Temperature-sensitive mutants of foot-and-mouth disease virus with altered structural polypeptides. II. Comparison of recombination and biochemical maps.

0022-538X/80/04-0067/06$02.00/0

Temperature-Sensitive Mutants of Foot-and-Mouth Disease

Virus

with Altered

Structural Polypeptides

II.

Comparison of Recombination and Biochemical Maps

ANDREW M. Q. KING,* WILLIAM R. SLADE, JOHN W. I. NEWMAN,AND DAVID McCAHON GeneticsDepartment, Animal Virus Research Institute, Pirbright, Woking, GU24 ONF, England

The structuralpolypeptides offoot-and-mouthdisease virus were digested with

Staphylococcus aureus V8 protease in the presence of sodium dodecyl sulfate.

Theprotease-resistant peptidesderived from temperature-sensitive mutants were

compared with those ofthe wild typeby electrofocusing in a polyacrylamide gel.

Covariation between the charge shifts of different peptides indicated that they

shared common sequences: onlyfive independent peptides in all were derived

fromVP1, VP2, and VP3, accounting forapproximately 50% of the polypeptide

sequences. In twoinstances,aminoacid substitutions that caused similar shifts in

theisoelectricpointwerefound to belocatedin different peptides. However, 15

mutants that possessed identical

shifts

in VP2 could not be distinguished by

peptide

analysis.

Thepolypeptides ofrevertantsableto grow atthenonpermissive

temperature werecompared withthose of the parentalmutants. By this test, 6 of

the12distinguishable classesof coat protein mutations were found to covary with

temperaturesensitivity.Inadditionto true revertants,several phenotypic

revert-ants which possessed a second charge change, either in a

different

structural

polypeptide orina

different

region of the samepolypeptide,were isolated. The

orientation of the recombination map was deduced from the loci of the coat

protein mutations.

Picornaviruses provide a unique example of genetic recombination between RNA molecules. Much workwasdevotedtoquantitating

recom-bination frequencies measured in crosses

be-tween markers for temperature sensitivity or

inhibitor resistance. This work culminated in recombinationmapsforpoliovirus and foot-and-mouth disease virus (FMDV or aphthovirus) which were

approximately

additive and linear (4,12, 15).

The problem of

relating

agenefunctionto a

map locus is

that,

thus

far,

no

genetic

markerin

either the

poliovirus

ortheFMDVmaphasever

been correlated witha

physicochemical

charac-ter in the viral RNA or in a virus-coded

poly-peptide. In the caseofpoliovirus, considerable

progress was made

by

observing physiological

defects ofmutantsunderrestrictive conditions

(5). Most of the

map's

mutationswere

assigned

tothatregionof thegenome,

approximately

40%

of the

total,

that codes for structural

protein.

However,many membersofthisclassofmutant

were also defective for other

functions,

for

ex-ample, RNA

synthesis.

The FMDV map

appeared

to be similar to

that of

poliovirus

inbothrespects. Aselectionof

18of thetotal70

temperature-sensitive (ts)

mu-tants atvarious map lociwerefoundtobeeither

acid labileor

cystine

dependent (13;

McCahon,

unpublished observations). Since both

charac-tersindicate defects in structural protein (14),

this result implied either that the entire map

represents only the coat protein region of the

genome or that many of the ts mutants have

second mutations affecting the coat proteins.

Twenty-nine mutants were examined for viral

RNA and

antigen

syntheses at the restrictive

temperature. Differences were demonstrated

among mutants,but therewas noclear

correla-tion with the map locus (R. A. J. Priston,

un-published observations). These results expose two limitations of the physiological approach: (i) it isimpreciseinthat, for example, individual structural

polypeptides

cannotbedistinguished; (ii)

pleiotropic

effects make it difficultto map

nonstructuralfunctions.

Electrofocusing

themutantgene

products

of-fers asolutiontothe

problem

of

physically

lo-cating missense mutations. The accompanying

paper (9) describes the

application

of

electrofo-cusingtothestructural

polypeptides

of FMDV

and itsmutants. In thispaper we describe the

propertiesofrevertantsandthe

electrofocusing

ofpeptidefragmentstodifferentiate among

mu-tations that cause similar

charge

changes.

