- Email: [email protected]
Paradoxical effects of sendai virus di rna size on survival: Inefficient envelopment of small nucleocapsids
VIROLOGY
165, 33 l-337
(1988)
Paradoxical Effects of Sendai Virus DI RNA Size on Survival: Inefficient Envelopment of Small Nucleocapsids GIAN G. RE’ Department
of Virology
and Molecular Received
Biology, December
AND
DAVID W. KINGSBURY
St. Jude Children’s 29, 1987;
accepted
Research
Hospital,
March
29,
Memphis,
Tennessee
38 10 1
1988
In the assembly of nonsegmented negative-stranded RNA viruses, such as Sendai virus, the envelopment process allows extensively deleted genomes to survive by transmission from cell to cell in virus particles. To assess the impact of the sizes of such defective-interfering (DI) genomes on their survival, we performed competition tests among various species. Among copy-back DI RNAs, a 450-base species was gradually eliminated from DI virions by a 1200-base species, and the latter was independently eliminated by a 2800-base species. In each case, the smaller RNA species was synthesized and encapsidated at least as efficiently as the larger species, revealing that the level of competition was at the envelopment step in virus assembly. In contrast to the results obtained with the copy-back DI RNAs, repeated high multiplicity passage of a family of four internally deleted RNAs eliminated all but the smallest species, comprising about 1600 bases. Both sets of findings can be reconciled by the hypothesis that the efficiency of o 198s DI nucleocapsid envelopment decreases progressively when the RNA is smaller than about 1600 bases. Academic
Press, Inc.
We have now performed competition experiments between Sendai virus DI RNA species that have homologous 3’ termini, either copy-back or genomic, to assess the effects of DI RNA size as a third variable. Unexpectedly, and in disagreement with conclusions from a previous study (Kolakofsky, 1979) we found that DI RNA species smaller than 1600 bases were gradually eliminated from virions in mixed infections with larger DI RNAs.
INTRODUCTION Defective-interfering (DI) genomes of negativestranded RNA viruses are replicated by the RNA synthesizing machinery supplied by nondefective helper virus (Perrault, 1981). If the rate of RNA polymerization is the same on all virus-specific templates, a small RNA molecule will be completed more quickly than a larger one. Huang and Baltimore (1970) proposed this explanation for the amplification of DI genomes in competition with their helpers. An in vitro correlate is the decreased replication of bacteriophage C&3 RNA in competition with fragments that retain RNA polymerase initiation sites (Mills et a/., 1967). By analogy, in mixed infections with two or more DI RNA species, larger species should eventually be outgrown by smaller species upon repeated passage. However, our previous studies of competition between Sendai virus DI RNA species uncovered at least two other factors that outweigh RNA size. The first of these factors is the possession of a genomic 3’ terminus, which is less efficient as a promoter of RNA synthesis than the copy-back 3’ terminus. (The latter is homologous to the 3’terminus of the positive-strand intermediate in virus genome replication.) The second factor is retention of transcriptional initiation and termination signals in internally deleted DI RNAs. Transcriptional activity also confers a selective disadvantage (Re and Kingsbury, 1986). ’ To whom
requests
for reprints
should
MATERIALS
AND
between
DI RNAs
Competition
METHODS
Properties of the Sendai virus DI RNA species used in this work are summarized in Fig. 1. The naturally occurring mixture of internally deleted RNAs of strain 7 (Re et al., 1985) and artificial mixtures of strains containing copy-back RNAs were serially passaged in 1 lday embryonated chicken eggs. Chorioallantoic cavities were inoculated with a total of 1 hemagglutinating (HA) unit of virus in 0.1 ml of phosphate-buffered saline. After 3 days at 35”, progeny virus was recovered from the allantoic fluid, and the viral RNA was isolated (Re and Kingsbury, 1986). Nucleocapsids
from chorioallantoic
membranes
Chorioallantoic membranes from infected eggs were washed twice with phosphate-buffered saline, cut into l-cm squares, and homogenized in 50 mM Tris-HCI, 150 mM NaCI, 0.5% sodium deoxycholate,
be addressed. 331
0042-6822/88
$3.00
Copyright 0 1999 by Academic Press. Inc. All rights of reproducton in any form reserved.
332
RE AND
KINGSBURY
1% (v/v) NP-40 (pH 8.0) in a large Dounce homogenizer. After removal of debris by low-speed centrifugation, nucleocapsids were isolated by centrifugation in a step gradient. After 90 min at 40,000 rpm in a Spinco SW40 rotor, nucleocapsids formed a visible band at the interface between 50% (v/v) glycerol (density 1.18 g/cm3) and 74% (w/v) sucrose in D20 (density 1.33 g/cm3) (Deshpande and Portner, 1984). Both solutions contained 10 mM Tris-HCI, 30 mM NaCl (pH 7.6). Nucleocapsid RNA was extracted by the same procedure used to extract RNA from virions (Re and Kingsbury, 1986).
Sendai Virus DI RNAs Bases
1-- 2,600 - 2,000 - 1,600
RNA analyses
I -- 1.200
RNA was terminally labeled with cytidine 3’,5’-[5’32P]bisphosphate by the action of bacteriophage T4 ligase, denatured with glyoxal, and electrophoresed in acid urea-agarose gels (Re et a/., 1983). In all cases, each lane received an equal amount of radioactive RNA. For blot hybridization, the RNA was transferred to nitrocellulose paper after electrophoresis and separate strips were hybridized with a 529-base cDNA probe representing the 3’ end of the NP gene and a loo-base cDNA representing the 5’ end of the L gene.
RESULTS Properties
of copy-back
RNA species
Three copy-back Sendai virus DI RNA species that differed markedly in size (Fig. 1) were examined in competition experiments. The largest, designated Rb, comprises about 2800 bases, almost 20°b as large as the parental virus genome. RNA Rb had originally been observed in the company of a slightly larger species, Ra (Amesse et al., 1982; Re et al., 1983), which was eliminated by repeated high-multiplicity passages of strain R (Re and Kingsbury, 1986). RNA 1 la was described as a single 1200-base species in Sendai virus strain 11 by Amesse et al. (1982). It was shown to have a copy-back 3’ terminus by Re et al. (1983). The smallest DI RNA used in this work, the 450base species designated Wa, has not been described before. We encountered it unexpectedly in a virus stock provided by Drs. W. Gorman and A. Pot-trier. Because it is markedly smaller than any Sendai virus DI RNA described previously, we considered it to be of special interest for the present studies. Blot hybridizations with cloned cDNA probes representing the 3’ terminus of the NP gene and the 5’ terminus of the L gene revealed only L-specific sequences (Fig. 2A), indicating that Wa is a copy-back RNA. The presence of the copy-back 3’-terminal sequence was confirmed by wandering-spot analysis (Fig. 2B) and by dideoxynu-
-
450
FIG. 1. Sendai virus DI RNA species used in the competition experiments The left panel depicts the four internally deleted genomes in strain 7 (Re et a/., 1985). In the right panel, the three copy-back DI RNA species found in strains R, 11, and W are shown. Black triangles, genomic 3’ends; black squares, copy-back B’ends; white squares, genomic 5’ ends; black segments, 3’-terminal NP gene sequences; shaded segments, 5’-terminal L gene sequences. Data are from Re et al. (1985), Re and Kingsbury (1986) and the present work (Fig. 2).
cleotide sequencing using a DNA primer complementary to the copy-back terminus (a gift of Dr. D. Gill; data not shown). A family of internally deleted DI RNAs was also present in our stock of strain W, as shown by the blot hybridizations with both NP-specific and L-specific probes in Fig. 2 (horizontal bars). These RNAs were present in small amounts and were not detected by 3’-terminal labeling (Fig. 2, lane W), so it is unlikely that they influenced the outcomes of subsequent competition experiments involving RNA Wa.
