Alignment of a restriction map with the genetic map of bacteriophage T4.

0022-538X/81/040115-10$02.00/0 Vol. 38, No. 1

Alignment

of

a

Restriction

Map with the

Genetic Map of

Bacteriophage T4

JIING KUAN YEE AND ROBERT C. MARSH*

Biology Program, University of Texas at Dallas, Richardson, Texas 75080 Received1October 1980/Accepted 31 December 1980

A restriction map of the bacteriophage T4 genome was aligned with the T4

genetic map. Included were the cleavage sites for BamHI, BglII, KpnI,

PvuI,

Sall, and XbaI. The alignment utilized the factthat the T4 genetic map had been

orientedpreviously withrespect to a T2/T4 heteroduplex map. DNA fragments

fromaBgII digestion ofcytosine-containing DNA from a T4dCTPase- denA

denB(rIIH23B)alc mutant werehybridized withfull-length chromosomal strands

of bacteriophageT2,and the heteroduplexeswere examined byelectron

micros-copy. From their lengths and patterns of substitution and deletion loops, the

heteroduplexes formed with6ofthe 13BglIl fragmentscould beunambiguously

identified andpositioned on the T2/T4 heteroduplex map. The ends of the T4 DNAstrands in the heteroduplexesdirectly identified thelocation of 10BglII cleavage sites. The remaining three BglII cleavagesites could be assigned to the

T2/T4 heteroduplexmapbasedontheir relativelocations on the restriction map.

Itwasalsopossibletoidentify thesourceof the DNAstrands(i.e.,T2 or T4) in

fourpreviously unassigned deletion loopsonthe T2/T4 heteroduplex. Among the

BglII fragments identified in heteroduplexes was the fragment containing the

rIIH23B deletion; this deletionwas used as the primary point of reference for

alignment of the T4 restriction map with the T2/T4 heteroduplex map and,

hence, with theT4geneticmap.

Since the isolation ofbacteriophage T4

mu-tantswhich allow thegrowth and packaging of

DNA containing

cytosine

instead of the

gluco-sylated

hydroxymethylcytosine

residues

nor-mally

present

(15),

T4 DNAhas beenamenable

toanalysis with restriction endonucleases.

Cir-cular

cleavage

mapsof the 166-kilobase

pair

T4

genomehave been

published

for restriction

en-zymes

BamHI,

BglII, KpnI,

SalI, SmaI,

and

XhoI (3, 6,

14),

and maps for

BglI, EcoRI,

HindIII,

and PstI have been

completed

recently

(2,12). In the

accompanying

report,the sites of cleavage by PvuI and XbaI were added to a

computer-fitted,

unified map of the

BamHI,

Bglll, KpnI,

and

SailI

cleavage

sites

(9).

Alto-gether,

morethan250

cleavage

sites have been

identifiedontheT4 genome.

We extend the usefulness of these maps

by

reportinghere an alignmentbetween our

com-puter-fitted restrictionmap and the T4genetic

map. Heteroduplexes wereformed between T4

BglII fragments

and

full-length

T2

chromo-somalstrandsand examinedby electron

micros-copy. Fromthepositionsof the ends of the T4

strands relative to known

T2/T4

nonhomolo-gies,the

Bglll

cleavagesitescould be locatedon

theT2/T4heteroduplexmap. This made

possi-ble an

alignment

between the restriction map

andthe T4 genetic map, asthe T2/T4 hetero-duplexmaphad been orientedpreviously with

respect tothe T4geneticmap(7). Thealignment

of therestriction and genetic mapsis fixed by

the

position

ofanrII deletion. Themapsappear

tobecorrelated within1to 2kilobases(kb)over

their

entirety.

O'Farrell

et al.

(12)

have

independently

aligned a T4 restriction map with the genetic

map,

primarily

by filter

hybridization

of

restric-tion

fragments

to clonedsegmentsofT4 DNA

knowntocarry

specific

genes.

Also,

Carlson

(2)

has determined a general alignment based on

the

cleavage

patterns of

cytosine-containing

DNAfromT4 alcmutantswhich carryvarious characterized deletions. Our resultsarein

agree-ment.

MATERIALS AND METHODS

Phage and bacterial strains.

Bacteriophage

T4

56-(amE51dCTPase-)denA(nd28)denB(ArIIH23B)

alc8 and its

restriction/modification-negative

hosts

Escherichia coli K803 (supE hsdSrgl

gal

met) and

B834 (supE+

ris

mB- galmet) were

provided by

L.

