Sizes of bacteriophage T4 early mRNA's separated by preparative polyacrylamide gel electrophoresis and identified by in vitro translation and by hybridization to recombinant T4 plasmids.

0022-538X/81/120772-18$02.00/0

Sizes of Bacteriophage T4 Early mRNA's Separated by

Preparative

Polyacrylamide Gel Electrophoresis and Identified

by In Vitro Translation and by

Hybridization

to

Recombinant

T4

Plasmids

ELTON T. YOUNG* AND ROSEMARY CRONE MENARD

Department of Biochemistry, University ofWashington, Seattle, Washington98195

Received 13February 1981/Accepted1July 1981

We determined thesizes ofspecific T4prereplicative mRNA's by preparative

polyacrylamide gel electrophoresis, andweused thefollowingtwotechniques to

identify specific gene transcripts: cell-free protein synthesis accompanied by

sodium dodecyl sulfate-polyacrylamide gelelectrophoresistodistinguish T4 poly-peptidesandhybridizationtorecombinant plasmids containing T4 DNA of known genetic composition. In our first analysis, the use of nonsense and in-phase

deletion mutants allowed unambiguous identifications of the functional

tran-scripts that encoded genes 32, rIIB, and rIIA. In addition, we identified the

functional transcripts that encoded genes 43, 45, 30, 39, and 52, the ,B-glucosyl

transferase gene, and the deletion293 region. A single peak of mRNA activity

that coded for gp43, gp39, gprIIA, /3-glucosyl transferase, and the polypeptide

encoded in the deletion 293 region was present; the other polypeptides were

encoded inmultiplemRNAspecies. gp46andgp32wereencodedbytwomRNA's,

andgp52andgprIIBwereencodedby three mRNA's. By hybridizing fractionated,

pulse-labeled earlyRNA to cloned restrictionfragmentsof T4DNA,weidentified

the same specific transcripts for genes 43, 52, and nIB. In addition, a

lower-molecular-weight RNA (presumably degraded mRNA)waspresentevenin

pulse-labeled RNApreparations. The distribution ofpulse-labeled RNAs that hybrid-izedto gene39,gene30, generIIA,gene40plusgene41,andgene42plus the

,B-glucosyl transferase gene indicated extensive degradation. We detected

cotran-scriptionofgenesrIlA and rIIB by rehybridization of RNA first annealed toan

rIIB plasmid and then eluted and annealed to anrIlA plasmid. The size

distri-butions of normalandchloramphenicol-treated RNAsthathybridizedtoplasmids

containing T4immediate earlygene30, gene39, gene40plusgene41, andgene

42plusthe f3-glucosyl transferase gene werenotsignificantly different. An understanding of the arrangement of T4

prereplicative transcriptionunits wouldincrease

our knowledge of the regulation of T4 early genes significantly. The current model of this arrangement (28) proposes that

chlorampheni-col (CAM)-resistant immediate early genes (transcriptsaresynthesizedafter infectionby T4 in the presenceofCAM)arepromoterproximal

to CAM-sensitive delayed early genes (tran-scriptsare notsynthesizedafter infection in the presenceofCAM)andthatquasi-late or middle-mode genes are regulated differently than im-mediateearly and delayed early genes. The

fac-torsthatinfluence the transition from

immedi-ate early transcription to delayed early tran-scription andsubsequentlytomiddle-mode

tran-scription are poorly understood. The Esche-richia coli termination factor rho is thought to be involved in the shift from immediate early

transcription to delayed early transcription (3,

5), and the T4 motgene productisinvolved in the expression of at least some middle-mode genes (18,20,27).

Various aspects of the model summarized

above have beentested. Themostdirect

predic-tionsarethat the sizes ofsomeimmediateearly transcriptswouldincreaseifdelayedearly

tran-scription were allowed and that

delayed early

transcripts would be the distalportionsof

poly-cistronic transcripts whose

promoter-proximal

portions would be immediate early transcripts.

This proposition was tested for the D2 region

(33) andfor the internalproteinsIPIIandIPIII,

buttheresultswereequivocal;the D2 RNA size

wasinvariant, whereasthe IP RNA seemedto

besomewhatlargerinthe absence of CAMthan

thepresenceof CAM.

Ingeneral, thephysicalsizes ofT4transcripts

77'2

on November 10, 2019 by guest

http://jvi.asm.org/

T4 EARLY mRNA SIZES 773

have beendifficulttostudybecauseof the labil-ity of the RNAs. The T4 tRNA'sare an

excep-tion to this. The presumed primary transcript synthesized in vitro has beenidentified,and the processing of this transcriptto mature tRNAs' has been studied (10). A similar study of T4 mRNA's that code for early proteins has not

been performed, although such a study would clearlybe relevanttoanymodel of regulation of

prereplicative transcription. In a few instances

the size of the functionaltranscript that codes

for an early protein has been identified by its ability to code foran enzymatic activity or an immunologically active polypeptide (1, 29). For the nonessential rII genes and the adjacent D

region, the RNA has been identified by deletion hybridization analyses (31, 32, 39). Inthis case

there is a discrepancy concerning the physical

sizes of the rIIA and B transcripts (33, 39). Wedetermined the sizes ofspecific T4 early transcripts bypolyacrylamidegel electrophore-sis. The RNAswereidentified after

electropho-resis by the following two techniques: in vitro translation and separation of the labeled poly-peptides by sodium dodecyl

sulfate-polyacryl-amide gel electrophoresis and hybridization of radioactive RNAtoclonedrestrictionfragments of T4 DNA. Analyses of functional early mRNA's revealed that a number oftranscripts

occurred as multiple species, which indicated

that thereweremultiplepromoters or

termina-tionsitesorthatpost-transcriptional processing

occurred. Some of these same multiple

tran-scriptscouldbeidentifiedbyhybridization, but the radioactive RNA species corresponding to

many genes were present as heterogeneous ar-rays ofsmallfragments,whichrepresented

de-gradedfragments of the active mRNA's.

MATERIALS AND METHODS

Biochemicals and media. Most reagents were purchased from standardsources.[3S]methioninewas preparedasdescribedbyCrawford andGesteland (6). Mouse RNAwas a gift from Ursula Storb, and 3H-labeled RNA from the midi-variant ofbacteriophage

Q8 (MDVRNA; 22)was agiftfrom FrankRyan.The

acrylamide used for gel electrophoresis was passed

through a membrane filter (Millipore Corp.) before

use.M9SandM9S.1media have beendescribed

pre-viously byBolleetal.(2).

Bacteria, phage, and plasmids. E. coli BE, a

nonpermissive host for ambermutants,wasusedas a

hostforallinfections. Plasmidsweremaintainedin E.

coli strain 802(suII+rk-mk+).

Thebacteriophagestrains withmultiplemutations

wereconstructedby standard phagecrosses(Table1).

TherIIA deletionmutationEM66andthe rIIB

dele-tionmutation 196have beencharacterized byA.Bolle

(personal communication). These mutations contain

wholly internal in-phasedeletions which correspond

to losses ofapproximately 300 and 100 amino acids

fromrIIA andrIIB, respectively.

The plasmids containing T4 restriction fragments

werefrom thecollections of Mattson et al. (19) and

Selzeretal.(34).Theproperties of these plasmidsare

shown in Table2.The T4restrictionfragments in the

original collection of Mattson et al. were cloned into

pCR1andweredesignated bytheprefix pVH. Mattson

and Van Houwe transferred these restriction

frag-ments to pBR322 and gave them new designations (the600 series), whichweused. Thedesignations of

the T4 recombinantplasmids that were derived from

the 600 series by cleavage with different restriction

enzymes than the enzymes used in the originalcloning

procedure contain asuffix letter (A, B, etc). Restriction

maps of the rII regions showing the origins of the

DNAfragments used in this work are shown in Fig. 1

(26a, 34).