Competition
between copy-back
RNAs
We began our study of competition between copyback Sendai virus DI RNAs with species Rb and 1 la. These genomes were present in different concentrations, relative to the standard helper virus, in separate virus strains. To eliminate any multiplicity bias and produce a starting inoculum for competition tests in which each DI RNA was equally represented, we infected embryonated eggs with mixtures of the two parental strains in various proportions. As shown in Fig. 3, eggs infected with 10 parts of strain 11, containing RNA 1 la, and 1 part of a passage of strain R,
SENDAI
VIRUS
DI RNA
333
SIZE
28S-18S-
-Wa FIG. 2. Copy-back nature of DI RNA Wa. (A) RNA blot hybridization analysis. Lanes M, S’terminally labeled ribosomal RNAs, V, terminally labeled RNA from strain W virions; NP and L, blots of unlabeled RNAs from strain W virions with 3ZP-labeled cDNA probes representing NP- and L-specific sequences, as described under Materials and Methods. Horizontal lines designate the small amounts of internally deleted DI RNAs present in this virus stock (see text). (5) Wandering spot 3’4erminal sequence analysis. Terminally labeled DI RNA Wa was isolated by gel electrophoresis, partially hydrolyzed with alkali, and subjected to two-dimensional polyacrylamide gel electrophoresis. The 3’2erminal sequence of Wa is compared with that of the copy-back DI RNA 1 1 a (Re et al., 1983).
containing RNA Rb (Re and Kingsbury, 1986) yielded the desired stock with equivalent amounts of the two DI genomes. After nine serial high-multiplicity passages, we were surprised to observe that the larger RNA species, Rb, outgrew 1 la (Fig. 3). This result challenges the idea that small size confers a replicative advantage on a DI RNA. However, our experimental protocol revealed only the proportions of the two RNA species in virions released from mixedly infected cells. It was possible that the replication of RNA 1 la was actually more efficient than the replication of RNA Rb, and Rb was competing with the assembly of 1 la into nucleocapsids or virions. To test this possibility, we examined intracellular DI RNAs in nucleocapsids from chorioallantoic membranes of mixedly infected eggs. As shown in Fig. 4, lane V, when the inoculum contained equal amounts of Rb and 1 1 a, the yields of the two species in virions were similar, as observed previously (Fig. 3, lane 2). However, in the intracellular nucleocapsid fraction, 11 a was markedly more abundant than Rb (Fig. 4, lane NC), indicating that the smaller 1 1a was replicated and encapsidated at least as efficiently as the larger Rb, but that encapsidated Rb molecules were incorpo-
rated much more efficiently into virions than encapsidated 1 1a molecules. These conclusions were supported and extended by the results of similar competition studies between copy-back DI RNA species 1 la and Wa. Starting with equivalent amounts of both genomes, 1 la eventually outcompeted the smaller RNA species, Wa, on repeated high-multiplicity passage (Fig. 5). Again, examination of the nucleocapsid fraction from mixedly infected chorioallantoic membranes revealed efficient replication and encapsidation of Wa, but more efficient assembly of the larger species, 1 1a, into virions (Fig. 6). These results reveal a progressive decrease in the ability of smaller nucleocapsids to enter virions. Competition within deleted RNAs
a family of internally
Sendai virus strain 7 possesses four well-characterized internally deleted RNA species (Re et a/., 1985; Hsu et al., 1985). As shown in Fig. 7, lanes 1 through 6, the largest and smallest species, 7a and 7c, respectively, outcompeted the intermediate species, 7d and 7b, early in a series of high-multiplicity passages. However, the smallest species, 7c, comprising about
334
RE AND
KINGSBURY
R:ll
M
-
MR
cc- 000
,
Passages
I
VNC
l
:&,fi2345678910
-50s 28S28S-
- Rb
18S,Rb 18S-
Ila lla
FIG. 3. Competition between copy-back RNAs Rb and 1 la. Acid urea-agarose gel electrophoresis of 3’-terminally labeled, glyoxaldenatured RNA from egg-grown virions. Lanes M, 28 S, and 18 S ribosomal RNA markers; R, progeny from passage 10 of strain R, from which RNA Ra, the companion of RNA Rb, had been eliminated (Re and Kingsbury, 1986); R: 11, progeny from eggs inoculated with passage 10 of strain R combined with strain 11 in the proportions shown; 17, progeny from one passage of strain 11; 2-l 0, progeny from successive passages (1 HA unit per egg) of the R: 11 1 :lO mixture.
1600 bases, became more prominent than 7a at the seventh passage, and was the sole survivor by passage 9 (Fig. 7). This outgrowth of the smallest RNA in strain 7 identifies a range of sizes of DI RNA molecules within which smaller molecules do indeed have a selective advantage over larger molecules. The 1600-base RNA 7c appears to approach the lower limit of this range, whereas our results with copy-back RNA species indicate that the 1200-base 11 a lies within a range of DI RNA sizes that are at a competitive disadvantage during virus assembly.
FIG. 4. Relative efficiencies of replication, sembly into virus particles of DI RNAs Rb fected with 1 HA unit of virus containing the in lane I. Three days later, virion RNA (lane cellular nucleocapsids in the chorioallantoic were examined by acid urea-agarose gel terminal labeling.
encapsidation, and asand 1 la. Eggs were inRb-1 1 a mixture shown v) and RNA from intramembrane (lane NC) electrophoresis after 3’-
of envelopment of DI nucleocapsids into virus particles. Our data indicate that Sendai virus nucleocapsids containing DI RNAs smaller than about 1600 bases (10% of the viral genome) are enveloped with decreasing efficiency as their sizes decrease. The efficient replication and encapsidation of DI RNA Wa shown in Fig. 6 rule out the possibility that mutations affecting these processes were responsible I 2 3 4 5 6 7 8 9 10 11 12 13 14 28S-
18S-
-1la DISCUSSION We have tested the hypothesis that small DI genomes of nonsegmented negative-stranded RNA viruses have a selective advantage over homologous RNA species of larger size due to a faster replication rate. Although our data are consistent with the idea that smaller RNAs are replicated more rapidly, we discovered another level of DI RNA reproduction at which small size conferred a disadvantage. This was the level
FIG. 5. Competition between copy-back RNAs 1 la and Wa. Acid urea-agarose gel electrophoresis of 3’-terminally labeled RNAs from egg-grown virions. Lanes 2 through 14, successive passages at 1 HA unit per egg of the virus mixture in lane 1, which contained equal amounts of RNAs 11 a and Wa, as described in the text.