Snyder (15). Originally, this T4 alc mutant was

thoughttocarry the 4.2-kbrIIH23 deletion,but in-steadit hasbeenfoundtocarryalongerrIIdeletion,

definedgeneticallyasextendingfromor neartheend 115

on November 10, 2019 by guest

http://jvi.asm.org/

of therIIAgeneinto theaclocus(8).We refer tothis

deletion asrIIH23B. Byelectronmicroscopyof het-eroduplexes, wefound that the rIIH23B deletion

re-moves5.8 kb and that itsstarting pointatthe endof the rIIA gene isindistinguishable from thestarting point forthe rIIH23 deletion.BacteriophageT2Land E. coliBwereobtained from W. Harm.

Growth of bacteriophage and isolation of DNA. T4 alc quadruple-mutant particles that

con-tainedcytosine-containingDNAweregrownby using firstE. coli K803 and thenE.coliB834,and theDNA

wasisolated by phenolextractionasdescribedinthe

accompanyingpaper(9).T2Lstockswereobtainedby

infectionofexponentially growingE.coli Bwithsingle plaques; Hbrothwasusedasthemedium;itcontains

(per liter)8gof nutrient broth (DifcoLaboratories),

5gofpeptone(Difco),5gofNaCl,and 1gofglucose. The T2Lparticles were purifiedby differential

cen-trifugation and stored in 20 mM Tris-hydrochloride (pH 7.8)-0.5mM EDTA.

T2/T4DNAhybridization.T4 cytosine-contain-ing DNA from the alcquadruplemutantwasdigested

withBglII (NewEngland Biolabs)asdescribed in the accompanying paper (9) and then extracted with phenol. The mixture of T4BglII fragmentswas

hy-bridizedwithfull-lengthT2 chromosomal strands

es-sentiallybythemethod of Davisetal.(4).This method minimizes breakage of the long chromosomal DNA strandsby usingalkalitosimultaneously disrupt phage particles and denaturethe DNA inthesametubeused

forthehybridization. To5x 109 T2 particles and0.2 ,ug of the T4BglII fragments in 45

Pi,

5plof 0.2 M EDTA (disodium salt)-1 MNaOHwasadded. After

10minatroomtemperature,the solutionwasadjusted

topH8.5with7

pl

of 1.5 M Tris-hydrochloride (pH 7.8)-0.2 M Tris base, and then 41 to 43

pl

of99% formamidewasadded.Renaturationofthe DNAwas

allowedtoproceedatroomtemperaturefor2hduring dialysis againstasolutioncontaining0.45MNaCl,0.1 MTris-hydrochloride (pH 8.5), 0.01 M EDTA, and40 to 50% formamide or for5 h without dialysis. The conditionsweredesignedtopermit 50% of the mole-culestoanneal.

Electronmicroscopy. Circular, single-strandedfd DNA (Miles Laboratories, Inc.) containing6,408

nu-cleotides (1) and double-stranded pBR322DNA (ob-tainedfromH.BoyerinE.coliRR1) containing4,362 basepairs (bp) (16)wereaddedtothe T2/T4 hybrid moleculesasinternal length standards. TheDNAwas

spreadontoamonolayer ofcytochromecin the

pres-enceofformamideasdescribedby Davisetal. (4) and thenpickedup oncarbon-coated grids.Thegridswere

rotaryshadowed with Pt and photographed with a

Siemans 1Aelectronmicroscopeatx10,000. Lengths ofDNA molecules weremeasured onprojections of

photographic negatives by using a linear integrator

(NumonicsGraphics).

RESULTS

T2/T4 aicmutantheteroduplex. The use

ofnonhomology loopsasmarkerstolocate the

ends ofBglII fragments ontheT2/T4

hetero-duplexmap, which inturnpermittedan

align-ment of the restriction and genetic maps,

re-quired verification of the published wild-type

T2/T4 heteroduplexmapfor the strains which

weused.Because of their circularpermutation,

T2 and T4 form large circular heteroduplexes

when denatured and reannealed.Measurements

made on suchheteroduplexes formed from T2L

andthe T4alc quadruple-mutantchromosomal

strandsyielded a map of substitution and

dele-tion loops (Fig. 1, outercircle) identical tothe

one constructed by Kim and Davidson (7) for

wild-type DNAs, except for an additional

sub-stitution-type nonhomology labeled G', which

weconsistently observed, and a 5.8-kb deletion

loop that resulted from the rIIH23B mutation

carried by the T4 alc quadruple mutant. As

pointed outby Kim and Davidson (7), this

het-eroduplex map is an idealized map. In individual

heteroduplexes, neighboring loops sometimes

merged to give a bigger loop. In other cases,

large substitution loops underwent additional base pairing, creating several smaller loops;

sometimes loops collapsed or became twisted

andappeared as stretches ofhomologousDNA.

