Isolation and properties of Moloney murine leukemia virus mutants: use of a rapid assay for release of virion reverse transcriptase.

0022-538X/81/040239-10$02.00/0 Vol. 38, No. 1

Isolation and Properties of Moloney

Murine

Leukemia

Virus

Mutants: Use of a Rapid Assay for Release of

Virion

Reverse

Transcriptase

STEPHEN GOFF, PAULA TRAKTMAN, AND DAVID BALTIMORE*

Centerfor Cancer Research and Department of Biology, Massachusetts Institute of Technology,Cambridge,

Massachusetts 02139

Received 25 November 1980/Accepted 7 January 1981

A rapid assay for retroviral reverse transcriptase activity released into the

culture medium by infectedcellswas developed. With the assay,4,000 clonally infected celllines could be tested in a few hours. We have adaptedthe assayfor

use as a screen for the detection of spontaneous viral mutants. Mutants of

Moloney murine leukemiavirus have been isolated which (i) produce a thermo-labilereverse transcriptase, (ii) are temperature sensitive forrelease of enzyme activity,or (iii) can only productively infect cells already producing gag-related polypeptides.The assay has also been useful for the isolation ofnonproducercells infected withvarious replication-defective transforming viruses.

Most knowledge aboutretroviruses has been gained by direct biochemical analyses of virion nucleicacids and proteins. Very little is known about thefunction of each virus-encodedprotein in the life cycle, largely because very few

mu-tantsmappedto specific proteins and affecting

specific functions have been isolated.In the

mu-rine system, there have been

temperature-sen-sitive mutants isolated with impairedgag and

gag-pol

cleavage

and reduced virus release at

390C (16, 19). Nonconditional mutants with al-tered protein phenotypes have been described

(15).Therealso have beenreportsofearly

mu-tants found to have temperature-sensitive

po-lymerase and RNase H function (17), and a nonconditionalpol mutant has been described recently (4).Finally,mutantswhichrelease par-ticles which cause no

plaque

formation in the

XC assay (XC- mutants) have been described

(5, 9, 10). Few of these mutants have been mappedonthe genome orascribedto

proteins,

and the lesion in most of the

early

and late

mutantsremains unknown.Moremutantshave

beendescribed and

mapped

inthe aviansystem. gag

processing

mutants have been

described,

and nonconditionalmutantsingag supportthe

idea that gagproteinsarerequiredforsuccessful

virion assembly (8). The behaviorofa mutant

inp15 confinmsitsprobableroleasthe protease responsible for gag and

gag-pol

cleavage (11).

envmutantshave beendescribed

(7,

14)

which

confirm the

following

about the env

glycopro-tein: its role in

determining subgroup specificity,

its necessary presence for viral

infectivity,

and

itsunimportanceinvirionassembly. Lastly, po-lymerasemutantswiththermolabileactivityare

numerous (1, 6), andmutantsinpolmaturation

have been discussed (2).

We have begunconstructingalibrary of

mu-tants of Moloney murine leukemia virus

(M-MuLV) blocked in various stages of the life cycle,which may beusefulinassigningfunctions to the known proteins. To detect mutants, a rapidassay for thevirus-encodedreverse

tran-scriptasewasdeveloped by which several thou-sand clones ofinfected cells could be

assayed

in anafternoon.

MATERLA1S AND

METHODS

Celis

and viruses. The NIH/3T3 cell line was

growninDulbecco's modifiedEaglemedium contain-ing 10% calfserum. M23 cells were derivedin this

laboratory (15).Thesecellscontainadefective

provi-rusof M-MuLV which encodesonlythe

Pr659M

pro-tein. M-MuLV clone 1was the source of infectious viruspreparationsandwasusedas ourwild-typevirus (3). Rauscher MuLVts29 (R-ts29)was agiftfromS. Aaronson (17). Viruswastitratedbythe XC plaque

assay (13). Samples to be assayed at temperatures other than

370C

were firstpreincubatedatthe indi-cated temperature for45min in mediumcontaining 20 mM HEPES

(N-2-hydroxyethylpiperazine-NV-2-ethanesulfonicacid), pH7.35, and thenappliedtothe cellsatthesametemperature.

