Alkaline DNase activity in cells infected with a temperature-sensitive mutant of herpes simplex virus type 2.

(1)

JOURNAL OF VIROLOGY, May1978,p.209-213 Vol. 26, No.2 0022-538X/78/00264209.00/0

Copyright01978 AmericanSocietyforMicrobiology PrintedinU.S.A.

Alkaline

DNase

Activity in Cells

Infected with a

Temperature-Sensitive

Mutant

of Herpes

Simplex

Virus Type 2

B. FRANCKE,1 H. MOSS,2M.C. TIMBURY,2ANDJ. HAY2*

TumorVirologyLaboratory,The Salk InstituteforBiological Studies,SanDiego,California92112,1and The Institute of Virology, University of Glasgow, Glasgow, ScotlandG115JR2

Received forpublication 18 October1977

BHK cells infected with the temperature-sensitive mutant ts13 of herpes

simplex virustype 2atanonpermissivetemperature lackthe alkaline nuclease

activity,which isinducedby themutant at apermissivetemperatureand by

wild-typevirusateither temperature. For ts13, enzyme activity could be induced by

atemperatureshifttopermissveconditions,butnotin thepresenceof

cyclohex-imide. Afterashiftfrompermissivetononpermissive conditions inthepresence

ofcycloheximide,theactivitywasstable inwild-type,butnotinmutant-infected,

cells.Afterextensivepurification, thewild-typenucleasewasfourfoldmoreheat

stableinthe presenceofsubstrate thanwasthemutantenzyme.Mixturesofboth

purified enzymes showed the predicted intermediate stabilities. The results

strongly suggest that theenzymeis virus codedand that themutant possessesa

lesioninthestructuralgeneof theenzyme.

Afterinfectionof cells in culture, herpes sim- gle medium with 10% (vol/vol) calfserum (Colorado

plex virus (HSV) inducesanumber ofenzymatic Serum Co.) and 10% (vol/vol) tryptose phosphate

activities related to DNA metabolism. Among broth (DifcoLaboratories)on5-cmplastic petridishes.

themare a new DNApolymerase andanuclease The virus used was HSV-2, strain HG 52 (wild

(5). Apartial purification of the nucleolytic ac- type), and a plaque-isolated temperature-sensitive

tivity (7) inclicated that itwas aDNA-specific mutant,

tsl3,

obtainedfroma

bromodeoxyuridine-mu-alkaline exonuclease andthatitsinduction and Virus stocks were prepared by infection of BHK

properties differentiated it from known cellular cellsat a multiplicity of10-2 PFU/cell and titratedon

nucleases. Preliminary experiments (4;H.Moss monolayers of BHK cellsunder an agarose overlay.

and J. Hay, unpublished data) showed that a Forexperimentalpurposes,BHKcellsin 5-cmpetri

DNase with properties very similar to those disheswereinfected with virus at 5PFU/cell,withan

reportedbyMonison and Keir(7)forHSVtype adsorption period of 1 h at

310C

or30

min

at390C,

1 was induced in cells infected with

wild-type

after whichtime5ml of fresh mediumwasadded. For

HSV type 2 (HSV-2) and

that,

of 13 HSV-2 temperature-shift experiments, the medium was

re-.meauesstvmuatstudied

(9), only

placed

with

5mlofmedium

prewarmed

tothe

respec-temperature-sensitive

mutants

ctyat

a

nonpemy

tivetemperature.Forexperiments involvinginhibition 1,

tsl3,

failed to induce theactivityatanonper- of protein synthesis by cycloheximide, the drug was missive temperature. A largeproportion ofthe present at 50yg/ml.At this concentration, the

incor-temperature-sensitivemutantsisolated for HSV poration of[3S]methionineby infectedcells was

re-(1, 8) exhibit a DNA-negative phenotype. The ducedto<2% of that of untreatedcontrol cultures.

ts13mutantof HSV-2 iscapableofsynthesizing Alkaline exonuclease assays from crude

ex-viralDNA at anonpernissive temperature (2); tracts.Extracts were prepared from 107cellsas

de-yet the amounts of DNA made vary consider- scribed by

Morison

and Keir (7) by

sonic

treatment

ably and aregenerally lowerthan those forwild- inanultrasonic water bath(Bransonic), yielding250

typinfectionsJ. Hyad , ,Francke,

unpub-

1l

ofcrude extract.The assay mixture (150

ul)

con-type

infections (J. Hay and B.