For example, several smaller loopswere often

observedinplaceof the Floop,and the

homol-ogous regions between loops V' and U' or

be-tweenloops H and S often remainedunpaired,

generatingloops larger than normal.

Heteroduplex mapping of the T4 Bgll fragments. The BglII restriction fragments

from atotaldigestion of T4alcmutant

cytosine-containing DNA(Table1)were hybridized with

full-length T2 chromosome strands and exam-inedbyelectron microscopy. Of 50well-spread heteroduplex molecules which were chosen for

measurement,one-halfcould beassigned to

spe-cific T4BglIIfragments.The otherseither could

notbe positioned unambiguously on the

heter-oduplexmap on thebasis of their loop patterns

or did not correspond in length to one of the

knownBglII fragments.Some examples in this

lattercategory were to be expected due to

ran-dom cutting of the ends of T4 chromosomes during phageDNApackaging.

For the initial measurementofthe length of

the T4 DNAstrandin aheteroduplex,the two

arms ofthe substitution loops were averaged,

and the lengths ofthe deletion loops were

di-vided inhalfunlessthe loops were already

as-signedtoT2 or T4DNA byKimand Davidson

(7). In the case of BglII fragments 5 and

7a,

onceheteroduplexeswere correlated with these

fragments,the measured lengths of the T4 DNA

strandswereadjustedupward by

assigrunent

of

previously unassigned single-stranded loops to

T4 DNA in order to more exactly match the

knownlengthsofthese fragments.

Figure2shows examples of theheteroduplex molecules that could be correlated with BglII

on November 10, 2019 by guest

http://jvi.asm.org/

ALIGNMENT OF T4 RESTRICTION AND GENETIC MAPS 117

0.

-FIG. 1. BgUlI restriction mapofT4(innercircle) alignedwith respecttotheT2/T4heteroduplexmap(outer

circle). The solidarrowsindicateBgUl cleavagesites locatedontheheteroduplexmapbyvisualizationof

heteroduplexes formedbetween T4

BgUII

fragmentsandfull-length T2chromosomal strands. The dashed

arrows indicate theBglII cleavagesites assigned to theheterodupkxmap on thebasis oftheirrelative

locationsontherestriction map. The circularheteroduplex mapwasdrawnbyaveragingthelengths ofthe twoarmsofsubstitutionloopsandtakingone-half ofthelengths ofdeletionloops. The letterdesignations

givento someofthenonhomologiesonthe

heteroduplex

mapby KimandDavidson(7)have beenretained;

theremainderofthenonhomologiesaredesignatedwithletters that carryprimes.Thesubscripts2and4are

usedtoindicate thesourcesofthesinglestrands in thenonhomologyloops, ifknown(72andT4,respectively).

Theful-length ofthe 5.8-kb rIIH23Bdeletion carried bythe T4 alcmutantisrepresentedas a segment

bisectingbothmaps.Its boundarieswereusedtofixthealignment ofthe maps.

fragments2a,2b,3, 4, 5,and7a.Figure3shows

theaverage

lengths

of theT4DNA strandsand

oftheindividualduplexandsingle-stranded

re-gions of all measured heteroduplexes. In the

middle oftheheteroduplex in Fig.2a, a

promi-nent deletion

loop

that corresponds in size to the5.8-kb rIIH23Bdeletion iseasily recognized. About400bp away isthe neighboring deletion

loop A,which wasassignedtoT4 DNAbyKim and Davidson (7) because theduplex segments

between loop Aand various rII deletionloops

often open up to form single substitution-type loops,the arms of which can be accounted for

only if loop A is T4 DNA. Missing from the

heteroduplex

in

Fig.

2ais

loop

I',whichshould

be visible between loopA4 and the left end of

on November 10, 2019 by guest

http://jvi.asm.org/

[image:3.494.54.436.63.458.2]

TABLE 1.

BglII

fragments

ofT4alcmutant cytosine-containing DNA

Fragment Size(bp)a

1 55,840 2a 17,530 2b 17,500 3 14,010 4 12,660 5 10,580 6 7,940 7a 5,540 7b 5,470 8a 4,410 8b 4,350 9 3,330 10 1,180

aSizes, as determinedbyagarosegel

electrophore-sis,wereadjustedslightlybyaleast-squares analysis

designed to yield a best-fit map for the fragments

producedbyBglII and five other restriction endonu-cleases(9).

the molecule.

Apparently

it

collapsed

in this

example, whichwas chosen for illustration

be-cause basepairingofthe segment between the

rIIH23B and A4 loops had occurred and all

otherregionswerewellspreadduring mounting.

