Formation, characterization and interfering capacity of a small plaque mutant and of incomplete virus particles of infectious bursal disease virus

297

Virus Research, 4 (1986) 297-309

Elsevier VRR 00244

Formation, characterization and interfering capacity of a small plaque mutant and of incomplete virus particles of infectious bursal disease virus H. Miiller,

H. Lange

Insiitut firVirologie, Justus-Liebig-Uniuersitiit

and

H. Becht *

Giessen, Frankfurter

Strasse 107, D-6300 Giessen, F.R.G.

(Accepted 6 December 1985)

Summary Serial undiluted passages of infectious bursal disease virus in chick embryo cells were accompanied by a von Magnus type fluctuation of infectivity in viral harvests and a gradual decrease of plaque size. From the 9th undiluted passage on, the whole virus population consisted of small plaque-forming virus. The small plaque size remained constant when subsequent infections were carried out at low multiplicities. Small plaque virus interfered with the replication of large plaque standard virus. The small plaque/low yield mutation favoured the generation of defective particles which could be separated from complete particles by their lower densities in CsCl-gradients, where six fractions became visible and could be analyzed separately. Most of the defective virus particles had lost the larger of their two dsRNA segments and showed an aberrant protein composition. They had a very low residual infectivity and were also able to interfere with the replication of complete virus. Infectious mutation,

bursal disease virus, multiplicity defective interfering particles

of infection,

small

plaque/low

yield

Introduction The non-enveloped particles of infectious bursal disease virus (IBDV) with an average diameter of about 60 nm possess a genome of double-stranded (ds) RNA arranged in two segments (Miiller et al., 1979). The capsid is composed of three

* To whom reprint requests should be addressed. 0168-1702/86/$03.50

0 1986 Elsevier Science Publishers B.V. (Biomedical Division)

298 groups of polypeptides with molecular weights (M,) around 90 kDa, 40 kDa to 50 kDa, and 30 kDa; at least the 40 kDa structural protein is derived from a larger precursor polypeptide. Virus preparations grown in chicken embryo (CE) cells contain a relatively high proportion of particles with low densities which form a ‘top component’ in CsCl-gradients. Most of the structural polypeptides of these largely non-infectious virus particles are composed of high M, polypeptides which were believed to represent precursor molecules of the regular structural proteins (Miiller and Becht, 1982). It was suspected that the conditions of virus growth in CE-cells, perhaps the mode of infection under von Magnus conditions (von Magnus, 1951) might be responsible for the formation of these incomplete particles. A reduction of virus yield and an increase in defective particles was noticed after serial passages of reovirus at high multiplicities of infection (MOI; Nonoyama et al., 1970; Schuerch et al., 1974; Ahmed and Graham, 1977). Therefore, passages of IBDV were carried out in CE-cells at high or low MOI, the virus produced under these conditions was characterized, and finally the capacity of these particles to act as defective interfering (DI) particles influencing the replication of standard virus was examined. In the course of these passages the simultaneous appearance of large and small plaques could regularly be observed. From a small plaque a variant strain had been established which, in contrast to the parental virus, had a slower replication rate in CE-cells and was not pathogenic for chicken (Cursiefen et al., 1979a) and which, therefore, could be used as a live vaccine (Cursiefen et al., 1979b). Since the conditions favouring the appearance of plaques with different diameters had not been assessed, the development of plaque sizes was followed during the passages. In particular, the structural properties and the interfering capacity of the virus derived from small plaques were investigated.