Twentytsmarkersof the FMDV recombination

mapareidentifiedasmissense mutations

affect-ing structural

polypeptides.

Since the

poly-67

on November 10, 2019 by guest

http://jvi.asm.org/

68 KING ET AL.

peptide-coding positions are known (6, 17), we

have therebydetermined the orientation of the

map.

MATERIALS AND METHODS

The origin of the FMDV mutants was described

previously (11-13).tsmutantsnumberedless than 100

wereinduced, and those numberedmorethan 100and

ts 03 were spontaneous. Growth, purification, and

trypsintreatmentof FMDV were asdescribed inthe

precedingpaper(9).

Peptide fingerprints. Virus pellets containing 50

jigofproteinweredisrupted by heatingto 100°C for

1min in50[L of 1% sodiumdodecyl sulfate (SDS)-2%

(vol/vol) 2-mercaptoethanol-0.01 M Tris-10% (vol/

vol)glycerol. Ten microliters of 1-mg/ml

Staphylococ-cusaureus V8protease (Miles Laboratories, Inc.) in

0.01 MTris-hydrochloride (pH 7.4)-i mM EDTAwas

added, andthemixturewasincubated for1h at37°C

before the reactionwasstopped byheatingto 100°C

for1 min. Half the preparationwasloadedpergel.

Electrofocusing. Electrofocusingwas asdescribed

for method (ii) of reference 9, exceptthe

electropho-resis was continued for the times indicated in the

figurelegends.

Determination of molecular weight. Molecular

weights ofpeptides were determined from a plot of

electrophoretic mobility in SDS-polyacrylamide gel

versus thelogarithm ofmolecularweight. Standards

usedwerethe intact formsofVP1,VP2 andVP3(each

with a molecular weight of 30,000), soybean trypsin

inhibitor (molecular weight, 22,500), myoglobin

(mo-lecular weight, 16,400), cytochrome c (molecular

weight, 12,400), and insulin (molecular weights, 2,600

and3,200). Amixtureof[35S]methionine-labeled and

unlabeled FMDV polypeptides was fingerprinted as

described above, and the gel was stained. Peptide

bands were cut out of the gel and washed twice in

waterfor10min and then in 0.125 M

Tris-hydrochlo-ride (pH 7.4)-2% 2-mercaptoethanol-0.1% SDS-10%

glycerol for 30 min. Each slice was loaded, with

markers in thesamesolvent,ontoan

SDS-polyacryl-amidegel (85 by6mm) containing 17.15%acrylamide

and0.35% N,N'-methylenebisacrylamide and

electro-phoresed with the Laemmli (10) discontinuous buffer

systemfor 6.5hat40V.TheSDS-polyacrylamide gels

werestained, slicedlongitudinally, dried, and

autora-diographed todetermine thepositions of the labeled

peptides.

Isolation ofrevertants. Two methodsfor isolat-ingrevertantswereused. In the first,whichwasused

withts 12, 16, 58, 107, and 109, testtube cultures of

BHKcellswereinfectedat37°C for30minandthen

washed with acid buffer, 0.1 M NaPO4 (pH6.2), to

destroy inoculum virus. After fresh medium was

added, the cultures were incubated at 37°C for a

further 75 min, followed by 4 h at 410C. The virus

yields were plaque assayed at 41°C. Plaques were

pickedaspresumptiverevertantsand clonedat410C.

Nomore thanonerevertantwasisolatedperculture.

The second method, which was used for the other

mutants listed in Table 1, except for ts 3, was a

modification ofthe procedureof Lake and Mackenzie

(11) for isolatingmutants.Each mutant wasassayed

by the agar-cell suspension technique(3), and plaques

were allowed to form at37°C for 16 to 20 h. An agar

layercontainingtetrazolium stainwasthen added to

theplates, and the incubationwascontinued for 20 to

24 h at41°C. Plaqueswithclearlydefined halos were

picked and cloned at 41°C. Seed stocks were prepared

in small monolayer cultures at 41°C. Working stocks

were prepared by a single passage of the seed stocks inmonolayercultures at37°C.