ted

hes

un-

tained

the

following

components (final

concentra-lisheddata). Thestudypresentedherewasun- tions): 50 mM

Tris-hydrochloride

(pH 9.0), 2 mM

dertakentodeterminewhetherts13 represents MgCl2, 10 mM 2-mercaptoethanol, 125 gAg of

3H-la-the structural gene for the viral alkaline nu- beled BHK DNA (1.6 x 103cpm/yg) per ml, and50

clease,as abasis forfurtherinvestigation of the iLd of crudeextract. Degradationof thelabeled DNA

potentialrole of thisenzyme duringviralrepli- wasmonitored byprecipitating theundigestedDNA

cation. with 5% trichloroacetic acid and counting the

precipi-tateaftercollectionontoglass-fiberfilters(Whatman MATERIALS AND METHODS GF/C) in toluene containing 5 g of_{per liter} _or _{by counting} _the _{soluble nucleotides}2,5-diphenyloxazole_as

Cellsand virus. BHK-21(C13)cels (MacPherson described by Morrison and Keir (7). Both methods and Stoker[6])weregrown inDulbecco-modified Ea- gaveidentical results. For each nuclease

determina-209

on November 10, 2019 by guest

http://jvi.asm.org/

(2)

210 FRANCKE ET AL. J. VIROL

tion, atime course with atleast three time points, z sampled induplicate, was carried out. The results are L presentedasmicrogramsof DNAdigestedin1minby 3

1 mgofprotein.

Purificationof alkalinenuclease.A total of 10 8 2_ cellswereinfected with wildtypeorts13for18 h at Z

320C. Cytoplasmicextracts werepreparedafterscrap- , _

ing the cellsinto the medium andpelletingthemat

1,200 x g. The pellets were washed in 0.15 M N 4

NaCI-0.02 MTris-hydrochloride(pH 7.5)at4°Cand X 3

suspended in 3 volumes of 10 mMTris-hydrochloride ia

(pH8.0)-i mMMgCl2-1 mMdithiothreitol. Nonidet 2; P-40wasaddedto0.2%(vol/vol), andthe wholewas,

homogenized with 50 strokes in a Dounce homoge- x

nizer. Aftercentrifugation for 30min at30,000rpm

05

6.0 6.5 70 75 no

8.s

90

and 40C in a Beckman 6OTirotor, the supernatant pH

was used as starfing material for purification. The

extract waschromatographedoncolumns of DEAE- FIG. 1. Effect of pH on the nuclease activity in Sephadex(PharmaciaFineChemicals, Inc.)andphos- extractsfrom tsl3ormock-infectedBHK cells. Ex-phocellulose (Whatman)andcentrifugedon aglycerol tractswerepreparedandassayed fornuclease activ-gradienttogivebetween 1-and>2,000-fold purifica- ityasdescribed in thetext, exceptthatTris-chloride tion with 20% (ts13)to30% (wildtype) recoveryof buffer was replaced with 50 mMN-2-hydroxyethyl

activity.Themethod ofpurificationand theproperties piperazine-N'-2-ethanesulfonic acidadjustedtothe of thepurifiedenzymewillbepublished elsewherein pH indicated with KOH andbroughttoequimolar

detail(Franckeetal.,manuscriptinpreparation). K+withKCI. Symbols: E uninfectedcell extract;0, Heatinactivation of thepurified nuclease. The ts13-infectedcellextract(31°C infectionfor18h);0, preincubation mixture contained thefoliowing:500 ,ug ts13-infectedcellextract(39°Cinfection for9h).

ofbovine serum albumin perml, 10 mM

,B-mercapto-ethanol, 1 mMEDTA, 10 mM Tris (pH 7.5, unless perature sensitive duringthe in vitroassay, at otherwiseindicated), 625 ,ug of BHK DNA (if present) leastup to 1h. Therate atwhichthesubstrate perml, and enough purified enzyme to degrade 6,g of DNA was degradedin vitro at390C wasabout DNA in 1 h at320C.Preincubation wascarried out at twice as fast as at

