Purification and properties of a bacteriophage T5-modified form of Escherichia coli RNA polymerase.

JOURNAL OFVIROLOGY, Apr.1975, p. 1042-1046 Copyright i 1975 AmericanSocietyforMicrobiology

Vol.15,No.4 Printed in U.S.A.

Purification and Properties

of

a

Bacteriophage

T5-Modified

Form

of

Escherichia

coli RNA

Polymerase

C. SZABO1 AND R. W. MOYER*

DepartmentofBiochemistry, College of PhysiciansandSurgeons,Columbia University, NewYork, New York 10032

Received forpublication 25 November 1974

Amodified form of Escherichia coli RNApolymerase that containsoneofthe

three T5-specific polypeptides known to interact with the host enzyme was

purified from bacteriophage T5-infected cells. The properties of this

T5-modi-fied enzyme appeared identical to those of the RNA polymerase derived from

uninfected non-colicinogenic cells and toafullyactiveenzymeisolated from

T5-infected ColIb+ cells after the limited in vivotranscription of T5 genesallowed

by the plasmid had ceased.

The RNA polymerase ofthehost cell Esche-richia coli isresponsibleforthesynthesisof the

three distinct classes of bacteriophage

T5-specific RNA formed during T5 infection. Byimmunoprecipitation studieswehaveshown

thataclassI(pre-early)T5-specific polypeptide

of 11,000 daltons and two classII (early)

poly-peptides of 90,000 and 15,000

daltons,

respec-tively, become associated with theRNA

polym-erase during the infectious cycle (C.

Szabo,

B.

Dharmgrongartama, and R. W.

Moyer,

Bio-chemistry, inpress). Wehavepresented datato

suggest that at least one of the three

polypep-tides, the 90,000-dalton species, tentatively

re-ferred to as aT5-specific "a'factor, is involved

in the in vivo regulation of class III (late)

transcription and is likely the product of gene

C2.Ifthehost E. colicell contains the

colicino-genic factor ColIb (ColIb+ cells), T5-specific

classIRNAis formednormally.

However,

after

the initiation of class IIphage RNA synthesis,

transcription is arrested and the infection aborts. As expected, infectedcolicinogenic cells show

only

the pre-early T5

peptide

of 11,000 daltons associated with the polymerase (C. Szaboetal.,inpress).

We have purified RNA polymerase from

T6-infected non-colicinogenic cells 20 min after

infection, using the standard procedures as

described by Burgess and Travers (2). Before

purification, the infected cultures had been

continuously labeled with [3H

Ileucine

under

conditions that label all three classes of T5

proteins, including those we find bound to the

polymerase. After each step ofpurification, a

portion of the partially purified enzyme was

'Present address: Lawrence Berkeley Laboratory, Uni-versityofCalifornia,Berkeley, Calif. 94720.

immunoprecipitated with antisera prepared

against theuninfected E. coliRNApolymerase holoenzyme, and the presence of the three

phage-specific proteins in the

immunoprecipi-tate was quantitated by analysis on sodium

dodecylsufate (SDS)-polyacrylamide gels. The

first step of the polymerase purification is the

precipitation of the enzyme by ammonium

sulfate such as we routinely used in our earlier

immunoprecipitation experiments (Szabo et

al., in press) and which, we have observed,

has no effect on the number or amount of the

phage polypeptides bound to the polymerase

(datanotshown). The RNApolymerase precipi-tated in the ammonium sulfate step is then

purified by DEAE-cellulose chromatography,

which removes all but the 15,000-daltonclass II

polypeptide from the enzyme (Fig. 1A). This remaining T5 polypeptide was rather strongly associated with the polymerase and remained

bound through a subsequent purification step

utilizing glycerol gradient centrifugation (Fig.

iB). After theglycerol gradient step of

purifica-tion, ourpreparationsofRNApolymerasewere

greater than 95'% homogeneous and contained

anequivalent amount ofthe 15,000-dalton

sub-unit per mol of enzyme as was originally

present in the immunoprecipitates prepared

from crudeextracts. We will refer to this

puri-fied form of RNA polymerase that contains a

single T5 polypeptide of 15,000daltonsisolated from infected non-colicinogenic cells as

RNAP(T5)(15), in agreement with our

sug-gestednomenclature (Szaboetal., in press).