SENDAI
VIRUS
M I VNC
28s
18s
,lla
-Wa FIG. 6. Relative efficiencies of replication, encapsidation, and assembly into virions of DI RNAs 11 a and Wa. Embryonated eggs were infected with 1 HA unit of virus containing the mixture of RNA species shown in lane I (egg passage 9). Three days later, virion RNA (lane V) and RNA from intracellular nucleocapsids in the chorioallantoic membranes (lane NC, egg passage 10) were examined by acid urea-agarose gel electrophoresis after terminal labeling. Ribosomal 28 S and 18 S RNAs (lane M) and DI RNA species 11 a and Wa are indicated.
for the inability of Wa to compete with RNA 1 la over multiple passages. The data in Fig. 6 were obtained at the 10th passage of the mixture of 11 a and Wa, when Wa was rapidly losing ground to its larger companion (Fig. 5). However, our examination of the replication and encapsidation of RNA 1 la in company with RNA Rb was made at an early passage (Fig. 4) so we cannot be sure that mutations affecting replication and encapsidation did not occur in 1 1a during this series of passages. Nevertheless, the fact that 1 la was not efficiently enveloped in the presence of Rb provides an adequate explanation for its outgrowth by RNA Rb. Although little is known mechanistically about the budding process by which paramyxovirus particles are assembled at the cell surface membrane, it is likely that we have been observing a consequence of limited membrane flexibility. Clearly, there is a limit to the degree of curvature that a patch of membrane can assume to form a spherical particle. Below that limit, no particles will be formed. Far above that limit, a wide
DI RNA
SIZE
335
range of particle sizes might be formed with equal efficiency. But as that lower limit of curvature is approached, the efficiency of particle budding would be expected to decrease, and smaller nucleocapsids would be discriminated against. Holland (1987) has considered similar constraints on the transmissibility of VSV DI RNAs. How, then, were such small Sendai virus DI RNAs able to survive when they were maintained under noncompetitive conditions, as in the preparation of working stocks of strains 1 1 and W? The situation is especially acute in the case of RNA Wa. It is only about 3% as large as the virus genome, probably close to the minimum size permissible for a viable DI RNA. Of its 450 bases, 50 at each terminus comprise signals for replication, leaving only 350 bases of L gene between them. A possible survival mechanism is that nucleocapsids containing RNAs Wa and 1 1a are too small to induce a virus bud on their own, but are incorporated as passengers into virions that are formed around genome-sized nucleocapsids. The well known pleomorphism of paramyxovirions encompasses the formation of virions containing multiple standard virus genomes or mixtures of standard virus genomes and DI RNAs (Hosaka et al., 1966; Kingsbury et a/., 1970). Even
Ml23456789 50s
28s 7a 7d 18s
7b 7c
FIG. 7. Competition among strain 7 (internally deleted) DI RNAs. Acid urea-agarose gel electrophoresis of 3’4abeled RNAs. Lanes M, 28 S, and 18 S ribosomal RNA markers; l-9, progeny from consecutive passages of strain 7 at 1 HA per egg. DI RNA species a, b, c, and d are indicated.
336
RE AND
cellular tRNAs and ribosomes are incorporated into virions (Kolakofsky, 1972; Kolakofsky and Bruschi, 1975). Budding in association with normal encapsidated RNA segments may also be involved in the transmission of the small influenza virus DI RNAs which are generated on repeated passages (Janda er a/., 1979). An alternative possibility is that encapsidated 1 la and Wa molecules are only partially impaired in their capacity to induce virus buds. In this case, more rapid rates of DI genome replication could compensate for decreased budding efficiency, enabling the RNAs to survive. There is precedent for our findings in the results of competition studies on DI RNAs of vesicular stomatitis virus (VSV), the well-studied model of the other family of nonsegmented negative-stranded RNA viruses, the rhabdoviruses. Perrault and Semler (1979) noted that a small copy-back VSV DI RNA (DI 0.10) did not compete well with either an intermediate-sized copy-back RNA (DI 0.22) or the large compound nontranscribing DI RNA, 0.50. The size of VSV DI 0.10 is 10% of the VSV genome, comparable to the size of Sendai virus DI RNA 1 la relative to its parental genome (9%). The competitive disability of DI 0.10 may therefore be similarly based on impaired budding during VSV particle assembly. A different case of outgrowth of small VSV copyback DI RNAs (DI-T and DI 0.45) by a larger species (DI 0.52) was reported by Rao and Huang (1982). In this case, the smaller DI RNAs lacked sequences proximal to their 3’ termini that may have been necessary for their encapsidation. As we have shown here, Sendai DI RNAs 1 la and Wa are not defective in encapsidation. In an earlier study of competition among Sendai virus DI RNAs, it was concluded that the size of an RNA had no bearing on its rate of replication or its ability to be assembled into virions (Kolakofsky, 1979). In that work, a single virus plaque was passaged undiluted five times in eggs before viral RNA was examined. Allantoic fluid from the 6th through 10th undiluted egg passages was then used to infect mammalian cells in culture, and intracellular encapsidated RNAs were analyzed by gel electrophoresis. Of the 1 1 DI RNA species observed, ranging in size from 670 to 7100 bases, there was some fluctuation in their relative abundances in the earlier passages, but a 1430base and a 3850-base species became predominant by the final passage. This trend toward the elimination of RNA species smaller than 1430 bases is consistent with the results we are reporting, and we believe that examination of additional virus passages in the previous work would have borne this out. We have now analyzed the effects of three factors
KINGSBURY
on Sendai virus DI RNA survival: the nature of the 3’terminal origin of RNA synthesis (genomic or copyback); the ability of the DI RNA template to be transcribed; and the role of the size of the DI RNA molecule. The relative importance of each of these factors is summarized in Table 1. The advantage of possessing a copy-back 3’ end is great, for two reasons. First, the copy-back terminus is a better promoter of RNA synthesis than its genomic counterpart (Re and Kingsbury, 1986). Second, genomes with copy-back 3’ends are transcriptionally inactive, allowing them to be engaged exclusively in replication. Perrault and Semler (1979) have noted the replicative advantage of nontranscribing VSV DI RNAs. Now, in our examination of RNA size, we have encountered contradictory effects. We saw a beneficial effect of decreasing size on replicative efficiency and assembly down to about 1600 bases, but a detrimental effect of decreasing size below this level. Therefore, we conclude that a Sendai virus DI RNA best suited for survival in a competitive situation will have a copyback 3’ terminus and comprise about 1600 bases. These deductions may have practical implications. DI viruses can influence the course of viral infections in vitro and in animals (Holland and Villarreal, 1974; Spandidos and Graham, 1976; Welsh et a/., 1977; Cave et a/., 1985; Atkinson e? al., 1986), and may be involved in the generation and maintenance of chronic viral diseases (Huang and Baltimore, 1970; Barrett et a/., 1986). From a medical standpoint, our results predict that very small paramyxovirus DI RNAs are unlikely to be responsible for persistent infections or as abundant as larger DI RNAs in clinical specimens, unless they can escape the restriction on extracellular transmission. For example, they might be spread through cell fusion or be maintained by coreplication within a proliferating cell population. There is still much more to be learned about properties of paramyxovirus and rhabdovirus DI RNAs that affect their ability to survive and interefere with the growth of the parental virus. The factors identified in Table 1 do not include all of the mechanistic possibilities or explain all of the observed phenomena. For
TABLE
1
FACTORS AFFECTING COMPETITION BETWEEN SENDAI VIRUS DI RNA SPECIES~ 1. copy-back 3’ terminus > genomic 2. Nontranscribing > transcribing 3. Size > 1600 bases B Factors the findings
3’terminus
are ranked in descending order of importance, of Re and Kingsbury (1986) and the present
based work.