Materials and Methods Cells CE-cells were prepared according to standard procedures and grown as monolayers on plastic Petri dishes (5 cm diameter; 2 to 3 x lo6 cells/dish when confluent) or in Roux bottles for large scale virus production. Virus, infection of cells, and titration of infectivity For the preparation of stock virus, 4- to 5-wk-old specific-pathogen-free chickens (Lohmann, Cuxhaven, F.R.G.) were infected with IBDV strain Cu-1 (Nick et al., 1976) by intrabursal application of the virus (Kaufer and Weiss, 1976). About 26 to 30 h later, when first signs of illness became obvious, bursal tissue was collected from infected birds and stored at -70°C as a 10% (w/w) suspension in phosphate buffered saline (PBS). CE-cells washed once with PBS were infected, incubated for 60 min at 39°C washed again three times and incubated further with 4 ml of Eagle’s minimal essential medium (MEM) for 20 h.

299 Infectivity titrations were performed as plaque assays (Nick et al., 1976); neutral red was added directly to the agar-overlay to give a final concentration of 1 : 5000, and plaques were finally scored after 3 days of incubation at 39°C. Purification of virus IBDV grown in CE-cells was concentrated from the culture medium by ultracentrifugation in a type 19-rotor (Beckman) at 19000 rpm, 4°C for 6 h. Cellular debris had been removed before from the virus-containing culture medium by low-speed centrifugation using a GSA-rotor (DuPont/ Sorvall). Virus pellets were resuspended in TNE-buffer (10 mM Tris-HCl, pH 7.2; 100 mM NaCI; 1 mM EDTA) by ultrasonication (Sonifier B-12, Branson Sonic Power Company, Danbury, CT, U.S.A.). Virus suspensions were clarified again (SS3Crotor, DuPont/ Sorvall; 20000 rpm, 4°C 20 min) and layered on top of CsCl-solutions in TNEbuffer with densities of 1.30 g/ml and 1.40 g/ml, and were then centrifuged for 3.5 h at 25 000 rpm, 4°C in a SW 28-rotor (Beckman). Clearly visible virus-containing fractions were collected, diluted 1 : 3 with buffer, and centrifuged again onto cushions of CsCl in a SW 41Ti-rotor (Beckman) for 2.5 h at 36000 rpm, 4°C. After this run, virus was collected again and adjusted to a mean density of 1.325 g/ml by the addition of CsCl or TNE-buffer. These preparations were centrifuged in a SW 60Ti-rotor (Beckman) at 40000 rpm, 4°C for 19 h. Brakes were not used at the end of the run. In the middle part of the tubes shallow gradients had formed under these conditions, and six virus-containing bands were clearly visible in a beam of light applied from the bottom of the tubes. They were collected separated by puncturing the tubes from the side with very thin hypodermic needles; their designation is 1 to 6 from top to bottom. Polyacrylamide gel electrophoresis (SDS-PAGE) Viral RNA was analyzed on 7.5% polyacrylamide slab gels prepared according to Laemmli (1970) after treatment of the IBDV particles with proteinase K in the presence of sodium dodecyl sulfate (SDS) as described elsewhere (Miiller et al., 1979). Structural polypeptides of IBDV particles were analyzed on 12.5% polyacrylamide gels as previously described (Muller and Becht, 1982) using the discontinuous SDS-gel system. After electrophoresis on polyacrylamide gels, viral RNA and structural polypeptides were visualized by silver staining using the method described by Follet and Desselberger (1983); staining in silver nitrate was for 90 min in the case of RNA and for 30 min in the case of proteins, respectively. Thereafter, the gels were photographed, individual lanes were cut out and the absorption of bands was measured at 580 nm with a Gilford 240 spectrophotometer equipped with a model 2520 linear transport accessory (Gilford Instrument Laboratories Inc., Oberlin, OH, U.S.A.). Electron microscopy and estimation of the number of virus particles Specimens were negatively stained with 1% (w/v) uranyl acetate and observed under a Siemens Elmiscope 101 at 80 kV and 40000 instrumental magnification.

300 The number of virus particles was estimated as described by Follet and Desselberger (1983) by counting virus particles in a mixture of purified virus particles and latex beads of known size and concentration (average bead diameter 91 nm; concentration 1.37 x 10” beads/ml; Balzers Union AG, Balzers, Liechtenstein).