Asample of theRNAP(T5)(15)from cells was

then analyzed on an SDS-polyacrylamide gel

and the total subunit pattern was compared

with that of a preparation ofRNA polymerase

purified from uninfectedcells. Coomassie

blue-1042

on November 10, 2019 by guest

http://jvi.asm.org/

VOL.15, 1975N

10 20 30 40 50 60 70

Slice number

FIG. 1. SDS-polyacrylamide gel electrophoresisof

'H-labeled T5-specific proteins bound to purified

RNA polymerase. A 100-mlsample ofunstarved E.

coliK-12W3110thy cells in T-MGM medium(6)was

infectedwith T5st(0), aheat-stabledeletionmutant

of T5+, at a multiplicity ofinfection of10.Leucine

wasthenaddedto afinal concentrationof0.1

,g/ml,

and a total of500

MlCi

of [4,5-3HJleucine (50 mCi!

Mumol) (New England Nuclear Corp.) was added in

aliquots to the infected culture at 1-min intervals

from 5 to 14minafterinfection.At15min, the [3H]-leucinewas chased with 0.2 mgofunlabeledleucine

perml,andafter20minofinfectionchloramphenicol

(0.1 mg/ml final) was added and the culture was

harvested by centrifugation. The 3H-labeled cells

were added to8gof identicallygrownand infected

unlabeled carriercells,and theRNApolymerasewas

purified bytheprocedure of Burgessand Travers(2)

and assayed aspreviouslydescribed (1). Aftereach

stepofpurification, aportion ofthe RNApolymerase

was immunoprecipitatedas described (Szabo etal.,

in press). The 'H-labeled, precipitated samples

wereanalyzedbySDS-polyacrylamide gel

electropho-stained polypeptide bands with mobilities

cor-respondingtothe ,B, ,B', a, and a subunits of the

hostenzyme were detected in bothpreparations

(Fig. 2). Since we have shown that all of the

radioactivity associated with the phage

poly-peptide of90,000 daltons was lost during this

method of purification (Fig. 1), the stained

polypeptide of 90,000 to95,000 daltons present in the RNA polymerase purified from infected cellswasprobably thehost a factor,whichhas a molecular weight of 95,000. From these results it

appearsthat

T5

infection does notpermanently

destroy the host a protein, and furthermore neither the association of the 15,000-dalton class II phage polypeptide nor any other

T5

modification permanently prevents the host a

factor from associating rather stronglywith the

enzyme.It is notknownatthistime whether the

binding of the

T5-specific

"a" factor and the host a factor are mutually exclusive.

We compared the in vitro transcriptional propertiesof

RNAP(T5)(15)

with those of three

other purified enzyme preparations: (i) RNA

polymerase from uninfected non-colicinogenic cells; (ii) RNA polymerase from uninfected

ColIb+

cells; and (iii)RNApolymerase obtained

from T6-infected ColIb+ cells that had ceased

all invivo transcription. The properties ofthe RNA polymerase purified from T5-infected ColIb+ cells seemed particularly pertinent for

this study. Althoughno

ColIb

factorprotein has been found associated with RNA polymerase

(Szabo

et al., in press), it is still quite

con-ceivable that the plasmid might antogonize

T5

transcription as aresultof acatalytic

modifica-tionofthepolymerase, which might be detected by an alteration inthe in vitro transcriptional

properties of the enzyme.The RNApolymerase

derived from each of the four preparations chromatographedonDEAE-celluloseas asingle peak that eluted at 0.19 M KCI and after low-salt glycerol gradient centrifugationwas at

least

95%

homogeneous. Fully active enzymes were recovered from all preparations including the one from infected

ColIb+

cells. All the purified preparations, when analyzed on

SDS-polyacrylamide

gels, appearedtohavethe host ar subunit (Fig. 2;

unpublished data).

A

sum-mary of the enzymatic activities of the four

resis as described by Weber and Osborn (10). The dataarepresented as 'H countsper minute pergel slice where theAbh and Ab, symbols indicate the positions in the gel of the stained immunoglobulin heavy (55,000 daltons) and light chains (25,000 dal-tons),respectively.(A)DEAE-cellulose-purifiedRNA polymerase; (B) low-salt glycerol gradient-purified RNApolymerase.

1043

on November 10, 2019 by guest

http://jvi.asm.org/

[image:2.507.49.239.53.446.2]

NOTES

SDS-polyacrylamide gel analysis ofpurified

RNA polymerase from late T5-infected

non-colicinogenic cells suggested that the enzyme

still contained the host a factor (Fig. 2A).