TABLE 1. BacteriophageT4mutants

Mutant Description

32amA453 44amN82 Amino-terminal ambermutation ingene32 (no identifiable poly-peptide);nogene 44 polypep-tidedetected

32amA453 44amN82 rIIdell241coversall knownpoint rIIdell241 mutations inrIIAandrIIBand deletespartof theD region;no polypeptides from genes 32, 44, rIIA, and rIIB

32amH18 44amN82 18,000-dalton polypeptide from gene32;nogene 44polypeptide rIIBdell96 Internal, in-phase deletion ingene

rnIBmakinga22,000-dalton rIIB polypeptide

rIAdelEM66 Internal, in-phasedeletion in gene rIlA makinga55,000-dalton polypeptide

32amH18 44amN82 18,000-dalton gp32; 22,000-dalton rIIBdell96 gprIIB;nogp44 identified 32amH18 44amN82 18,000-dalton gp32;22,000-dalton

rIIBdell96 gprIIB;55,000-daltongprIIA;no rIIAdelEM66 gp44

TABLE 2. T4 recombinant plasmids

Plas- T4 Restriction

mid gene(s)a site(s)used Reference(s) mid

~~~for

cloning

621 52 RI 19

622 43 RI 19

624 42,,fgt RI 19

625 30 RI 19

626A 39 HindIII 19;Young, unpub-lisheddata

627 40,41 RI 19

pTB35 rIIA HindIII 34

pABI rIIA HindIII, RI* 34;Selzer, unpublished data

pTB1O rIIAB HindIII 34

pABIV rIIB RI*,HindIII 34;Selzer, unpublished data

aNoothergenes havebeen identified by marker rescue tests.Plasmid621(gene 52)containsadditional DNA which hasnot beenidentified.Theplasmidscontaining genes 43, 30, 39, rIlA (pTB35 andpABI),and rIIB (pABIV) have been shownbymarker rescuetests tobewholly internal to the gene indicated.The T4rIIA and rIIBrestrictionenzymefragments usedin this work areshowninthe mapsinFig.1.Thelength of theT4fragmentinplasmid pTB1Ois 873 basepairs (26a).

VOL. 40,1981

on November 10, 2019 by guest

http://jvi.asm.org/

rIEA H

3B

H RI i RI RI- H RI W

GENE52 w RI

pTB I

[image:3.500.82.416.58.157.2]

pTB10

FIG. 1. Mapsof T4 rIIA and rIIB restriction enzymefragments. The maps are drawn approximately to

scale. H, HindIII.

Nucleic acidpreparation.T4 DNA was prepared

frompurifiedT4 phage, and plasmid DNA was pre-pared by minor modifications of the cleared lysate procedure of Katz et al. (16), as described elsewhere

(42). Theethidium bromide and residual protein that

remained after centrifugation in CsCl were removed

by twophenolextractions.

Both radioactive RNA and nonradioactive RNA

were isolated from T4-infected E. coli BE cells, as

describedelsewhere (42).

Preparative electrophoresis of RNA on

poly-acrylamide gels. We used the preparative electro-phoresis apparatus of Hagen and Young (14), with modifications (13, 26). Denaturation of RNA samples with dimethyl sulfoxide before electrophoresis has

beendescribedpreviously(13, 26).Briefly,each RNA

sample dissolved in Tns-hydrochloride (pH

7.5)-EDTA wasminxedwith spectroscopy grade dimethyl

sulfoxidesothat the final concentrations in the sample

were90%dimethylsulfoxide, 10 mM Tris-hydrochlo-ride, and 3 mM EDTA. Then the sample was

incu-bated for5min at370C,cooled to room temperature,

andloadedonto agel. The percent recovery of RNA

activityfromthegelswasusually30to100%.

Cell-freeproteinsynthesis, electrophoresisof

polypeptides, anddensitometricanalysisof

au-toradiograms.AnS30extractwaspreparedfrom E.

coli MRE600 and used asdescribedpreviously (26).

Radioactive polypeptideswereanalyzedasdescribed

by Pachl andYoung (26).

Hybridizationtofilter-bound DNA. Filters

con-tainingimmobilizedT4phageorplasmidDNAwere

preparedand usedasdescribed elsewhere (42).Ifthe

annealed RNAwas tobe eluted foruseinasecond

experiment,the firstannealingwasperformedin 0.03

M sodiumcitrate-0.3 M sodium chloride(pH 7.0)-50%

formamide. The RNA was eluted by heating the

washedfilter in a solutioncontaining10mMTris(pH

8.8) and0.2mM EDTA for1minat1000Cand then

rapidly coolingthe solution. The filterwasremoved,

50

Ig

of tRNAwasadded,the solutionwasincubated

with5ug ofRNase-free DNase for15minat300Cand

phenol extracted, and the RNA was recovered by

ethanolprecipitation (42).Elution of the RNA after

the filtersweretreated with RNasewasperformedas

describedelsewhere(42).

RESULTS

Sizes

offunctional

early

transcripts.

RNA

was extracted 12 minafter infection ofE. coli

BE(su-)

with T4 32amA453 44amN82, as

de-scribed above.(Ambermutants weredesignated bygenenumber, followed by alleledesignation; for example, the ambermutation A453isingene 32,and themutantcontaining thismutationwas

designated 32amA453.)About 400 ,ugofpurified RNA was denatured by treatment with 90% dimethyl sulfoxide before continuous-elution preparativeelectrophoresison apolyacrylamide gel (13) (Fig. 2). The radioactive

polypeptides

synthesized in vitro by the fractionated RNA

were analyzedby sodium dodecyl

sulfate-poly-acrylamide

gel electrophoresis. The

identifica-tion of rIIB mRNA activity wasfacilitated by

using phage containing nonsense mutations in genes 32and44andinfectingansu-host. This allowedgprIIBtobeidentified andquantitated in the absence of comigrating gp32 and gp44

(23).(gprIIB,gp32,andgp44 arethe polypeptide products ofgenesrIIB,32,and44,respectively.)

Using

the double amber mutant had an

addi-tional advantage. Because bothgp32 and gp44 wererequired forDNAsynthesis and hence late

T4 RNA synthesis, this mutant did not make

any late RNA which could interfere with the identification ofearly mRNA species. Figure 2

showsanautoradiogram of the dried gel. Without

considering

the

identity

of the indi-vidual proteins synthesized by the fractionated RNA,weobservedtwodifferenttypesof behav-iorfor the mRNAspecies identified by in vitro translation. Some mRNA's

migrated

at one or

more

unique

rates,asshown

by

thepresenceof

mRNA activities for their cognate proteins in

one or more unique sets ofadjacent fractions collectedfromthegel. Other mRNA'sappeared

to bepolydisperse; their activities were spread rather continuously across the

gel.

For some

RNA

species

this

"tailing" appeared

to result

fromamajor

peak

whosemRNAactivity trailed throughout the

remaining

fractions.However,in other casesthemRNA activitywasdistributed fairly

uniformly

over abroadrange ofmolecular weights. The polydispersity could have arisen

for one ormore reasons, including

incomplete

washing from the elution chamber,

aggregation

of theRNA,

physical

size

heterogeneity

of the mRNA, and comigration of different

polypep-- pTS35 b Asi

.0.

pABI

on November 10, 2019 by guest

http://jvi.asm.org/

T4 EARLY mRNA SIZES 775

V . a 9 - - _1-_. -.

A~~~~~~~~~~f

FIG. 2. Autoradiogramof T4 early polypeptides synthesized byfractionated32-44-RNA. A400-pgsample

of RNA isolated12minafterinfection of E. coli BE with T4 32amA453 44amN82 was denatured with 90%

dimethylsulfoxide for3minat37°C and thenfractionatedon apreparative gel (diameter, 5 cm)containing

2.25% acrylamide.TheRNAwaselectrophoresed at 15 mA for11 h, and the eluted RNA was concentrated by

ethanolprecipitation and then translated inacell-freeprotein-synthesizing system containing

[3S]methio-nine. The radioactivepolypeptides wereseparated by electrophoresis on a discontinuous sodium dodecyl

sulfate gel system containing equal volumes of 12.5, 10, and 7.5%acrylamidelayered sequentially over one

another. Thestacking gelcontained 3%acrylamide. After electrophoresis for3hat25mA, thegelwastreated

withdimethylsulfoxideandsubjectedtofluorographyasdescribed in thetext. Thepolypeptides synthesized

in vitrofromunfractionatedRNAs derivedfromcellsinfectedwithphagesofthe genotypesindicated above

theslotsareshownatthesidestoallowidentificationof specific T4polypeptides. The molecular weight scale

atthe topwasderivedfrom the 16S and 23S rRNA's present on this gel and from these and additional

molecularweightstandardsseparatedonanidentical gel (seeFig. 8). The two gelsusedin this analysis did

notseparatethe T4polypeptides identically. gp43 was the upper band in both gels. wt, Wild type.

tidesencodedby mRNA's of different but

over-lapping sizes. For those mRNA's which hadan

intensepeak ofactivity followed by trailing,we

assumed that the apparent

heterogeneity

was

primarilyanartifact caused

by incomplete

elu-tion from the elution chamber and efficient translation of the

late-eluting

fractions in the cell-freesystem, as hasbeen observed with T7 mRNA(25, 26).

Several proteins could be identified

readily

aftertranslation of the fractionated

RNA,

based

ontheirrelative

electrophoretic

mobilities,

their

knownmolecular

weights,

and

analyses

ofamber

mutants (24, 40). The

autoradiogram

in

Fig.

2

shows thegenes thatcoded for these

proteins;

they includedgene43,gene46,gene 30

plus

gene 39, gene 52,gene32, gene

rlIB,

andgene IPIII.

gp3O

and gp39

comigrated

onthe

gel

shown in

Fig. 2;in otheranalyses (see Fig. 6) these gene

products were separated from each other, and

their individual mRNA activities could be iden-tified. The RNA forrIIA couldnotbeidentified

unambiguously.

Therewere tworeasonsforthis.

gprIIA wassynthesized poorlyinvitro,and two different mRNA fractions coded for a protein

that wasapproximatelythesamesize asgprIIA.

These two mRNA species were identified in

fractions34 to 36 and39 to 60 of the

autoradi-ogramshowninFig.2.Theprotein encoded by themRNA that eluted in fractions34 to36had

the same

mobility

as the

polypeptide

that was

encoded

by

the deletion 293 (del 293) region (24).gprIIB appearedtobeidentified unambig-uouslyonthegel sincenogp32wasmade when unfractionated RNAwasused

(Fig.

2).However,

even

partial suppression

of the amino-terminal

gene 32 amber mutation (A453) could cause

synthesis

of

enough

gp32 toallow ittobe

con-fused with

gprIIB.

To overcome this

problem

and to

identify

the mRNA for

rILA,

we

per-formedtwoexperiments.Inthe firstexperiment RNA from the

triple

mutant T4 32amA453 44amN82rIIdell241was

analyzed

after fraction-ation on a

preparative

polyacrylamide gel.

In this mutant both rIIA and rlIB were deleted completely.The translationproductswere

ana-lyzedasdescribedabove (Fig. 3).NorIIB poly-peptide wasfound on the gel, whereas mostof

the otherpolypeptides with molecular weights

less than 60,000 whichwerepresentontheFig.

2 gel were identified, and their mRNA's

mi-grated at

approxiimately

the same rates as on

thegelshown inFig.2.Anexceptiontothis was the gene 52 message activity (see below). This

VOL. 40,1981

on November 10, 2019 by guest

http://jvi.asm.org/

[image:4.500.50.448.66.246.2]

FIG. 3. Identification of rIIB message activity by analysis of RNA from anrIIdeletionmutant. RNA was extracted 12 minafterinfection of E. coliBEby T4 32amA453 44amN82rIIdell241andanalyzed as described

in thelegendtoFig.2andinthe text. The polypeptidessynthesized in vitroby unfractionated RNA isolated

12minafter infection by wild-typeT4 (T4 wt) T4 32amA453 44amN82, and T4 32amA453 44amN82rIIdell241

areshown in the threeslots on theleft. The elution positions of the 16S and 23S rRNA markers are indicated

onthe top ofthe autoradiogram. The open circle on the figure indicates the expected position of therIIB polypeptide.

was additional evidence that the polypeptide

labeledgprIIB in Fig. 2 wasthe product of the rIIB gene. Unfortunately, nopolypeptides with molecular weights of more than 60,000 were made in thesecond experiment. Thus,the

pre-sumptive rIIA message in Fig. 2 could not be

identified because neither of the two potential

rIIApolypeptideswassynthesized.

Only the middle-sized (Mr, 1.2 x

106)

gp52 message activity was present in the RNA

iso-lated after infection with the deletion mutant

rII1241, whereas three peaks of gp52 message

activity were detected when RNA from wild-type T4 was analyzed. The right endpoint of del1241 was estimated to be 0.7 kilobase from the right or 3' end of rIIB (7, 8). Since it has beenestimated that rIlB and gene 52 are sepa-rated by about 3.5 kilobases (7, 40), the right endpoint ofdel1241 is about2.8kilobasesfrom gene52.None of the gene52messagesobserved waslarge enoughto startwithin thisregionand continuethrough gene52.

A morepositive identificationof

rnIB

mRNA andidentificationofrIlAmRNAwereachieved byusingin-phasedeletionmutants. de196 isan

internal, in-phase deletion within rlIB that pro-ducesashortenedpolypeptidewithamolecular weight of 22,000 (A. Bolle, personal communi-cation). delEM66 lies within rIIA, and coding also remains in phase in this case, so that an

rIIA polypeptide with a molecular weight of about 55,000 issynthesized (Bolle, personal

com-munication). These mutations wereintroduced

into the double amber mutant 32amH18

44amN82 to produce triple and quadruple

mu-tantswith thefollowinggenotypes:T432amH18

44amN82 delEM66, T4 32amH18 44amN82 del196, and T4 32amH18 44amN82 del196 delEM66. The gene 32 amber allele used for these experiments (H18) had the advantage that it produced a recognizable amber fragment with

a molecular weight of 18,000, so the gene 32 messagecould also beidentified. Figure4shows the identification of these polypeptides on a

sodiumdodecylsulfate-po;yacrylamide gel.The truncated gene 32, rIIA, and rnIB polypeptides

wereidentifiedeasilyandmigratedinregionsof the gel thatwere free of other radioactive

poly-peptides,asshownbythesamplIsisolated from thewild-typeandmutant-infected cells.

RNA from the triple mutant 32- 44- rIIB del196 was fractionated by preparative

poly-acrylamidegel electrophoresisessentiallyas de-scribed in the

legenid

toFig.2. Theeluted RNA

wastranslated inacell-freeprotein-synthesizing

system, and the radioactive polypeptides were

separated bysodium dodecyl

sulfate-polyacryl-amide gel electrophoresis. Figure 5 shows an

autoradiogram of the driedgel.TherIIBdell96

polypeptide and the gene 32 amber fragment

H18 could be identified easily. The rIIBdell96

messageactivitymigratedmorerapidlythanthe wild-type rIlB message activity, as expected. Thiswasevidentby the relativepositions of the first rIIB message peaksand the positionofthe 16SrRNA(Fig.2and5). The molecularweights

and the differencesbetweenwild type anddell96

arediscussed below.gprIIAappearedtobe

pres-entinfractions68 to 74,whichalso containeda

small peak of rlIB message activity. Gene 43

message activity migrated slightly faster than theputativepolycistronicrlIA-B message.

on November 10, 2019 by guest

http://jvi.asm.org/

[image:5.500.63.457.75.229.2]

T4 EARLY mRNA SIZES 777

s:3 ti' L W

t \- (. \ \

rzJ

tT

C\ -n C\ (\J (\J C %

\\\ 1, 1 1 1/

Gene

43

-r

lA-qo.

0,1X

_

*

rIIA

delEM66-

a

.*" a ..;.:i,.~ "

32?

_*

_.11--.2

r M/