on
SENDAI
VIRUS
example, in the family of internally deleted Sendai virus DI RNAs of strain 7, none of the rules in Table 1 explains why the second largest member, RNA 7d, has always been the least abundant species, or why the largest member, 7a, was particularly resistant to replacement by smaller species (Fig. 7). These may be cases where mutation has altered the efficiency of DI RNA replication or encapsidation. The likely location of such mutations is in the 5’termini of these RNAs, since their 3’ termini are faithful copies of the parental sequence (Re et al., 1985). ACKNOWLEDGMENTS We thank Sandra Farmer for skillful technical assistance, Dr. Exeen Morgan for the /VP-specific cDNA probe, Dr. Dalip Gill for the L-specific cDNA probe and primer, and Drs. Wendy Gorman and Allen Portner for Sendai virus strain W. This work was supported by Research Grant RG 1 142 from the National Multiple Sclerosis Society, by Cancer Center Support Grant CA 21765 from the National Cancer Institute, and by American Lebanese Syrian Associated Charities of St. Jude Children’s Research Hospital.
REFERENCES AMESSE, L. S., PRIDGEN, C. L.. and KINGSBURY, D. W. (1982). Sendai virus DI RNA species with conserved virus genome termini and extensive internal deletions. Virology 118, 17-27. ATKINSON, T., BARRET, A. D. T., MACKENZIE, A., and DIMMOCK, N. J. (1986). Persistence of virulent Semliki Forest virus in mouse brain following co-inoculation with defective interfering particles. J. Gen. Viral. 67, 1189-l 194. BARRE-, A. D. T., CROSS, A. J., CROW, T. J., JOHNSON, J. A., GUEST, A. R., and DIMMOCK, N. 1. (1986). Subclinical infections in mice resulting from the modulation of a lethal dose of Semliki Forest virus with defective interfering viruses: Neurochemical abnormalities in the central nervous system. 1. Gen. Viral. 67, 1727-l 732. CAVE, D. R., HENDRICKSON, F. M., and HUANG, A. S. (1985). Defective interfering virus particles modulate virulence. f. Viol. 55, 366-373. DESHPANDE, K. L., and PORTNER, A. (1984). Structural and functional analysis of Sendai virus nucleocapsid protein NP with monoclonal antibodies. I/irology 139, 32-42. HOLLAND, I. J. (1987). Defective interfering rhabdoviruses. In “The Rhabdoviruses” (R. R. Wagner, Ed.), pp. 297-360. Plenum, New York. HOLLAND, J. J., and VILLARREAL, L. P. (1974). Persistent noncytocidal vesicular stomatitis virus infections mediated by defective T parti-
DI RNA
SIZE
337
cles that suppress virion transcription. Proc. Nat/. Acad. Sci. USA 71, 2956-2960. HOSAKA, Y.. KITANO, H., and IKEGUCHI, S. (1966). Studies on the plemorphism of HVJ virions. virology 29, 205-221. Hsu. C.-H., RE. G. G., GUPTA. K. C., PORTNER, A., and KINGSBURY, D. W. (1985). Expression of Sendai virus defective-interfering genomes with internal deletions. Virology 146, 38-49. HUANG, A. S., and BALTIMORE. D. (1970). Defective viral particles and viral disease processes. Nature (London) 226, 325-327. JANDA, J. M., DAVIS, A. R., NAYAK, D. P.. and DE, B. K. (1979). Diversity and generation of defective interfering influenza virus particles. Virology 95, 48-58. KINGSBURY, D. W., PORTNER, A.. and DARLINGTON, R. W. (1970). Properties of incomplete Sendai virions and subgenomic viral RNAs. Virology 42, 857-87 1. KOLAKOFSKY, D. (1972). tRNA nucleotidyl transferase and tRNA in Sendai virions. J. Viral. 10, 555-559. KOLAKOFSKY, D. (1979). Studies on the generaiton and amplification of Sendai virus defective-interfering genomes. Virology 93, 589-593. KOLAKOFSKY, D., and BRUSCHI, A. (1975). Antigenomes in Sendai virions and Sendai virus-infected cells. Virology 66, 185-l 91. MILLS, D. R., PETERSON, R. L., and SPIEGELMAN, S. (1967). An extracellular Darwinian experiment with a self-replicating nucleic acid molecule. Proc. Nat/. Acad. Sci. USA 58, 217-224. PERRAULT, 1. (1981). Origin and replication of defective intefering particles. Cur. Top. Microbial. Immunol. 93, 151-207. PERRAULT, J., and SEMLER, B. L. (1979). Internal genome deletions in two distinct classes of defective interfering particles of vesicular stomatitis virus. froc. Nat/. Acad. Sci. USA 76, 6 191-6195. RAO, D. D., and HUANG, A. S. (1982). Interference among defective interfering particles of vesicular stomatitis virus. J. Viral. 41, 210-221. RE, G. G., GUPTA, K. C., and KINGSBURY, D. W. (1983). Genomic and copy-back 3’ termini in Sendai virus defective interfering RNA species. J. Viral. 45, 659-664. RE, G. G., and KINGSBURY, D. W. (1986). Nucleotide sequences that affect replicative and transcriptional efficiencies of Sendai virus deletion mutants. /. Viral. 58, 578-582. RE, G. G., MORGAN, E. M., and KINGSBURY, D. W. (1985). Nucleotide sequences responsible for generation of internally deleted Sendai virus defective interfering genomes. Virology 146, 27-37. SPANDIDOS, D. A., and GRAHAM, A. F. (1976). Generation of defective virus after infection of newborn rats with reovirus. J. Viral. 20, 234-247. WELSH, R. M., L&MPERT, P. W., and OLDSTONE, M. B. A. (1977). Prevention of virus-induced cerebellar disease by defective interfering lymphocytic choriomeningitis virus. /. Infect. Dis. 136, 391-399.