Results Generation

of a small plaque-variant

by serial undiluted passages

In a first series of experiments the effects of various dilutions of the inoculum on virus yields were tested. The starting material was a homogenate of the bursa Fabricius of IBDV-infected birds. Virus had been purified through three consecutive plaque passages and propagated in CE-cells. When this culture medium (titer 1 X 10’ plaque-forming units (PFU)/ml) was inoculated onto CE-cells either undiluted or in dilutions of lo-‘, lop2 and 10m3, which corresponded to about 10 to 0.01 PFU/cell, it became obvious already after the first passage that diluting the inoculum resulted in higher virus yields (Fig. 1). Upon further passages virus titers varied considerably, the highest drops and subsequent rises of virus titers occurring after about every fourth undiluted passage which corresponds to a typical von Magnus phenomenon. Figure 1 also shows that after 12 consecutive passages at dilutions of lop3 the titers remained at a relatively constant level close to 10’ PFU/ml. The infection of CE-cells with a diluted virus inoculum resulted in larger plaque diameters than inoculation with undiluted culture medium. From the ninth passage on large plaques (LP) had a constant mean diameter of about 5 mm while after undiluted passages plaques only reached diameters from 1 mm to a maximum of 3 mm (SP). For the preparation of seed virus for virus production on a large scale, LP-virus and SP-virus from the 11th diluted or undiluted passage was therefore

0.01 0 2 1

c 1

1

’

6 8 NUMBER

10 OF

1

’

12 11 16 PASSAGES

1

18

(

20

1

Fig. 1. Virus yields after diluted and undiluted serial passages of IBDV in chick embryo (CE) cells. For successive passages, culture media were either diluted 1 : 1000 with MEM, or were left undiluted, and 0.2 ml of each was used to infect CE-cells. After incubation at 39’C for 20 h infectivity titers were determined by plaque assays on CE-cells.

301

1.29 g/ml

LP -VIRUS

_

1.30

_

1.33 g/ml

SP-VIRUS

Fig. 2. Formation of virus bands by centrifugation of IBDV in linear CsCl-gradients as described in Materials and Methods. Virus was produced by infecting CE-cells with LP-virus at a MO1 of 0.01 PFU/cell, and SP-virus at a MO1 of about 0.5 PFU/cell. The bands in the gradient were visualized by applying a beam of light from the bottom of the tubes. The electron micrographs show IBDV particles present in both variants with an irregular morphology accumulating at a buoyant density of 1.29 g/ml (fraction number 1) and at buoyant densities of 1.30 g/ml to 1.33 g/ml (fraction numbers 2 to 6) negatively stained with uranyl acetate. Particles have diameters of about 60 nm.

again plaque-purified three times. The plaque size of the SP-variant remained constant when it was carried through eight consecutive passages at dilutions of 10-3; the titers, however, scaled up considerably. On the other hand, passages with undiluted LP-virus resulted in a gradual drop of virus titers and a decrease in plaque sizes (data not shown). Physico-chemical characteristics of LP- and SP-virus particles When LP-virus or SP-virus was grown in CE-cells and was purified by gradient centrifugation, six clearly visible bands formed in linear CsCl-gradients with mean densities of 1.325 g/ml, and mere inspection of the tubes showed that the highest amount of LP-virus particles accumulated in virus fractions 5 and 6, corresponding to buoyant densities of 1.32 g/ml and 1.33 g/ml, respectively, whereas in tubes containing SP-virus particles fraction 2 in the upper part of the gradient at a buoyant density of about 1.29 g/ml was the most prominent one (Fig. 2). The upper band with the lowest density contains empty shells and irregular masses, whereas virus particles in the other bands with higher densities have the regular morphology of IBDV particles. Figure 3 summarizes the relative distributions of virus particles of the two variants throughout the gradients. From the optical densities at 260 nm and 280 nm it is evident that the relative amount of RNA increases towards the

302 3.0-

*

2.5.