Inother examples of thisheteroduplex, loop I'

was visible eitheras a

deletion-type loop

oras

the

asymmetric

substitution-type loop

shownin

Fig. 3.

Ateach end of theheteroduplex regioninFig.

2aisasingle-strandedtail of T4

DNA;

the single-stranded T2 DNAextends overamuch

longer

distance,

past the

edges

of the

micrograph.

Cleavage

ofT4alcDNA

by BglII

within theH

nonhomology region would have generated the 1.32-kbtail ofT4 DNA attheright end of the heteroduplex. Howthe tail ofT4 DNA which

was seen attheleft end of all examples of this

heteroduplex was generated isnot directly

ap-parent. Itcould have resulted fromcleavage by

BglIl

withintheH' deletion

loop,

which was not

assignedtoT2 or T4 DNA (7).However, as an

examination of the end of the heteroduplex formed with the adjoining

BglII

fragment

showed (Fig. 2dand 3; seebelow), the H'loop

wasapparentlyT2DNA,andcleavageofthe T4

DNA occurred 560 bp beyond the H' deletion

loop. Thus,the tail of T4 DNAappears to have

resulted from afailure of the T4DNA to base

pair with T2 sequences beyond the H'2 loop,

probably due to insufficienthomology and the

lack of a well-paired neighboring region, as

wouldbe found in acomplete T2/T4

heterodu-plex.

Altogether,

the sequences

assigned

to T4 in

the

heteroduplex

shown in

Fig.

2a were12.36±

0.09kblong (Fig.

3),

indicating

that the T4 DNA

strandwasderived from

BglII

fragment

4,which

was12.66kblong,asdeterminedbyagarosegel

electrophoresis.

On the BglII restriction map of the T4 alc

mutantDNA, one end of BglII fragment 4

ad-joins the smallestBglIIfragment(fragment10),

andbeyond that is fragment 2b. The

heterodu-plex that corresponds to this fragment is shown in Fig. 2b. It contains the seven nonhomology

loopsfrom K'toO',which extend from near the

end ofBglII fragment 4 in the H substitution

loop in a counterclockwise direction. A small,

eighthloop was present between M' andN',but

it did not occur regularly in T2/T4

heterodu-plexes and thus was notdesignated with a letter.

Given the fact that deletion loops 0 and P

represent T4 DNA (7), the T4 strand in this

heteroduplex was 17.97 ± 0.21 kblong (Fig. 3),

compared with 17.50 kb for BglIIfragment2b

from agarose gels.

The right end of BglIIfragment2b was

gen-eratedbycleavage 190 bp beyond loopO',within

a region of T2/T4 homology. At the left end,

cleavage occurred betweenloopsS andR,

leav-ing space forBglII fragment 10 between

frag-ments 2b and 4 (Fig. 1). Note that inthe

heter-oduplex shown inFig2b thestrand of T2 DNA

terminated within the region of homology

be-tween loops K' and J', leaving the T4 DNA

unpaired over the relatively short distance to thehomology region betweenloops S andR.

Proceeding around the restriction map from

BglII fragment2b, the nextfragmentsfor which

unambiguous

heteroduplexes

were observed

were fragments 7a and 5, both of which are

shownhybridized tothe same T2 DNA strand

inFig.2c.Theintervening fragment 7bisabsent,

andexcept for this particular example, its

cor-respondingsequencealong the T2 DNAwould

have remained undetected, as it contains no

observablenonhomologies. Inagarose gels,

frag-ment7b was 5.47kblong; inFig. 2cthe

corre-spondingT2single-stranded regionwas 5.39 kb

long.

The region to which BglII fragment 7a was

hybridized covered deletionloop V' and

substi-tution loop U'. The 1.39-kb loop V' was not

assigned to either T2 or T4 by Kim and

David-son,butmustbeT4 DNA ifthe measured length

of the T4 strand in theheteroduplexis to match

the length ofBglII fragment 7a. Including V',

the T4 fragment was 5.76 kb long (Fig. 3); in

agarosegels,fragment 7a was 5.54 kb long. The

counterclockwise.end

of fragment 7a could be

placed 1.85 kb beyond loop V'. The other end waslocatedabout 1kbpast the U'substitution

loopin aregion whichnormallyisbase paired in

full-length

T2/T4 hybrids but remained

un-paired inthisheteroduplex.