31°C,

and extracts fromcells

450C and was terminated bychillingthe mixtureto infected with tsl3 at

310C

did not differ

from

40C.To assay forresidualactivity, the niixture was

w*ld-type-infected

cell extracts in this respect

dilutedfivefold, restoring theconditionsofthestan- (

dard nucleaseassay described above. Since the protein (data notshown). Theapparent lack of mvitro contentof thepurified enzymeswasnotmeasurable, temperature sensitivity ofthe tsl3-induced ac-these results arepresented as the percentage of resid- tivitywill be discussed below.

ualactivity after preincubation for the times indicated, In contrast, when tsl3-infected cells were

with theactivity ofasamplenotexposed to 450C as shiftedto anonpermissivetemperatureinvivo,

100%. Ifkept at 40C, the preincubation mixture had the alkaline nuclease activity behaved in a tem-no effect on theenzymaticactivity, a control that was perature-sensitive fashion (Fig. 2). Both wild performed with each of the inactivation experiments type and

tsM3

inducedsimilaramounts of

activ-describedbelow. ity at

310C.

A shift to

390C

hadlittleor noeffect

onwildtype,but resulted inareduction of the

RESULTS activity in

ts13-infected

cells(Fig. 2a and c).The

One of thedistinguishingfeatures ofthe HSV- kinetics of theturnoff of the nucleasewas

vari-induced nuclease is its high pH optimum (7). able,dependingmostmarkedlyonthetiming of

When either uninfected or tsl3-infected cells theshiftto390C(see alsoFig. 3). Forareduction

wereassayedfornuclease activitybetweenpH to 50% of thepreshift level, between 30and 90

5.5 and 9.0, the results shown in Fig. 1 were min at390Cwasrequired, whereas after 3 h the obtained. The crude extracts used in this exper- activityreached backgroundlevels in allcases. iment showactivity at acidpHand atalkaline Inthereverseexperiment,shifting to a

permis-pH,yetlittleactivityatpH7.The acidnuclease sive temperature after 9 h at310C, tsl3-infected

appearstobeof cellularorigin,since it isfound cells were capable of expressing the nuclease.

inuninfectedcells,whereas thealkaline nuclease Thekineticsof the turn-onexhibited no

notice-appears to be virus induced. The absence of the ablelag period. Again, ashiftfrom 39 to 310C

latter from cells infected with Msl3 at anonper- had little effect on the activity present in

wild-missive temperature (390C) is unique for this type-infected cells (Fig. 2b and d). To test mutant.All other mutants of HSV-2expressed whethertheappearance ofalkalinenuclease in

somealkaline nuclease after infection at

390C

ts13-infected cellsafter a shift to a permissive

(Moss and Hay, unpublished data). Figure 1 temperature wasduetode novosynthesis or to

showsthat theenzymaticactivitywasnot tem- theactivation of an inactive enzyme, the

http://jvi.asm.org/

[image:2.509.270.458.77.223.2]

(3)

VOL 26,1978 DNase ACTlVITY IN CELLS INFECTED WITH HSV-2 211

imentshown in Fig. 3 was performed. ts13-in- ifcycloheximidewaspresent,indicatngthatits

fected cells,lackingtheactivitydue to a shift to appearancerequired the

synthesis

of new

pro-39°C for 3 h, were capable ofproducing the tein and that the preexisting activity became

nuclease againat31°C at anytimeduringthe irreversibly inactivated during the period at

replicative cycle. At later times,induction was 39°C.The accelerated appearance of nuclease in

pooreranddelayed,probablyreflectingthestate celLs

kept

at

39°C

from the time of infection and

of the infected cellsatthis time.However,atno shiftedto31°Cat3h suggests that the mutation

time didtheactivityreturnafterashift to31°C does not involve any steps in

early

tanscrip-tional control but rather affects the enzyme

it-self. To compare the heatlabilityof the

mutant-a)