1

2 FRACTION

3

1

5

6

NUMBER

Fig. 3. Relative distributions of protein and RNA content in IBDV particles separated by gradient centrifugation. Here the optical densities (OD) at 260 nm (light columns) and 280 nm (dark columns) of the six virus bands becoming visible in the final self-forming gradient are shown. The density of the gradient within the range of the six fractions is shown as an insert in B. (A) LP-virus; (B) SP-virus.

bottom of the tubes, which is correlated with an increase in infectivity. The relative differences of infectivity among the six fractions became particularly striking when the ratios of infectivity per number of physical particles were compared. It is evident from Table 1 that for the LP-variant the particle/PFU ratio was about lo4 in the low density fractions, but close to 10 in fractions 5 and 6 with high densities. When the average size of plaques formed by virus particles in the individual fractions were compared, there was a slight but steady increase in plaque diameters from top to bottom of the gradients (Table 1). All fractions of the SP-variant contained smaller proportions of infectious virus than the corresponding fractions of the LP-variant. Interference of SP-virus with the formation of large plaques In a further series of experiments the capacity of the SP-variant for interfering with the replication of LP-virus, virus yields and formation of plaques, was tested. LP-virus particles accumulating in fractions 5 and 6 of CsCl-gradients were adjusted to a MO1 of 0.01 PFU/cell. This dilution of seed virus was chosen because it had become evident from the results partly presented in Fig. 1 that these conditions were optimal for highest yields of LP-virus. The results of mixed infections with purified virus particles of the SP-variant presented in Table 2 show that SP-virus caused a considerable depression of virus yields and a decrease in plaque sizes. The highest concentrations of SP-virus in mixed infections resulted in values which were almost identical to those obtained with the SP-variant alone, and its influence gradually declined as the inocula of the interfering virus were diluted.

303 TABLE

1

DETERMINATION OF BUOYANT DENSITIES. PLAQUE OF LP-VIRUS AND SP-VIRUS PARTICLES SEPARATED IBDV variant

a

Fraction

number

h

Buoyant

SIZES AND PARTICLE/PFU-RATIOS BY GRADIENT CENTRIFUGATION

Number

density

of particles/PFU

’

Plaque size ’ (mm)

&/ml) LP-virus

1 2 3 4 5 6

1.294 1.298 1.303 1.314 1.328 1.330

5.1 x 104 3.6 x lo4 5.9x10’ 1.2X101 2.8 x 10’ 2.5~10’

4.4 4.5 4.8 5.0 5.1 5.0

SP-virus

1 2 3 4 5 6

1.293 1.299 1.303 1.313 1.328 1.332

2.3x 10’ 2.1 x 10” 4.6~10” 6.3 x 10’ 1.1 Xl02 1.0x102

2.2 2.2 2.5 2.8 3.0 2.9

“ Generation of the two variants forming large plaques (LP-virus) or small plaques (SP-virus) by diluted or undiluted serial passages of IBDV in CE-cells is described in the text. h Fractions were prepared as described in the legend to Fig. 3. ’ Particle numbers were estimated by mixing virus samples with latex beads of known size and concentration and counting under the electron microscope. Infectivity was determined by plaque assay. d Mean diameter of at least 50 plaques. TABLE

2

INTERFERING Inoculum LP-virus MO1 0.01 _ _

_

0.01 0.01 0.01 0.01 0.01 0.01 0.01

ACTIVITY

h SP-virus MO1 _ 100 10 1 0.1 0.01 0.001 0.0001 100 10 1 0.1 0.01 0.001 0.0001

OF SP-VIRUS

WITH

THE REPLICATION Plaque size d

Virus yield ’ (PFU/ml)

(mm)