Rapidreversetranscriptaseassay. Cloneswere

prepared by infection ofNIH/3T3 cellsat a

multiplic-ityofapproximately0.3PFU/cellfor 2 hat37°C.The

cellswerethentrypsinized,counted,and seeded into

96-wellcloning traysat 0.3cellsper well(inavolume 239

on November 10, 2019 by guest

http://jvi.asm.org/

of0.2mlperwell).The cloneswereallowedtoexpand

for10to 12days.Areplicatingdevice (Flow

Labora-tories, Inc.) wasusedtoremove10,ulof the

superna-tantmedium from eachwelland add it to 50

pi

ofa

reaction cocktail aliquoted into disposable 96-well trays(Dynatech Laboratories).The cocktailgavethe followingfinal concentrations: 50 mM Tris-hydrochlo-ride, pH8.3; 20mMdithiothreitol;0.6mMMnCl2;60 mM NaCl; 0.05% Nonidet P-40; 5 jig of oligodeoxy-thymidylicacidperml;10jig ofpolyriboadenylicacid

perml;and 10,uMa[nP]dTTP (specific activity,1Ci/

mmole).Thereactionwasincubatedatthe

appropri-atetemperature for 1 to 2h,and 10 pd of the mixture

wasspotted directlyontosheets ofdryDEAEpaper

(DE-81; Whatman, Inc.) ina96-spot grid, using the

mechanical replicator. The paper was then washed withgentle rockingatroomtemperature three times

in 500 ml of 2x 0.3 M NaCl-0.03 M sodium citrate for 15min each and then twice in 500 ml of95% ethanol.

Thepaper wasdried andexposedtoX-rayfilm (XR-5)at-70°Cwithanintensifyingscreen(Dupont

Light-ning Plus).

Infection of replicate trays ofcells.Virus inthe culture medium above clones of cellswaspassedonto recipient cell lawnsasfollows. The recipient96-well trayswereseeded with 0.1 ml ofacellsuspension (104

cells/ml)in mediumcontaining2,ugofPolybreneper

ml. After 1day,10lO ofthe medium in eachcloning

well ofa tray of donor cells was transferred to the recipient trays, usingthe replicating device; the

in-fectedcells were incubated for 1 day, and then the

medium was changed to normal medium (lacking Polybrene).After 1moreday,therecipientcells were

assayedforreversetranscriptase bytherapid

proce-dure.

Conventional reverse transcriptase assays.

Theconventionalreversetranscriptaseassay(12)was carried outbyincubating 50

pl

of the culturemedium to be assayedwith 100jil ofacocktail to give final

concentrationsasgivenabovefor the rapidassayand

by countingthetrichloroaceticacid-precipitable

ma-terial after filtrationthrough membrane filters

(Mill-poreCorp.). Assaysat32and 39°Cwereperformed by

preincubationof the viruswith0.1M Tris-hydrochlo-ride, pH7.5,attheindicatedtemperatureforvarying times followed byaddition of the remaining

compo-nentsof the cocktail.

Immunoprecipitationofvirus-specific proteinsand

gel electrophoresisof theproteinswere asdescribed

previously (18).

RESULTS

A rapid screen for release of reverse

transcriptaseactivity. Theprotocolfor

assay-ingindividual clonesofcells for releasedreverse

transcriptase activity is shown in Fig. 1. Cells

wereinfected withapreparation of wild-type

M-MuLV and replated into cloning wells; after

growth, partof the culture mediumwas added

to a reverse transcriptase exogenous reaction

mixture with detergent and incubated, and a

portionwasthenspottedontoDEAEpaper.The

RAPID REVERSE TRANSCRIPTASE SCREEN

INFECT CLONE CELLS

VIRUS ~ CELL GROW UP

STRUS MONOLAYER (2-3weeks)

REPLICATE

REPLICATE TRAYOF 1oaI O(32P-TTP-RTCOCKTAIL

(hwbet.2hr. 37C) SAVE CLONES

WASH AUTORAIOGRAPH |..

DEAEPAPER

X-RAYFILM

FIG. 1. Outline of the procedure for assaying

clonedcell linesforreleaseofviralreverse transcrip-tase. Cells were infected with virus, cloned, and grownfor2 to3weeks. Ten microlitersof the culture medium was removed and mixed with an enzyme cocktail.After incubation,the[32P]DNA productwas

boundtoDEAE paper, whichwaswashed and

ex-posedtoX-rayfilm.

paperwas washed andexposedto film for

auto-radiography.

Atypicalassayon 10 traysisshown in Fig. 2.

In the experiment, approximately half of the

wells contained cells(412 of 960);approxiimately

one-third of these (125 of 412) were releasing

reverse transcriptase enzyme. Duplicate assays

of the same cells were reproducible (data not

shown;seebelow).Although the number of cells

in each well at the time of assay was highly

variable, there was no obvious correlation

be-tweenthe number ofcells and the intensity of

the signal. Many of the very dark signals were

produced by only a few cells (<103), whereas

many of the faint signals were produced by

confluent wells (containing 104 cells or more).

Weconcluded that there was a wide range in the

level of enzyme released by cell lines, as has

been notedby others (15). The sensitivity of the

assaywasveryhigh; as few as 100 cells produced

positive signals. The assay shown was completed

inafew hours of work, and the fih shown was

a 12-h exposure.