165, 33 l-337
(1988)
Paradoxical Effects of Sendai Virus DI RNA Size on Survival: Inefficient Envelopment of Small Nucleocapsids GIAN G. RE’ Department
of Virology
and Molecular Received
Biology, December
AND
DAVID W. KINGSBURY
St. Jude Children’s 29, 1987;
accepted
Research
Hospital,
March
29,
Memphis,
Tennessee
38 10 1
1988
In the assembly of nonsegmented negative-stranded RNA viruses, such as Sendai virus, the envelopment process allows extensively deleted genomes to survive by transmission from cell to cell in virus particles. To assess the impact of the sizes of such defective-interfering (DI) genomes on their survival, we performed competition tests among various species. Among copy-back DI RNAs, a 450-base species was gradually eliminated from DI virions by a 1200-base species, and the latter was independently eliminated by a 2800-base species. In each case, the smaller RNA species was synthesized and encapsidated at least as efficiently as the larger species, revealing that the level of competition was at the envelopment step in virus assembly. In contrast to the results obtained with the copy-back DI RNAs, repeated high multiplicity passage of a family of four internally deleted RNAs eliminated all but the smallest species, comprising about 1600 bases. Both sets of findings can be reconciled by the hypothesis that the efficiency of o 198s DI nucleocapsid envelopment decreases progressively when the RNA is smaller than about 1600 bases. Academic
Press, Inc.
We have now performed competition experiments between Sendai virus DI RNA species that have homologous 3’ termini, either copy-back or genomic, to assess the effects of DI RNA size as a third variable. Unexpectedly, and in disagreement with conclusions from a previous study (Kolakofsky, 1979) we found that DI RNA species smaller than 1600 bases were gradually eliminated from virions in mixed infections with larger DI RNAs.
INTRODUCTION Defective-interfering (DI) genomes of negativestranded RNA viruses are replicated by the RNA synthesizing machinery supplied by nondefective helper virus (Perrault, 1981). If the rate of RNA polymerization is the same on all virus-specific templates, a small RNA molecule will be completed more quickly than a larger one. Huang and Baltimore (1970) proposed this explanation for the amplification of DI genomes in competition with their helpers. An in vitro correlate is the decreased replication of bacteriophage C&3 RNA in competition with fragments that retain RNA polymerase initiation sites (Mills et a/., 1967). By analogy, in mixed infections with two or more DI RNA species, larger species should eventually be outgrown by smaller species upon repeated passage. However, our previous studies of competition between Sendai virus DI RNA species uncovered at least two other factors that outweigh RNA size. The first of these factors is the possession of a genomic 3’ terminus, which is less efficient as a promoter of RNA synthesis than the copy-back 3’ terminus. (The latter is homologous to the 3’terminus of the positive-strand intermediate in virus genome replication.) The second factor is retention of transcriptional initiation and termination signals in internally deleted DI RNAs. Transcriptional activity also confers a selective disadvantage (Re and Kingsbury, 1986). ’ To whom
requests
for reprints
should
MATERIALS
AND
between
DI RNAs
Competition
METHODS
Properties of the Sendai virus DI RNA species used in this work are summarized in Fig. 1. The naturally occurring mixture of internally deleted RNAs of strain 7 (Re et al., 1985) and artificial mixtures of strains containing copy-back RNAs were serially passaged in 1 lday embryonated chicken eggs. Chorioallantoic cavities were inoculated with a total of 1 hemagglutinating (HA) unit of virus in 0.1 ml of phosphate-buffered saline. After 3 days at 35”, progeny virus was recovered from the allantoic fluid, and the viral RNA was isolated (Re and Kingsbury, 1986). Nucleocapsids
from chorioallantoic
membranes
Chorioallantoic membranes from infected eggs were washed twice with phosphate-buffered saline, cut into l-cm squares, and homogenized in 50 mM Tris-HCI, 150 mM NaCI, 0.5% sodium deoxycholate,
be addressed. 331
0042-6822/88
$3.00
Copyright 0 1999 by Academic Press. Inc. All rights of reproducton in any form reserved.
332
RE AND
KINGSBURY
1% (v/v) NP-40 (pH 8.0) in a large Dounce homogenizer. After removal of debris by low-speed centrifugation, nucleocapsids were isolated by centrifugation in a step gradient. After 90 min at 40,000 rpm in a Spinco SW40 rotor, nucleocapsids formed a visible band at the interface between 50% (v/v) glycerol (density 1.18 g/cm3) and 74% (w/v) sucrose in D20 (density 1.33 g/cm3) (Deshpande and Portner, 1984). Both solutions contained 10 mM Tris-HCI, 30 mM NaCl (pH 7.6). Nucleocapsid RNA was extracted by the same procedure used to extract RNA from virions (Re and Kingsbury, 1986).
Sendai Virus DI RNAs Bases
1-- 2,600 - 2,000 - 1,600
RNA analyses
I -- 1.200
RNA was terminally labeled with cytidine 3’,5’-[5’32P]bisphosphate by the action of bacteriophage T4 ligase, denatured with glyoxal, and electrophoresed in acid urea-agarose gels (Re et a/., 1983). In all cases, each lane received an equal amount of radioactive RNA. For blot hybridization, the RNA was transferred to nitrocellulose paper after electrophoresis and separate strips were hybridized with a 529-base cDNA probe representing the 3’ end of the NP gene and a loo-base cDNA representing the 5’ end of the L gene.
RESULTS Properties
of copy-back
RNA species
Three copy-back Sendai virus DI RNA species that differed markedly in size (Fig. 1) were examined in competition experiments. The largest, designated Rb, comprises about 2800 bases, almost 20°b as large as the parental virus genome. RNA Rb had originally been observed in the company of a slightly larger species, Ra (Amesse et al., 1982; Re et al., 1983), which was eliminated by repeated high-multiplicity passages of strain R (Re and Kingsbury, 1986). RNA 1 la was described as a single 1200-base species in Sendai virus strain 11 by Amesse et al. (1982). It was shown to have a copy-back 3’ terminus by Re et al. (1983). The smallest DI RNA used in this work, the 450base species designated Wa, has not been described before. We encountered it unexpectedly in a virus stock provided by Drs. W. Gorman and A. Pot-trier. Because it is markedly smaller than any Sendai virus DI RNA described previously, we considered it to be of special interest for the present studies. Blot hybridizations with cloned cDNA probes representing the 3’ terminus of the NP gene and the 5’ terminus of the L gene revealed only L-specific sequences (Fig. 2A), indicating that Wa is a copy-back RNA. The presence of the copy-back 3’-terminal sequence was confirmed by wandering-spot analysis (Fig. 2B) and by dideoxynu-
-
450
FIG. 1. Sendai virus DI RNA species used in the competition experiments The left panel depicts the four internally deleted genomes in strain 7 (Re et a/., 1985). In the right panel, the three copy-back DI RNA species found in strains R, 11, and W are shown. Black triangles, genomic 3’ends; black squares, copy-back B’ends; white squares, genomic 5’ ends; black segments, 3’-terminal NP gene sequences; shaded segments, 5’-terminal L gene sequences. Data are from Re et al. (1985), Re and Kingsbury (1986) and the present work (Fig. 2).
cleotide sequencing using a DNA primer complementary to the copy-back terminus (a gift of Dr. D. Gill; data not shown). A family of internally deleted DI RNAs was also present in our stock of strain W, as shown by the blot hybridizations with both NP-specific and L-specific probes in Fig. 2 (horizontal bars). These RNAs were present in small amounts and were not detected by 3’-terminal labeling (Fig. 2, lane W), so it is unlikely that they influenced the outcomes of subsequent competition experiments involving RNA Wa.