BglIIfragment5was 10.58 kblongas

on November 10, 2019 by guest

http://jvi.asm.org/

ALIGNMENT OF T4 RESTRICTION AND GENETIC MAPS 119

a) T2/T4

Bg/II

fragment 4

H4

b) T2/T4

B_/II

fragment 2b

N'

_Wof P4 v ",'IV

FIG. 2. ElectronmicrographsandtracingsofheteroduplexDNAsformedfrom full-lengthT2chromosomal strands and strandsfrom BglIIfragments ofT4.Thearrowsdesignatetheendsofthe T4 DNA strandsand,

thus,BglII cleavagesites.Nonhomologyloopsaredesignated bylettersasdescribed in thelegendtoFig.1.

minedbyagarosegels. Toaccountforthislength

inthecorrespondingheteroduplex inFig. 2c, we

assigned the long arm of substitution loop W'

and the deletion

loop

X'toT4. Altogether, the

T4 strandintheheteroduplex thenwas 10.42 ±

0.42 kb

long

(Fig. 3).

Bgll

cleavage siteswere

located2.47kbbeforeloopW'4 and1.85kbafter

loopX'4.

Turning

tothe restriction fragments located

on the clockwise side of

BglII

fragment 4,

ex-amples of the heteroduplexes assigned to

frag-ments 3 and 2a are shown in Fig. 2d and e,

respectively. The heteroduplex in Fig. 2d

con-tained the

readily

recognized

pattern

of

non-homologies between loops D' and G' and

con-tained a T4 strand with an average length of

14.12±0.13kb(Fig.3). From agarosegels,

BglII

fragment 3 was 14.01 kb

long.

Note that the

right endof the

heteroduplex

region

which

ad-joined

fragment

4was

fully duplex and,

accord-ingtoits 6.02-kbdistance from

loop G',

termi-nated before the H' deletion

loop,

which was

6.23kb from

loop

G'. This

supports

the earlier

assignmentsof

loop

H'toT2 DNA and the

BglII

cleavage site at the end of

fragment

4 to the

homologous

region

beside the H'

loop

rather

thantothe DNAwithin the H'

loop.

However,

our

assignment

of

loop

H'toT2 DNAcan

only

be tentative. The left end of the

heteroduplex

withBglII

fragment

3 is located 2.80 kb from

loopD'.

The

heteroduplex

in

Fig.

2e contained the

nonhomologies

C'

through

N4,

along

with an

unletteredsubstitution

loop

near

N4,

whichwas

on November 10, 2019 by guest

http://jvi.asm.org/

[image:5.494.104.393.73.450.2]

c)

T2/T4B/II fragments5and 7a

d) T2/T4 Bg/II fragment 3

_o_~~~~~~~~~~~~~~~~

_Ec

FIG. 2cand d.

e)

T2/T4

Bg/II

fragment 2a

FIG. 2e. 120

B/

on November 10, 2019 by guest

http://jvi.asm.org/

[image:6.494.114.399.43.678.2]

17.26+0.21(6)

1.28

t r, '8

JLt1. . ,058

, I

0.22 2.01

B' C'

0.48

0.18074:I0i3 o22 0.21 (2b) i n r- r-

1.1

r-'

Kiu l

R J' K' L'

17.97 t0.21 (4)

0.53

0.55 0.33 0.94 0.23

5.43 1'37M'1102 I,Q19

II I :I II %

Li L-I I L

0.10 0.33 0.69

02 N' 0'o

14.12 0.13 (5)

1.51

0.261 0.57 0.28

r , e

2.80 "11.12' 1.1 6.02

(3) r- r-- rJ

U- L.- L..-j Li

0.42 0.51 0.80

D' E' F' G'

L 12.36 0.09(6)

0.28 1.10 1.32 0.66 20.38

.44' 182' 45.32

(4 )

