddevoid of lambda receptor.

Such results regarding the Shigella strains

ex-tend the observation of Gemski et al. (5) and strongly suggest that, in Shigella as in E. coli

K-12, synthesis of the enzymes of maltose

metabolism andofthe lambdareceptorprotein is regulated by asingle positive regulatorgene,

which would be inactive in most naturally occurring Mal- strains. Extracts of several Mal+ strains of Salmonella and Levinea areall

devoid oflambdareceptoractivity.

The results in Table 2 demonstrate that the

activity found inextractsofE.coli andShigella

strains can be inhibited by an antiserum

di-rected against the lambda receptor ofE. coli

K-12. The activityinsuchextractsisthus likely

dueto aprotein similartothe lambdareceptor

ofE. coli K-12.

Ascan beseen inTable 1, inactivation ofXV

by extracts of several E. coli and Shigella

strains does not require thepresenceof

chloro-form or ethanol. Whether this results from a property ofthe lambda receptorprotein itself,

which wouldbeslightly different in such strains

from what it is in E. coli K-12, or from the

TABLE 2. Inhibitionoflam bdareceptoractivity from

variousstrainsby antiserumdirectedagainstthe lambdareceptorfrom E.coli K-12a

Untreated Extract treated extract withantiserum Bacterialstrains Pla- Pla- Pla-

Pla-quesat quesat quesat quesat

t =O t= 12 t=O t = 12

min min min min

Noextract 490 436 555 560

Escherichia coli K-12 250 53 548 515 Hfr G6

E.coli66.71 338 55 515 445

E.coli 67.71 179 29 502 354

E.coli 19.72 278 47 481 423

E.coli 13.73 283 46 488 294

Shigella sonnei 22.70 526 41 470 457

S.sonnei 3.74 560 40 432 357

aThe antiserum against lambda receptorwasused

at a 1/30 dilution in saline. E. coli extracts were diluted 100-fold, and S. sonnei extracts 10-fold, in 10-2 M Tris buffer, pH 7.5. These diluted extracts were incubated for 1 h at 37C with 100 ul ofsaline (untreated extracts) or with 100

pl

of the diluted antiserum (extract treated withantiserum). The vol-umes werethencompleted to 1 ml with 10-2 M Tris bufferand, at time (t) =0min, 1ml of a suspension of

XVh in 10-2 M MgSO4 was added. One-hundred-microlitersamples were withdrawnimmediately (t= 0min)and 12 min later (t = 12min) andaddedto100 uI ofindicatorbacteria (CR63)for

plating.

presence of some other component in the ex-tracts still remainsto befound.

Absence of receptor activity in extracts of strains other than E. coli and Mal+ Shigella could be due either to an actual absence of receptor in these strains, or to a lack of extrac-tion ofthe receptor by 1% sodium cholate and 2

X 10-3M EDTA. In

fact,

as seeninTable

1,

the amount of

protein

extracted varies

depending

on the strain. However, no obvious correlation was foundbetweenthe amount of

protein

found in the extracts and the presence or

absence

of receptor

activity.

Also,

even sonic extracts of two Salmonella strains and five Mal- Shigella strains, supplemented with 1% sodium-cholate and 2 x 10-3 M EDTA, were found totally

devoid of lambdareceptor

activity. (Extracts

of E. coli K-12 HfrG6 prepared in this way have the same activity as regular cholate EDTA extracts.)

Consequently

it would seem that absence of receptor activity in cholate EDTA extracts of a strain really reflects

absence

of active receptor in these strains, rather than a lack ofextractibility.

Accessibility

of

lambda

receptor in vivo.

Several of the strains listed in Table 1, mainly

J.VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/

(5)

those

which

were

shown above

to

have

receptor,

were tested

for

their ability

to

adsorb

phage

lambda.

As seen in

Table

3,

all

Shigella

sonnei

strains, as well as some ofthe E.

coli

strains,

do

adsorb

XVh, as would be expected since they

have receptor. Most of the E. coli strains, however, do not adsorb the phage, although

their

extracts contain receptor

activity.

This

result could

beexplained if, insuchstrains, the

receptor

was not accessible to the phage. An

independent way of studying the accessibility of

the

receptor in vivo consists of testing its ability to react with antibodies. Contrary to what

happens with

antibodies directed against

sev-eral other

surface antigens, antibodies against

lambda receptor

do not induce an easily

detect-able clumping

of bacteria. Some differential

clumping

of sensitive versus

lambda-TABLE 3. Adsorption of phage lambda and of anti-lambda receptor antibodies on someenterobacteria

Bacterial strains AdsorptionofXVhmeas- Adsorption ofXVdetected Antibodylabeling uredby centrifugationa byelectron

microscopyb

(immunofluorescence)c

Escherichia coli K-12 HfrG6 E.coli K-12 pop 1730

E. coli

ML30

E. coli 13.73 E.coli 14.72 E.

coli

18.72

E.coli 63.71 E.coli 66.71

E.coli 67.71 E.coli 16.72 E.