Competition
between copy-back
RNAs
We began our study of competition between copyback Sendai virus DI RNAs with species Rb and 1 la. These genomes were present in different concentrations, relative to the standard helper virus, in separate virus strains. To eliminate any multiplicity bias and produce a starting inoculum for competition tests in which each DI RNA was equally represented, we infected embryonated eggs with mixtures of the two parental strains in various proportions. As shown in Fig. 3, eggs infected with 10 parts of strain 11, containing RNA 1 la, and 1 part of a passage of strain R,
SENDAI
VIRUS
DI RNA
333
SIZE
28S-18S-
-Wa FIG. 2. Copy-back nature of DI RNA Wa. (A) RNA blot hybridization analysis. Lanes M, S’terminally labeled ribosomal RNAs, V, terminally labeled RNA from strain W virions; NP and L, blots of unlabeled RNAs from strain W virions with 3ZP-labeled cDNA probes representing NP- and L-specific sequences, as described under Materials and Methods. Horizontal lines designate the small amounts of internally deleted DI RNAs present in this virus stock (see text). (5) Wandering spot 3’4erminal sequence analysis. Terminally labeled DI RNA Wa was isolated by gel electrophoresis, partially hydrolyzed with alkali, and subjected to two-dimensional polyacrylamide gel electrophoresis. The 3’2erminal sequence of Wa is compared with that of the copy-back DI RNA 1 1 a (Re et al., 1983).
containing RNA Rb (Re and Kingsbury, 1986) yielded the desired stock with equivalent amounts of the two DI genomes. After nine serial high-multiplicity passages, we were surprised to observe that the larger RNA species, Rb, outgrew 1 la (Fig. 3). This result challenges the idea that small size confers a replicative advantage on a DI RNA. However, our experimental protocol revealed only the proportions of the two RNA species in virions released from mixedly infected cells. It was possible that the replication of RNA 1 la was actually more efficient than the replication of RNA Rb, and Rb was competing with the assembly of 1 la into nucleocapsids or virions. To test this possibility, we examined intracellular DI RNAs in nucleocapsids from chorioallantoic membranes of mixedly infected eggs. As shown in Fig. 4, lane V, when the inoculum contained equal amounts of Rb and 1 1 a, the yields of the two species in virions were similar, as observed previously (Fig. 3, lane 2). However, in the intracellular nucleocapsid fraction, 11 a was markedly more abundant than Rb (Fig. 4, lane NC), indicating that the smaller 1 1a was replicated and encapsidated at least as efficiently as the larger Rb, but that encapsidated Rb molecules were incorpo-
rated much more efficiently into virions than encapsidated 1 1a molecules. These conclusions were supported and extended by the results of similar competition studies between copy-back DI RNA species 1 la and Wa. Starting with equivalent amounts of both genomes, 1 la eventually outcompeted the smaller RNA species, Wa, on repeated high-multiplicity passage (Fig. 5). Again, examination of the nucleocapsid fraction from mixedly infected chorioallantoic membranes revealed efficient replication and encapsidation of Wa, but more efficient assembly of the larger species, 1 1a, into virions (Fig. 6). These results reveal a progressive decrease in the ability of smaller nucleocapsids to enter virions. Competition within deleted RNAs
a family of internally
Sendai virus strain 7 possesses four well-characterized internally deleted RNA species (Re et a/., 1985; Hsu et al., 1985). As shown in Fig. 7, lanes 1 through 6, the largest and smallest species, 7a and 7c, respectively, outcompeted the intermediate species, 7d and 7b, early in a series of high-multiplicity passages. However, the smallest species, 7c, comprising about
334
RE AND
KINGSBURY
R:ll
M
-
MR
cc- 000
,
Passages
I
VNC
l
:&,fi2345678910
-50s 28S28S-
- Rb
18S,Rb 18S-
Ila lla
FIG. 3. Competition between copy-back RNAs Rb and 1 la. Acid urea-agarose gel electrophoresis of 3’-terminally labeled, glyoxaldenatured RNA from egg-grown virions. Lanes M, 28 S, and 18 S ribosomal RNA markers; R, progeny from passage 10 of strain R, from which RNA Ra, the companion of RNA Rb, had been eliminated (Re and Kingsbury, 1986); R: 11, progeny from eggs inoculated with passage 10 of strain R combined with strain 11 in the proportions shown; 17, progeny from one passage of strain 11; 2-l 0, progeny from successive passages (1 HA unit per egg) of the R: 11 1 :lO mixture.
1600 bases, became more prominent than 7a at the seventh passage, and was the sole survivor by passage 9 (Fig. 7). This outgrowth of the smallest RNA in strain 7 identifies a range of sizes of DI RNA molecules within which smaller molecules do indeed have a selective advantage over larger molecules. The 1600-base RNA 7c appears to approach the lower limit of this range, whereas our results with copy-back RNA species indicate that the 1200-base 11 a lies within a range of DI RNA sizes that are at a competitive disadvantage during virus assembly.
FIG. 4. Relative efficiencies of replication, sembly into virus particles of DI RNAs Rb fected with 1 HA unit of virus containing the in lane I. Three days later, virion RNA (lane cellular nucleocapsids in the chorioallantoic were examined by acid urea-agarose gel terminal labeling.
encapsidation, and asand 1 la. Eggs were inRb-1 1 a mixture shown v) and RNA from intramembrane (lane NC) electrophoresis after 3’-
of envelopment of DI nucleocapsids into virus particles. Our data indicate that Sendai virus nucleocapsids containing DI RNAs smaller than about 1600 bases (10% of the viral genome) are enveloped with decreasing efficiency as their sizes decrease. The efficient replication and encapsidation of DI RNA Wa shown in Fig. 6 rule out the possibility that mutations affecting these processes were responsible I 2 3 4 5 6 7 8 9 10 11 12 13 14 28S-
18S-
-1la DISCUSSION We have tested the hypothesis that small DI genomes of nonsegmented negative-stranded RNA viruses have a selective advantage over homologous RNA species of larger size due to a faster replication rate. Although our data are consistent with the idea that smaller RNAs are replicated more rapidly, we discovered another level of DI RNA reproduction at which small size conferred a disadvantage. This was the level
FIG. 5. Competition between copy-back RNAs 1 la and Wa. Acid urea-agarose gel electrophoresis of 3’-terminally labeled RNAs from egg-grown virions. Lanes 2 through 14, successive passages at 1 HA unit per egg of the virus mixture in lane 1, which contained equal amounts of RNAs 11 a and Wa, as described in the text.