RESULTS

Location of amino acid substitutions. Ta-ble 1 lists the 24 ts mutants of FMDV (strain 0

Pacheco) thatpossessedvariant structural

poly-peptides. All70 ts mutants of the recombination

map were screenedbyelectrofocusing forshifts

in the isoelectric points ofVP2, VP3, VP4 and

themajor fragmentofVP1, VP1L,producedby

trypsintreatmentof the virus(1).For 29 of these

mutants, thesearch was extended to intactVP1,

but no additional variants were found. Also

screened werethreets mutants not onthemap,

oneofwhich, ts3, was found to bevariant; its

properties arealso included in Table 1. Both of

the mutants that carried the guanidine

resist-ance (gs+) marker, tsgs+ 03and tsgs+ E18, had

normal structural polypeptides. The

accompa-nying paper (9) describes all of the classes of

charge shifts that could bedistinguishedonthe

TABLE 1. Summaryofelectrofocusingmutations

observedinFMDVstructuralpolypeptides

Score of revert-Location ants

Class of amino with

of mu- Mutant acid with

tation substitu-

wild-tion' type

poly-peptide

a ts 12 VPl/b 0/3

b ts 16 VP1l3b 0/3

c ts 40 VP1a 2/4

d ts 58 VP1a 0/3

e ts103 VP1a 1/4

f ts 109 VP1a 0/3

g ts 3 VP2-yb ND'

h ts65, 63, 33, 58, 62, VP2,8 11/16d

47, 59, 37, 42, 23,

61, 66, 44, 41, 67

ts 22 VP2,8 0/8

ts 109 VP2a 3/3

k ts 28 VP3a 1/7'

1 ts 107 VP3a 2/3

aa,

/,

and

-y

specifypeptides producedby the action

ofS. aureus V8protease in the presence of SDS.

bThesesubstitutions arelocated in polypeptide

seg-mentsthat areabsent from peptidefingerprints.

'ND, Not determined.

dTotal for this class of mutation. Inaddition, four

suppression reversions were found that affected the

samepolypeptide.

'Also foursuppression reversionsaffecting VP 1.

J. VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/

FMDV RECOMBINATION MAP: COAT PROTEIN MUTATIONS 69

basis of magnitudeordirection. In several cases,

different mutants exhibited identical charge

shifts. Thus, the variant VP2 of ts 109 was

indistinguishable fromthatof ts 22, the

differ-encebetween the VP1Lof ts 16 and that of ts

109wasuncertain,and,most notably, there were 15 identical cases of altered VP2, designated

"class h" in Table1.

To discriminate among as many different

classes of mutationalevents aspossible, the

var-iant polypeptides were subjected to a simple form ofpeptide fingerprinting,adapted from the

method of Cleveland et al. (2). Virus samples were denatured with SDS and, before

electro-focusing, were

digested

withS. aureus V8

pro-tease. The resulting fingerprints are shown in

Fig.1.

Atotal ofseven prominentand reproducible

peptideswerederived from VP1,VP2, and VP3.

Theorigin ofeach peptide could be inferredby aninspectionofmutantfingerprints. For

exam-ple, only one peptide, VPla, is derived from VP1, and it is clear which peptide when the fingerprints ofts 103 (gel

f)

and wild-type ts+

(gel

b)are

compared.

Thistechnique enabledus to distinguish those VP1 mutants that are lo-catedwithin this peptide (ts40, 58, 103,and109)

from those thatare not(ts 12and 16). Figure 1

also shows that the

6,000-dalton

VPla is

con-tainedentirely within VP1L.