~~~~~~ ...l 4- 4 S Wa

:e,h-' :§PA

-':.

FIG. 4. T4polypeptides synthesizedinvivoby

us-ingwild-type (wt)andmutantphage. CulturesofE. coli BE infected with T4phage were labeled with

[3S]methionineateither8to12 minor20to24min

(infectionwithwild-type phage only [slotlabeled T4

wt-16-20']). The radioactive polypeptides were

ex-tracted andseparatedon a 10%polyacrylamide by

electrophoresis for5hat25mA. Therelevant

geno-typesareshownatthetop;thegenescorresponding tothemarkedpolypeptidesareshownonthe left. ever, therewasaconsiderable amountof back-ground due to translation of endogenous mRNA's in the S30 extract, which interfered with theidentification of minor T4 polypeptides,

such as rIIA. Most of the other T4 mRNA's identified in Fig.2could also beidentified in Fig.

5,and the migrationratesof these mRNA'swere

approximately thesame onthe twogels. Some of thesemessages areidentifiedbygenenumber (Fig. 5).

Ashortened

rILA

polypeptide should be

syn-thesized more

readily

in the cell-free system,

therebymaking the rIIAmessageeasierto iden-tify. To do this, RNA extracted from cells

in-fected with the

quadruple

mutant32-44-del196

delEM66

was fractionated, and the mRNA

ac-tivitywas

analyzed

asdescribed above. The rIIB del196, rlIA

delEM66,

and 32amH18 polypep-tideswere identifiedeasily. The mRNA activi-ties for

rIlA

delEM66 and rIIB del196

polypep-tides bothcomigratedat a ratethatrepresented

an

RNA

with a molecular

weight

of

approxi-mately1.1 x 106. The rIIB del196message activ-ity was alsopresent in lower-molecular-weight fractions,as observed inFig. 2and5. Noother

message

activity

forrIIAdelEM66wasdetected.

An

analysis

of thedistribution ofgene32,rIIA,

andrIIB messageactivities is

presented

below.

Quantitation

of

functional

early

mes-sages.Tocomparetheamountsof

protein-syn-thesizing

activity

in the different mRNA's cod-ing for thesamepolypeptide andtoexamine the

apparentcomigration of rIIA and rIIBmessage

activitiesmore

carefully, autoradiograms

ofgels such as the one shown in Fig. 2 were scanned

with amicrodensitometer. The results of these

measurements areshowninFig.6forgenes 43,

de1293,

52,

nIB,

46, 30,and39 (from Fig. 2and

otherdata not

shown)

and in Fig. 7 for genes

rIIBdell96 and rIIAdelEM66(from the analysis of RNAs extracted from cells infected with the

mutant 32- 44- rIIBdell96 rIIAdelEM66). Of these nine genes,

only

genes43,

del293,

39, and

rIIA were represented bysingle messageswith uniquesizes. There weretwo

peaks

ofmessage

activity

for genes 32and46and three

peaks

of

message activity for genes rIIBdel196 and 52.