Isolation of mutants coding for thermo-labile reverse transcriptase. To detect

on November 10, 2019 by guest

http://jvi.asm.org/

[image:2.494.266.453.110.314.2]

ISOLATION AND PROPERTIES OF M-MuLV MUTANTS 241

FIG. 2. A typical assay of cloned cell lines for

releasedreversetranscriptase.Ten traysofcellswere

prepared and assayed as outlined in Fig. 1. The

[32P]DNA boundtoDEAEpaperwasexposedto

X-rayfilm for12h.

tantsof M-MuLVwhich encodedathermolabile

reverse transcriptase activity, cloneswere

pre-paredin 20traysasdescribedabove and allowed

to expandfor 10 days at 370C. The cells were

thengrownfor 2moredaysat320C.Portionsof

culture medium were removed and added in

duplicate to an incomplete cocktail mix which

lacked thetemplate, primer,andnucleoside

tri-phosphate.Afterpreincubation for 1 hat45or

250C, the template, primer, and dTTP were

added,andthe reactionwasallowedtoproceed for2h. AnX-rayfilmexposureoftwopairsof

traysis showninFig.3A. Ingeneral, the clones

released temperature-resistant enzyme which

showedsimilaractivityatthetwotemperatures;

a few, however, released enzyme with a much

reduced activity at450C. Avery fewseemed to

be coldsensitive. A total of 13 clones were

se-lected (11 temperaturesensitive and 2 cold

sen-sitive), grown into larger cultures, and reassayed

as described above. Of these 13, 10 were

repro-ducibly thermolabile (Fig. 3B); their behavior

wassimilar to that of the R-ts29, as previously

described (17). No consistently cold-sensitive

clone wasdetected.

Biological propertiesofthe reverse tran-scriptase mutants. Eachofthe 10putatively

mutant clones was grown at

320C,

and each

culture medium was harvested. The released virions were preincubated for 45 min at 32 or

390C

and then titrated in XC plaque assays perfornedatthesame temperature. Of the 10,

5 never gave XC plaquesunder any condition

and presumablywereeither uninfectiousdue to the mutation or were XC-negative variants of M-MuLV (5, 9, 10). Ofthe 10, 3 were XC posi-tive, but showednothermolabilityover that of the controlwild-type virus (whichlost 50 to 80%

of itsinfectivityat390C, relativeto32°C, in this

experiment). The XC titers of representative members of this group (6H3 and TSP-11F8)areshown inTable1.Theremaining two

mutants (TSP-11A4 and TSP-6E9) were

ob-viously more thermolabile than the wild type was in the XC assay; in separate experiments, they showed a 40- to 200-fold reduced titer at

thehigher temperature (Table 1). They also had

lower titers than did the wild type after

320C

incubation. Thus, these two mutants, chosen onlyonthe basisofthe invitroreverse

transcrip-tase assay, were also thermolabile by plaque

assay.

Kinetics ofinactivation ofreverse tran-scriptase. Toconfirmthequalitative resultsof therapidassay,conventionalreverse

transcrip-taseassays wereused. Virion harvestswere

col-lected from the mutant cell linesand preincu-bated for

varying

times in buffered mediumat

twotemperatures, and then exogenous reverse transcription was allowed to proceed for 1 h.

Three conditionsweretested:

(i)

preincubation

andassay at

250C;

(ii) preincubation andassay

at

390C;

and (iii) preincubation at

390C

and

assayat

250C. Examples

ofinactivation curves

for four of themutants,

along

with controlcurves

ofwild-typeM-MuLV andR-ts29, areshown in

Fig.4.

Wild-type viral reverse

transcriptase

was

in-activatedapproximately50% in this

experiment

during preincubation at 25°C; itshowed some

additional loss when

preincubated

at 390C. In

other experiments,the loss of

activity

wasless

but was

generally

10 to 50% and was

slightly

VOL. 38, 1981

on November 10, 2019 by guest

http://jvi.asm.org/

[image:3.494.48.241.63.453.2]

242 GOFF, TRAKTMAN, AND BALTIMORE

FIG. 3. IsolationofceU clonescontainingmutantM-MuLV genomesencodingthermolabilereverse tran-scriptase. (A)Ten traysofclonedcellswereassayedintheduplicateat25and450C.

Autoradiograms of

two

of these traysareshown. A totalof13clonesshowingdifferentialactivitywerechosen

from

thisscreen.

(B)

Repeat ofthe assay on the 13 clones chosen from the initial screen. Ofthe 13, 10 were

reproducibly

thermolabile.R-ts29 isincludedfor comparison (lowerright sample).