Three

peptides

are derived from VP2: thea

peptide

isaltered in thecaseofts' 109

(Fig.

lm),

whereas the other two bands are shifted

coor-dinately ints 22(gel k) and classh mutants (gel

i). Thesetwohomologouspeptides are therefore

designated VP2,B. Thus, there are three

distin-guishable locations foraminoacidsubstitutions

in VP2: the a and,B peptides and segments of

VP2 that are not represented at all in the

fin-gerprintsof Fig. 1. ts 3 (gelh) and a suppressor

ofts 37 (seebelow)arein the third category.

ts 28

(gel

n) and 107(gel o), thetwomutants

altered inVP3, have amino acid substitutionsin

the same (a) pair of homologous peptides,

al-though the shifts were of differing magnitude.

The moreacidic(lower) of the pair co-runs with

the lower of thetwo

VP2f8

bands. Aprominent

peptide toward the low-pH end of thegradient was also assigned to VP3: in Fig. 1 it is

desig-nated

VP3,8.

None of the 0 Pacheco mutants

exhibited an alteredVP3,B; however, VP3

mu-tantsoftwootherserotype0-strains of FMDV

were found to have a variant peptide at this

isoelectricpoint (datanotshown).

Molecular weights were determined for the

peptides derived from VP1, VP2, and VP3 by electrophoresis on an SDS-polyacrylamide gel (Table 2). The seven peptides leave much of VP2,VP3,and,especially,VP1unaccounted for.

Electrofocusing the SDS-peptide complexes

to-ward

the

cathode afterpreparing the sampleas

described by O'Farrell et al. (16) resulted in fingerprints thatweresimilartothose of Fig.2,

but inverted (data not shown). However, no

I-

*.4,

2-' 1 a Jr doa * . .j3

*

|!_

_ *m _- _ dW*

3.6- -w .T.

*1 -a

. ,., I r w A .

a b c de hg h

j

* / m n o

FIG. 1. Peptide fingerprints ofFMDV mutantswith altered structuralpolypeptides. Virus polypeptides

weredisruptedin SDS andelectrofocusedtoward theanodeforIh at 200V, followedby 6.5 h at400V, after

either(a)notreatmentor(btoo)digestionwith S.aureusV8protease.(aand b)ts+;(c)

ts'

previously treated

withtrypsinto cleave VP1 toVP1L;(d)ts12;(e)ts16;(f)ts103; (g) ts 40; (h) ts 3; (i) ts 37;(j)ts37R50; (k) ts

22; (1)ts58; (m)ts109; (n)ts107;(o)ts 28.

A.

3p-

4-* - -9 VOL. 34,1980

4

on November 10, 2019 by guest

http://jvi.asm.org/

[image:3.508.105.400.410.617.2]

[image:4.508.59.248.94.178.2]

70 KING ET AL.

TABLE 2. Molecularweights

of

peptides generated

byproteolysis in SDS solution

Peptide Mol wt

VP1a 6,000

VP2a 10,000

VP2,83 (upper)' 10,000

VP2,8 (lower) 10,000

VP33a (upper) 11,000

VP3a (lower) 16,000

VP3,B 6,000

Upper (mostbasic)of thetwoVP2,Bbands inFig.

1.

additionalpeptideswerefound.Furthermore,no

diffusible peptides were revealed by rapidly

staining gels immediately after electrofocusing.

The smallpolypeptide, VP4,wascomparatively

resistant to proteolysis under the conditions

used forfingerprinting.

Peptide fingerprints show that the variant

formsof VP2 in ts22and109,although

indistin-guishable in the intact form, have mutations

affectingdifferentpeptides.Hence, the two

mu-tations cannot be sistersand arelistedseparately

inTable 1.Similarly,the VP1 mutations ofts 16

and 109 wereshowntoaffect differentregionsof

VP1. However, fingerprints failed to

discrimi-nate among any of the class h mutations of

Table 1: all 15 exhibited identical shiftsin the

same 10,000-dalton peptides derived from VP2. Twoexamples,ts 37and58, are shown inFig. 1.

Thus, the 26 coat protein mutations were

re-solved into 12 classes (Table 1, a through 1),

distinguishable by electrofocusing the variant

polypeptides and theirfragments.