The

largest

RNA from therIIB

region

was not

detected in the RNA isolated from rIIB+-in-fected cells

(Fig.

6). Gene 30 message

activity

migrated heterogeneously, with no apparent

well-definedpeak of

activity.

gp3O

andgp39were notseparatedonthegel shown in

Fig.

2;thus, it

appeared that the gene 39 messageactivity was also heterogeneous. Numerous other message

activitieshavingoneof these types of behavior were observed (Fig. 2, 3, and 5). The genetic origins of most of these were not identified. A

conspicuousprotein havingamolecular

weight

of 20,000 to 25,000 was translated from four different messengers having molecular weights of0.25x 106,0.50x 106,0.8 x

106,

and1.0x

106.

We foundavery similar distribution of

messen-ger activity fora

protein

that had an identical

rle-

;

196

-Vol- 40, 1981

on November 10, 2019 by guest

http://jvi.asm.org/

[image:6.500.47.238.76.529.2]

OI.

~4 . -. * .~-. s- -. ...s - S

FIG. 5. AutoradiogramofT4earlypolypeptides synthesized by fractionated32t 44 rIIBdell96 RNA. RNA isolated 12minafter infection ofE.coli BE byT4 32amH1844amN82 rIIBdell96wasfractionatedona5.0-cm

2.0%,' acrylamide-0a5%agarosegel. The RNA wasdenatured before electrophoresis by incubatingit in 80%0 dimethyl sulfoxide for10minat37C The eluted RNA wasconcentratedbyethanolprecipitationandwas

translatedin a cell-freeprotein-synthesizingsystem as described in the text. The radioactivepolypeptides

wereseparatedona10%sodium dodecylsulfate-polyacrylamideget Thegelwasdried,andanautoradiogram

wasmade. The slotsatthe sides contained the radioactiveproteins programmed by unfractionatedRNAs isolatedafter infection ofE.coliBEbythephagestrainsshown above the slots. The genescorrespondingto

themarkedpolypeptidesareshown onthe sides. The elutionpositions of168 and 238 rRNA'sareindicated

atthetop. wt, Wildtype.

Molecular weight x10-6

20. ypu 0

I0 ~~~~80

0 ~~~~~~60gp39

50 d

40

gpd

40

30

20 20

0

23034 842465054 22 2630 343842469 Frochon number

FIG. 6. Densitometricanalyses ofT4mRNA

activ-ities.Autoradiogramssuchasthose shown inFig.2, 3, and5 were scanned with a Joyce-Loebi double-beam recording microdensitometerequippedwith a

model JL-20013integrator.Theexposureschosenfor

analysis provided a linear response between film

darkeningasmeasuredonthe recorder and

radio-activity in thesample. Thegeneproducts analyzed

areindicatedonthefigure.

molecular weight when T4 late RNA was

ana-lyzed (42). This protein had the appropriate

molecular weight to be IPIII, but this identifi-cationwas notverified.

Sizes of functional early messages. RNA molecularweightcanbeestimatedfrom the time it takes an mRNAto be eluted from a gel be-cause there isalinearrelationship betweenlog

molecular weight and thereciprocal of the elu-tion time (13, 26). Thepreparative gelsused in

thisstudywerecalibratedwith RNAmarkersof

knownmolecular weights,ranging fromatRNA with a molecular weight of 2.5 x 104 to 28S

rRNAfrommouseliver cells (molecularweight

1.7 x

106).

These markers definedacurvefrom which the molecular weights of RNAs of

un-knownsize could be determined.Figure8shows

a curve for marker RNAs fractionated on a

preparative gel containing 2.25%

polyacryl-amide. Except fortRNA, all of the markerslay

on astraight line. Mostof the RNAs identified migrated in the linear region. The molecular weights of theT4mRNA's identified aboveare

shown in Table 3,together with the molecular weights ofthecorrespondingpolypeptides.

on November 10, 2019 by guest

http://jvi.asm.org/

[image:7.500.67.460.64.313.2] [image:7.500.67.258.424.583.2]

T4 EARLY mRNA SIZES 779

180

t 60

I6

~40

- ' 4

20 iv

^ C .o

l50

~40-

30-

,,10-2

Froction number

FIG. 7. DensitometricanalysesofrIImRNA

activ-ities. RNA isolatedafterinfection ofE.coliBEby T4

32amH18 44amN82 rIIBdell96 rIIAdelEM66 was

separatedon a2.25%polyacrylamide gel. TheRNA

was not denatured with dimethyl sulfoxide. After

electrophoresis and elution, the RNA wasanalyzed

asdescribedinthelegendtoFig.2and in thetext.

The radioactive polypeptides were separated on a

10%polyacrylamide gel, from which an

autoradi-ogramwasmade.TherIIBdel196, rIIAdelEM66,and

32amHl8polypeptideswereidentifiedby comparison

with the corresponding polypeptides synthesized by

unfractionated RNA,whichwerepresentonthesame

gel. Densitometric analyses ofthese three

polypep-tideswereperformedas described in the legendto

Fig. 6. The RNAmolecularweightscale atthe top

wasderivedfromanidenticalgelonwhichamixture

ofyeast and E. coli RNAswasfractionated.

ures2, 3,and5containmolecular

weight scales,

from which themolecular

weights

of other RNA species could be determined. Table3also shows the molecularweight of the mRNA that coded

forT4

,B-glucosyl

transferase

(designated

mfigt);

thisinformationwas obtained by measuring

f8-glucosyl

transferase enzyme

activity

after RNA

fractionated by preparative

gel

electrophoresis

was

translated (data

not

shown).

This RNAwas

extracted from cells infected with a T4 mutant

containing adefectivea-glucosyl transferase to

avoid having to distinguish between the two enzymeactivities.

Table3shows theexcess codingcapacitiesof the genes and thus provides an indication of

whether the mRNA's were monocistronic or

polycistronic.The excesscoding capacityof the

polycistronic

rIIAB

transcriptwasnotrII-coding capacity. The size of this polycistronic RNA

10 Z Q5

St 03

i

0.1

.05

th .03 .02

.01 .02 .03 .04 .05

/elution position 1D6 .07 .08

FIG. 8. Plot of logarithm ofmolecularweight

ver-susthereciprocalsoftheelutionpositions ofthe RNA

standards. Amixturecontaining

double-stranded3H-labeled Q/8MDV-1 RNA and nonradioactivemouse,

yeast,and E. coli RNAswasdenatured withdimethyl

sulfoxideandseparatedby electrophoresisat15mA

ona2.25%polyacrylamidegel. The elutionpositions

of the radioactive RNA speciesweredeterminedby

countingaportion of each sample inaliquid

scintil-lationcounter. Theelutionpositions ofthe

nonradio-active rRNA'sweredeterminedwithachartrecorder

monitoringthe elution buffer.Molecularweights of

1.7x 106, 1.3x 106, 1.07x 106, 0.7x 1(1,0.56x 106,

0.148x106, and0.074x106wereassumedformouse

28S,yeast26S, E. coli 23S,mouse18S,E. coli16S (17),

Q83MDV-1 double-stranded(ds)andQ/8 MDV-1 sin-gle-stranded(ss)RNAs(22), respectively. The

double-stranded QI) MDV-1 RNA, which waspresent in smallquantities, presumably represented

nondena-tured molecules.

decreased with each rII deletion introduced into the phage, and the excess coding capacity

re-mained approximately constant, indicating a

roughly quantitativeagreementbetween the size oftherII deletionandthedecrease in sizeofthe polycistronic RNA.

Sizes ofpulse-labeled early transcripts. The isolation and characterization of cloned fragments of T4 DNA (19, 34, 37, 38) provided

a convenient source of specific hybridization

probes for T4 messages (42). By hybridizing

pulse-labeled, fractionated RNA to filters

con-taining early genes, the sizes of both the func-tional and the nonfunctional transcripts could be measured. By annealing specific gene frag-mentswithpulse-labeledRNA rather than with

continuously labeled or unlabeled RNA, we

hopedtodetectfull-lengthtranscripts.

