JOURNAL OF VIROLOGY, OCt. 1976, p. 1-8Copyright C 1976 American Society for Microbiology

Vol. 20, No. 1Printed in U.S.A.

Homologous Interference of Lymphocytic ChoriomeningitisVirus: Detection and Measurement of Interference Focus-

Forming UnitsMIRCEA POPESCU,' HORST SCHAEFER, AND FRITZ LEHMANN.GRUBE*

Heinrich-Pette-Institut fur experimentelle Virologie und Immunologie an der Universitdt Hamburg,2000 Hamburg 20, Federal Republic ofGermany

Received for publication 12 May 1976

Lymphocytic choriomeningitis (LCM) virus defective interfering (DI) particlesform foci of protected cells in a monolayer under an agarose-containing overlaymedium. Foci originate from one cell dually infected with at least 1 interferencefocus-forming unit and infectious virus. As a result, an interfering factor isproduced and released which interacts with neighboring cells, thereby protect-ing them against cytopathic lysis by challenge virus. The property of individualLCM virus DI particles to induce countable foci has been made the basis of aquantitative assay that is comparable in every respect to the plaque assay ofinfectious virus and is much more sensitive and probably more accurate thanother procedures used to measure LCM virus DI particles. LCM virus waspassaged, undiluted, 10 times in cell cultures. When yields were analyzed as toconcentrations of PFU and interference focus-forming units, both entities werefound to fluctuate with the pattern expected from theoretical considerations.

Defective viral particles with the ability tointerfere have been shown to accumulate dur-ing replication of many different kinds of vi-ruses (3), including lymphocytic choriomenin-gitis (LCM) virus (10, 11). In addition to in-hibiting virus replication, LCM virus defectiveinterfering (DI) particles inhibit cytopathic ef-fects (CPE). In this report it is shown that inmonolayer cell cultures individual DI particlesinduce foci that consist of cells protectedagainst the cytolytic activity of infectious virus.This property has been made the basis for anassay that allows the quantitative determina-tion of DI particles in LCM virus preparations.

MATERIALS AND METHODSCell cultures. L cells, and in a few experiments

HeLa, BHK-21, and Vero cells, were used. Growthmedia consisted of Eagle minimum essential me-dium supplemented with non-essential amino acidsand 5% calf serum. The pH was maintained at 7.2with 0.14% NaHCO3 under 5% C02 at 37°C. Plasticpetri dishes, 3.5 cm in diameter, were seeded with 3x 105 cells in 2 ml of medium. Cultures were alwaysused within 24 h, when the numbers of cells did notexceed 6 x 105 and monolayers were still incomplete.

Virus. All experiments were done with the WEstrain ofLCM virus (8). Highly cytolytic (challenge)virus was prepared by 12 intracerebral passages in

I Permanent address: Institutul de Virusologie 'StefanS. Nicolau," Sos. Mihai Bravu nr. 285, Bucharest, Ru-mania.

the mouse brain. It was propagated in L cells andharvested 20 h after infection. Virus rich in DI parti-cles was obtained by undiluted passage in L cells(see Fig. 3). To eliminate infectious virus from prep-arations, these were irradiated with UV light underconditions known to inactivate PFU but to leave DIparticles functionally intact (10). Infectious viruswas titrated by counting plaques on L cell monolay-ers. The overlay medium contained 0.55% agarose,and the total time of incubation was 5 days. Thecells were then fixed with formaldehyde and stainedwith crystal violet (7).

Infective center assay. Virus in 0.1 ml of mediumwas spread onto L cells and allowed to adsorb for 60min at room temperature. The cells were thenrinsed twice and covered with 2 ml of prewarmed(370C) medium to be incubated at 370C for 6 h. Theywere washed again three times, dispersed with tryp-sin, concentrated by centrifugation, and counted.Serially diluted cells were dispensed to L cell mono-layers, which were kept for 30 min at 370C. Thesewere covered with 2 ml of medium containing 0.8%methyl cellulose cP 400. After 4 days at 370C,plaques were visualized and counted as for theplaque assay (see above). Numbers of PFU thatwere not cell associated were determined in parallel,and infective center counts were corrected corre-spondingly.