TABLE 1. XCplaqueassayonthermolabilereversetranscriptasemutants'

Expt 1 Expt2

Virus Temp('C) Ratio(32'C/ Temp (°C) Ratio (32'C/

32 39 39°C) 32 39

390C)

Wildtype 3x105 5x104 6 5x 104 2x 104 2.5

TSP-6H3 5x10' 2x101 2.5 5x10' 2x10' 2.5

TSP-11F8 9xi05 2x105 4.5

TSP-6E9 5 x 103 <5x101 100 4 x103 2 x10' 200

TSP-11A4 2 x 103 5x10' 40 2x 103 2 x10' 100

aVirus harvestswerecollectedat32°C. The viruswaspreincubatedattheindicatedtemperaturefor45min

and then titratedatthesametemperature.Tableentriesaretiters inPFU/ml.

greater during preincubation at the elevated

temperature. The activity, when assayed at

390C,

wasconsiderably reduced,independent of the time ofpreincubation. R-ts29 activity was

stable at

250C,

butextremely sensitive to

prein-cubation at

390;

1hreducedthe activity 20-fold

toundetectable levelsevenwhenthe assay was performedat25°C. Essentiallynoactivity could

bedetected in any assay carried out at

390C.

The reversetranscriptase released by each of

the mutant clones was morethermolabile than

the wild-type enzyme. The enzymes produced

bytwoof the clones(TSP-11A4andTSP-6E9)

were extremely temperature sensitive; like R-ts29,

they

were rapidly inactivated by

preincu-bationat39°C (Fig. 4). Theyalsoshowed almost

on November 10, 2019 by guest

http://jvi.asm.org/

[image:4.494.103.400.79.385.2] [image:4.494.54.453.453.548.2]

ISOLATION AND PROPERTIES 243

WILD TYPE

100 100

50 5 3 450

*1

0

0 15 30 50

TSP 6E9

00- _ 100

0

\\ o --,

50_ .s 50

0 15 30 50

R-TS21

TSP 6H

.A

9 the virions released in that period were har-vested andbanded

by isopycnic centrifugation.

.~--_ Whenthe labeled proteins in the virions were

>0

analyzed

by

polyacrylamide

gel electrophoresis

(Fig.

5),

cellsinfected with wild-type virus were

found to have released large amounts of virion

proteins

atbothtemperatures; the most abun--- dant protein byfarwasp30,a productof the gag

50 gene. Neither clone 6E9 nor clone

TSP-13 6H3 was temperature sensitive for release of

° virions (data not shown), although both

pro-duced athermolabilereversetranscriptase

activ-ityand one, inaddition,wastemperature

sensi-tive for plaque formation (Table 2). Thus, re-verse transcriptase mutants are not always af-...._. fected in release. Two other clones, however, f

(TSP-11A4

and

TSP-11F8)

were temperature

on

1 30 50

TSP IIA4 TSPIIF8 o o

100__100 o o

50 50

0 15 30 50 0 15 30 50

Preincubation (minutes)

FIG. 4. Heat-inactivation curvesforreverse tran-scriptasereleasedfrom wild-typeM-MuLV-infected cells, R-ts29-infected cells, andfourmutant clones. Culture medium fromeach clone waspreincubated

for varyingtimes and thenassayed forafixedtime

(1 h). The conditions were: bothpreincubationand assayat390C(-); preincubationat39°Candassay

at25°C (A); and bothpreincubation andassayat

25°C (0).Eachofthefourclones releasedareverse

transcriptaseactivitymorethermolabile than thatof

the wildtype.

noactivity when assayed at 390C. Two others

wereless sensitive topreincubation,with

activ-ity fallingslowly during preincubationandassay

at 390C. Clone TSP-11F8 produced enzyme

which was particularly resistant to a 30-min

preincubation at 390C if the assay was then

performedat250C. Thus,theenzymefromthis

clone was not irreversibly inactivated by the

preincubation. It is noteworthy that the two

viruses containing the most thernolabile

en-zymes were the same twowhich showed

tem-peraturesensitivityinthe XCplaqueassay

(Ta-ble2).

Associated phenotypic changes in

tem-perature-sensitive reverse transcriptase

mutants. Four of the mutant cell lines were

testedfor additional effects of themutationon

theassemblyofvirions. Cellswerelabeled

[image:5.494.54.243.57.355.2]

over-night (at32or390C) with[35S]methionine,and

TABLE 2. Summary ofproperties of four thermolabile reverse transcriptase mutantsa

Raidrse

Rapidre-

~transcrip-

Reverse

Vhion

Vrs_Vr

_transcrip-

verse XC_XCa8yassay tase_t8 heat_et rl_release taseassay

inactiva-TSP-6H3 tsb +(very low) ts +

TSP-11F8 ts + ts ts

TSP-6E9 es ts ts +

TSP-11A4 Ls ts ts ts

aData weretaken from Table 1 and Fig. 3 through 5. ts,Temperature sensitive.