Radioactive T4 early RNA was prepared as

IF

4s\

VOL. 40,1981

on November 10, 2019 by guest

http://jvi.asm.org/

[image:8.500.248.443.53.287.2] [image:8.500.48.236.62.294.2]

TABLE 3. RNAand polypeptide molecular weights of T4 mRNA activities

Excess coding

miRNA designa- M,of RNA opoyeida capacityof

Gene tion

(x106)

Ml ofpolypeptide mRNA (no.of

aminoacid resi-dues)

43 m43 1.2 112,000 200

del293 mdel293 1.0 103,000 70

46 m46a 0.95 71,000 300

m46b 1.5 71,000 850

39 m39 0.82 64,000 240

52 m52a 0.65 51,000 200

m52b 1.0 51,000 550

m52c 1.3 51,000 850

32 m32a 0.52 36,000 200

m32b 0.75 36,000 430

,Bgt mnfgt 0.67 46,000 250

rII transcripts

rIIABpolycistronicb rIIA+ rIIB 1.5 86,000+33,000 420

rIIA+rIIBdell96 1.3 86,000 +22,000c 320

rIIAdelEM66rIIBdell96 1.1 55,000 +22,000c 400

rIIBa. 0.55 33,000 250

rIIBb+ 0.88 33,000 580

rIIBdell96a 0.45 22,000 250

rIIBdell96b 0.78 22,000 580

aSee references24and 39.

bThe excesscodingcapacity is non-(rIIA+rlIB).

Bolle,personal

communication.

The Mr of gprIIA was

smaller

than thevalue of 95,000 reported by

O'Farrell

etal., (24), but it agreed with the data reported here.

described elsewhere

(42)

and was fractionated

by

electrophoresis

oncontinuous-elution

prepar-ative

polyacrylamide

gels

as described above. Thefractionated RNAwas

hybridized

to nitro-cellulose filters

containing

immobilized

plasmid

DNA. Each filter contained DNAfrom a

plas-mid

containing

a

single

T4restriction

fragment

ofknown

genetic

origin.

Insome cases, restric-tion fragments

wholly

internal to a gene were

used sothat the

transcript

could be associated

unambiguously

with one gene. In other cases,

therestriction

fragment

usedasa

hybridization

probecontainedpartsoftwo

adjacent

genes. In

thesecases,the

transcripts

couldcomefromone

gene orfrom bothgenes.

Figure 9 shows the results obtained

by

per-forming

hybridizations

to

plasmids

containing

gene 43, gene52, gene

rIIB,

gene

rIIA,

gene39,

gene 40plusgene41, gene 42

plus

the

/-gluco-syltransferase gene

(designated ,Bgt),

and gene

30.

Significant

fractions of the RNase-resistant

radioactivityrepresenting the RNAs from gene 43, gene 52, gene 42

plus

gene

18gt,

and generIIB (Fig.8athroughcandg) hadthesamemolecular weights as the functional mRNA's described above. For

example,

there was a

peak

of RNA

activitythat

hybridized

to aninternal

fragment

ofgene 43 which hada molecular

weight

of1.1

x

106.

However, most of the RNA

complemen-tarytogene43 wasmuchsmallerand

migrated

heterogeneously at molecular weights ranging from 0.1 x 106 to 0.8x 106. Aportion of the gene 52messagemigrated in discrete peaks with

mo-lecular weights of0.60x

106

and0.85 x

106.

The sizes of these two species were similar to the sizes of two of the three functional gene 52 messages(Table 3). rIIB RNAs that migratedas

discrete

peaks

withmolecular weights of0.50 x

106

and0.80 x

10c

alsocorresponded

closely

in sizetothetwolower-molecular-weightrIIB mes-sagesdetectedbyin vitrotranslation.Morethan

50% of the gene 52 and rIIB messages were

smallerthan the discretepeaks observed,as was

thecaseforgene 43.

A

polycistronic

rIIAB message was detected

byinvitro translation (Fig. 6).Wethought that this would beamajor species in RNA thatwas

pulse-labeledat anearlytimesince it isbelieved

to be the initial transcript for rIIB (31). How-ever, norIIB RNAthatcorrespondedin molec-ular weight to the rIIA-B polycistronic RNA detected by in vitro translation was resolved

(Fig.9c). Hybridizationto aplasmid containing

aninternal HindIIIrestrictionfragmentor

rIlA

also failed to detectalargerIIAtranscript (Fig. 9d). Infact, we detected very

little

RNA

com-plementarytorIIA,probablybecause there was

less rIlA RNA than RNAs for the other early

genes studied (41). RNAs complementary to

gene 39, gene40plusgene 41, gene42plusgene

on November 10, 2019 by guest

http://jvi.asm.org/

[image:9.500.65.461.81.322.2]

T4 EARLY mRNA SIZES 781

16S 23S 30540 ___165 __23S

20 20

10 16 22 26 22 28 34

FIG.9.racionaionofuls-laee T4eal

4N an 8eetofsecfcT Nsb hybii

gene30 genes40/41t

2 4

4fE-oiB y idtpf 4a ecie intheet

2- 4-~~~~~

010 1622283440 010 6 222834 40 Fraction number

FIG. 9. Fractionation ofpulse-labeled T4 early

RNA and detectionofspecific T4 RNAs by

hybridi-zation to recombinantplasmids. RNA waslabeled

with[3H]uracilbetween3and6minafter infection

ofE.coliBEbywild-typeT4asdescribed in thetext.

Thelabelingwasterminatedat 6min by adding cold

uracilandKCN,and the cellswererapidly disrupted

by hot sodiumdodecylsulfatelysis(14). About 50

jig

of RNA containing 5 x 106 cpm was subjected to

electrophoresis on a 2.25% polyacrylamide gel as

described in thelegendtoFig.2and in the text. The

RNAwas nottreated withdimethyl sulfoxide.

Frac-tions(1 ml;20minof elution time at a flow rate of

0.05ml/min)werecollected. Asmall portion of each

samplewasprecipitated with trichloroacetic acid to

determine the distribution of total radioactive RNA,

and0.50-ml portionsoffractions10through 45

(rep-resenting RNA species in the molecular weight range from0.025 x106to 2 x106)were incubated for 18 h at

67'C in siliconized scintillation vials containing

plasmid DNA immobilized on nitrocellulose filters

(see text). After 18 h, the scintillation vials were

chilledrapidly, and thefilterswereremoved, washed

several times with2x SSC(lx SSC is 0.15M NaCl

plus 0.015Msodium citrate), treated with RNase,

washedseveral moretimes with2xSSC, dried, and

counted withaliquid scintillation counter. The

back-ground radioactivity that hybridized to a control

filter whichwaspresent ineach vialand contained

only pBR322 DNAwassubtracted. This value ranged

from 99 cpm in fraction 10 (the peak of total

radio-activity elutedfromthegel) to 30cpm in fractions

containing high-molecular-weight RNA (molecular

weight,>106). The T4 recombinant plasmids used in

,Bgt, and 30 migrated

heterogeneously (Fig.

9e

through h). The molecularweights of the RNAs

thathybridized tothese genesranged from2.5 x 104to about 1.5 x 106. There were somehigh

points in the distributions ofspecificRNAsfor

gene 39, gene 40plusgene 41, gene 42plusgene

,Bgt,

andgene30, butmostof the

peaks

were not

reproducible whenadifferent RNA

preparation

was analyzed. The peak of mRNA that was

complementarytogene 42plusgene

Blgt

in frac-tions18through20,whichrepresentedanRNA molecular weight of0.6x

106,

wasthesamesize

asthe mRNA that coded for

f3-glucosyl

trans-feraseinvitro (Table3;

unpublished data).

Sizes of

pulse-labeled early

transcripts

after

denaturation

with

glyoxal.

Electropho-resisofRNAscompletelydenatured by

glyoxal

(21) provided betterestimates of RNA molecu-lar weights than analysis of native RNAs and should have diminished artifacts caused by RNA-RNA aggregation.

Radioactively

labeled

T4 RNA wastreated withglyoxal and fraction-atedbyelectrophoresisasdescribed

previously.

Theeluted RNAwas

hybridized

tofour different

plasmids containing restriction fragments of

genesrIIA,

nIB,

43, and 52.Controlexperiments showed thatglyoxalated RNA hybridized with the same efficiency as non-glyoxalated RNA, probably because the glyoxal adductswere

un-stable and came offthe RNA at the elevated

temperatureusedforhybridization (Fig. 10). Glyoxalated RNAs complementary to genes rIIA,rIIB, 43, and 52 migratedmore heteroge-neously than the native RNAs. However, we

observed the same range ofmolecular

weights

with the denatured RNA as with the native RNA (whichwasdenatured with

dimethyl

sulf-oxide beforeelectrophoresis). Gene43RNA mi-gratedveryheterogeneously. The distribution of

RNAthathybridized togene 43 extendedto a

maximum molecular weight of approximately l

ax

106: this was

similar

to the value of 1.1 x

106 obtained whennative RNAwasused. Denaturation with

glyoxal

increased the

het-erogeneity of the migration rates of all four

RNAs

examined;

rIIB RNA was the least

af-fected.Inother

experiments

weobserved similar increasedheterogeneity with otherearlyRNAs

(complementarytoplasmids

containing

gene30,

gene 40

plus

gene41,gene 42

plus

gene8gt, and

gene 39).Nevertheless, thesame discrete RNA

thehybridizations and the T4 genes whichthey

con-tainedwere asfollows: p622 containing gene43(a);

p621 containing gene52(b);containing pABIV

con-taining generIIB(c);pABI containing gene rIIA(d);

p626Acontaining gene39(e); p627containinggenes

40and41

(t);

p624containing genes42and/8gt (g);

p625containing gene30(h).