I\ -._

0.35 5.80 2.45

H2 I rIIH23B H2

2~~~~ ~ ~ ~ ~ ~~~~~04

*- 5.76 5.39 10.42 t0.42( 3) H

0.49

2.03 1.39 0.56 0.71

IIV

(70-7b-5)} H J

~~~

~~18.5

~

2.47 :,} 4.83 \ 6 8

'J v u v

0.80-0.94 0.12

U' W2

FIG. 3. MapsoftheT2/T4BglII fragment heteroduplexes. The heteroduplexesarenumbered accordingto

the T4BglII fragment which they contain. Distancesaregiven in kilobases foreach duplex and

single-strandedregion. Theaveragelength±standard deviationfor eachT4strand isgiven above eachmap,and

thenumber of heteroduplexes measured is given inparentheses. The two values given for the length of substitutionlooplFintheheteroduplexwithBglII fragment7arepresentthelengths of thetwoarmsof the

loopinfullT2/T4heteroduplexes.Heretheupper armisnotshown duetothelack of base pairingbeyond the

loopin theheteroduplexwiththeBglII fragment.

onlyoccasionally observed and isnotshown in

Fig. 3. Theaveragelengthof the T4strandwas

17.26 ±0.21kb, comparedwithavalue of17.53

kb forBglll fragment2ameasuredwithagarose

gels. The rightend of theheteroduplex region

terminated 580 bp beyond loop C'. This was

within60bpof thelocationassignedtothesame

BglII cleavage siteby measuringfromloopD'

in theheteroduplexwithBglII fragment3.The

leftend of theheteroduplexwithBglll fragment

2awaslocated1.28kb beforeloop N4,within 170

bp ofthelocationassignedtothisBglII cleavage

site by measuring from loop C" in

heterodu-plexes formed withpartof theneighboring BglII

fragment1 (datanotshown). Several

heterodu-plexes containing this end ofBgll fragment 1

were found, although no unambiguous

hetero-duplexes containingthe entire 56-kbBgll

frag-mentwereobserved.

Alignment of the T4 restriction and

ge-neticmaps.Altogether,theheteroduplexes

be-tween T2 DNA and the T4 BglII fragments

permitted 10 ofthe 13 Bgl cleavage sites on

the T4 alc mutant DNA to be located onthe

T2/T4 heteroduplexmap.Theexactlocation of

thelargerIIH23B deletionontheT2/T4

heter-oduplex map and withinBgJII fragment4 was

also defined.By using this deletionas apointof

0.74 1.28:L

I

0.72

A'

on November 10, 2019 by guest

http://jvi.asm.org/

[image:7.494.51.446.63.451.2]

reference and the BglII cleavage sites to

cor-rectly orient theT4BglII restrictionmapwith

respect to theT2/T4heteroduplexmap,thetwo

maps werealigned,asshown inFig.1.Thesmall

differences observed between the locations of thecleavage sitesontherestrictionmap(Fig. 1,

inner circle) and the locations of the cleavage sites on the heteroduplex map (Fig. 1, arrows

pointingtoward theoutercircle) reflect the fact

that the restriction map is a strictly physical

mapof the T4 genome, whereas the

heterodu-plex map was constructed by averaging the

lengthsof the substitution looparmsandhalving thelengths of thedeletion loops. This procedure

was necessary because not all

loops

could be

assigned to T2 orT4. Over

long distances,

the contributions fromT2 andT4average out,but

overshortdistances,unequalamountsof DNA attributable to T2 or T4 can cause the BglII sites on the heteroduplex map to be shifted relativetothephysical restrictionmap.

AsFig. 1shows,twoofthethreeBgllI

cleav-agesitesnot

directly

locatedby the heteroduplex analysis fallatthe ends offragments that

con-tain no nonhomology loops, i.e., fragments 8b

and 9. These two neighboring sites could be

locatedontheT2/T4heteroduplexmapby

mea-suring from the

BglII

cleavage site located

be-tween

fragments

5and9.The last

BglII

site falls

betweenfragments8aand6.Itslocationonthe heteroduplexmapcouldnotbe determined

ex-actly duetothepresence of unassigned deletion loops in both of these

fragments;

however, by measuring from the

BglII cleavage

site at the end offragment7a,it could be definedaslying

at one end or the other of a 330-bp interval,

which is adistance equal tothe unassigned S' deletion

loop

in

fragment

8a.

The final alignment ofthe BglII restriction mapwith the T4geneticmapis shown in Fig.4.

Also includedarethe cleavage sites forthefive

otherrestrictionenzymesmapped in the

accom-panying report (9). The alignment is based on

theprior orientation of the T2/T4 heteroduplex

map with the T4

genetic

map as described by

Kim and Davidson(7). The rII locus was used

astheprimary referencepoint for thealignment, with the start of the rIIH23B deletion being

placedattheleftend of the rIIA gene. TherIIA

gene has beenestimatedtocontain 2,300 bp by

electron microscopy ofheteroduplexes formed

