Vol. 18, No. 1 Printed in U-SA.
JoURNAL or Vmowoy, Apr. 1976, p. 365-369 Copyright C 1976 American Society for Microbiology
Transcriptase Activity and Genome Composition of Defective Influenza Virus WILLIAM J. BEAN, JR.,' AND ROBERT W. SIMPSON* Waksman Institute of Microbiology, Rutgers University-The State University of New Jersey, New Brunswick, New Jersey 08903
Received for publication 28 August 1975
Eight genome RNA segments are present in both normal and von Magnustype influenza virus preparations and all species are transcribed by the virionassociated polymerase. Although both the RNA polymerase activity and the amount of the three largest RNA segments are reduced in defective influenza virus preparations, these reductions do not appear to be great enough to account for the much greater loss of infectivity. The production of noninfectious particles of influenza virus by passage of undiluted allantoic fluid from infected eggs was first described by von Magnus (14). Although it had been reported the total amount of nucleic acid found in these defective particles was less than that found in complete virions (1), the elucidation of the segmented nature of the influenza genome and the demonstration that the proportion of the largest RNA segment was reduced in incomplete preparations (7, 11) suggested that the noninfectious particles lacked the largest genome segment. However, Pons and Hirst (11) reported that influenza virus subjected to a single undiluted passage showed a 20-fold reduction in infectivity, but no detectable change in the amount of the largest RNA species. This implicated the existence of some other defect that would lead to the loss of the large RNA species after additional high-multiplicity passage. Earlier work in this laboratory (5) suggesting that von Magnus-type virus had a reduced level of RNA polymerase activity and the finding that defective particles of vesicular stomatitis virus lack an active transcriptase (12) suggested that an impaired polymerase might be responsible for the noninfectious nature of the defective particles. This study was initiated to determine if any correlation could be found between the loss of infectivity and the polymerase activity or genome composition of defective influenza populations. Virus strains AO!WSN and A,/WS were propagated in either embryonated chicken eggs or in primary chicken embryo fibroblast cultures as previously described (3). Defective virus preparations were produced by undiluted passage of 1-ml volume of either allantoic fluid or 1 Present address: St. Jude Children's Research Hospital, Memphis, Tenn. 38101.
cell culture medium. Radiolabeled WSN virus was made by the addition of either 20 1ACi of [3H]uridine/ml or 10 /Ci of 32P/ml to the culture medium. Al virus was purified by polyethylene glycol precipitation followed by sucrose equilibrium centrifugation (3). The virion-associated RNA polymerase activity was assayed as described previously (3). The production of defective virus particles after serial passage of undiluted egg or tissue culture fluids with either the WS or WSN strain of influenza virus was monitored by calculating the log of the ratio of infectivity to hemagglutinin titer. These biological characteristics and the corresponding RNA polymerase activities are shown in Table 1. In all cases the amount of RNA polymerase activity of defective preparations was less than that of the corresponding control virus preparation, but these decreases (two- to threefold) were very small when compared with the decreases in infectivity. The greatest decrease in infectivity occurred in the third passage of egg-grown WSN virus. In this case the specific infectivity was reduced by a factor of 10W while the polymerase activity was lowered by a factor of 2.7. Our analysis (data not shown) of the RNA segments of the chicken embryo fibroblastgrown WSN von Magnus preparation using cylindrical polyacrylamide gels by the method of Bishop et al. (4), showed a clear reduction in the amount of the largest RNA size class similar to that shown previously (7, 11). Thus, in the normal virus preparation the large RNA peak represented 48% of the total RNA and this was reduced to 24 and 17%, respectively, in second and third undiluted passage virus. Since a 2/3 reduction in a single RNA species did not appear to be sufficient to explain the much greater loss in infectivity, the possibility was 365