320 390 320 390 320 390 320 390

FIG. 5. Analysis ofvirionproteins released into the culturemediumbyclonesexpressingthermolabile

reversetranscriptase. Cells were labeled with

["SI-methionineat 32or390C, and the virions released into the medium werepurified. Theproteins were

separated by electrophoresis through polyacrylamide gels and detected byfluorography. Theposition of

the viralprotein p30 is indicated.

CI-w

0

0

38, 1981

on November 10, 2019 by guest

http://jvi.asm.org/

[image:5.494.257.441.277.582.2] [image:5.494.278.421.401.580.2]

TRAKTMAN,

sensitive for release of virus (Fig. 5); much less

p30 wasdetected in harvests atthe high

tem-perature. Their behavior was similar to that of

M-MuLV ts3 (Fig.5). In asimilarway, the

R-ts29-infected line was temperature sensitive for release (data not shown). Thus, a number of independentreversetranscriptasemutants have

beenfound tobe,inaddition, defectivein their

abilitytoassemble orreleasevirions.

Isolation ofmutants temperature sensi-tive for release of reverse transcriptase. The reverse transcriptase assay could also be

used todetectandisolatemutations in retroviral

proteins other than the reverse transcriptase

itself. To detectclones with conditional

muta-tionsin any gene affecting assemblyorrelease

ofvirus, we chose to assaythereverse

transcrip-tasereleasedateach of two temperatures.Cells

were infected as above and distributed to 10

traysofcloningwells.Thecloneswere expanded

for20daysat320C, and the culture mediumwas assayedat250C (withoutapreincubation); then the mediumover the clones was changed, and after 2 daysat 390C, a second portion was as-sayed at 39°C. Comparison of the two assays revealed approximately 11 clones (out of 200

producing reversetranscriptase)whichreleased

less activityatthe elevatedtemperature; seven

of these provedtobeconsistently temperature

sensitive for release upon retesting. No

con-sistentlycold-sensitive clones weredetected. Preliminary characterization ofmutants

temperaturesensitive forrelease. The

viri-ons of seven clones which were temperature

sensitive for release were tested biologically. Culture mediumwasharvested fromeach ofthe

sevenclones after incubation for 1dayat

320C.

These harvests were then titrated in an XC

plaque assay carried out at either 32 or

390C.

Wild-type virus showed a high titer at either

temperature (Table 3). Four of the mutant

clones either released viruswhichwas XC

neg-ative in the assay ateither temperature or

re-leased a verylow titerof viruswhich was

unde-TABLE 3. XCplaque assay on temperature-sensitive virus harvested at32°C

Virus

Tempa

(°C)

Ratio

(320C/

32 39

Wild type 4 x105 2 x105 2 TS1OC7 5 x 104 <103 >50

TS 1E6 8x 104 3 x103 27

TS4D2 1.4x 104 2x103 7

a

Virus

harvestswere

collected

from

cell

lines after

1dayofgrowthat320C.Thevirus waspreincubated attheindicated temperaturefor45minandtitrated

atthesame temperature. Table entries aretiters in

PFU/ml.

tectable inourassays.Theremaining three

mu-tantclones released virions whichwerecapable

of producing plaques at 320C but which

pro-ducednoorfewer plaques when assayedat390C

(Table 3). Thus, the three producer cell lines carrying these viruses were temperature

sensi-tive for release of virion reverse transcriptase,

and, moreover, the virus released at the

permis-sive temperature was itselfunabletocomplete

aninfectionatthe nonpermissive temperature.

The viral proteins synthesized by the seven

cell lines were analyzed by metabolic labeling

with[3S]methionine, immunoprecipitation,and

polyacrylamide gel electrophoresis. The

prod-uctsof thegag gene

(Pr65`)

and the envgene

(Pr80env) weredetected in all of the clones; the

onlyabnormalitywas alarge overproduction of

Pr8Oenv at the elevated temperature by one of

the XC-negative clones (TS 9F11) (Fig. 6). No

other gross changes in the size orlevel of

pro-duction of the viral proteins were detected

among themutants.

Isolation of mutants in thegag gene.

Us-M 320 390 320 390

I-Pr8Oenv

-Pr659°9

p30-FIG. 6. Envelopeglycoprotein produced by the TS

9F1 Iclone. The indicatedcells were labeled at 32 or

39°C with [35S]methionine, andextracts were

pre-pared. Viral envelope glycoprotein was immunopre-cipitated. The proteinswereanalyzed by

electropho-resis andfluorography.

VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/

[image:6.494.265.426.319.597.2]

ing therapid assay,clonalisolates of M-MuLV canbetested for theirabilitytoreplicate intwo different host cells. Ifone of the hosts already

carries adefective provirus which provides one

oftheviral functions, thenmutants inthat func-tion wouldbe detected ashostrange mutants,

able to grow onthe hostcarrying the function

and not onnormal cells. Hosts expressing the

env function could not be used to isolate env

mutants because they are not superinfectible,

but hosts producinggag orpolproteins could, in principle, be used toisolate mutants in the gagorpolgene.One such hostcell wasavailable

to us in the form of the cell line M23 (15), a

derivativeof NIH/3T3 cells carryingadefective M-MuLV

provirus

and synthesizing only Pr659"9. Preliminary mapping (unpublished data)suggeststhat theprovirusinM23 contains a1.4-kb deletionmappingintheenvgene,near the 3' endof the viralgenome.Thiscell line has

been found tosynthesizeashortviralRNA(15)

which

apparently

directs synthesis of only

Pr65V"';

nopol- orenv-relatedprotein has been detected. Because this host cell is superinfecti-ble, in

principle

itwould bepossible to isolate

gag mutantswith thishost, which wouldgrow

by complementation but which wouldnot rep-licateinnormal cells.

Accordingly, M23 cells were infected with wild-typeM-MuLV anddistributedto 10

cloning

trays asdescribed above. After15days at

370C,

the cells were assayed for release of reverse

transcriptase;

at the same time, culture fluids werepassedtofreshM23cellsand in duplicate tofreshNIH/3T3cells.After 2days, these

cells

were also

assayed.

Two clones were detected which released

transcriptase activity

and re-leased virus which

productively

infected M23 cells but whose

yield

did not infect

NIH/3T3

cells

(Fig. 7).

Analysis ofthe viralproteins synthesized by thetwoclones showednoalterations from wild-typecontrols (datanotshown). Analysis of the

virusreleased from theseclones by XC plaque

assay,however, showedadistinct

difference

be-tweenthe clones and controls andconfirmed the

hostrangephenotype of thesetwomutants

(Ta-ble 4). The supernatants were titrated on two hosts: M23 cells and NIH/3T3 cells. Wild-type virus, as well as wild type virus grown on M23

cells,

showed identical titers on these two hosts.

The virus releasedbythe two clones showed a

hightiteronM23andavery low titeron

NIH/

3T3 cells (Table 4). In addition, the mutant

plaque numberson M23cells gave a linear

re-sponsetodilution;their

plaque

numberson3T3

cells, however, dropped off very rapidly with

dilution of the stock, suggesting that multiple

hits were required to initiate an XC plaque. The data were consistent with the notion that two genomes were present in the virus preparations

and that both must infectthe same cell to

pro-duce a plaque. The virus would yielda higher

andlinear titeron M23 cells because thesecells alreadycontainedone of the required viruses as anendogenousprovirus.

Isolation

of transformed nonproducer cell lines.A commonprobleminthe analysisof

replication-defective

transforming viruses is the

isolationofatransformedcell clone which is free

of helper virus.Wemadeuseof the rapid reverse transcriptaseassay to facilitatethis procedure. Amixed stockof AbelsonMuLVwith its helper M-MuLV was collected from a producer cell. NIH/3T3

cells

wereinfected at a low

multiplic-ity (0.7 PFU/cell; 0.15 focus-forming units per cell),trypsinized, and clonedinmicrotiter trays. Aftergrowth for13days,the clones were assayed forreleasedreverse transcriptaseactivity.Those clones scored as negative for release ofreverse transcriptasewerethen scored

visually

for mor-phological

transformation.

In thisway, several transformed nonproducer

lines

wererapidly iso-lated. In one representative experiment, 4% of thetotal clones showed the desired phenotype.

DISCUSSION

The reverse transcriptase assay described is

rapid andinexpensive; several thousand clones can be screened in a

day

by

one person. The

assayis

remarkably sensitive,

allowing

detection

ofactivityfrom asfewas a 100cells.The screen isdirectly useful forcloningvirusesandfor the isolation of

transformed nonproducer

clones fromamixed virus stock ofa

transforming virus

and its

helper.

In

addition,

the screen is

ex-tremely

helpful

in theidentification of variant clonesproducing

temperature-sensitive

or host

range mutantsof retroviruses.Wehaveisolated

asmall number ofmutantsof threetypes;more

of each class andof other classescould

easily

be isolatedfor further

analysis.

Aremarkablefinding

implicit

in oursuccessful isolation ofthesemutantsisthe

high

frequency

of spontaneous mutations which arise in puta-tively

wild-type

stocksofM-MuLV.