Although the a subunit of RNA polymerase is

required for the efficient transcription of T4

DNA, transcription

proceeded quite

effectively

with calf thymusDNA in thecompleteabsence

of the a subunit. Thus, the ratio of template

activities with T4 and calf thymus DNA has

been defined as a measure of functional a

subunit activity (8). This ratio with enzyme

containing the a subunit derived from

unin-fected cellswas approximately 2, and theratio

did not appear to change either as a result of

infection with T5orfromanyeffect of the ColIb

plasmid (Table I).These datamaybecompared

with those obtained for the RNA polymerase

lacking theasubunitthatwasisolated after T4

infection and had an activityratio with T4 and

calf thymus DNA of 0.2. Our results suggest

[image:3.507.70.250.78.473.2]

that each of these preparations contains a

FIG. 2. SDS-polyacrylamide gel electrophoresis of

RNApolymerase purified throughalow-saltglycerol

gradient. The preparations were electrophoresed on

7.5%SDS-polyacrylamide gels for8 hat8mApergel.

After electrophoresis the protein bands were

visual-ized with Coomassie blue. (A) RNA polymerase

purified from uninfected non-colicinogenic E. coli

K-12 W31IOthy-; (B) RNA polymerase purified from

T5-infectednon-colicinogenic E. coli K-12 W31IOthy

cells after20min ofinfection.

purified enzyme preparations, utilizing various

templates, is presented in Table 1. All four

preparations yielded approximately the same

total numberofenzymeunits (data notshown) and had similar specificactivities with agiven

DNAtemplate. T5 DNA was foundtogive the

maximum in vitroactivity.

TABLE 1. Template activitiesof RNApolymerase purified from uninfected and T5-infected cellsa

Spact(U/mgwith

variousDNA Ratio templates

Enzymesource T4/

T5b T4 Calf T5/ calf

T5b T4c thy-dT4 ty

musd ty

mus

Uninfected E. coliK-12

W3110thy. ... 1,215 987 475 1.2 2.1 T5-infected E. coliK-12

W31l0thy. ... 965 749 361 1.3 2.1 Uninfected E. coliK-12

W3110thy- (ColIb) 995 723 351 1.3 2.1 T5-infectedE.coli K-12

W3110thy- (ColIb) 1,184 1,045 514 1.1 2.0

aRNApolymerasefromeachsource waspurified through thelow-salt glycerol gradient step as describedby Burgess and Travers(2)togreaterthan95%homogeneity.Theassay mix contained (in 0.25 ml) Tris-hydrochloride (0.04 M), MgCl2(0.01 M), EDTA(0.1 mM), dithiothreitol (0.1 mM), KCl(0.15M), potassiumphosphate(0.4mM), bovineserum albumin (0.5 mg/ml), undegraded DNA, UTP, GTP, CTP, and ATP (0.1 mM each) (P-L Biochemicals), and RNA polymerase(2 to 54gper assay) at a final pHof7.9.Each assay also contained 2 MCi of [8-1C 1ATP (New England Nuclear Corp.). After incubation for 10 min at 37 C, the reaction was terminated bythe addition of 2.5 ml of cold trichloroaceticacid(5%, wt/vol)containingsodium pyrophos-phate(0.1mM).The samples were allowedtostand on ice for 15minand the precipitate was collected onglassfiber filters (Schleicherand Schuellno. 29), washed with a solution of trichloroacetic acid (2%, wt/vol) that contained sodium

pyrophosphate (1.01M), dried, and counted. h 10MgofundegradedT5 DNAwasusedperassay.

10MgofundegradedT4 DNA was usedperassay. d 25 ug ofcalfthymusDNAwas used per assay.

1044 J. VIROL.

.A..

on November 10, 2019 by guest

http://jvi.asm.org/

[image:3.507.271.463.330.477.2]

VOL. 15, 1975

functionalhostfactor.Additionalevidence

sup-porting this hypothesis is the effect ofKCI on

the in vitro enzyme activity. Host core

polym-erase activity (lacking the a subunit) is only

negligibly affected by KCI, whereas the

holo-enzyme activity is stimulated by KCl up to a

concentration of 0.2 M (9). All of our

prepara-tions of RNA polymerase required 0.2 M KCI

formaximum activity.