VOL. 40,1981

on November 10, 2019 by guest

http://jvi.asm.org/

[image:10.500.50.245.54.345.2]

782 YOUNG AND MENARD

S-Z

16s 23s generllAa

b

A4 gere52

6 F

2

'7 C: 11 '2. l^ A7 C7. 1 1

7 ID -523 31Y3 47 5 7 1'

Fractionnumber

FIG. 10. Fractionation of

pul

RNA denatured by glyoxal and

T4 RNAsbyhybridizationtorec

TheRNA preparation described

9wasanalyzed inanidenticalZ

the RNAwastreated with1M

gl

sulfoxidein 10 mMKPO4 buffer

at50°Cbeforeelectrophoresison

0.5%agarosegel. (a)pABI (gene

52). (c)pABIV(generIIB). (d) p6

species was often discerned

RNAswiththesemolecularw

werenotdueto aggregation.

CotranscriptionofrIIAa

etal. (31) detected cotranscr

rIIBbyusingdeletion hybridi

Wedetectedcomigration ofn

rIIA and rIIB polypeptidesbu duciblydetect radioactiverII

this molecular weight. The fa active rIlA orrIIB RNA tha

expected position of a polyc

message could have been du

down of full-length transcript detectlinked rIIA-rIIBtransc:

mids containing restriction fi

rII region (Table 2) (35, 39

early RNA was hybridized

containsarestriction fragmen

3' end ofrIIB, the 5' end of

region in between which isni

(7). After hybridization to

pTB17DNA, the RNA was E

izedtoeitherapTB17filteroi

pTB35DNA. pTB35containE

restrictionfragment. Wecoulo

detectlinkedrIIA-B transcri] cedurewasused, although th

bridizedefficientlyto

pTB17;

possible explanation for thi

small amountof polycistroni

comparedwith thesmaller b

rIIBtranscripts(Fig. 6),whicl

pleted more effectively forti

16s 23s DNA in theinitial hybridization.However, when gene r

lB

we

hybridized

the RNA firstto

pTB35

and then to pTB17, we also failed to detect linked

rIIAB

transcripts.

The most likely reason for the failure to detect

linked

rIIAB

transcripts with pTB35 and pTB17 was that there were very few intact polycistronic

d transcripts. The T4 DNA fragments in pTB35

gene43 andpTB17 are separated in the T4 genome by about 0.9 kilobase. A cleavage of the polycis-tronic rIIAB RNA anywhere in this region, either in vivo, during isolation, or during the first

hybridization,

would have unlinked the 5 23 31 55

rIIA-

and

rlIB-specific

RNAs. If this reasoning se-labeled T4 early is

correct,

linked

rIlAB

transcripts might

have detection of specific been detected more

easily

by using

rIIA-

and

ombinantplasmids.

nIB-specific

DNA

fragments

that were more

in the legend to Fig. closely linked on the T4 genome.

Selzer

(unpub-manner, except that lished data) obtained recombinant DNA clones

1yoxal-50%dimethyl from the rIIAB junction region by partial RI*

(pH

7.0) for 60min digestion of a HindIII

fragment

spanning the

a2.0%acrylamide-

rIIAB

junction region. The recombinant clones

rIIA).

(b)

p621 (gene obtained

by

this procedure included a clone

(gene 43). which had only

rIIA

sequences (pABI) and an-other clone which had only

rIIB

sequences

1,

suggesting that (pABIV) on the plasmid. Recently, this region eights existed and was sequenced (26a), and a promoter sequence (Pribnow box) and the presumed start site of

md

rIIB.

Schmidt therIlB middle-mode RNA were located in a iption ofrIIA and DNA region of

rIIA

that lie between the two ization techniques. fragments cloned in pABI and pABIV. These nRNA activity for two plasmids should hybridize only to an

rIIA-itcould not repro- specific RNA and an

rIIB-specific

RNA, respec-'A or

rIlB

RNA of tively, because pABI lies to the left (5') of the ilure to find radio-

rIIB

middle-mode promoter.

it migrated at the Cotranscription of

rIIA

and

rIIB

was detected :istronic rIIA-rIIB by rehybridizing RNA first hybridized to pABI

ie to rapid break- or pABIV and then subjecting the preparation s. In an attempt to to a second annealing. In the second hybridiza-ripts, we used plas- tion reaction radioactive RNA that had been ragments from the eluted from pABI was able to hybridize to ). Radioactive T4 pABIV, and RNA that had been eluted from to pTB17, which pABIV was able to hybridize to pABI (Table 4). It that includes the The control experiment with RNase-treated hy-Fgene 52, and the brids showed that the abilities of RNAs re-ot deleted by saA9 covered from pABI and pABIV to hybridize to filters containing pABIV and pABI, respectively, depended on eluted and hybrid- RNA sequences that were not in a DNA-RNA a filter containing hybrid form during the first annealing. None of ed aninternal rIIA the radioactive RNA that was eluted from pABI d not reproducibly and pABIV was able to hybridize in the second pts when this pro- annealing to pABIV and pABI, respectively, if Le eluted RNA hy- the first hybrids were treated with RNase,

al-asecond time. One though the RNase-treated hybrids did yield

is failure was the RNAs that rehybridized to plasmids containing c rIIAB transcript sequences identical to those used in the first

out

more abundant annealing. An additional control showed the hwould have com- specificity of hybridization. The same low level he complementary of background radioactivity was observed with

J.

VIROL,

on November 10, 2019 by guest

http://jvi.asm.org/

[image:11.500.59.253.51.213.2]

TABLE 4. Cotranscription of rIIA and rIIB

RNasetreat- Inputfor

Amt

(cpm)annealed in secondhybridization %Reannealedtofilter

RNasetreat-Input for

~~with:

containiing:C

Plasmidusedin mentafter secondan-

with:_______ning:'

firstannealing0 first anneal- nealing pAI ABV p6(gn pB22 rA nB

ing (cpm) (priAI) (IIB) 12) pBR322 rIIA rlIB

pABI(rIIA) + 416 134 30 21 32 25 0

pABI(rIIA) - 1,460 551 275 30 46 36 17

pABIV(rIIB) + 290 30 110 26 24 0 27

pABIV(rIIB) - 408 116 224 24 35 21 48

aThe initial input radioactivity was approximately

106

cpm of

[3H]RNA

labeled from 3 to 6 min after

infection by wild-type T4 at300C,andhybridization was for 30min at 40°C in a final volume of 150ul containing

50%formamide, 0.03 M sodium citrate, and 0.3 M sodium chloride.

bEach second hybridization reaction mixture was incubated for 72 h under the conditions described in

footnoteafor theinitial hybridization. The hybrids were not treated with RNase after the second reaction.

'Abackground of 30 cpmwassubtracted from

all

rawdata.

filters containing either the vector alone (pBR322)ora

plasmid

containing unrelated T4

gene 12.

The relative amounts of the

polycistronic

rIIABtranscript and the non-polycistronic rIIB transcript could be estimated from the data in Table 4. It appeared that at least 50% of the elutedrIlA transcripts extended into rIIB and

that about 50% of the

rIlB

transcripts labeled between 3 and 6 min after infection included

rIIA sequences; that is, they were transcribed from therIIA promoter.Thesevalueswerevery

rough estimates since the

hybridization

effi-ciency inthe first

annealing

wasunknown and in the second

annealing

not all of the RNA

present wasrecoveredashybrid.

Sizes

of

immediate

early transcripts

syn-thesized in the presence and absence of

CAM.T4

protein synthesis

is

required

in vivo

for the

synthesis

of allprereplicative transcripts (12, 30). The average sizes of IPII and IPIII transcripts made in thepresence of CAMwere less thanthe sizes of these

transcripts

madein

the absenceofCAM(1).Thiswas

interpreted

to mean that CAM

prevented

these transcripts

from

being

elongated

intoneighboringgenes. An

alternative

interpretation

wasthatCAM caused

a breakdown ofthe RNA so that transcripts shorter than

full-length

transcripts were de-tected. The data discussed above indicated that

many T4 RNAswerepresent

primarily

in

tran-scriptsshorter thanfull-length transcripts even in the absenceofCAM.

We examined the sizes ofnormal transcripts and transcripts from cells (CAM RNA)

CAM-treated from immediateearlygene 30, gene 39, gene 52, gene42plusgene

,fgt,

andgene40plus

gene41byhybridizingfractionated RNAstothe

appropriate

plasmids.