Thermola-bile and

temperature-sensitive

mutants were

found in screens ofonly200 to 300

clones;

the

frequency of recovery of mutants was in the

range of 3 to 5%. Thishigh

frequency

of

appear-ance of mutations has been noted

previously

(15). Some step in the life

cycle

of

M-MuLV,

and

presumably

many other

retroviruses,

must be highly

mutagenic;

likely

candidates for this step are the

transcription

of the

provirus

into

on November 10, 2019 by guest

http://jvi.asm.org/

246 GOFF, TRAKTMAN, AND BALTIMORE

FIG. 7. Isolationofhost rangemutants.M23 cellswereinfectedwithM-MuLV, cloned, andassayedfor reversetranscriptase(left panels). The virusreleased by these cellswasthen passedtofresh M23 cellsor to

freshNIH/3T3 cells. After2days, these cells were also assayed (middle andrightpanels). Clonesreleasing

viruscapable ofinfecting M23 cells but notNIH/3T3 cells were selected for further study.

thegenomic RNAand thereversetranscription of the RNA into circular double-stranded DNA.

Bothof these stepsarethoughttobe carriedout

by nonproofreading polymerases and maywell havehigherror rates.Theremaythus be several

mutations introduced into each retrovirus

ge-nomeeach time it ispassedfrom celltocell.

The mutantsisolated in ourfirst screen are probablydefective in the pol gene itself because

the enzyme is thermolabile in vitro by two

as-says. It isnevertheless possible that mutations

inadjacent viral proteins could reduce the

sta-bilityofthe enzyme. It isinterestingthatthepol

mutantsoften,butnotobligatorily,havedefects

in assembly and release of virions. These

mu-tants raise the

possibility

thatthe

polymerase

proteinor aprecursorof the enzyme is involved

in virion assemblyatthe cell surfaceasis also suggested by other data (P. Traktman and D. Baltimore,manuscript inpreparation).

The mutants recovered in thesecondscreen

could potentially lie in any viral gene, or even

potentiallyinacellular gene, neededfor

assem-bly; the high frequencyof appearance of these

mutations would argue for a viral origin, and

those that have a temperature-sensitive yield

mustbe in the virus. Theonly gross alteration

inamajorviralproteinwastheoverproduction

of theglycoproteinPr8Oenvatthenonpermissive

temperature inone clone. All of the other

on November 10, 2019 by guest

http://jvi.asm.org/

[image:8.494.110.397.77.451.2]

ISOLATION AND PROPERTIES OF M-MuLV MUTANTS 247

TABLE 4. XCplaque assay onvirus preparations harvested from clonedcell linesa

Celllawn Virus prepn Dilution

NIH/3T3 M23

Wildtype grownonM23 cells 1:100 TMTCb TMTC TMTC TMTC

1:300 500 500 500 500

1:1,000 200 200 200 200

1:3,000 100 100 100 100

1:10,000 75 47 37 45

1:30,000 22 16 12 15

1:100,000 6 6 4 3

Clone 8G4 1:100 100 200 TMTC TMTC

1:300 23 13 500 500

1:1,000 2 1 150 150

1:3,000 2 2 64

-1:10,000 0 0 30

1:30,000 0 0 6

1:100,000 0 0 0 1

Clone M16 1:100 100 100 TMTC TMTC

1:300 13 17 300 300

1:1,000 100

-1:3,000 0 0 43 40

1:10,000 0 0 14 12

1:30,000 0 0 4

-1:100,000 0 0 0 1

Wild-type virus grown on 1:1,000 105 57

NIH/3T3cells 1:10,000 13 4

1:100,000 1 0

1:1,000,000 0 0

a

Virus

harvests fromthe indicated

cell

lines were diluted and used to infect either NIH/3T3

cells

orM23 cells. Table entriesareplaquescounted per6-cmdish.

bTMTC, Too manytocount.

tants presumably either carry point mutations

which inactivate the function of a particular

protein without altering its size or alter

unde-tected minorproteins.

Mutations of our thirdclass,which are

com-plementedbythe defectiveprovirusof M23cells,

presumably lie in the gag gene. In M23 cells

these mutants replicatenormally, but inNIH/

3T3 cellsthey donot.Apparentlyrecombination

did notoccur in the initial infection becausein

preliminary experiments two RNA sizes were

still detectable in the cells: full-size RNA and

the deleted M23 RNA (unpublished

data).

Sim-ilarly,among theraresuccessfullyinfected

NIH/

3T3 cells therewere some reverse

transcriptase-positive clones which also carried both RNAs.

Thus bothgenomescanbereleased and

appar-ently must be transmitted to a newly infected

cell if that cell istobecomeaproducerof virus.

The virus thus showsahightiteronM23cells,

which already provides one of the needed

ge-nomes, and avery low titer on

NIH/3T3

cells.

The fact that the mixed virus stock shows

mul-tihit kinetics on NIH/3T3 cells suggests that

both genomes

rarely

are

packaged

intoone

vir-ion or, morelikely, that both genomes in a virion

arerarely convertedtofunctionalproviruses.

Weareattemptingtoclonethesemutantsfree

of their M23helper byscreening NIH/3T3cells

infectedat alowmultiplicityforthesynthesis of

viralproteins. We can thenanalyze themutants

biochemicallyand determine the exact nature of

the defect in thegag gene.