366
NOTES
J. VIROL.
RNA was electrophoretically eluted from each strip and freed of contaminating acrylamide as described by Young and Young (15). Probes for In vithe eight complementary RNA species were Lg tro made by covalently linking the genome then Virus pool polymIU/m HAmlU/ RNA segments to filter disks as described by AU erase Saxinger et al. and Miller et al. (10, 13). Actc Direct hybridization of in vitro-synthesized 500 4.62 6.8 x 108 16,824 WSN/CEF transcriptase product RNA to the filter-immonormal bilized genome RNA segments was complicated 440 2.84 5.6 x 106 8,192 WSN/CEF by the presence of an excess of virion template VMP2 290 2.6 x 104 1.40 1,024 WSN/CEF RNA in the original reaction mixtures. To preVMP3 vent the labeled transcriptase product RNA 520 5.48 6.1 x 108 2,048 WSN/EE from preferentially annealing to the virion normal 190 1.20 6.5 x 104 RNA in solution, rather than to the appropriate 4,096 WSN/EE VMP3 filter-immobilized genome RNA segment, the 5.83 1.1 x 1010 1,820 16,824 WS/EE self-devised continuous flow hybridization sysnormal tem described in Table 2 was used. The anneal790 2.99 8.0 x 106 8,192 WS/EE VMP3 ing of viral complementary RNA, generated by a Normal or defective (von Magnus) preparations were the transcriptase of normal and defective vigrown in embryonated eggs (EE) or chicken embryo fibro- rions, to the individual filter-immobilized geblast (CEF) cultures as described in the text. VMP2 and nome RNA segments is shown in Table 2. The VMP3, second and third von Magnus (undiluted) passage, results indicate that all resolvable RNA species respectively, in the host system indicated; normal, low mul- are transcribed in vitro by both normal and tiplicity passage virus; HAU, hemagglutinin units. b IU, Infectious units per milliliter determined for WSN defective virus preparations. as plaque-forming units on CEF monolayers or for WS as The exact reason for the loss of infectivity of egg-infectious units (mean infective dose of titer per milli- defective influenza particles remains an open liter x 0.699). question. Neither the reductions in the amounts c Picomoles of uridine 5'-monophosphate incorporated per milligram of viral protein after 60-min incubation at of the larger RNA species nor the relatively 32 C. small decrease in the RNA polymerase activity appear to be great enough to account for the considered that the large RNA peak actually significantly higher reduction in infectivity obrepresented a composite of two or more species, served. Although we cannot rule out the possione of which might be preferentially lost in the bility that certain RNA segments may not be defective particles. An effort was therefore fully functional in vivo, all of the eight resolvamade to develop a polyacrylamide gel system ble RNA species were able to initiate transcripthat would provide better resolution of the tion in vitro. An attractive explanation for the noninfeclarge RNA species. Figure 1 shows the separation of 32P-labeled WSN genome segments on a tious nature of von Magnus virus is suggested 2.5% slab gel into eight species including three by the results of Ada and Perry (1) who demonin the high-molecular-weight size class. A strated that the total amount of RNA in von marked reduction in all three of the large Magnus particles was less than that in normal RNAs was evident but there did not appear to virions. The RNA analysis presented in this be a selective loss of any one RNA species in the report and genetic evidence for the existence of von Magnus virus preparations. eight recombination-complementation groups Although we had found that defective influ- (8) suggest that the minimum number of RNA enza virus preparations show only a nominal segments in an infectious particle must be loss of the in vitro RNA polymerase activity eight. It is obvious that if less than this critical associated with infectious virions (Table 1), it number of segments are packaged in any single was of interest to determine if all RNA seg- virion, even if all segments are present in the ments present in normal and defective particles population, these particles by themselves are actually transcribed in vitro. To accomplish would be noninfectious (11, 14). The incorporathis, the genome RNA from 7 mg of purified 32p_ tion of less RNA per particle in defective prepalabeled WSN virus was run on a slab gel as rations would readily explain all of the findings described in Fig. 1, and strips containing the of this investigation. However, one would preeight individual RNA segments were cut from dict that a viral population containing all of the the dried gel by using the autoradiogram as a virion genetic information packaged in subgenreference template. After rehydration, the omic amounts would show multiplicity-dependTABLE 1. Properties of normal and defective influenza virus preparations used in this study
NOTES
VOL. 18, 1976
367
A BC D
40001"%
C) lo C]
-
%4mmop
12 -
t
-.
%'
4mmu'2
.3
A4 ~~~~
(9 14-
z
16-
.*-' 7 4*
8
(I) 18FIG. 1. Autoradiogram of 32P-labeled WSN RNA genome segments, resolved by polyacrylamide gel electrophoresis. Virus pools were grown in the presence of 32P and purified as described in the text. Viral RNAs were resolved on polyacrylamide slab gels (0.5 by 10 by 18 cm) containing 2.5% acrylamide, 0.125% methylene bis acrylamide, 0.1 % ammonium persulfate, and 0.1% NNN',N'-tetramethylethylenediamine. Acrylamide and bis acrylamide (Eastman) were recrystallized from chloroform and acetone, respectively. The gel buffer and electrode buffer contained 0.4 M Tris (pH 7.4), 0.02 M sodium acetate, 0.002 M EDTA, and 0.1% sodium dodecyl sulfate (SDS) (4). Slabs were cast and run in a vertical glass cell by using a modified Buchler apparatus (model 3-1071) equipped with a pump to continuously circulate the electrode buffer between the upper and lower reservoirs. Purified 32P-labeled virus was lysed with SDS (2%) and electrophoresed for 14 1/2 h with a constant current of 50 mA. The gels were then dried with heat and vacuum and autoradiographed on Kodak NS-2 X-ray film. (A) Normal WSN, log plaque-forming unitslhemagglutinin units (log PFUIHA = 4.94); (B) WSN VMP2 (log PFU/HA = 2.61) (C) WSN VMP1 (log PFU/HA = 4.17); (D) normal WSN (log PFU/HA = 4.97).