Determination of the interfering dose. Cultureswere infected with 500 to 1,000 PFU contained in 0.1ml of medium together with the material to betested, and adsorption was for 30 min at 370C. Incu-bation under a medium containing 0.8% methyl cel-lulose and staining with crystal violet followed theprotocol previously described for the infective center

1

on April 27, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

2 POPESCU, SCHAEFER, AND LEHMANN-GRUBE

assay. When CPE had developed, i.e., on day 4, cellprotection was assessed with an empirical scaleranging from + + + + to +. By our definition, one

interfering dose is that amount of interfering mate-rial that exerts a slight but unquestionable reduc-tion of CPE corresponding to + in a 3.5-cm L cellmonolayer culture. This procedure is similar to theone proposed by Welsh and Pfau (11).Making reduction of CPE the basis for an inter-

ference assay is dependent on a general associationof both parameters. The relationship between yieldand CPE has been investigated under a variety ofexperimental conditions, with the consistent resultthat inocula sufficiently high to reduce CPE also ledto a corresponding reduction of the infectious yields.An illustration for this statement may be found inFig. 2.

Total numbers of DI particles. Particles with theability to interfere with virus replication in a partic-ular host cell may be estimated from the effect theyhave on infectious yields during one cycle of multi-plication. A formula was developed by assuming (i)all-or-none one-hit kinetics of interference, (ii) Pois-son distribution of both DI particles and PFU, and(iii) equal numbers of cells in experimental andcontrol cultures:

YE (1 ) n

Yc (1 - e-m(E))

where YE is 16-h yield in terms of PFU from experi-mental culture (infected with known numbers ofPFU and unknown numbers of DI particles), Yc isyield of PFU from control cultures (infected underconditions where infection of one cell with both PFUand DI particles is highly improbable), m(C) is PFUper cell (multiplicity of infection) in control culture,m(E) is PFU per cell in experimental culture, and n

is number of interfering particles per cell. Totalnumber of DI particles in the inoculum may becalculated by multiplying the value for n with thenumber of cells in the culture. With the necessaryadjustments, this formula may also be used to calcu-late total numbers of DI particles on the basis ofreduction of infectious centers.

All-or-none one-hit kinetics was inferred by Welshet al. (10) from dose response data and has beenconfirmed by us in single-cell experiments, of whichone is described (Table 1). The results show thatcells either produce the expected number of infec-tious units (9) or produce none at all ifblocked by DIparticles. Also, the numerical relationship betweenPFU and DI particles leaves little doubt that one DIparticle is capable ofexerting the effect. The validityof this statement was tested statistically. Cultureswere infected with mixtures consisting of infectiousvirus at three concentrations and DI particles ob-tained by UV irradiation at different dilutions, andtotal yields of infectious virus per culture were de-termined after one cycle, corresponding to 16 h ofincubation (5). The data (not shown) are compatiblewith one-hit kinetics and hence do not contradict theassumption expressed above, namely, that the indi-vidual DI particle is capable of blocking the cell.

The numbers of DI particles were also calculatedfrom the reduction they effected on infective centers.