2.3~10s

4.2

7.3 x 105 4.1 x 10h

2.1 2.3

3.2 x 2.7 x 2.4x 3.8 x 7.8 x

10’ 10’ 10’ 10’ 10h

2.2 2.4 2.2 2.1 2.3

2.5~10” 8.3 x 10h 2.6 x 10’ 4.5 x 10’ 5.1 x 107 1.8 x 10’ 2.5 x lo*

2.2 2.1 2.1 2.3 3.0 4.2 4.2

OF LP-VIRUS Reduction

’

in yield ’

_

_ _

98.9 96.4 88.7 80.5 77.9 21.7 0

a Generation of the two variants forming large plaques (LP-virus) or small plaques (SP-virus) by diluted or undiluted serial passages of IBDV in CE-cells is described in the text. h IBDV particles of both variants with the closest ratio of physical to infectious particles (buoyant densities 1.32 g/ml to 1.33 g/ml; Table 1) had been prepared by centrifugation in CsCi-gradients. ’ Infectivity was assayed 20 h after infection of CE-ceils with either virus or both viruses simultaneously. Mean of duplicate cultures. d Mean diameter of at least 50 plaques. e Percent of LP-virus.

304 SP-virus at a multiplicity lower than 0.01 PFU/cell was no longer capable of inhibiting the growth rate of LP-virus in mixed infections. The virus population produced under these conditions formed a variety of plaques of all sizes, including large plaques of the original diameter of 4.2 mm. The average of plaque diameters listed in Table 2 expresses the tendency towards large plaque formation. Analysis of RNA segments by SDS-PAGE SDS-PAGE analysis of the dsRNA encapsidated in the IBDV particles from all fractions of CsCl-gradients showed that the two segments migrated the same distance in the gels and were therefore almost equal in size. This was the case for analogous fractions of LP-virus and for SP-virus. However, while virus particles with buoyant densities of 1.32 g/ml to 1.33 g/ml (fractions 5 and 6) contained approximately equimolar amounts of either segment, low density-particles from

A

B 123456

1

2

Fig. 4. SDS-polyacrylamide gel electrophoresis of the two RNA segments of the SP-variant. Conditions for virus production and gradient centrifugation were the same as for Fig. 2. Equal numbers of physical particles (2 x 10”) from each fraction were pelleted at 50000 rpm for 60 min at 4°C in a SW 60 Ti-rotor (Beckman). After SDS/proteinase K-treatment RNA was analyzed on 7.5% polyacrylamide gels, visualized by silver staining (A) and monitored at 580 nm (B) as described under Materials and Methods. Migration in the electric field was from top to bottom in (A), and from left to right in (B). Numbers refer to gradient fractions as for Fig. 2.

305

Fig. 5. SDS-polyacrylamide gel electrophoresis of structural proteins of LP-virus (A) or SP-virus (B). Virus particles accumulating in the six fractions of the density gradient described in Fig. 2 were collected and pelleted at 50000 rpm for 60 min at 4°C in a SW 60 Ti-rotor (Beckman). After boiling for 2 mm in electrophoresis buffer (60 mM Tris-HCl, pH 6.8; 2% SDS; 25% glycerol. 2 M urea; 5% 2-mercaptoethanol; 0.1% bromophenol-blue) polypeptides were separated on 12.5% polyacrylamide gels from top to bottom and visualized by silver staining. Numbers on top of the figures refer to gradient fractions as in Fig. 2. Numbers at the side correspond to apparent M, in kDa. LP-virus analyzed in (A) was grown in only half the number of cells than SP-virus.