1s+, 31@C

C)

tiI3,

31iC

induced nuclease with the wild-type enzyme

after shifting infected cells to a nonpermissive

temperate,theexperimentshown in .4 was

E t 1 t E I .ts13 NUCLEASE ACTIVITY

X- 311,C

I -~~~~~~~~~~~~~~1.5-*+39"C

o~~~~~~~~~~~~~~~~ ~~~~ 0-431*,C

S-+31' C+cycloheximide

°-E 0 EvL 2

-(b)ts+, 39)c d) ts)3, 39(c(

El~~~~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~ ~~~~1.0

zS

CP~~~~~~~~~~~

0.5-0 612 1624 6 1218S24C

(3) (6) (9) (12) (3) (6) (9) (12) HOURS AFTERINFECTION

HOURS AT 31-C, (39"C) FIG. 3. Behavior of the

nuclea

activity

in tsl3-FIG. 2. Effect of tenperatureon theinductionof infectedcells.tsl3-infectedceUswereshiftedto 39°C alkalinenuckaseactivity.Four seriesofdishes were at 0 h or at6f 12,and 18 h after infectionat 31°C infectedwithwild-tpeHSV-2orwith tsl3 and incu- (indicatedbyt) At 3 haftertheshift-up,eachsetwas batedat31°C(0)or39°C(@).At the timesindicated, shiftedback to 31 C(indicated byl),andthe recovery samples were taken for extract preparation and ofthenuckaseactivitywasfollowedin theabsence DNase assayat31°C,asdescribed in the text, atpH orpresenceof50pgof cycloheximideperml, added

9.0. At tines indicated by arrows, samples were atthetimeof shift-down. Experimentaldetailswere

shiftedupto390C (t)orshifteddownto31°C(1). asdescribedinthekgendtoFig.2.

a)ts+ b) ts 13

X 331C

0.* 391C

,, N 391C + Cycloheximide

If,~~~~~~~~~~~~~~~~~~~~~~~~I

1.0 /

X£

E /

O6 6 121x21

0

HOURS AFTER INFECTION

FIG. 4. Effect of temperature shift-up on preexisting nucleaseactivity. Two series of dishes were infected withwild type ortsl3at31°C.At 12 and 18h,thecultures were shifted to39"C(indicated byt). and half of them receivedcycloheximide at the time of the shift-up. Experimental details were as described in the legend to Fig. 2.

http://jvi.asm.org/

[image:3.509.78.225.169.369.2] [image:3.509.262.453.215.362.2] [image:3.509.137.378.458.630.2]

(4)

212 FRANCKE ET AL. J. VIROL.

performed. When protein synthesis was in- TABLE1. Inactivationofwild-type(ts+) andtsl3

hibitedatthe time of shiftto390C in wild-type- nucleaseactivitiesat450Ca

infected cells, the level of nuclease activity in- Inactivation conditions

t1/2

(min)

dicatedaheat-stableenzyme(Fig.4a). The

ts13-induced activitywasheatlabile,asnoted earlier, pH DNA ts+ tsM3

whichwasslightlyexaggerated by thepresence 9.0 - 8(6.5)b 9(16.6)c

of cycloheximnide (Fig. 4b). This experiment also + 16(61.7) 16(78.7)

arguesagainstapotentialinhibitor of preexisting

nuclease induced byashiftto390C, which could 7.5 - 9(77) 8(112)

havegoneunnoticed inthe typeofexperiment + 136(79) 36 (108)

shown in Fig.2 duetocontinued de novo syn- a The preparation of purified enzymes and

preim-thesis in wild-type-infected cells. cubation conditions were as described in the text.

The results reported so far are compatible Preincubation was for 15, 30, 60, 90, and120minat

with the notion that the tsl3 mutation corre- 450C, andt1/2wasdetermined graphically after

plot-spondstothestructuralgenefor theviral alka-

ting

logio of residual activity againsttime.

line nuclease. Since crude extracts from tsM3- bNumbers in parentheses are initial enzyme

activi-infected cells failed to showtemperature sensi-

ties

(at zero

time)

expressed as counts per minute x

tivity whenassayed in vitroat390C(seeFig. 1), e Even at0°C,the enzymeactivities were unstable

theheat lability of purifiedenzymepreparations in theabsence of DNA at pH 9.0.

fromwild-type- and tsl3-infected cellswas

com-pared. During the purification (as outlined 10

above) the acid exonuclease activity was

re-moved during the first chromatography step

(only the alkaline activity bound to DEAE- X

Sephadex). Both enzyme preparations were 70 o A (

morethan2,000-foldpurifiedover asoluble cy- (a)

toplasmic extract, and the final preparations

contained too little protein to measure. The 50- (b)

nuclease activities for both preparations were 0

I- +A

linearly

proportional

to the volume of enzyme , ts ts3

usedper assay(datanotshown).Table1shows 4 x 3 0 (c)

thatpreincubationoftheenzymesby themselves w A 2

at450Cresulted inarapid loss of activity, with 4 30 03 2 0

nomeasurable difference between wildtypeand d 0 0 3

tsl3, but in the presence of substrate (DNA), z

the wild-type activity was 3.5 to 4 times more ae (d)

stable than the tsl3 enzyme. Interestingly, the

stabilizing effect of the substrate was onlyob- 0

served at close to neutralpH (7.5) and not at

the pH optimum for enzymatic activity (9.0).