TABLE 1. Effect of DI particles on LCM virusproduction by individual L cells

Avg no. of No. of Cups withKind of cellsP cups as-vcells/cup' ps as-sayed

A. Interfered 1 24 0ca. 8 13 0

B. Productively 1 19 9'infected ca. 8 14 14d

a Petri cultures containing 5.2 x 105 L cells were infectedwith (A) 2 x 106 PFU plus 1.6 x 105 IFU, corresponding to atotal number of approximately 8 x 106 DI particles (forexplanation, see text); (B) 1.7 x 106 PFU plus 2.4 x 103 IFU,corresponding to a total number of approximately 1.2 x 105DI particles.

b Cells were washed, incubated for 6 h, washed again,and dispersed with trypsin, as has been described in connec-tion with the infective center assay. They were counted,and volumes of 20 gl of medium containing either 0.4 or 8cells were distributed into individual cups of microtestplates. After 30 min, trays were examined with a micro-scope and cups that had received an average of 0.4 cells andcontained one and only one cell were marked. After 8 h ofincubation, media from all individual cups were assayed forPFU with entirely negative results. Cups were then filledagain with 20 iLl of prewarmed medium, and after furtherincubation for 8 h the contents from each cup, together with50 /l of medium with which they had been rinsed, wereadded to individual assay cultures; plaques were counted 4days later.

c Numbers of PFU released by single cells were: 168, 263,243, 159, 346, 341, 230, 116, and 280.

d Cytolytic plaques were too numerous to allow counting.

When the values thus obtained were compared withthe numbers of DI particles as calculated on thebasis of yield reduction, the latter were slightlylower (see Table 5). However, taking into accountthe methodological complexities of both methods,the agreement is quite satisfactory, and this weconsider valuable additional evidence strengtheningour assumption that the method of one-cycle yieldreduction measures, indeed, all interfering particlesas they express themselves under our experimentalconditions.

RESULTSInterference focus-forming unit (IFU).

When interfering particles are added to L cellmonolayer cultures together with highly cyto-lytic infectious virus, zones of protected cellsappear. These resemble morphologically thecomet-shaped LCM virus plaques described byHotchin (2), with the obvious difference thatthe latter are made up of lysed or partly lysedcells whereas the comets caused by interferingparticles consist of intact cells. When spread ofvirus is prevented by, for example, agarose inthe medium, foci consisting of islands of viablecells develop which contrast with a back-ground of lysed monolayer cells. After stain-ing, these foci can easily be counted (Fig. 1).Comets or foci, respectively, also form in mon-

J. VIROL.

on April 27, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

INTERFERENCE FOCUS-FORMING UNITS OF LCM VIRUS 3

FIG. 1. Foci in an L cell monolayer caused by IFU. Diameter of the petri plate, 3.5 cm.

olayer cultures of all permanent cell linestested, i.e., HeLa, BHK-21, and Vero cells.The procedure for detecting IFU is as fol-

lows. Cultures containing not more than 6 x105 cells are infected with 0.1 ml of seriallydiluted test material together with 0.1 ml ofchallenge virus containing (in addition to var-iable but low numbers ofIFU) approximately 5x 104 PFU. After 30 min at 37°C, the cells arecovered with medium containing 0.4% aga-rose, and incubation is at 37°C for 2 days. Toaccelerate lysis of unprotected cells, 1.5 ml ofliquid medium is then added to the agarose-containing overlay and incubation is contin-ued for 2 more days. Formaldehyde, 1.5 ml ofa10% solution, is added, and cultures are left atroom temperature for 2 h. The overlay is re-moved and areas of nonlysed cells are stainedwith a 2% solution of crystal violet in 10%formaldehyde. For the calculation of titers,the numbers of IFU added with the challengevirus are subtracted.Development of an IFU assay. The first

question to be answered was whether IFU aredistributed according to the Poisson series.Foci were counted in 98 cultures that had been

inoculated under identical conditions withmaterial from one pool. The mean and thesingle standard deviation were calculated as53.8 and 10.4, respectively. Theoretically, amean of 54 has 99% confidence limits at 37 and76, and lower or higher values should not havebeen observed more than once in this trial; infact, values below 37 were counted six timesand the value of 76 was exceeded once. A re-peat experiment in which counts from 60plates were evaluated led to a mean of 123.6IFU/culture, with a standard deviation of 14.7.In this case, three observed values fell outsidethe 99% confidence limits expected for a meanof 124. From these results we conclude that,just as is the case with virus particles in gen-eral, IFU are Poisson distributed, althoughsome deviation is apparent.Application of this assay depends on predict-

able reductions of counts with dilutions of thematerial to be titrated. In Table 2 the resultsof one of four experiments that were analyzedin this respect are presented. In each case therelationships between numbers of foci andconcentration were found to be linear.To obtain some knowledge concerning the