coli

19.72

E.coli 02

E.

coli

04 E.

coli

06 E.coli 07 E.coli010

E. coli 018 E.coli 028 E.coli 075

E.

coli 081

E.

coli

083

Alkalescensdispar4.74

A.dispar 25.73

Shigella

sonnei 22.70

S.sonnei20.73

S.sonnei 25.73 S.sonnei 3.74 S.sonnei 28.73 S.sonnei27.70

S.sonnei28.70

S.sonnei 29.70 S.sonnei 30.70 S.sonnei 46.70

Shigella

dysenteriae

1.30.74

Shigella

flexneri

1.15.73 Shigella boydii1.35.74 SalmonellatyphimuriumLT2 Salmonella ohio

Salmonellaanatum

+

+ +

+ + + +

+ +

+

± +

+

+ + + + + + + + -4-+

+

+ +

+

aAbout 5 x 103 active

AVh

particles were added toabout 2 x 108 bacteriasuspended in 1 ml of10-2 M

MgSO,. The same amount ofphage was added to a control tube containing 1 ml of 10-2 M MgSO,but no bacteria.Aftera 15-minincubationat 37C,thesuspensionswerecentrifugedfor5min at 10,000 rpmand the

supernatantwasassayedfor activephageparticles.+,Indicatesthat the supernatantcontainedless than 10% phageasfoundin thecontroltube; -, indicates that thesupernatantcontainedmorethan 50%phageasfound inthecontroltube.

"Thebacteriawereeither devoidofadsorbedphageparticles (-)orcoveredwithatleast100particles (+). cThebacteriawereeitherclearlyfluorescent (+)oralmostinvisible (-) underthemicroscope.

683

on November 10, 2019 by guest

http://jvi.asm.org/

[image:5.507.59.453.193.592.2]

(6)

resistant bacteria upon addition of suitably

diluted antiserum

can be

detected

under the

microscope,

but this reaction does not have a

sufficient degree of

reliability.

For this reason

immunofluorescence

was

preferred,

because

it

gave

unambiguous results.

Out of 14 strains of

bacteria which

did not

adsorb the

phage

al-though they have lambda receptor, 12 cannotbe

labeled

with the

antibodies,

a result favoring

inaccessibility

of the receptor in such strains

(Table 3). On the other

hand,

outof 15

E.

coli

and

Shigella strains which

do

adsorb the

phage

14 reactwith the

antibodies.

Plating

efficiency

of

phage lambda

on

vari-ous E. coli and

Shigella strains.

All the strainslisted in

Table

1had firstbeen tested for their

ability

tosupport

growth

of

phage

lambda,

by cross-streaking

against XVand

XVh.

Only

E.

coli K-12 HfrG6 gave a

positive

result.

Once

it

became

apparent that several of the strains

could

adsorb

the

phage

a furthercheck of their

ability

to support its growth seemed desirable.

No

plaques

of XVor

XC

157 could

be

obtained

on

any ofthe sevenstrains ofE.

coli which

adsorb

lambda,

nor on four out of'the 10strains of

S. sonnei

which also adsorb the

phage.

The

effi-ciency

of

plating

of

phage lambda

on these

strains must

be

less than 10-8 or 10-9.

On

the

six

remaining

S.

sonnei

strains,

however,

plaques

of XV and

XCI57

were

obtained,

at

frequencies ranging

between

10-5and 10-7.

Experiments

are in progress to

determine

whether the rare plaques

obtained

in thiscase

are formed

by lambda

mutants present in the

phage

stock, or

by wild-type

particles which

escape a restriction system.

DISCUSSION

Active

lambda

receptorcan

be

extracted f'rom most Mal+ strain of E.

coli

and

Shigella.

On the other

hand,

no

lambda

receptor was found in

extracts of several Mal- strains of'

Shigella,

a result suggesting that regulation of maltose metabolism and lambda receptor synthesis is thesame in

Shigella

as it is in E.

coli

K-12 (7). Lambda receptor activity could not be found in extracts of several

strains

of'

Salmonella

and

Levinea, although

all the strains

tested

were

Mal-'. No searchhasbeen done in such bacteria

f'or a protein cross-reacting

immunologically

with lambda receptor. Such a protein may be present, still endowed

with

a f'unction in

mal-tose transport andchemotaxis towards maltose,

but

unable toinactivate phage lambda. Extracts of' several strains of' E.

coli

and

Shigella inactivate wild-type lambda even

in

absence of chloroform or

ethanol,

contrary

to

what is found withextractsofE. coli K-12. This observation may be of

importance

for the

study

of theinteraction between

phage

lambda and its

receptor.