Pispa and Buchanan (7) have studied the

specificity of the in vitro transcription of T5

DNA by host RNA polymerase purified from

uninfected cells. They found that the host a

enzyme transcribed primarily class I and class

IIsequencesand onlyasmall amount of classIII

sequences. We have shown that the net

en-zymatic activity of purified RNA polymerase

doesnotappear tobe altered either by the ColIb

plasmid or by T5 infection, However, RNA polymerase from late T5-infected cells, under

theproperconditions, mayrecognizeand

tran-scribe predominantly

late T5 gene sequences,

which are not efficiently read by the enzyme fromuninfectedcells. Likewise, RNA

polymer-ase purified from T5-infected ColIb+ cells

pos-sessesnormal net activity, althoughthe

possi-bilityexiststhatthis enzyme may have lost the

abilitytotranscribeclassII T5 genesequences.

To test thesepossibilities, thespecificityofthe

in vitro transcription of native T5 DNA by all

four of the RNA polymerase preparations we

havedescribedwasexamined by hybridization-competition analysis.

3H-labeled T5 RNAthat was synthesized in

vitro washybridizedto

T5

DNA in thepresence

ofunlabeled T5 RNA

synthesized

in vivo. The competitionofthelabeledin vitroRNA with an

excess of the in vivo RNA isolated at various

timesafterinfectionmeasuresthe proportionof

each class of RNA synthesized in vitro. The

specificity of the in vitro transcription is thus determined,sinceeach temporalclass of in vivo

T5 RNA istranscribedfrom aunique set of T5 gene sequences. Figure 3 illustrates the results ofthehybridization-competition analyses, and the data show that all four of the enzyme

preparations described above

transcribed

T5 DNA with the same specificity. Unlabeled

(A)Urkabeled3minCompetitorRNA (B) Unlabeled 8minCompetitor RNA (C)Unlabeled20min CompetitorRNA

100 200 300 400 500 100 200

300

400 500 100 200 300 400 500

[image:4.507.54.446.340.495.2]

/.g

Competitor RNA

FIG. 3. Hybridization-competition analysis ofT5 RNA synthesized in vitro by differentpreparations of purifiedRNApolymerase. Labeled in vitro RNA and unlabeledinvivoRNA werepurifiedby the hotphenol

method described by Lembach and Buchanan (4). Unlabeled competitor preparations were isolated from

infectedcellsat3, 8,and 20min afterinfection. [3HJRNA wassynthesizedinvitro byaslight modification of

the standard RNApolymeraseassaysystem of Burgess (1). Each mixture contained Tris-hydrochloride (0.04

M),MgCl2 (0.01 M),EDTA(0.1mM), dithiothreitol(0.1mMCalbiochem),KCI(0.2 M),potassium phosphate (0.4 mM),bovineserumalbumin(0.5 Mg/ml),nativeundegradedT5 DNA (20 ug), UTP, GTP, CTP,and ATP (0.1 mMeach) (P-L Biochemicals), and RNA polymerase(10 Mg) inafinalvolume of0.5ml. TheRNA was

labeled by the addition of [5,6-3H]UTP (New England Nuclear Corp.) at a final specific activity of 125 mCi/mmol. The reaction was carried out at 37 C for 15 min before isolation of the RNA. [3H]RNA was synthesized in vitro by RNA polymerase isolated from either uninfected non-colicinogenic E. coli K-12 W3110thy- (0), or colicinogenic E. coli K-12 W3110thy- (ColIb+) (A) and T5-infected non-colicinogenic E. coli K-12 W3110thy- (0) orcolicinogenic E. coli K-12 W3110thy- (ColIb+) cells (A). The labeled RNA

washybridizedtoT5DNA that had been adsorbedto membranefilters(MilliporeCorp.) (0.3 MgofDNAper filter) asdescribedbyCooper(3).Eachhybridizationmixture contained['HJRNA (10,000counts/min),0.3 Mg ofT5DNA,andthe indicatedamountofunlabeled T5 in vivo competitorRNA isolated at: (A) 3min; (B) 8

min;and(C) 20minafterinfection. 100

80 .0 I

c

C. c

.a

0.

0

C

L-c

0-0) 0~

60F

40F

20

I - . .

.--I --- -- -I ---I - - -I .-- -- -2 lr%lf%IIfA^^ Am^^I I aON^ 129%^ Ar%^ wr%^Il

1045 NOTES

I

I

on November 10, 2019 by guest

http://jvi.asm.org/

NOTES

phage RNA isolated 3 min after infection

con-sisted ofonly class I species andwas partially

competitive with the [3H]RNA synthesized in

vitro, indicatingthat at least 25% of the in vitro

RNA in each case wascomposedofclass I gene

sequences (Fig. 3A). Phage RNA isolated 8 min

after infection consists ofboth class I and class

II genesequences (5). Whenthis RNAwasused

as the competitor, only 20% of the in vitro

[3H]RNA remained in

hybrid

form

(Fig.