VOL. 20, 1976

on April 27, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

4 POPESCU, SCHAEFER, AND LEHMANN-GRUBE

TABLE 2. Relationship between dilution andnumber ofIFU counted

No. of IFU in 2 culturesDilution factor

Observed Expected + single SDa

3 256 263 ± 16.24 172 197 ± 14.06 146 131 ± 11.48 101 99 9.9

12 74 66 8.116 65 49 7.024 28 33 5.732 20 25 5.0

a Assuming a Poisson distribution; SD, standarddeviation.

reproducibility of the assay results, one poolwas independently titrated five times. Thevalues obtained varied between 9.0 x 104 and1.6 x 105 IFU/ml. A similar variation wasobserved with numbers of PFU titrated inparallel.Mechanism of focus formation. In close

analogy to measuring infectious centers, thedirect IFU assay as just described has beenmodified to measure the number of interfer-ence focus-forming centers. Cells from cul-tures infected with virus containing IFU arethoroughly rinsed, dispersed with trypsin,counted, and distributed to new monolayercultures. Challenge virus is then added, andcultures are incubated and foci are made visi-ble as described above. It could be shown thatclose to 100% of the developing foci were cellassociated. Also, a direct relationship wasfound to exist between numbers of distributedcells and numbers of foci (data not shown). Wemay conclude that each focus is initiated by anindividual interfered cell.The question was asked whether a focus

represented an area of the monolayer pro-tected by spread from the center to the periph-ery of some interfering factor or whether afocus developed by clonal expansion of oneprotected cell. Thymidine at a concentrationthat blocked cell division but did not affect cellviability (2 mM) was included in the agarose-containing overlay medium, and IFU were as-sayed as described. In two experiments thenumbers of IFU were not diminished as com-pared with the controls, which shows that afocus may develop under conditions in whichcell multiplication is virtually absent. How-ever, foci under thymidine appeared to be lessdense, and we interpreted this to mean thatcell multiplication did play a role. That DIparticle-protected cells may multiply was con-firmed in a further experiment in which thedevelopment of such cells into clones was in-

vestigated. Cultures were infected, and singlecells at a concentration of one per sample weredispensed into microcups. They were incu-bated for 10 days and clones were counted. Inparallel, the same suspensions were distrib-uted into cups containing amplifier cells, andinfection was determined in these after 4 daysof incubation. When evaluating the results, itmust be borne in mind that with an average ofone cell per sample, according to Poisson dis-tribution, we should expect 36.8% of the cupsto have received one cell and 26.4% to havereceived more than one cell. With an averageof 0.5 cell/sample, these values should be 30.3and 9.0%, respectively. Taking this into ac-count, we see (Table 3) that the cloning effi-ciency of cells protected by DI particles ap-proached 100%. In contrast, of the cells calcu-lated not to have been protected by DI parti-cles, approximately 50% did not form coloniesand a similar proportion was shown to haveproduced infectious virus. Evidently, infectionunder these conditions results in cell destruc-tion, whereas protection by DI particles allowscell multiplication to proceed normally. Itshould also be pointed out that cells protectedby DI particles did not release detectablequantities of infectious virus, which confirmsour previous conclusion that interference inthis system is an all-or-none phenomenon.Next we wanted to know when spread of

the interfering factor causing the formation offoci occurred. Cultures infected with IFU plusinfectious challenge virus or with infectiousvirus only were incubated under liquid over-lay medium. This was replaced by agarose-containing medium after 0, 16, 20, and 24 h ofincubation. On day 4, PFU and IFU assays,respectively, were performed. The zero-timecultures contained foci or plaques only. In thecultures in which the liquid medium had beenexchanged against an agarose-containing me-dium at 16 h, approximately 20 and 25% ofcomets had developed in addition to foci orplaques, respectively, and when the mediumwas exchanged after 20 and 24 h comets pre-dominated in both groups of cultures. Fromthese results it may be concluded that the cell-protecting material is released before h 16,and probably at about the same time wheninfectious virus is released.