Other experiments (41)

had shown thatplasmids containingthesegenes

hybridized to T4 RNAsynthesized inthe pres-ence ofCAM. Our results are shown in Fig. 11.

The RNAs wereglyoxylated before

electropho-resistoprevent

aggregation

and to

provide

more

reliable

estimates

of molecularweights.Asnoted above,denaturation ofnormalT4RNAledto a moreheterogeneousmigration

pattern.

The

mo-lecular weight distributions of the

complemen-tary normal and CAM RNAs were distinctly different only for gene 40 plus gene 41. The

molecularweight of the CAM RNA forgene 40

plusgene41waslessthan the molecularweight of the normal RNA for gene 40 plus gene 41. However,the sizes ofCAM RNAs synthesized

fromgene30, gene 39,andgene 42plusgene,Bgt were notdifferent than the sizes of the normal

RNAs

complementary

tothese genes. The

max-imum

molecular

weights,

as

well

astheaverage

molecular

weights,

were

approximately

thesame in normal and CAM RNAs complementary to

these genes. For genes 30 and 39 only

small

fractions of the radioactive RNAs made either

in the presence or the absence of CAM were

largeenoughtobefull-length transcripts.

We also studied the size distribution of

func-tionalT4mRNAinapreparation of CAMRNA.

Nonradioactive RNAwasisolated 10minafter infection by T4, and CAM was added 5 min

before infection. The RNAwaselectrophoresed

on a preparative RNA gelandanalyzed as

de-scribed in the legendto Fig.2. No polypeptide with a molecular weight greater than about 21,000was

synthesized

(41). All of theT4 CAM

RNA

polypeptides

were

synthesized

by mRNA's

having molecularweights of0.1 x 105to0.2 x

105 (datanotshown).Althoughnospecificgene

productswereidentified,many of thesmall

poly-peptides codedforbyCAM RNAs had thesame

mobilities as the

small

polypeptides coded for

by normal RNAs. However, in normal RNAs someof the

small

polypeptideswereencodedby large mRNA's, as shown in Fig. 2, whereas in CAM RNAs no largeT4 mRNA'swere

identi-fied bytranslation offunctional mRNA's.

VOL. 40,1981 T4 EARLY mRNA SIZES

783

on November 10, 2019 by guest

http://jvi.asm.org/

I,

0c

Fraction number

FIG. 11. Analysis of fractionatedT4CAM-treated

RNAsby hybridizationtorecombinantplasmids.

Ra-dioactive RNAwaspreparedbylabelingcells with

[3H]uracilbetween 3 and6min after infection byT4

52amEA118 either in the absenceorpresenceof200 pgofCAMpermladded5min before infection.The

RNA wasisolated andanalyzed byelectrophoresis

andhybridizationasdescribedinthelegendtoFig.

9 and in the text. The RNA was denatured with

glyoxal before electrophoresis as described in the

legendtoFig.10 andfractionatedbyelectrophoresis

on a 2.0% acrylamide-0.5% agarose gel; 0.9 ml of

every other 1-mlfraction was hybridized tofilters

containing the following plasmid DNAs: (a) p625

(gene 30),noCAM; (b) p625 (gene 30), plusCAM; (c)

p626A (gene 39), no CAM; (d) p626A (gene 39), plus

CAM; (e) p627 (genes40and41), no CAM; (D)p627

(genes40and41), plusCAM; (g) p624 (genes42and

,8gt),noCAM; (h) p624 (genes42 andfigt), plusCAM.

DISCUSSION

With a few notable exceptions, T4 mRNA's

are unstable, as are mostprocaryotic mRNA's.

Nevertheless, estimatingthemolecularweights ofspecific T4transcriptsunder different

condi-tions, such as different times after infection, duringinfections with putative regulatory mu-tants(tsGl,regA),andduringinfections of

mu-tant host cells (RNase I-, RNase III-), should

provide information important to an

under-standingofthe arrangement of T4transcription

units and

synthesis,

processing, and

degradation

ofT4 mRNA.

Functional T4 mRNAwas detected by

frac-tionating

total RNA from T4-infected cells on

preparative polyacrylamide

gels, followed

by

translationinvitro toidentify specific

polypep-tides

and, indirectly,

specific mRNA's. Sincewe used autoradiography of one-dimensional

so-dium

dodecyl

sulfate-polyacrylamide gels to

identify

T4

polypeptides,

only themost

promi-nentT4early proteins could be identified easily. Theminimumsizeofthe mRNAdetected had

tobesufficienttocode for theintact

polypeptide

chain,

since that was our means of

identifying

themRNAactivity.Oursizeestimates of the T4

RNAswerebasedon comparisonsoftheir

mo-bilities

(more

properly, their mobilities and

elu-tion rates) with the mobilities of a series of RNAs with known molecular weights. For T7

early

andlatemRNA's, this method provideda

fairly accurate measurement of molecular weights (26).Wehavenoreason to believe

other-wise for T4RNAs. For four of thenine mRNA's

whose sizes are showninTable 3, we observed

asingle messenger peakofapproximately

mon-ocistronicsize. In each casetheapparent

molec-ularweight exceeded the coding capacity.Some

extra nucleotides were presumably present in

theflankingregions, but the apparentexcessof

200 to600nucleotideseven forthose RNA

spe-ciessuspectedtobemonocistronic (suchasthe gene 43 message) might indicate that our size estimateswere toolargeforall ofthemRNA's.

Forexample, the onlygene 43 message activity

that was detected had a molecular weight of approximately 1.2 x 106. Sincegp43 hasa

mo-lecularweightof 1.15 x

105,

amessengerwith a

molecularweightofapproximately 1.0 x 106 is requiredtocode for thegp43polypeptide. There

wasverylittle ifanygene 43 mRNAactivityof

higher molecular weight, as might have been

expected if this RNA had been derived from a

larger precursor. However, rapid processing to

thesize observed couldhave prevented the iden-tification ofalargegene 43transcript.

Severalearly messagesdisplayedtwo ormore

peaksofactivity. Gene52messageactivityhad

peaks representing RNAs with molecular weightsofapproximately0.65 x 106, 1.0 x 106, and1.3 x 106. Thesmallestofthesecouldhave

beenamonocistronicgene 52message since gp52

has amolecularweight of about55,000.The two

largergene 52 messagescould have been poly-cistronic messages. The bestway to determine this and to determine which other genes are

linkedontheputativepolycistronic messages is

to useplasmids containinggene 52 andadjacent

regions. Only thegene 52 mRNA with a

molec-ular weight 0.8 x 106 was observed in RNA

on November 10, 2019 by guest

http://jvi.asm.org/

[image:13.500.63.256.54.355.2]

VOL. 40,1981

extracted after infection with the rII deletion

mutant del1241. The deletion endpoint nearest gene 52isveryfar fromgene52. This

observa-tion isinterestingbuthas not beenreproduced.

Itis alsointeresting that another deletioninthis

sameregion affects gp52synthesis. The deletion

saA9 (7) causes overproduction ofgp52 specifi-cally (Mattson and Bolle,personal communica-tion). Inthis casethe effect of the deletion on gp52 expression could be due to transcription

from therIIBpromoter(Mattson,personal

com-munication),although itmaybe thatsequences

in betweenrIIB and gene 52 have an effecton gene 52

transcription.

Gene32isrepresented bytwoactive mRNA species. These have

approximately

thesame

mo-lecularweightsas two

peaks

of stable radioactive RNA activity that were observed after a

long

chase in the presence of

rifampin

(H. Krisch,

personal communication).

In thiscase the two

peaks of radioactive RNAwereassociated with

gp32 message activity. In both our studies and those of

Krisch,

most ofthe

gp32-synthesizing

activitywas associated with the lower-molecu-lar-weightgene 32mRNA.Itisalso noteworthy thattheRNA extracted fromcellsinfected with

agene 32amber mutant contained a full-sized

gene 32 message.

Unlike the messages just described, several mRNA activities migrated very heteroge-neously. One such mRNA coded for gp3O; the

minimum molecular

weight

ofthis mRNA was

about 0.60 x

106,

and its maximum molecular weightwas1.5x 106to2.0 x 106. Several

differ-entmechanisms couldgive risetomRNA's that display heterogeneous

electrophoretic

mobili-ties. They could be derived fromlarger precur-sors which are

degraded

randomly

into

frag-ments, someof which canstill functionas

tem-plates, or they could be transcription products of genetic regions lacking strong termination signals. Therearealso otherpossibilities.

Hybridizationtoclonedrestrictionfragments containing known T4 genes was

performed

in orderto

identify specific

T4 mRNA's

directly.