3). Therefore, the greatest proportion of the RNA

synthesized in vitrowith all four enzyme

prepa-rations consisted of early sequences. All three

classesofphage RNA werepresent inthe RNA

isolated 20 min after infection.Although classI

RNA wasnotsynthesizedfrom8 to20 minafter

infection, hybridizablesequences ofclass I RNA

were present inthe cells, eventhoughthis class

ofRNA was no longer metabolically active. All

the [3H ]RNA was displacedfromthe hybrid by

20-min competitor RNA, indicating that all

RNA species synthesized in vitroare also

pres-ent in infected cells. All four enzyme

prepara-tions transcribed class I and class II sequences

more efficiently than class III sequences, since

onlyabout 105% of the in vitro RNA consisted of class III sequences.

The similar catalytic properties of these four

preparations fail to suggest any independent

role for the T5 polypeptide of 15,000 daltons

that is found associated only with the RNA

polymerase purified from infected

non-colicino-genic cells.Furthermore, the isolation of a fully

active preparation from infected ColIb+ cells

with properties similar to those for control preparations suggests that either any

plasmid-directed inactivation of RNA polymerase

activ-ity toward T5 DNA is readily reversible or that

the inactivation of transcription may be the

indirect consequence of some other

plasmid-directed modification ofcellularproperties.

(This work is taken in part from a thesis

submitted

by C.S. to the Faculty of Pure

Science;

ColumbiaUniverstiy, inpartial

fulfill-mentofthe

requirements

for thePh.D.

degree.)

ThisinvestigationwassupportedbyPublicHealth Service grant AI-10302from the National Institute ofAllergy and Infections Diseases and grant VC- 142 from the American CancerSociety.C.S.wassupported byPublicHealth Service training grant GM00255 from the National Institute of General Medical SciencestotheDepartmentof Biochemis-try. R.W.M. is supported by Public Health Servicecareer

developmentawardGM34354 from the National Institute of General Medical Sciences.

We acknowledge, with thanks, the gift of antisera

pre-pared against RNA polymerase noloenzyme from E. coli BprovidedbyP. R.Srinivasan and B.Dharmgrongartama.

LITERATURE CITED

1. Burgess, R. R. 1969. Anewmethod for the largescale

purificationofEscherichia colideoxyribonucleic acid-dependentribonucleic acidpolymerase.J. Biol. Chem. 244:6160-6167.

2. Burgess,R.R.,andA. A.Travers.1971.DNAdependent RNA polymerases (EC2.7.7.6), p. 851-853. In G. L. Cantoni and D. R. Davies(ed.), proceduresin nucleic acid research, vol. 2. Harper and Row Publishers,

NewYork.

3. Cooper, T. G. 1972. Procedures for uniform binding of DNA to nitrocellulose filters. Anal. Biochem. 47:481-486.

4. Lembach, K.J., andJ.M.Buchanan. 1969. The relation-shipofproteinsynthesistoearly transcriptioneventsin

bacteriophageT4-infected Escherichia coli B. J. Biol. Chem.245:1575-1587.

5. Moyer, R. W., and J. M. Buchanan. 1969.Patternsof RNAsynthesisinT5-infected cells. Proc. Natl. Acad. Sci.U.S.A.64:1249-1256.

6. Moyer, R. W., and J. M. Buchanan. 1970. Effect of calcium ions on synthesis of T5-specific proteins. J. Biol.Chem.245:5897-5903.

7. Pispa, J. P., and J. Buchanan.1971.Synthesisof

bacte-riophage T5-specific RNA in vitro.Biochim. Biophys.

Acta247:181-184.

8. Schachner, M., W. Seifert, and W. Zillig. 1971. A correlation of changes in host and T4 bacteriophage

specificRNAsynthesiswithchangesofDNA depend-entRNApolymerasesinEscherichia coliinfectedwith bacteriophage T4. Eur.J. Biochem. 22:520-528. 9. Stevens, A. 1974. Deoxynucleic acid dependent

ribonu-cleic acid polymerase from two T4 phage-infected

systems.Biochemistry 13:493-502.

10. Weber, K., and M. Osborn. 1969. The reliability of molecular weight determinations by dodecyl sulfate-polyacrylamide gel electrophoresis. J. Biol. Chem. 244:4406-4412.

1046 J.VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/