Although

all wild strains of' E. coli tested

during

the course of this work

synthesize

a

lambda receptor molecule active in

vitro,

they

do not support

growth

of the

phage.

In many

casesthis is due to

inaccessibility

ofthe

recep-tor in

vivo,

since thebacteria donot adsorb the

phage

and do not react with

antireceptor

anti-bodies. In suchbacteria the receptor is

likely

to

be buried under the

lipopolysaccharide

(5).

However, although

the receptor is accessible in

atleast one

rough

strain, i.e.,

E. coliK-12

(14),

it is also accessible inseveral smooth strains of

well-characterized

0-antigenic specificity.

A

possible interpretation

would be that the

lipo-polysaccharide

of the different strains have

0-specific

chains ofdifferent sizes which inturn

havedifferent effectsonthe

accessibility

ofthe receptor.

In

sevoral

strains of E.

coli

and

Shigella

on

which

phage

lambda can

adsorb,

it stillcannot

grow.

Why

this is sohas not been

investigated.

The

phage

may be unable to

inject

its DNA

(16),

its DNA may be restricted (2, 12),

and/or

someofthe

phage

gene

products

may be unable

tointeract

correctly

withsomespecific

products

present in these bacteria (1, 6).

Conceivably,

elucidation of' what blocks

phage

development

in such bacteria could

yield

interesting results

concerning

the interaction of the

phage

with its

host.

Particularly

suitable for such studies would be the strains of'S. sonnei shown hereto

plate

the

phage

at a low but measurable

fre-quency.

ACKNOWLEDGMENTS

We gratefully acknowledge the skillful technical help of Madeleine Jolit, Celestine Derval, and Christine Charle-Marsaines. We express our thanks to Claude Frehel for examiningourpreparationsunder theelectronmicroscope.

This work wassupported by grants from the Centre

Na-tionalde la RechercheScientifique, theDelegationGenerale

a la RechercheScientifique et Technique, and the National

Institutesof Health.

LITERATURE CITED

1. Baron, L. S., I. R. Ryman, E. M. Johnson, and P. Gemski,Jr.1972.Lytic replicationofcoliphage lambda in Salmonella typhosa hvbrids. J. Bacteriol. 110:1022-1031.

2. Bertani, G., and J. J. Weigle. 1953. Host controlled

variation inbacterial viruses. J.Bacteriol.65:113-121.

3. Dessein, A., and M.Schwartz. 1974. Is there a pullula-naseinEscherichiacoli?Eur. J.Biochem.45:363-366. 4. Edwards,P.R., and W. H.Ewing. 1972.Identificationof

enterobacteriaceae, 3rd ed. Burgess Publishing Co., J. VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/

(7)

BACTERIOPHAGE LAMBDA RECEPTOR

Minneapolis.

5. Gemski, P., Jr., J. A. Alexeichik, and L. S. Baron. 1972. Behavior of coliphage lambdainShigella flexneri2a.J. Virol. 10:668-674.

6. Georgopoulos, C.P., R. W. Hendrix, S.R.Casjens, and A. D.Kaiser. 1973. Host participationinbacteriophage lambda head assembly.J.Mol. Biol.76:45-60. 7. Hofnung, M. 1974. Divergent operons and the genetic

structureofthe maltose B region in Escherichia coli K12. Genetics76:169-184.

8. Lederberg, E. M. 1951. Lysogenicity in E. coli K12. Genetics 36:560.

9. Lederberg, J. 1947. Gene recombination and linked segregation inEscherichia coli. Genetics 32:505-525. 10. Lowry, 0. H., N. J. Rosebrough, A. J. Farr, and R.J.

Randall. 1951. Protein measurement with the Folin phenolreagent.J. Biol.Chem. 193:265-275.

11. Luria, S. E., and J. W. Burrous. 1957. Hybridization between Escherichia coli and Shigella. J. Bacteriol. 74:461-476.

12. Meselson, M., R. Yuan, and J. Heywood. 1972. Restric-tion and modificationofDNA. Annu. Rev. Biochem. 41:447-466.

13. Nairn, R. C. 1969. Fluorescent protein tracing, 3rd ed. Livingstone, Edinburgh.

14. 0rskov, I., and F. Orskov. 1960. The H. antigenof the "K12" strain. A new E. coli H antigen: H48. Acta Pathol. Microbiol.Scand. 48:47.

15. Randall-Hazelbauer, L., and M. Schwartz. 1973. Isola-tionofthe bacteriophage lambdareceptorfrom Esche-richia coli. J.Bacteriol. 116:1436-1446.

16. Scandella, D., and W. Arber. 1974.AnEscherichia coli mutant which inhibits the injectionofphageADNA. Virology58:514-525.

VOL. 15, 1975