ACKNOWLEDGMENIS We thank Owen Witte forhelpfuldiscussions.

ThisworkwassupportedbycontractN01-CP-53562 from theDivision of Cancer CauseandPrevention,National Cancer Institute, and by Public Health Service grant CA-14051(core granttoS.E.Luria)from the NationalCancer Institute. S.P.G. isa Jane CoffinChilds Memorial Fund Fellow. D.B. isan

AmericanCancerSocietyResearch Professor. LITERATURE CITED

1. Alevy, M. C. andP. K.Vogt.1978.Tspolmutantsof

avian sarcomaviruses:mappingand demonstration of single-cyclerecombinants.Virology87:21-33. 2. Eisenman,R.N.,W. E.Mason,and M.Linial. 1980.

Synthesis and processing ofpolymerase proteins of wild-typeandmutantavianretroviruses. J.Virol. 36: 62-78.

3. Fan,IL, andM. Paskind. 1974. Measurement ofthe sequencecomplexityofclones ofMoloneymurine leu-VOL. 38, 1981

on November 10, 2019 by guest

http://jvi.asm.org/

[image:9.494.50.448.87.375.2]

kemia virus 60 to 70S RNA: evidence for a haploid genome. J. Virol. 14:421-429.

4. Gerwin, B. J., A. Rein, J. G. Levin, R. H. Bassin, M. B.Benjers, S. V. S. Kashmini, D. Hopkins, and B. J.O'Neill. 1979. Mutant of B-tropic murine leukemia synthesizing an altered polymerasemolecule. J. Virol. 31:741-751.

5. Hopkins, N., and P. Jolicoeur. 1975. Variants of N-tropic leukemia virus derived fromBALB/c mice. J. Virol. 16:991-999.

6. Linial, M., and W. S. Mason. 1973. Characterization of twoconditionalearly mutants of Rous sarcoma virus. Virology 53:258-273.

7. Mason, W., and C. Yeater. 1977. A mutant of Rous sarcomaviruswith a conditional defect in the determi-nant(s) of viral host range.Virology 77:443-456. 8. Mason,W.S.,C.Yeater, T. V.Bosch, J. A. Wyke,

and R. R. Friis. 1979. Fourteentemperature-sensitive replication mustants of Rous sarcoma virus. Virology 99:226-240.

9. Rapp, U. R., and R. C. Nowinski. 1976. Endogenous ecotropic mouse type C virusesdeficient in replication andproduction of XC plaques. J. Virol.18:411-417.

10.Rein, A.,B.I.Gerwin,R. H.Bassin, L. Schwaim, and G. S.Schialovsky. 1975. Areplication-defective var-iant of Moloney murine leukemiavirus. I. Biological

characterization.J.Virol.25:146-156.

11.Rohrschneider,J.M., H.Diggelman, H. Ogura, R. R. Friis, and H. Bauer. 1976. Defective cleavage of a precursor polypeptide in atemperature-sensitive

mu-tant of aviansarcoma virus.Virology 75:177-187. 12. Rosenberg, N., and D. Baltimore. 1978.Theeffect of

helper virus on Abelsonvirus-induced transformation oflymphoid cells. J. Exp. Med.147:1126-1141.

13. Rowe, W.P.,W.E.Pugh,and J.Hartley.1970.Plaque assaytechniques for murine leukemia viruses.Virology 42:1136-1139.

14. Scheele, C. M., and H. Hanafusa. 1971. Proteins of helper-dependent RSV. Virology 45:401-410. 15. Shields,A.,0. N.Witte,E.Rothenberg,and D.

Bal-timore. 1978.High frequency of aberrantexpression of Moloney murine leukemia virus in clonal infections. Cell 14:601-609.

16.Stephenson,J.R., and S. A. Aaronson.1973. Charac-terizationoftemperature-sensitive mutantsof murine leukemia virus.Virology 54:53-59.

17.Tronick,S.R.,J.R.Stephenson,I.M.Verma, and S. A. Aaronson. 1975.Thermolabile reverse transcriptase of a mammalian leukemia virus mutant temperature sensitive in its replication and sarcoma virus helper functions. J. Virol.16:1476-1482.

18.Witte,0.N.,N.Rosenberg,M.Paskind,A.Shields, and D. Baltimore.1978.Identification ofanAbelson murine leukemia virus-encoded protein present in transformed fibroblast and lymphoidcells.Proc. Natl. Acad. Sci.U.S.A. 75:2488-2492.

19.Wong,P. K.Y.,L. J.Russ,and J. A.McCarter. 1973. Rapid selecti e procedure forisolation of spontaneous temperature-:?nsitive mutants ofMoloney leukemia virus.Virology 51:424-431.

on November 10, 2019 by guest

http://jvi.asm.org/