It also appeared important to know when,in relation to IFU, infectious virus had to beadded for foci to develop. Data in Table 4 showthat IFU cause the formation of foci whenchallenge virus is added up to 12 h before anduntil 16 h after IFU, although the time limitsfor maximal counts appear to be narrower.

Sensitivity of IFU assay. The newly devel-

J. VIROL.

on April 27, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

INTERFERENCE FOCUS-FORMING UNITS OF LCM VIRUS

TABLE 3. Determination of the ability of single cells protected by DI particles to multiply and formcountable colonies

% Calculated % Observed

Type of cells0 b Cells protected by Colonies/cup Cups with vi-Cells infected DI particles rustone >one

A. Interfered 0.00 >99.99 37 24 0B. Productively infected 76.49 23.51 32 3 40.4

Uninfected control 0 0 36 14 0a As in Table 1. In addition, uninfected cultures served control purposes.b Cells were incubated and dispersed as described in connection with the infective center assay (see text).

They were adjusted to one cell/0.05 ml of medium, and these volumes were dispensed into 192 wells of micro-cell culture plates per type of cells. They were incubated for 10 days (0.025 ml of medium added twice), andcolonies were counted.

c Cells were dispersed as described above into 480 wells for each type. Immediately afterwards, 7 x 103amplifier cells in 0.025 ml of medium were added; incubation was for 4 days. Each microculture was thenexamined for infection on the basis of both development of CPE and release of complement-fixing antigen(1). Controls were included to allow quantification of the amount of free virus as well as an assessment of thereliability of the criteria chosen for demonstrating infection. Where suitable, data were corrected corre-spondingly. The percentage of infected microcultures is expressed on the basis of cups calculated to havereceived one or more cells.

oped procedure has been compared in terms ofsensitivity with other methods used in thislaboratory and elsewhere. In a number of ex-

periments interfering dose and IFU were nu-merically compared. From the results of 10independent parallel titrations, one interfer-ing dose was calculated to correspond to 220IFU. One experiment in which additional pa-

rameters were determined is presented in Fig.2. As has already been mentioned, the resultsshow a high degree of correlation between re-

duction of CPE and reduction of infectiousyield. Not surprisingly, the optical impressionof complete protection is apparent rather thanreal, because further reduction of infectiousyields occurred at concentrations of virus

higher than were necessary to cause protec-tion graded + + + +. On the other hand, oneinterfering dose (designated +), which in thisparticular experiment corresponded to 280IFU, did not significantly reduce the yield.

In other experiments total numbers of DIparticles as calculated from reduction of infec-tive centers and/or yield were determined inparallel with numbers ofIFU. In each case theformer exceeded the latter by factors varyingbetween 10 and 100. One observation of thiskind is shown in Table 5. In this case thedifference was, on the average, 15-fold.Evidence has been presented showing that

interference foci develop by spread of some

interfering material rather than by expansionof a clone. Thus, foci can only develop ifenough cells remain free of cytopathogenicchallenge virus such as to allow spread of pro-tection to noninfected neighbors. Therefore,

TABLE 4. Efficiency offocus formation as influencedby the interval of time between addition ofIFU to

cells and challenge with infectious virus

Time (h) of addition of chal- % Focus formationblenge virus to assay culturesrelative to time of addition A B

of IFU"-16 '5 '7-12 20 14-8 105 94-4 105 1110 100 100

+4 83 76+8 68 44+12 44 ND+16 41 68

a Minus sign denotes infectious challenge virusbefore IFU; plus sign denotes challenge virus afterIFU.