SENDAI
VIRUS
M I VNC
28s
18s
,lla
-Wa FIG. 6. Relative efficiencies of replication, encapsidation, and assembly into virions of DI RNAs 11 a and Wa. Embryonated eggs were infected with 1 HA unit of virus containing the mixture of RNA species shown in lane I (egg passage 9). Three days later, virion RNA (lane V) and RNA from intracellular nucleocapsids in the chorioallantoic membranes (lane NC, egg passage 10) were examined by acid urea-agarose gel electrophoresis after terminal labeling. Ribosomal 28 S and 18 S RNAs (lane M) and DI RNA species 11 a and Wa are indicated.
for the inability of Wa to compete with RNA 1 la over multiple passages. The data in Fig. 6 were obtained at the 10th passage of the mixture of 11 a and Wa, when Wa was rapidly losing ground to its larger companion (Fig. 5). However, our examination of the replication and encapsidation of RNA 1 la in company with RNA Rb was made at an early passage (Fig. 4) so we cannot be sure that mutations affecting replication and encapsidation did not occur in 1 1a during this series of passages. Nevertheless, the fact that 1 la was not efficiently enveloped in the presence of Rb provides an adequate explanation for its outgrowth by RNA Rb. Although little is known mechanistically about the budding process by which paramyxovirus particles are assembled at the cell surface membrane, it is likely that we have been observing a consequence of limited membrane flexibility. Clearly, there is a limit to the degree of curvature that a patch of membrane can assume to form a spherical particle. Below that limit, no particles will be formed. Far above that limit, a wide
DI RNA
SIZE
335
range of particle sizes might be formed with equal efficiency. But as that lower limit of curvature is approached, the efficiency of particle budding would be expected to decrease, and smaller nucleocapsids would be discriminated against. Holland (1987) has considered similar constraints on the transmissibility of VSV DI RNAs. How, then, were such small Sendai virus DI RNAs able to survive when they were maintained under noncompetitive conditions, as in the preparation of working stocks of strains 1 1 and W? The situation is especially acute in the case of RNA Wa. It is only about 3% as large as the virus genome, probably close to the minimum size permissible for a viable DI RNA. Of its 450 bases, 50 at each terminus comprise signals for replication, leaving only 350 bases of L gene between them. A possible survival mechanism is that nucleocapsids containing RNAs Wa and 1 1a are too small to induce a virus bud on their own, but are incorporated as passengers into virions that are formed around genome-sized nucleocapsids. The well known pleomorphism of paramyxovirions encompasses the formation of virions containing multiple standard virus genomes or mixtures of standard virus genomes and DI RNAs (Hosaka et al., 1966; Kingsbury et a/., 1970). Even
Ml23456789 50s
28s 7a 7d 18s
7b 7c
FIG. 7. Competition among strain 7 (internally deleted) DI RNAs. Acid urea-agarose gel electrophoresis of 3’4abeled RNAs. Lanes M, 28 S, and 18 S ribosomal RNA markers; l-9, progeny from consecutive passages of strain 7 at 1 HA per egg. DI RNA species a, b, c, and d are indicated.
336
RE AND
cellular tRNAs and ribosomes are incorporated into virions (Kolakofsky, 1972; Kolakofsky and Bruschi, 1975). Budding in association with normal encapsidated RNA segments may also be involved in the transmission of the small influenza virus DI RNAs which are generated on repeated passages (Janda er a/., 1979). An alternative possibility is that encapsidated 1 la and Wa molecules are only partially impaired in their capacity to induce virus buds. In this case, more rapid rates of DI genome replication could compensate for decreased budding efficiency, enabling the RNAs to survive. There is precedent for our findings in the results of competition studies on DI RNAs of vesicular stomatitis virus (VSV), the well-studied model of the other family of nonsegmented negative-stranded RNA viruses, the rhabdoviruses. Perrault and Semler (1979) noted that a small copy-back VSV DI RNA (DI 0.10) did not compete well with either an intermediate-sized copy-back RNA (DI 0.22) or the large compound nontranscribing DI RNA, 0.50. The size of VSV DI 0.10 is 10% of the VSV genome, comparable to the size of Sendai virus DI RNA 1 la relative to its parental genome (9%). The competitive disability of DI 0.10 may therefore be similarly based on impaired budding during VSV particle assembly. A different case of outgrowth of small VSV copyback DI RNAs (DI-T and DI 0.45) by a larger species (DI 0.52) was reported by Rao and Huang (1982). In this case, the smaller DI RNAs lacked sequences proximal to their 3’ termini that may have been necessary for their encapsidation. As we have shown here, Sendai DI RNAs 1 la and Wa are not defective in encapsidation. In an earlier study of competition among Sendai virus DI RNAs, it was concluded that the size of an RNA had no bearing on its rate of replication or its ability to be assembled into virions (Kolakofsky, 1979). In that work, a single virus plaque was passaged undiluted five times in eggs before viral RNA was examined. Allantoic fluid from the 6th through 10th undiluted egg passages was then used to infect mammalian cells in culture, and intracellular encapsidated RNAs were analyzed by gel electrophoresis. Of the 1 1 DI RNA species observed, ranging in size from 670 to 7100 bases, there was some fluctuation in their relative abundances in the earlier passages, but a 1430base and a 3850-base species became predominant by the final passage. This trend toward the elimination of RNA species smaller than 1430 bases is consistent with the results we are reporting, and we believe that examination of additional virus passages in the previous work would have borne this out. We have now analyzed the effects of three factors
KINGSBURY
on Sendai virus DI RNA survival: the nature of the 3’terminal origin of RNA synthesis (genomic or copyback); the ability of the DI RNA template to be transcribed; and the role of the size of the DI RNA molecule. The relative importance of each of these factors is summarized in Table 1. The advantage of possessing a copy-back 3’ end is great, for two reasons. First, the copy-back terminus is a better promoter of RNA synthesis than its genomic counterpart (Re and Kingsbury, 1986). Second, genomes with copy-back 3’ends are transcriptionally inactive, allowing them to be engaged exclusively in replication. Perrault and Semler (1979) have noted the replicative advantage of nontranscribing VSV DI RNAs. Now, in our examination of RNA size, we have encountered contradictory effects. We saw a beneficial effect of decreasing size on replicative efficiency and assembly down to about 1600 bases, but a detrimental effect of decreasing size below this level. Therefore, we conclude that a Sendai virus DI RNA best suited for survival in a competitive situation will have a copyback 3’ terminus and comprise about 1600 bases. These deductions may have practical implications. DI viruses can influence the course of viral infections in vitro and in animals (Holland and Villarreal, 1974; Spandidos and Graham, 1976; Welsh et a/., 1977; Cave et a/., 1985; Atkinson e? al., 1986), and may be involved in the generation and maintenance of chronic viral diseases (Huang and Baltimore, 1970; Barrett et a/., 1986). From a medical standpoint, our results predict that very small paramyxovirus DI RNAs are unlikely to be responsible for persistent infections or as abundant as larger DI RNAs in clinical specimens, unless they can escape the restriction on extracellular transmission. For example, they might be spread through cell fusion or be maintained by coreplication within a proliferating cell population. There is still much more to be learned about properties of paramyxovirus and rhabdovirus DI RNAs that affect their ability to survive and interefere with the growth of the parental virus. The factors identified in Table 1 do not include all of the mechanistic possibilities or explain all of the observed phenomena. For
TABLE
1
FACTORS AFFECTING COMPETITION BETWEEN SENDAI VIRUS DI RNA SPECIES~ 1. copy-back 3’ terminus > genomic 2. Nontranscribing > transcribing 3. Size > 1600 bases B Factors the findings
3’terminus
are ranked in descending order of importance, of Re and Kingsbury (1986) and the present
based work.