fractions in the upper parts of the gradients had disproportionate RNA contents (Fig. 4): in fractions 3 and 4 there were reduced amounts of the slower migrating segment when compared to the faster migrating one, and in fractions 1 and 2, the more slowly migrating segment was almost missing. Figure 4 presents the results of experiments performed with the RNA extracted from the same number of physical particles of the SP-variant; essentially identical results were obtained with the LP-virus. Analysis of structural proteins of LP-virus

and of SP-virus

by SDS-PAGE

The polypeptide patterns of LP-virus particles and of SP-virus particles from the six fractions visible in CsCl-gradients were compared by electrophoresis in polyacrylamide gels (Fig. 5). Virus particles present in fractions 5 and 6 of both variants showed the typical polypeptide composition of standard IBDV particles and were, therefore, designated as ‘complete’, Virus particles with lower densities, however, which accumulated in fractions 1 and 2 of SP-virus preparations, showed marked differences. Numerous protein bands with a M, between 90 kDa and 50 kDa could

306 TABLE

3

INTERFERING ACTIVITY OF INCOMPLETE COMPLETE VIRUS PARTICLES a IBDV variant

LP-virus

Inoculum (Number

h of particles/cell

IBDV PARTICLES

Virus yield ’ r) (PFU/ml)

WITH THE REPLICATION

Plaque size ’

Reduction

Complete

_ 5 000 1000 100

100 100 100 100 100

x.1 x 105

2.9

_

5 000 1000 100

100 100 100

2.2 x 10s 4.1 x IO5 5.x x 10”

2.2 2.2 2.3

12.9 49.4 28.4

SP-virus

in yield ’

(mm)

Incomplete

6.6 x 10’ 1.2x107 2.2 x 10’ 3.6 x lo7

OF

5.0 4.6 4.1 4.6

X1.X 66.1 45.5

Complete IBDV particles correspond to particles accumulating in fractions 5 and 6, and incomplete to particles in fractions 1 and 2 in self-forming C&-gradients (see Table 1). Infection of CE-cells was as in Table 2. Infectivity was tested after 20 h by plaque titrations. Mean of duplicate samples Mean diameters of at least 50 plaques. Percent of complete virus particles. Particle numbers were estimated by mixing virus samples with latex beads of known size and concentration and counting under the electron microscope.

be seen. In the medium size group, there was an excessive quantity of a protein with a M, of about 46 kDa besides the usual 48/49 kDa polypeptides, but the protein with a M, of 40 kDa, one of the main structural proteins, was almost absent. Instead, an additional polypeptide with a M, of 36 kDa was seen which was never observed in complete particles of fractions 5 and 6. The other structural polypeptides of the standard type with M, of about 32 kDa were also missing in these low density particles, or they could be detected in trace amounts only. On the other hand, virus particles recovered from the relatively small fraction 2 of LP-virus preparations consisted mainly of polypeptides with 50 kDa and 46 kDa whereas the 36 kDa-protein was inconsistently present in some gels in minute amounts. Interference of incomplete particles with the replication of complete particles separated in the same grudient The structural characteristics of the incomplete low density particles accumulated in fraction 2 suggested that they correspond to the DI type. Their potential influence on the replication of complete standard IBDV particles in fraction 5 was therefore tested by mixing varying amounts of particles from fraction 2 (between 100 and 5000 particles per cell) with a constant number of particles from fraction 5 (100 particles per cell) before their inoculation onto CE-cells. As shown in Table 3, increasing numbers of incomplete particles resulted in decreasing virus yields; furthermore, plaque diameters were reduced significantly. Essentially the same results with respect to virus yields and plaque size were obtained with LP-virus or the SP-variant.