b Two preparations with IFU were used: A, IFUfreed of PFU by UV irradiation; B, IFU enrichedrelative to PFU (IFU/PFU, -2) by undiluted pas-sage in cell culture. ND, Not determined.

the multiplicity of infection must not exceed acertain value, which appears to be 0.1. Con-versely, a focus only forms if a cell receivesPFU, in addition to IFU, within certain limitsof time (see above). Hence, the underestima-tion of IFU can, at least in part, be explainedby the procedure used for their quantification.To test this conclusion, the sensitivity of theIFU assay was increased by increasing thenumber of doubly infected cells. Monolayer Lcells were infected with LCM virus contain-ing, in addition to known quantities of IFU,known quantities of PFU. Interference focus-

5VOL. 20, 1976

on April 27, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

6 POPESCU, SCHAEFER, AND LEHMANN-GRUBE

8 -848 HOURS

1 1

76~~~~~~~

C-)

LD0~

0~

2

-3 -2 -1

DIWTION OF INTERFERNG

MATERIAL (LOG10)

FIG. 2. Relationship between reduction ofyield ofinfectious virus and depression of CPE in L cellcultures (interference) and number ofIFU initiallyinoculated. IFU at a concentration of8.77 x 105/mlwere freed of infectious virus by UV irradiation.They were serially diluted 3.16-fold (log1,o 0.5), and0.1 ml was added to L cell cultures concomitantlyinfected with 103 PFU of virus. After 14 and 48 h ofincubation, cells from individual cultures were dis-rupted by sonication and total numbers ofPFU fromeach culture were determined by plaque assay. Inparallel cultures maintained under a methyl cellu-lose-containing overlay medium, protection againstviral CPE was determined and graded + to + + + +.One interfering dose causes protection ofcells graded

forming centers were determined as has beendescribed and were compared with IFU ini-tially added. The results show (Table 6) thatfocus-forming centers exceeded IFU as ini-tially added by factors ranging from 2.3 to11.9. However, the differences were lowerthan the differences between IFU and totalnumbers of DI particles, which were deter-mined in parallel when the conditions weresuitable.Nature of IFU. IFU are readily neutralized

by antiserum directed against all immuno-genic components of the LCM virus. Also, incentrifugation experiments they differ onlyslightly from the infectious virus. However, inmarked contrast to infectious LCM virus, IFUare highly resistant to UV irradiation, a prop-erty that has also been found for LCM virus DI

particles by Welsh et al. (10). When IFU wereadded to monolayers without challenge virus,they quickly disappeared; at most, 10% of theinput could be recovered when such cells weresonically treated after an incubation period of16 h.The nature of the factor that causes an is-

land of cells to be protected against the cyto-lytic action of challenge virus is not yet under-stood. From the way the material spreads un-der liquid medium (formation of "comets"), itappears to be particulate. However, at thetime when release of cell-protecting materialis well under way, namely, after 20 h of incu-bation, numbers ofIFU in the culture were notfound to be increased. Furthermore, after 5days of incubation, that is, at a time when focihad developed to completeness, the number ofIFU recoverable from the cells making up thefoci was even less.

Foci have been consistently found to be neg-ative for antigen, as can be revealed by immu-nofluorescence. They are cytolytically de-stroyed when superinfected with vesicular sto-matitis virus but not when superinfected withcytolytic LCM virus.

Fluctutation of infectious virus and IFUduring undiluted virus passages. To obtain ad-ditional information with regard to the inter-

TABLE 5. Comparative determination ofIFU and DIparticles

Inoculum (x 103) DI particles (x 103) in inocu-lum calculated from:a

PFU IFU Infective Yieldcenters

784 102 2,041 1,266248 32 697 58324.8 3.2 57 46

a Calculation based on reduction of infective cen-ters or yields, respectively. Yield and number ofinfective centers from cultures infected with 2.48 x103 PFU (and 0.32 x 103 IFU) taken as control val-ues.