Revertants. Revertants were selected for

their abilitytogrow at thenonpermissive

tem-perature.Electrofocusingwasthenperformedto

determine whether the variantpolypeptidehad

alsoreverted.Table 1shows that for thedouble

mutants ts 58 and 109 it is the VP2 alteration

that covaries with temperature sensitivity and

not the alteration in VP1. No covariation was

observedfor three tsmutants, ts12, 16,and 22,

althoughaminimum of three revertants of each

was studied. The mutation within ts 3 had not

beenmapped and was not tested. Theremaining

classes oftsmutation inTable 1 each scored at

least one instance ofreversion in their variant

polypeptide. Five examples only from among the

class h mutationswere tested for covariant

re-version. They all scored positive, and it is

as-sumed that all 15members of this class covary

with temperature sensitivity. On the basis of

theseresults, 20 of the 26 electrofocusing

muta-tionslisted in Table 1 were assigned to the coat

protein region of the genome. These 20

muta-tions fell into six physically distinguishable

classes (Table 1, c, e, h, j, k, and 1).

This test does not distinguish between a true

reversion and a suppression reversion that

hap-pens to restore thewild-type chargeby a

substi-tution elsewhere in thepolypeptide.The

distinc-tion is not important since bothprovide positive

evidence for covariation. Such compensating

charge shifts might be revealedbyfingerprinting

the revertant polypeptide. However, when this

was done (e.g., revertants of class h and k

mu-tations), no hidden suppressors were found (data

notshown).

It is of interest that several suppression

re-vertants were found among those rere-vertants in

which the wild-type charge on the polypeptide

was not fully restored. The suppressors were of

twotypes, examplesof which areshown inFig.

2. ts37R50possessedamutation thatsuppressed

ts 37and that compensated only partlyfor the

chargeshift in VP2.Fingerprintsof ts 37 and ts

37R50 were identical (Fig. 1), showingthat the

ts mutation and its suppressor affect different

regions of thepolypeptide. For the VP3variant,

ts 28, suppression revertants that possessed a

charge shiftin another polypeptide, VP1, were

isolated (Fig. 2).

DISCUSSION

Anovel method ofpeptidefingerprintingwas

used to distinguishamong similar

electrofocus-ing mutations. The technique was also useful for

identifying suppression reversions and for

set-tingalimit of 0.3kilobase on thephysical

sepa-ration of the 15 class h mutations. A

determi-

4.-FIG. 2. Suppressionreversions. Virus was treated

withtrypsin, and after disruption with SDS the

poly-peptides were electrofocused toward the anode for 1

hat200V,followed by 5 h at 400 V. (a) ts+;(b)ts28;

(c)ts28and ts+,co-electrofocused; (d) ts28R30;(e)ts

28R30 and ts+;

(t)

ts37; (g) ts 37 and ts+;(h) ts 37R50;

(i)ts37R50 and ts+.

J. VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/

[image:4.508.286.423.421.600.2]

FMDV RECOMBINATION MAP: COAT PROTEIN MUTATIONS

nation of the position of each peptide in the

polypeptide sequence would open the way to

"proteasecleavagemapping" of missense

muta-tions by electrofocusing. However, it should be

noted that approximatelyhalf of the

informa-tionalcontent was lost by digestion tofragments

thatwere too small to detect or too acidic or

basictofocus.

It might be expected that phenotypic

rever-sions, inwhich a second mutationcompensates for thets defect, would predominate over true

reversions (8). Hence, a lack ofcovariation

be-tween tsandelectrofocusingmutations does not

prove that the mutations are independent,

whereas a single instance of covariation was

taken as sufficient evidence that the two

char-acters werecaused by a single mutation.

For several phenotypic revertants,

suppres-sion mutationswererevealedas asecond charge

change in one of the structural polypeptides. One class ofsuppressors that altered the a

pep-tide of VP1 suppressed the ts defect of ts 28,

located in VP3a. This suggeststhat these two

regions of the polypeptide arefunctionally

as-sociated.