For a reproducible difference in

t42-

at 450C

between the two preparations, it was further _

necessarytokeepthe salt concentrationatless lC 0 o 60 90

than 30 mM KCI. Bovine serum albumin, al- MINUTES AT 450C

thoughused in theexperiments shown,wasnot

requiired

for the stability of theenzyme,nor did FIG. 5. Heatstabilityofwild-tbpe. and

tsl3-puri-itqaffec

the

stati45

or

the

enzymatic

activit

fiedalkalinenucleases.

Preincubation

at 45°Cwas

it

affec(

tde

ta

at45 orthe

enzymatiC

activity in thei presenceofDNAatpH7.5asdescribed in the

at 310C (data not shown). Because potential text. Each preincubation mixture(200(1)contained contaminants in theenzymepreparations,caus- 15idof

enzyme.

Both enzymes had beenadjustedto

ing the different heat labilities, could not be equalactivityper volume before mixing. The 100%

excluded, the mixing experiment shown in Fig. values(i.e.,notexposedto45°C) were asfollows(in

5 was performed. Both preparations were ad- IHcountsperminuteX103 acidsolubleafter 60 min

justed to equal activitiespervolume and then of incubation at32°Cunder nuclease assay

condi-preincubated at450C in the presence of DNA tions):13.5 (a), 13.7 (b), 13.2

(c),

and 13.4 (d). (a) and

by themselves and as 2:1 enzyme mixtures of (d) weredrawn to best fit theexperimentalpointsx

wildtypeto mutantor mutant to wildtype.The and0,respectively. (b) and(c)representtheoretical

wldx

typetomutant or mutanttowildtype.The curves derivedfrom (a) and (d) by adding thetwo

experimentalpoints for the mixtures fell closely inputactivitiesat 1/3 or 2/3, respectively, corrected on the theoretical lines constricted from the for thesurvivingactivity for eachtime pointduring

best-fitlines for thetwoenzymesalone.It,there- theinactivation.

http://jvi.asm.org/

[image:4.509.267.458.273.518.2]

(5)

VOL. 26, 1978 DNase ACTIVITY IN CELLS INFECTED WITH HSV-2 213

fore, appears that the increased heat lability of the infectivity of virus preparations made at a

the ts13 nuclease in the presence of DNA atpH permissive temperature was sensitive to heat

7.5 is a property of the enzyme itself and not underconditions of wild-type stability. Thus, in

caused bya contaminant. The failure ofcrude this mutant, we can define two

temperature-extractsfrom ts13-infected cellstoshow temper- sensitivefunctions, one related to the virus

par-aturesensitivity duringanin vitro assay is pos- ticle and likely to be a structural polypeptide

siblyaconsequence of thehigh pH required for and the other the virus nuclease activity. It is

the assay. conceivablethatthesetwochangesinphenotype

are the result of a single genetic change. If,

DISCUSSION however,we aredealingwithadouble

temper-The data given hereshow that atemperature- ature-sensitivemutant,thenwecannotat

pres-sensitive mutantof HSV-2 induces an alkaline ent say which of these is lethal. We haverecently

nuclease activity that is temperature sensitive isolated revertants oftsM3,and their examination

both invitro and in vivo. Shifting infectedcells shouldhelpto resolvesomeof these issues.

in vivo from a permissive to a nonpermissive

temperature results in irreversible inactivation ACKNOWLEDMENTS

ofthe enzyme'sactivity,asdocumented by the We thankJohn Subak-Sharpe for his interest in thiswork,

requirement forprotein synthesisfor thereap- MajorieStrobel-Fidlerfortechnicalassistance,andthe Inter-pearanceofactivity after shifting backto aper- national Union Against Cancer (UICC) fora