on
SENDAI
VIRUS
example, in the family of internally deleted Sendai virus DI RNAs of strain 7, none of the rules in Table 1 explains why the second largest member, RNA 7d, has always been the least abundant species, or why the largest member, 7a, was particularly resistant to replacement by smaller species (Fig. 7). These may be cases where mutation has altered the efficiency of DI RNA replication or encapsidation. The likely location of such mutations is in the 5’termini of these RNAs, since their 3’ termini are faithful copies of the parental sequence (Re et al., 1985). ACKNOWLEDGMENTS We thank Sandra Farmer for skillful technical assistance, Dr. Exeen Morgan for the /VP-specific cDNA probe, Dr. Dalip Gill for the L-specific cDNA probe and primer, and Drs. Wendy Gorman and Allen Portner for Sendai virus strain W. This work was supported by Research Grant RG 1 142 from the National Multiple Sclerosis Society, by Cancer Center Support Grant CA 21765 from the National Cancer Institute, and by American Lebanese Syrian Associated Charities of St. Jude Children’s Research Hospital.
REFERENCES AMESSE, L. S., PRIDGEN, C. L.. and KINGSBURY, D. W. (1982). Sendai virus DI RNA species with conserved virus genome termini and extensive internal deletions. Virology 118, 17-27. ATKINSON, T., BARRET, A. D. T., MACKENZIE, A., and DIMMOCK, N. J. (1986). Persistence of virulent Semliki Forest virus in mouse brain following co-inoculation with defective interfering particles. J. Gen. Viral. 67, 1189-l 194. BARRE-, A. D. T., CROSS, A. J., CROW, T. J., JOHNSON, J. A., GUEST, A. R., and DIMMOCK, N. 1. (1986). Subclinical infections in mice resulting from the modulation of a lethal dose of Semliki Forest virus with defective interfering viruses: Neurochemical abnormalities in the central nervous system. 1. Gen. Viral. 67, 1727-l 732. CAVE, D. R., HENDRICKSON, F. M., and HUANG, A. S. (1985). Defective interfering virus particles modulate virulence. f. Viol. 55, 366-373. DESHPANDE, K. L., and PORTNER, A. (1984). Structural and functional analysis of Sendai virus nucleocapsid protein NP with monoclonal antibodies. I/irology 139, 32-42. HOLLAND, I. J. (1987). Defective interfering rhabdoviruses. In “The Rhabdoviruses” (R. R. Wagner, Ed.), pp. 297-360. Plenum, New York. HOLLAND, J. J., and VILLARREAL, L. P. (1974). Persistent noncytocidal vesicular stomatitis virus infections mediated by defective T parti-
DI RNA
SIZE
337
cles that suppress virion transcription. Proc. Nat/. Acad. Sci. USA 71, 2956-2960. HOSAKA, Y.. KITANO, H., and IKEGUCHI, S. (1966). Studies on the plemorphism of HVJ virions. virology 29, 205-221. Hsu. C.-H., RE. G. G., GUPTA. K. C., PORTNER, A., and KINGSBURY, D. W. (1985). Expression of Sendai virus defective-interfering genomes with internal deletions. Virology 146, 38-49. HUANG, A. S., and BALTIMORE. D. (1970). Defective viral particles and viral disease processes. Nature (London) 226, 325-327. JANDA, J. M., DAVIS, A. R., NAYAK, D. P.. and DE, B. K. (1979). Diversity and generation of defective interfering influenza virus particles. Virology 95, 48-58. KINGSBURY, D. W., PORTNER, A.. and DARLINGTON, R. W. (1970). Properties of incomplete Sendai virions and subgenomic viral RNAs. Virology 42, 857-87 1. KOLAKOFSKY, D. (1972). tRNA nucleotidyl transferase and tRNA in Sendai virions. J. Viral. 10, 555-559. KOLAKOFSKY, D. (1979). Studies on the generaiton and amplification of Sendai virus defective-interfering genomes. Virology 93, 589-593. KOLAKOFSKY, D., and BRUSCHI, A. (1975). Antigenomes in Sendai virions and Sendai virus-infected cells. Virology 66, 185-l 91. MILLS, D. R., PETERSON, R. L., and SPIEGELMAN, S. (1967). An extracellular Darwinian experiment with a self-replicating nucleic acid molecule. Proc. Nat/. Acad. Sci. USA 58, 217-224. PERRAULT, 1. (1981). Origin and replication of defective intefering particles. Cur. Top. Microbial. Immunol. 93, 151-207. PERRAULT, J., and SEMLER, B. L. (1979). Internal genome deletions in two distinct classes of defective interfering particles of vesicular stomatitis virus. froc. Nat/. Acad. Sci. USA 76, 6 191-6195. RAO, D. D., and HUANG, A. S. (1982). Interference among defective interfering particles of vesicular stomatitis virus. J. Viral. 41, 210-221. RE, G. G., GUPTA, K. C., and KINGSBURY, D. W. (1983). Genomic and copy-back 3’ termini in Sendai virus defective interfering RNA species. J. Viral. 45, 659-664. RE, G. G., and KINGSBURY, D. W. (1986). Nucleotide sequences that affect replicative and transcriptional efficiencies of Sendai virus deletion mutants. /. Viral. 58, 578-582. RE, G. G., MORGAN, E. M., and KINGSBURY, D. W. (1985). Nucleotide sequences responsible for generation of internally deleted Sendai virus defective interfering genomes. Virology 146, 27-37. SPANDIDOS, D. A., and GRAHAM, A. F. (1976). Generation of defective virus after infection of newborn rats with reovirus. J. Viral. 20, 234-247. WELSH, R. M., L&MPERT, P. W., and OLDSTONE, M. B. A. (1977). Prevention of virus-induced cerebellar disease by defective interfering lymphocytic choriomeningitis virus. /. Infect. Dis. 136, 391-399.