Fivecharge changemutationslisted in Table

1 failed to covary with temperature sensitivity in reversion studies and are therefore omitted

from the discussion of the recombinationmap.

Infour of these cases, thereare additional rea-sons for believing that their alterations are

causedby second mutations, independentof the

tscharacter. ts 58and109each havetwovariant

polypeptides, only one of which covaried with

the tslesion. Preliminary recombinationstudies

show thattsandcharge change markers ofts16

and 22 canbe separated. For example, ts+

re-combinants, produced by crossing ts 22 with

viruses having mutations that map to its left,

invariablypossessthe alteredVP2ofts 22 (data notshown). The statusofonemutant, ts 12, is

stilluncertain, but its inclusion wouldnotaffect theinterpretationwhichfollows.

Comparison

of biochemical and

recom-binationmaps. FMDV RNA hasatotallength of8.0kilobases. All of theknown geneproducts

P20VP VP2

VPt3

VP1 __

are encoded along a 7.5-kilobase sequence

bounded

on oneside

by

a

polycytidylic

acidtract near the 5' end of the virus RNA and on the

other side by the 3'-terminal polyadenylic acid

tract (6, 7, 17). Reversion studiesidentified20 ts

mutations that covariedwithalterationsinVP1,

VP2, or VP3(Table1). These threepolypeptides

are encoded together near the 5' end of the

genome, as shown in Fig. 3. Within this set of

mutations, there is noapparentcorrelation

be-tweenindividualmaploci and

polypeptide-cod-ingregions. However,thedistribution ofthe 20

chargechange lociis

significantly

nonrandomin

that they all lieat, or tothe left of,a recombi-nationfrequencyof1.25%,nearthecenterof the

map. It follows that the recombination map

representsthe genome oriented with its 5'

ter-minusontheleft.

The

recombination

data from which themaps

ofboth FMDV and

poliovirus

wereconstructed

arenotoriously subjecttovariation. Ofthe 70ts

mutations of the FMDV map, five loci at the

extreme left, includingtwo charge change

mu-tants, ts 28and65, were

singled

out

by

McCahon

et al. (15) as being

particularly speculative.

If theseare

eliminated,

a

relationsbip

betweenthe

recombination frequency and the

genomic

dis-tance becomes clear.

Figure

3 shows the map

lociof the five different classes ofchargechange mutation. The h

class,

represented by

mutants

withidentical substitutions in the

fB

peptide of VP2, is

designated

hin

Fig.

3.gs+ and ts

03

are

standard reference

points

of the map (15). h specifiesa

position

between0.8and1.6kilobases on the biochemical map, and since it was

mapped

independently

14times, itslocus is

ac-curately

known. The standard deviation ofthe h

loci,

+0.2% recombination

frequency,

is a

measureof themapping

precision

forindividual

markersinthecoat

protein region.

Figure 3 shows a

possible

alignment of the

biochemical

and recombination maps. Recom-bination frequency values coincide with

poly-peptide-coding regions

for all of the mutants

except ts103, whichlies

approximately

one

stan-darddeviation from the

boundary

ofthe

VP1-P52

S GENOME DISTANCE (kb)

0 1 2 3 4 5 6 7

I I 1 I

'.5 I I

1.0 0.5

% RECOMBINATION FREQUENCY

3'

I

t

03

FIG. 3. Tentativealignmentoftherecombinationmap(lower scale)with the biochemicalmap(upper scale) oftheFMDV genome. Genetic markersareplottedatthelocipreviouslydeterminedbyMcCahon et al. (15).

h isplottedatthemeanlocusoftheclass hmutations(seetext).kb,Kilobase.

71

VOL. 34, 1980

IL

K

I'A .7

1

Ile

on November 10, 2019 by guest

http://jvi.asm.org/

[image:5.508.121.407.553.636.2]

72 KING ET AL.

coding region. Since ourinformation islimited toagroup of proteins encoded at one end of the genome, it is notpossible toestimate precisely

howmuchofthe genomeisrepresentedby the

recombination map.