Yamagiwa-Yosh-mlssvetemperature. Inaddition,thewild-type idaMemorialInternationalCancerstudygranttoJ.H.

missivetemperature.In addition, the wild-type ThisworkwassupportedbyPublicHealth Servicegrants

activity isstable intheabsence ofproteinsyn- CA-15088 and CA-14195 awarded by the National Cancer

thesisat anonpermissive temperature, whereas Institute.

themutantactivity is not. Incubation of purified

enzymes invitroat45°C showsafourfold-faster LITERATURE CITED

inactivation rate for the ts13-induced enzyme 1. Benyesh-Melnick,M., P. A.Schaffer,R. J.Courtney,

overthewild-type enzyme. These findings pro- J.Esparza and S.Kimura.1974.Viralgenefunctions

vide the firststrongevidence that thenuclease expressed and detected by temperature-sensitive

mu-activity

_activityis coded directly bythe virus genome,

_{~~~~~~~~~~~~~Symp.}

tants of herpes_Quant._{Biol. 39:731-746.}simplex virus. Cold Spring Harbor

although they donotruleoutthe

possibility

that 2. Halliburton,I.W., and M.C.Timbury.1973.

Charac-partoftheenzymeprotein iscontributedby the teristics oftemperature-sensitive mutants ofherpes cell. The accelerated appearance ofactivityata simplexvirus type 2. Growth and DNA synthesis.

Vi-permissive

temperatureaftera

period

of3hat rlg 46-8

3.

Hafliburton,

l.W.,and M. C.Timbury.1976.

Temper-a nonpermissive temperature indicates that ature-sensitive mutants of herpessimplexvirus type2:

early steps in the infectious cyclearenotaffected descriptionof threenewcomplementationgroupsand

by the mutation, as would be expected for a studiesontheinhibitionof hostcellDNA synthesis.J.

lesion in the structural gene ofa late function. 4

Gen.

Virol. 30:207-221.

4.Hay,J.,H.Moss,and L. W. Halliburton. 1971.

Induc-Halliburton and Timbury (2) have reported tion ofdeoxyribonucleic acid polymerase and deoxyri-that tsl3 makes significantly less virus DNA bonuclease activities incellsinfected withherpes

sim-thandoes wild typeat anonpermissivetemper- plex type II. Biochem. J. 124:64P.

ature,but isnotDNAnegative.Ourownstudies 5. Keir,H.M.,andE. Gold. 1963. Deoxyribonucleicacid

onDNA synthesis agree with this assessmet . nucleotidyltransferaseanddeoxyribonucleasefrom

cul-on DN>A synthlesis agree with this assessment tured cellsinfectedwithherpessimplexvirus.Biochim.

andwouldsuggestthat HSV-2 canreplicateits Biophys.Acta 72:263-276.

DNA,atleasttosomeextent,inthe absence of 6- MacPherson,L.A., and M.G. P. Stoker. 1962.Polyoma thegreatmajorityof thisnucleaseactivity.Sev- transformationof hamstercellclones-aninvestigation

eral_eralpossibilities_pssibiitiesarise from this iresut._{rise rom tis reslt. Prhaps}Perhaps of_16:147-151.geneticfactorsaffecting cell competence.Virology

the virus needsonlytrace amountsofexonucle- 7.Morrison,J.M.,and H. M. Keir.1968. Anew

DNA-aseactivityfor its replication,or, alternatively, exonuclease incellsinfectedwith herpes virus:partial

suchDNAasissynthesizedints13-infectedcells purificationand properties of theenzyme. J.Gen.Virol.

atanopermissvetemprature sdefecive in 3:337-347.

at a nonpermssive temperature IS defective in 8. Subak-Sharpe, J.H.,S. M.Brown, D. A.Ritchie,M.

some way. The DNase activity could alsorep- C.Timbury,J. C. M.Macnab, H.S.Marsden,and

resent a dispensable viral function, atleast in J. Hay. 1974. Genetic and biochemical studies with the system used to isolate and test the virus herpesvirus. Cold SpringHarbor Symp. Quant. Biol.

39:717-730.

mutant. 9. Timbury,M. C. 1971. Temperature-sensitive mutants of

Earlierobservations on tsM3 (3) revealed that herpessimplex virus type 2. J. Gen. Virol.13:373-375.

http://jvi.asm.org/