307 Discussion It is evident from Fig. 1 that after infection of CE-cells with IBDV at a MO1 of 1 to 10 PFU/cell the periodic fluctuations in virus titers follow the well known von Magnus pattern (von Magnus, 1951; Cooper and Bellet, 1959; Huang et al., 1966). However, during these passages the observation was remarkable that plaque sizes decreased from each passage to the next until a rather uniform small plaque size was reached after about ten passages. This type of small plaque was already visible as a relatively small fraction among plaques during the initial undiluted passages. Diluting the inoculum favours the formation of virus particles capable of forming large plaques, because their yield and replication rate is considerably higher than that of the small plaque variant as previous results (Cursiefen et al., 1979a) had demonstrated. This means that the relatively small proportion of SP-virus which has a chance to be formed under the conditions of replication at low MO1 does not grow to titers high enough to reach the dilution steps in plaque assays where individual plaques become visible. Passages in CE-cells at high MOI, however, favour their enrichment because they can exert an inhibitory effect on the formation of LP-virus which finally results in a total repression of large plaques in a high-passage virus population. This interference could clearly be assessed by the results presented in Table 1. Simultaneous infection of CE-cells with purified particles of LP-virus and of SP-virus did not only cause a drop in virus titers but also a reduction of plaque size. This small plaque mutant must be stable because upon further passages at low MO1 virus titers increased but plaques of the large type never reappeared. One can conclude from these observations that SP-virus only partially meets the characteristics which have been attributed to DI particles (Huang and Baltimore, 1977; Holland et al., 1980; Perrault, 1981; Dimmock, 1985). The interfering capacity of the SP-variant is one of the effects of a mutation which so far can only be described phenotypically as small plaque/low yield. In contrast to the definition of DI particles reviewed by these authors, SP-virus represents a self-replicating entity whose propagation does not depend on a helper virus and whose defectiveness only concerns the rate and efficiency of its replication. ‘Complete’ IBDV particles of LP-virus and the SP-variant accumulating in fractions 5 and 6 of CsCl-gradients were almost all infectious, whereas a ten-thousand-fold smaller specific infectivity was measured in the fractions with lower buoyant densities of both variants as an expression of their defectiveness. One can assume that this drastic drop in the ratio of physical to infectious virus particles is mostly the result of their inconsistent RNA contents. It is clear from Fig, 4 that in these fractions there are virus particles without any RNA and particles where one RNA segment, mostly the larger one, is missing. A particularly high proportion of such incomplete particles (in the gradient fractions 1 and 2) is produced by the SP-variant after the infection of CE-cells at high MOI. Their polypeptide pattern shows that there is an accumulation of large polypeptides, which are most likely precursor molecules, at the expense of proteins with M, of 40 kDa and 32 kDa which in standard particles function as the main

308

structural polypeptides. Instead, a new polypeptide with the molecular weight of 36 kDa became regularly visible in particles from fraction 2 of SP-virus. One could imagine that the mutation leading to the formation of small plaques has an effect, perhaps only a side effect, on the structure of a precursor protein where another domain becomes sensitive and accessible to processing by cellular proteases besides the usual cleavage site. The incomplete virus particles in fraction 2 of the gradient depressed the harvest of complete virus from fraction 5 when CE-cells were infected simultaneously with a defined number of particles from either fraction of the same gradient. This means that defective particles from the same virus preparation disturb the replication of complete virus, and since the inhibitory effect is proportional to the number of physical particles in the incomplete fraction, one can assume that those particles with a deleted RNA segment are the main cause of this interference. A similar situation seems to exist among reoviruses where repeated passages at high MO1 promoted the formation of defective particles. Although it had not been possible to isolate DI particles of reovirus in pure form, it could be shown that one of their RNA segments is missing (Nonoyama and Graham, 1970; Schuerch et al.. 1974). Ahmed and Fields (1981) found that besides this type of deletion DI particles contain numerous mutants, including plaque size/low yield mutations. These authors proposed that in contrast to nonsegmented viruses interference by reovirus DI particles is not determined by genes which are deleted but by the expression of mutant genes and incorporation of defective proteins into progeny virus. The SP-variant of IBDV fits into this model and underlines the significance of plaque size mutations. Furthermore, the protein with the M, of 36 kDa, which only appeared in the defective particles of fraction 2 in the gradient, most likely represents such an incorrect protein whose incorporation into virus particles could result in an unbalanced encapsidation of the two segments of the genome into progeny virus.