Pulse-labeled RNA was used to increase the fraction ofnascentand

newly synthesized

tran-scripts. Forseveral

early

T4 genesafunctional

mRNA had been identified unambiguously by

translation in vitro (genes 43, 52, 39,

rnIB,

and

rIIA),and we had ahybridization probetoassay

specificallyanddirectlyfor theRNA (thegene 52probemay have contained asmallamountof non-gene 52 DNA). For genes 43, 52, 39, and rIIB the radioactive RNAs that were able to

hybridizetothecorrespondingplasmidshad size distributions that included peaks representing RNAswith the same molecular weights as the

functional mRNA's detected. However, there

T4 EARLY mRNA SIZES 785

wasalsoa considerableamount of

lower-molec-ular-weight RNA, presumablyrepresenting

de-graded mRNA. In thecaseoftherIIAmessage,

essentially allof thehybridizing RNA that was

detected was smaller than the functional

mRNA.However, most significant was the fail-ure toobserve pulse-labeledRNA with a higher

molecularweight than the molecular weight of

thefunctional mRNA's fromthe same gene. It isunlikelythat this failure was due to the pres-enceofonly incomplete transcriptsin the

pulse-labeled RNA since a 3-min pulse should have

beenlong enoughto

allow

synthesis of an RNA

witha molecular weight of 2 x

106

to 3 x

106

(considerably larger than the transcripts

de-tected). However, to test this hypothesis

di-rectly,RNA waslabeled from0 to 10 and from 2to 4minafter infectionwith T4, the RNA was

isolated immediately afterlabeling was

termi-nated, and the distribution ofgene 43message was measured by hybridization to a gene

43-containingplasmid (datanot shown). The RNA

labeled from 0 to 10 min after infection had a

distribution ofgene

43-annealing

RNAvery sim-ilartothat showninFig. 9forRNA labeled from 3to 6min. Inthe RNApulse-labeled from2 to 4 minafter infectionthere was relatively more gene 43 message in the peak with amolecular weight of 1.2 x

106.

Higher-molecular-weight

gene 43 RNA was notdetected in either RNA

preparation. This suggested that the gene 43 message with a molecular weight of 1.2 x

106

represented thetranscription unit forgene 43 or

thatprocessing occurred

during

transcription,

as

appears tobe thecasefor

bacteriophage

T7.

Several different mechanisms could have

pro-duced the

multiple transcripts

from othergenes,

bothidentified (52,

nIB,

32,

IPIII)

and uniden-tified. These

transcripts

could have arisen

by

processing ofa larger

transcript,

as is the case

with theT7

early region primary transcript

and several T7 late

transcripts. Alternatively, they

couldhave been

synthesized

from

independent

promoters and a common termination site. There arewell-documented

precedents

for this possibility. Early T7 transcription by E. coli

RNA

polymerase

andlate T7

transcription by

T7 gene 1RNA

polymerase

utilize

multiple

pro-moters and a common termination signal to generate all of the

early

messages and several

latemessages,respectively (9, 11,26).The

mul-tiple transcripts of the small DNA

phages

are

also derivedfrom

multiple

promoters anda com-mon termination site

(35).

Synthesis

of RNA

transcripts froma

single

promoter and

multiple

termination sites could also generate multiple

mRNA

species.

An

example

of sucha

transcrip-tion unit occurs in

bacteriophage

lambda. The

transcripts

promoted

from

PI

and Pr terminate

on November 10, 2019 by guest

http://jvi.asm.org/

at ti and

t,

unless the lambd,- N protein is present.InthepresenceofNprotein,

transcrip-tion continues into more distal genes. No T4 proteinwith N-likeactivity has been identified,

althoughanti-terminationhas beensuggestedas

anexplanationfor thesynthesisofdelayed early

RNAbyreadthroughfromadjacentimmediate early regions (30, 31).

Thesize of the RNA from the rIIregion is of

particular interest. rIIB message activity was

detected by thesynthesis of either thewild-type

rIIB polypeptide or the synthesis of a shorter rIlB polypeptide coded for by a mutant rIIB gene containing an internal in-phase deletion

which removed approximately 300 base pairs.

rIIA message activitywas more difficultto de-tect,butwasidentifiedunambiguously by using

anrIlAmutantcontaininganinternal in-phase

deletionwhich reducedthemolecularweight of the rIIA polypeptide by about 30,000 (from 85,000to 55,000) in ourgel system. A multiple

mutant containing both rII in-phase deletion

mutations, a gene 32 amber mutation which

producedanidentifiableamberfragment,anda

gene 44 amber mutation allowed us to detect

and measure both rIIA mRNA and rIlB mRNA,

as well as gene 32 mRNA in the same RNA

preparation. Threepeaks of rIIBdeletion mes-sageactivityweredetected;thesecorresponded

to molecular weights of approximately 0.45 x

106,

0.78x 106,and 1.3x 106 (designatedmrIIBa, mrIIBb, and mrIIAB, respectively). The two

smaller rIlB messages were alsoidentified

un-ambiguously in wild-type rII RNA (T4 32amA453 44amN82) (Fig. 2) and in RNA

con-tainingwild-type rIlAbutrlIB deletionmessage (T4 32amH1844amN82) (Fig. 5). The high-mo-lecular-weight rIlB message (mrIIAB)

comi-grated with message activity forrIIAdelEM66. Asimilar comigration ofrIIA andrIIB message activitieswasobservedwhenweused RNA

con-tainingeitherwild-typerIIAandrIlB messages,

rIIB

deIrlIA

wild-type message, or

rnIB

wild-typerIIAdeletionmessage. Inthecase of wild-typeRNA,the message activitiesforgprIIBand for gprIIA migrated more slowly, representing

higher molecularweights than when RNAs from thedeletionmutants were examined. The find-ing that the molecularweight of thelargest

rnIB

message was decreased by a deletion in rIIA proved that thetwomessage activitieswere on

thesamemolecule.

Hybridizationto aclonedrestrictionfragment

of

rnIB

confirmed the existence of two of the functional messages just described above. An rIIBprobe detected peaks of rIlB RNAat

mo-lecularweights ofapproximately 0.5 x 106 and

0.8 x

106,

approximately the same size as ob-served forfunctional rIIB message. There was

also aconsiderable amount of lower-molecular-weight rIIB RNA,presumablyrepresenting de-graded rIIB RNA. No high-molecular-weight

polycistronic rIIAB RNA was detected

repro-ducibly with either the rIIA-specific probe or

the nIB-specific probe. Essentially all of the rIIA-specific RNA was present in a heteroge-neousdistribution withmolecular weights rang-ing from 2.5 x 104 to 1.0x

106,

suggesting that rIIAmRNA wasbroken down more rapidly than rIIB mRNA. Polycistronic rIIAB transcripts weredetectedbyrehybridizingradioactive RNA first hybridized either to a plasmid containing the 5'-terminal end of rIIB or to a plasmid

containingthe 3'terminusof rIIA.

Schmidt and co-workers (31) originally

re-ported apolycistronic rIIAB transcript, aswell

as a monocistronic rIlB transcript, based on

studies in which hybridizationto phage DNAs

containing various rII deletions was used. Our

results confirm the existence of both a polycis-tronicrIlAB transcript and a smaller, possibly

monocistronic rIIB transcript. In addition, we

foundathirdrIIB message thatwas not

antici-pated. Other investigators havestudiedthesize of rII RNAbyusing deletion hybridization tech-niques. Sederoffetal. (33) foundrII RNA that

sedimentedinabroad distributionrepresenting

molecularweights rangingfrom 2.5 x 104to 1.2

x106.These authorsdid not distinguish between

rIIA-specificandrIIB-specific sequences. This is the same molecular weight range that we

ob-served when we used hybridization to specific rIlA and rIIB probes. On the other hand,

Wit-mer(38) claimedthat he detected an

rnIB

tran-scriptwithamolecularweight of0.43 x 106, an

rIlA transcriptwith a molecular weight of 0.71 x 106,andapolycistronicrIIABtranscript with

amolecular weight of1.1 x

106.

Onlythe poly-cistronicrIIABtranscriptwasdetected when T4 RNAsynthesizedinvitrowasanalyzed by

Wit-mer (38). Therewas nolower-molecular-weight

RNA present in the RNApreparations of

Wit-merdespite repeatedincubation of the RNA at

an elevated temperature during hybridization

beforeanalysisof the RNAbysucrosegradient

sedimentation. We cannot account for the dis-crepancies betweenourresults and those of Wit-mer (38). In particular, we never detected a

monocistronic rIlAtranscript,norhavewe ever examinedanRNApreparationthatdidnot

con-tain significant amounts of lower-molecular-weight rIIA-specific RNA and rIIB-specific

RNA.

Itis thought that CAM inhibitssynthesis of

delayed earlyRNAby preventingelongation of

transcriptsfrom immediateearlygenes into

de-layed early genes. If thiswere the case, CAM-treated RNA transcribed from immediate

early

on November 10, 2019 by guest

http://jvi.asm.org/