withvarious rII deletion mutants (7). This is in

line with length

estimates

based on the

molec-ularweightof the rIIA protein (13).

DISCUSSION

BglII

cleavesthecytosine-containing DNA of

T4amE51 nd28

'rIIH23B

alc8

into 13fragments,

6of whichwehaveidentifiedinheteroduplexes with full-length T2 chromosomal strands. The ends of these

heteroduplexes

specifiedthe loca-tionsof10of theBglII

cleavage

sitesontheT2/

T4heteroduplexmap,providing the framework for (i) theassignment of the remaining 3BglII cleavage sitesonthe

heteroduplex

map,

(ii)

the identificationof the source of DNA infourmore

of the nonhomologyloopsontheT2/T4 heter-oduplex map, and (iii) the orientation of the restriction mapwith respect to the T4 genetic

map. The final

alignment

between the

restric-tion and

genetic

mapswasfixed

by

the location of the startof the

rIIH23B

deletionatthe end of the rIIAgene.

The accuracywith which the ends of the T4

DNA strands in the heteroduplexes with T2 could beidentified, and thus theaccuracywith which the

BglII

restriction sites could be

as-signedtothe

T2/T4 heteroduplex

map,canbe gauged by

examining

thosecaseswhere

hetero-duplexes

werefound forneighboring restriction

fragments (e.g.,

fragments

1and2a or

fragments

2a and 3). Inthese twocases, the independent assignment of the restriction sitesvariedby

only

60and170nucleotide

pairs.

This isinlinewith

thelimit ofresolution

expected

for

heteroduplex

mapping by electron

microscopy.

Because our alignment of the T4restriction

map with the

genetic

map is fixed at the rII

locus, itsaccuracy overthe

entirety

of themap

depends upon the extent to which the

genetic

map itself represents true

physical

distances along theDNA.

Actually,

thecorrelation should bequite good,asthe

genetic

map in

Fig.

3isnot a classical map based only on recombination frequencies, which canvaryfrom region to

re-gion within the genome. For many

loci,

the geneticmaprepresentsphysical distances deter-mined by

measuring

thefrequency with which

two

given

genes are carried together on the

incomplete

chromosomes found insmallT4

par-ticles (10, 11). The positions of the remaining

genes were

interpolated

from recombination

fre-quencies.

The distance inkilobasepairs between

the rII and lysozyme genes was measured

di-rectly byelectronmicroscopyofdeletion

heter-oduplexes (7). Thesetwoloci arenearly opposite

one another on the circular genetic map and

thus have served tofix distances along the T4

genetic map at two widely separated points.

Between these points some distortion can be

anticipated, but itshould not be too great.

The location ofthe single BamHI cleavage

site canbe taken as an indication of thedegree

with which the restriction and genetic maps

coincide outside the rII and lysozyme regions.

Our

alignment places

the BamHI site within

gene8, verynearitsjuncturewith gene 7. This

on November 10, 2019 by guest

http://jvi.asm.org/

\

P,,r-~~~~~~~A

oe~ ~ ~

FIG. 4. Restriction mapof T4 DNA forBamHI, BglII, KpnI, PvuI, SalI, and XbaI aligned with the T4

genetic map. Theunifiedmapof cleavagesitesforthese restriction endonucleaseswasconstructedbyMarsh

andHepburn (9). Thegeneticmapisanupdatedversionofthe mapofWood and Revel(18), withexpansion

ofthe distancebetween genes30and54in accordancewith the mapofMosig(11). Genes whoseexactlocations areunknownarebracketed.Shadedareasdenote theestimated sitesofgenes,basedonheteroduplexmapping

and the molecularweightsofpolypeptideproducts (18).Thealignmentofthe maps isfixed bythe boundaries

oftherIIH23B deletion carried bythe T4 alcmutantusedtoconstructthe restriction map. The scalejust

insidethegeneticmapshowsthe distance in kilobasepairsaround the map; the totallengthis166kb(9).

is in exact agreement with Wilson et al. (17), who located the BamHI site by digestion ofa

clonedEcoRIfragmentknown to carry genes 7

through9.

Hanggi andZachau (5) have mapped indetail

asegment of the T4 genome that extends from

gene 35throughgene 63. Ouraligrunent places

oneXbaI cleavagesitewithinthe end of gene 34

proximalto gene 35 and another in the middle

ofgene 63, in exact agreement with the locations

determined by Hanggi and Zachau (5). These

authors also identifiedaSaIIsite near theXbaI

site in gene 34, as we did, and showed that a

second

SailI

site lies within the

dihydrofolate

reductase

(frd)

gene. Toaccommodate this

as-signment,thefrdgene mustbemoved about 1