TABLE 6. Determination of interference focus-forming centers (IFC) formed in L cell cultures

inoculated with varying numbers ofIFUInoculated culture (x 103) Counted/cul- Rtio

ture (x 103) as (IFC/IFU)PFU IFU IFC

1,070 (2.52)a 31.6 117 3.7844 (2.00) 4.73 56.2 11.9214 (0.50) 6.32 42.0 6.6169 (0.40) 0.945 4.84 5.121.4 (0.05) 0.632 3.29 5.216.9 (0.04) 0.095 0.22 2.3

a Multiplicity of infection is given in parentheses.

J. VIROL.

on April 27, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

INTERFERENCE FOCUS-FORMING UNITS OF LCM VIRUS

action between IFU, infectious virus, andcells, yields from 10 virus passages performedwith undiluted inocula were analyzed (Fig. 3).As a rule, high concentrations of PFU wereassociated with high concentrations of IFUand, when such materials were used to initiatethe next passage, reduced yields of either typeof particles resulted. However, the ratio be-tween PFU and IFU varied considerably, andin materials from two passages the valueswere below 1.

Yields corresponding to low concentrations ofIFU, i.e., from passages 2, 5, 7, and 9, inducedCPE, whereas harvests corresponding to theremaining passages all left the cells of inocu-lated cultures visibly unharmed.

DISCUSSIONUnder certain experimental conditions nu-

merous viruses are known to generate defectiveparticles that inhibit the replication of the ho-mologous infectious virus (3). In accordancewith a proposal by Huang and Baltimore, theyare usually called DI particles (4). HomologousLCM virus interference caused by DI particlesis a two-faceted phenomenon; to the extent thatreplication of infectious virus is reduced, devel-opment of CPE is suppressed. This dual natureof LCM virus interference is also the basis ofour observation that LCM virus DI particlesmay function as IFU. All our data are compati-ble with the assumption that interaction be-tween at least 1 IFU with at least 1 PFU withinone cell leads to the formation of a focus. Themolecular mechanisms of this cooperation arenot yet known, and even the mode of action ofIFU is only partly understood. After dual infec-tion ofa cell with both PFU and IFU, a factor isreleased that renders the cells with which itmakes contact refractory to CPE as well as toproduction of infectious progeny virus. Fromthe way it spreads under liquid medium (for-mation of"comets"), we assume that this factoris also particulate. However, it appears to bedifferent from IFU, because their numberswere not increased in cultures at a time whenthe cell-protecting factor was being released. Afinal decision as to whether IFU and thespreading material, which protects cellsagainst the cytolytic destruction by challengevirus, are different entities has yet to be made.Notwithstanding the limitations of knowl-

edge, focus formation by IFU has allowed thedevelopment of a new quantitative assay thathas proved a valuable tool for studying thenature and mode of action of LCM virus DIparticles. The close association between the twoparameters of LCM virus interference, reduc-

7F

0-

LL

0I

,-fLiD-C)I

O 1 5 10NUMBER OF PASSAGES

FIG. 3. Effects ofpassaging LCM virus undilutedin L cell cultures on concentrations ofPFU and IFUin the yields. Cultures were infected and incubatedfor 48 h at 37°C. Cells and media from two cultureswere harvested and treated with an ultrasonic disin-tegrator, and 0.1 ml of this material was used undi-luted to infect new cultures. Remaining virus wasstored at - 70°C to be titrated in parallel at the com-pletion of the passage series.