Acknowledgements

This study was supported by the Deutsche Forschungsgemeinschaft (Sonderforschungsbereich 47). We are obliged to Dr. K. Wahn for the help with the electron microscope. We thank Ingrid Staiger and Claudia Haas for excellent technical assistance.

References Ahmed, R. and Graham, A.F. (1977) Persistent infection in L cells with temperature sensitive mutants of reovirus J. Virol. 23, 250-262. Ahmed, R. and Fields, B.N. (1981) Reassortment of genome segments between reovirus defective interfering particles and infectious virus: construction of temperature-sensitive and attenuated viruses by rescue of mutations from DI particles. Virology 111, 351-363. Cooper, P.D. and Bellett, A.J.D. (1959) A transmissible interfering component of vesicular stomatitis virus preparations. J. Gen. Viral. 21. 485..-497.

309 Cursiefen, D., Kaufer, I. and Becht, H. (1979a) Loss of virulence in a small plaque mutant of the infectious bursal disease virus. Arch. Virol. 59, 39-46. Cursiefen, D., Vielitz, E., Landgraf, H. and Becht, H. (1979b) Evaluation of a vaccine against infectious bursal disease virus in field trials. Avian Pathol. 8, 341-351. Dimmock, N.J. (1985) Defective interfering viruses: modulators of infection. Microbial. Sci. 2, l-7. Follet, E.A.C. and Desselberger, U. (1983) &circulation of different rotavirus strains in a local outbreak of infantile gastroenteritis: Monitoring by rapid and sensitive nucleic acid analysis. J. Med. Viral. 11, 39-52. Holland, J.J.. Kennedy, S.I.T., Semler, B.L., Jones, C.L., Roux, L. and Grabau, E.A. (1980) Defective interfering RNA viruses and the host cell response. In: Comprehensive Virology, Vol. 16 (FraenkelConrat, H. and Wagner, R.R., eds.), pp. 137-192. Plenum Press, New York. Huang, AS. and Baltimore, D. (1977) Defective interfering animal viruses. In: Comprehensive Virology, Vol. 10 (Fraenkel-Conrat, H. and Wagner, R.R., eds.), pp. 73-116. Plenum Press, New York. Huang, AS., Greenwalt, J.W. and Wagner, R.R. (1966) Defective T particles of vesicular stomatitis virus. I. Preparation, morphology and some biologic properties. Virology 30, 161-172. Kaufer, 1. and Weiss, E. (1976) Electron-microscope studies on pathogenesis of infectious bursal disease virus after intrabursal application of the causal virus. Avian Dis. 20, 4833495. Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T 4. Nature (London) 227, 680-684. Miiller, H. and Becht, H. (1982) Biosynthesis of virus-specific proteins in cells infected with infectious bursal disease virus and their significance as structural elements for infectious virus and incomplete particles. J. Virol. 44, 384-392. Mtiller, H., Scholtissek, C. and Becht, H. (1979) The genome of infectious bursal disease virus consists of two segments of double-stranded RNA. J. Viral. 31, 584-589. Magnus, P. von (1951) Propagation of the PR8 strain of influenza virus in chick embryos, 3. Properties of the incomplete virus produced in serial passages of undiluted virus. Acta Pathol. Microbial. Stand. 29, 156-181. Nick, H., Cursiefen, D. and Becht, H. (1975) Structural and growth characteristics of infectious bursal disease virus. J. Virol. 48, 261-269. Nonoyama, M., Watanabe, Y. and Graham, A.F. (1970) Defective virions of reovirus. J. Virol. 6, 226-236. Perrault, J. (1981) Origin and replication of defective interfering particles. Curr. Top. Microbial. Immunol. 93, 151-196. Schuerch, A.R., Matsuhisa, T. and Joklik, W.K. (1974) Temperature-sensitive mutants of reovirus. VI. Mutants ts 447 and ts 556 particles that lack either one or two L genome RNA segments. Intervirology 3, 36-46. (Manuscript

received

16 September

1985)