kb closertogene 32 than is shown on thegenetic

map inFig.3.Thisshouldrepresent the variance

that canbeexpectedbetween the restriction and

geneticmapsas

aligned

in

Fig.

3.

Usingdifferent

approaches

and different T4

mutants, twoother groups of workers have

in-dependently aligned restriction maps with the T4genetic map. Carlson

(2)

hasexamined the 123

on November 10, 2019 by guest

http://jvi.asm.org/

[image:9.494.44.444.72.459.2]

124 YEE AND MARSH

BglI, KpnI, andSailIcleavage patterns of

sev-eralT4deletionmutants todeterminea

general

alignment

between the maps. O'Farreil et al.

(12) have constructed a T4 restriction map which contains the

EcoRI, HindIII,

and PstI cleavage sites and have aligned it with the T4 genetic mapby hybridization of restriction

frag-ments toclonedsegmentsof the T4genomeand

by correlationof restriction

fragment

sizes with thepattern offragments

produced by digestion

of the cloned DNAs. In both cases, there was

agreement with the

alignment

which wefound

for therestriction and

genetic

maps.

ACKNOWLEDGMENTS

This investigation was supported by Public Health Service research grant GM 23608 from the National Institute of Gen-eral Medical Sciences and by Public Health Service Research Career Development Award CA 00490 to R.C.M. from the National Cancer Institute.

LITERATURE CMD

1. Beck, E.,R.Sommer,E.A.Auerswald,C.Kurz,B. Zink, C. Osterburg,H.Schaller,K.Sugimoto,H. Sugisaki, T. Okamoto, and M. Takanami. 1978. Nucleotide sequence of bacteriophage fd DNA. Nucleic Acids Res.5:4495-4503.

2. Carlson,K. 1980. Correlation betweengenetic map and map ofcleavage sites for sequence-specific endonucle-asesSall,KpnI,BgII,and BamHI inbacteriophageT4 cytosine-containing DNA. J. Virol. 36:1-17.

3. Carlson,K., and B. Nicolaisen.1979.Cleavagemap of bacteriophage T4 cytosine-containing DNA by se-quence-specificendonucleasesSalI and KpnI. J. Virol. 31:112-123.

4. Davis, R. W., M.Simon,and N. Davidson. 1971.

Elec-tron microscope heteroduplex methods for mapping regions of base sequence homology in nucleic acids. MethodsEnzymol. 21D:413-428.

5. Hanggi,U.J., andH. G. Zachau. 1980. Isolation and characterization of DNAfragments containing the di-hydrofolate-reductase gene of coliphage T4. Gene

9:271-285.

6. Kiko, H.,E.Niggemann,and W.Ruger. 1979. Physical mapping of the restriction fragments obtained from bacteriophage T4 dC-DNA with restriction endonucle-asesSmaI, KpnI and BglII. Mol. Gen. Genet. 172:303-312.

7. Kim,J.-S., and N. Davidson. 1974. Electron microscope heteroduplex study of sequence relations of T2,T4,and T6 bacteriophage DNAs. Virology 57:93-111. 8. Kutter, E., A. Beug, R. Sluss, L. Jensen, and D.

Bradley. 1975. The production of undegraded cytosine-containing DNA by bacteriophage T4 in the absence of dCTPase and endonucleases II and IV, and its effect on T4-directed protein synthesis. J.Mol. Biol. 99:591-607. 9. Marsh, R.C., and M. L. Hepburn. 1981. Map of restric-tion sites on bacteriophage T4 cytosine-containing DNAfor endonucleases BamHI, BgilI, KpnI, PvuI, Sall, and XbaI. J. Virol. 38:104-114.

10. Mosig, G. 1968. A map of distances along the DNA molecule ofphage T4. Genetics 59:137-151.

11. Mosig,G.1976.Linkage map and genes of bacteriophage T4, p. 664-676. InG. D. Fasman (ed.), Handbook of biochemistry and molecular biology, 3rd ed. Nucleic acids, vol. 2.CRC Press,Cleveland, Ohio.

12. O'Farrell, P., E.Kutter, and M. Nakanishi. 1980. A restriction mapof the bacteriophage T4 genome. Mol. Gen. Genet. 179:421-435.

13. O'Farrell,P.Z.,L. M.Gold,and W. M.Huang.1973. The identification ofprereplicative bacteriophage T4 proteins. J. Biol. Chem. 248:5499-5501.

14. Ruger, W., M. Neumann, U. Rohr, and E. Nigge-mann. 1979. Thecomplete maps of BgIII, Sall and XhoI restriction sitesonT4dC-DNA. Mol. Gen. Genet. 176:417-425.

15. Snyder,L.,L. Gold,and E. Kutter. 1976. Agene of bacteriophage T4 whose product prevents true late transcription on cytosine-containing T4 DNA. Proc. Natl. Acad. Sci. U.S.A. 73:3098-3102.

16. Sutciffe, J.G. 1979. Completenucleotide sequence of theEscherichia coliplasmid pBR322.ColdSpring Har-borSymp. Quant. Biol. 43:77-90.

17. Wilson, G. G., R.L.Neve, G. J.Edlin, and W. H. Konigsberg.1979.The BamHirestriction site in the bacteriophage T4 chromosome is located in or near gene 8.Genetics93:285-296.

18.Wood,W.B., and H. R. Revel. 1976. The genome of bacteriophage T4. Bacteriol. Rev. 40:847-868.

J. VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/