tion of replication of infectious virus and sup-pression of CPE, has previously been made thebasis for an assay (11), and initially we mea-sured LCM virus interference by using a simi-lar procedure. Although inspection of a cultureplate and assessment of CPE as the basis forestimating interference is convenient and use-ful for many purposes, it is nevertheless insen-sitive and inaccurate when detailed questionsare to be answered. Determining DI particlesfrom the effect they have on either infectivecenters or infectious yield is also open to criti-cism. When the number of cells receiving DIparticles is less than one-half the number ofcells in the culture, reduction of both infectivecenters and infective yield becomes minimaland may be grossly influenced by experimentalvariation. Thus, the limit of reliability of thismethod corresponds to approximately 2 x 105DI particles. Analysis of our data with the aimof comparing the sensitivities of the three pro-cedures under consideration reveals that the

5h

4F

7VOL. 20, 1976

on April 27, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

8 POPESCU, SCHAEFER, AND LEHMANN-GRUBE

IFU assay is over 100 times more sensitive thandetermination of the interfering dose. As con-trasted with ascertaining the number of inter-fering particles on the basis of yield reduction,the sensitivity of the new assay is higher bymore than three orders of magnitude.The high degree of correlation between IFU

on the one hand and reduction of infectiousyield and CPE on the other is illustrated in anexperiment in which yields from 10 undilutedvirus passages were analyzed as to formation ofIFU, formation of PFU, and induction of CPE.As a rule, numbers of IFU increase togetherwith numbers of PFU under conditions whereCPE develops. Alternating passages are char-acterized by decline of both IFU and interfer-ence potential. A full understanding ofthe com-plex pattern observed requires a more detailedstudy of the numerous parameters involved.Even so, we may assume that a dynamic inter-action between infectious virus, interferingparticles, and host cell is the cause of the cycli-cal variation and that our observation is a fur-ther example of a feedback phenomenon pro-posed on theoretical grounds by Huang (3)and verified experimentally for vesicularstomatitis virus (6).

ACKNOWLEDGMENTSWe thank Magdalina Kallay for excellent technical as-

sistance.This work was supported by a research grant (BCT 90)

from the Bundesministerium fur Forschung und Technolo-gie, Bonn.

LITERATURE CITED1. Gschwender, H. H., and F. Lehmann-Grube. 1975. Mi-

cromethod for the titration oflymphocytic choriomen-ingitis virus in cell cultures. J. Gen. Virol. 26:205-208.

2. Hotchin, J. 1972. Slow viruses and neurological dam-age. Monogr. Hum. Genet. 6:172-181.

3. Huang, A. S. 1973. Defective interfering viruses. Annu.Rev. Microbiol. 27:101-117.

4. Huang, A. S., and D. Baltimore. 1970. Defective viralparticles and viral disease processes. Nature (Lon-don) 226:325-327.

5. Lehmann-Grube, F. 1971. Lymphocytic choriomeningi-tis virus. Virology monographs, vol. 10. Springer-Verlag, Vienna.

6. Palma, E. L., and A. Huang. 1974. Cyclic production ofvesicular stomatitis virus caused by defective inter-fering particles. J. Infect. Dis. 129:402-410.

7. Popescu, M., and F. Lehmann-Grube. 1976. Diversityof lymphocytic choriomeningitis virus: variation dueto replication ofthe virus in the mouse. J. Gen. Virol.30:113-122.

8. Rivers, T. M., and T. F. M. Scott. 1936. Meningitis inman caused by a filterable virus. II. Identification ofthe etiological agent. J. Exp. Med. 63:415-432.

9. von Boehmer, H., F. Lehmann-Grube, R. Flemer, andR. Heuwinkel. 1974. Multiplication of lymphocyticchoriomeningitis virus in cultivated foetal inbredmouse cells and in neonatally infected inbred carriermice. J. Gen. Virol. 25:219-228.

10. Welsh, R. M., C. M. O'Connell, and C. J. Pfau. 1972.Properties of defective lymphocytic choriomeningitisvirus. J. Gen. Virol. 17:355-359.

11. Welsh, R. M., and C. J. Pfau. 1972. Determinants oflymphocytic choriomeningitis interference. J. Gen.Virol. 14:177-187.

J. VIROL.

on April 27, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from