Ourphysical interpretation of the FMDV re-combination map shown in Fig. 3 must be

re-gardedastentative. However,itprovides a use-ful theoretical basis for future recombination studies, which we hope will lead to a better

description of genetic recombination in RNAat

themolecular level.

LITERATURE CITED

1. Burroughs,J.N.,D. J.Rowlands,D. V.Sangar, P.

Talbot, and F. Brown. 1971. Further evidence for

multiple proteins in the foot-and-mouth disease virus

particle.J. Gen. Virol. 13:73-84.

2. Cleveland, D.W., S. G.Fischer, M. W. Kirschner,

and U. K. Laemmli.1977.Peptide mapping bylimited proteolysis in sodiumdodecylsulfate andanalysis by

gelelectrophoresis.J. Biol. Chem.252:1102-1106.

3. Cooper, P. D. 1961. Animproved agar-cell suspension

plaque assay forpoliovirus:somefactorsaffecting effi-ciencyofplating. Virology13:153-157.

4. Cooper, P. D. 1968. Ageneticmap ofpoliovirus

temper-ature-sensitivemutants.Virology35:584-596.

5. Copper,P. D. 1977. Genetics ofpicomaviruses,p.

133-207.In H. Fraenkel-Conratand R. R.Wagner (ed.),

Comprehensive virology, vol. 9. Plenum Publishing

Corp., NewYork.New York: Plenum Press.

6. Doel, T. R.,D. V. Sangar,D. J. Rowlands, and F.

Brown. 1978.Areappraisalof thebiochemical map of

foot-and-mouthdisease virus RNA. J. Gen. Virol.41:

395-404.

7. Harris,T.J.R.,and F. Brown.1976.Thelocationof

thepoly(C) tract in the RNA offoot-and-mouth disease virus. J. Gen. Virol.33:493-501.

8.Jarvik, J., and D. Botstein. 1975. Conditional-lethal mutations that suppress genetic defects in morphogen-esis byaltering structural proteins. Proc. Natl. Acad. Sci. U.S.A. 72:2738-2742.

9. King, A.M.Q.,and J. W. I. Newman. 1980. Tempera-ture-sensitive mutants offoot-and-mouth disease virus withaltered structural polypeptides.I.Identification by electrofocusing. J. Virol. 34:59-66.

10. Laemmli, U. K. 1970. Cleavage of structural proteins

during theassemblyofthehead of bacteriophage T4. Nature(London)227:680-685.

11. Lake,J., and J. S. Mackenzie. 1973. Improved

tech-niquesfor the isolation of temperature-sensitive mu-tantsoffoot-and-mouth disease virus. J. Gen. Virol. 12: 665-668.

12. Lake,J.R.,R. A. J.Priston,and W. R.Slade.1975. A geneticrecombination map of foot-and-mouth disease virus. J. Gen. Virol.27:355-367.

13. Mackenzie, J. S., W. R. Slade, J. Lake, R. A. J.

Priston, J. Bisby,S. Laing, and J. Newman. 1975. Temperature-sensitive mutants of foot-and-mouth dis-easevirus: theisolation of mutants and observations on their properties and genetic recombination. J. Gen. Virol.27:61-70.

14.McCahon, D.,and P. D.Cooper.1970.Identification of

poliovirus temperature-sensitive mutants having

de-fects in virusstructural protein. J. Gen. Virol. 6:51-62.

15.McCahon, D.,W. R.Slade,R. A. J.Priston,and J. R.

Lake. 1977. Anextendedgenetic recombination map

forfoot-and-mouth disease virus. J. Gen. Virol. 35:555-565.

16. O'Farrell,P.Z.,H. M.Goodman,and P. H. O'Farrell.

1977. Highresolutiontwo-dimensionalelectrophoresis ofbasic aswell asacidic proteins.Cell 12:1133-1142.

17.Sangar,D.V.,D. N.Black,D.J.Rowlands, andF.

Brown. 1977. Biochemical mapping of the foot-and-mouth disease virus genome. J.Gen. Virol. 35:281-297. J. VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/