368
NOTES
J. VIROL.
TABLE 2. Annealing of normal and von Magnus transciptase product RNA (cRNA) to individual genome RNA segments of WSN virus immobilized on filters Counts/min annealed for genome segment:
Virus used as source of cRNA
None
1b
2
3
4
5
6
Normal WSN
283 (3.4)r
666 (8.0)
815 (9.8)
822 (9.9)
833
(10.0)
716 (8.6)
158 (1.8)
463 (5.3)
752 (8.7)
772 (8.9)
759 (8.7)
652 (7.5)
718 (8.3)
(11.5)
WSN VMP1
7
8
744
905
(8.9)
(10.9)
630 (7.6)
998
891 (10.3)
WSN VMP2
208 609 931 847 827 623 1021 1158 1228 (2.3) (6.6) (10.1) (9.2) (9.0) (6.8) (11.1) (12.6) (13.3) a WSN genome RNA segments were separated by polyacrylamide gel electrophoresis by using the conditions described in Fig. 1. The segments were eluted, purified, and attached to phosphocellulose filters (0.15 Ag/filter) as described in the text. Transcriptase product RNA was prepared in 625 Al of polymerase reaction mixtures containing 500 ,ig of purified virus (Lowry determination) and five times the normal amount of 3H-uridine 5'-triphosphate. Reactions were run for 2 h and the product RNA was purified as described by Bishop et al. (4). Aliquots of product RNA synthesized by normal, VMP1, and VMP2 WSN (about 5,000 counts/min) were dissolved in 300 Ml of 3 x SSC, 50% formamide (13). The solution was continuously cycled through a length of silicone tubing which was maintained at 95 C to melt any templateproduct complexes which formed in solution, and then through a chamber containing the individual filterimmobilized genome RNA segments which were maintained at annealing temperature (37 C). A detailed description of this hybridization system will be published separately (Bean and Simpson, manuscript in preparation). After 16 h the filters were removed and the amount of product RNA retained by each filter was determined. The RNA retained on each set of filters represents 77, 71, and 81% ofthe input radioactivity, for normal, VMP1, and VMP2, respectively. b Genome RNA segment on filter. c Percentage of total input counts/min.
ent reactivation, a phenomenon which has not been demonstrated with von Magnus virus (2). In this respect, it should be noted that Hirst and Pons (9) have presented evidence that a large fraction of virions in normal WSN populations apparently contain less than a complete set of viral RNA segments. When the particles of such populations were allowed to aggregate the infectivity was significantly increased. Thus, these workers have provided evidence that several incomplete particles in "normal" myxovirus populations can initiate a complete infectious cycle when introduced into the same cell at the same site. The possibility that von Magnus particles might be reactivated under these conditions deserves consideration. This study was supported by funds from Public Health Service research grant AI-09124 from the National Institute of Allergy and Infectious Diseases. W. J. Bean, Jr. was supported as a Predoctoral Research Intern by funds from grant AI-09124 and as a Postdoctoral Fellow by the Charles and Johanna Busch Foundation. The technical assistance of Sheila Mazar, Rebecca M. Shamy, and Kathleen Strzelec is gratefully acknowledged. LITERATURE CITED 1. Ada, G. L., and B. T. Perry. 1956. Influenza viral nucleic acid: relationship between biological characteristics of the virus particle and properties of the nucleic acid. J. Gen. Microbiol. 14:623-633. 2. Barry, R. D. 1961. The multiplication of influenza vi-
rus. 1. The formation of incomplete virus.
3. 4.
5.
6.
7.
8.
9.
10.
11.
Virology
14:389-397. Bean, W. J., and R. W. Simpson. 1975. Virion associated transcriptase activity of influenza recombinant and mutant strains. J. Virol. 16:516-524. Bishop, D. H. L., J. F. Obijeski, and R. W. Simpson. 1971. Transcription of the influenza ribonucleic acid genome by a vision polymerase. II. Nature of the in vitro polymerase product. J. Virol. 8:74-80. Chow, N., and R. W. Simpson. 1971. RNA-dependent RNA polymerase activity associated with virions and subviral particles of myxoviruses. Proc. Natl. Acad. Sci. U.S.A. 68:752-756. Company, R. W., N. J. Dimmock, and H. Meier-Ewert. 1970. An electron microscopic study of the influenza virus-infected cell, p. 87-108. In R. D. Barry and B. W. J. Mahy (ed.), The biology of large RNA viruses Academic Press Inc., New York. Duesberg, P. H. 1968. The RNA's of influenza virus. Proc. Natl. Acad. Sci. U.S.A. 59:930-937. Hirst, G. K. 1973. Mechanism of influenza recombination. I. Factors influencing recombination rates between temperature sensitive mutants of strain WSN and the classification of mutants into complementation-recombination groups. Virology 55:81-93. Hirst, G. K., and M. W. Pons. 1973. Mechanism of influenza recombination. II. Virus aggregation and its effect on plaque formation by so-called noninfectious virus. Virology 56:620-631. Miller, N. R., W. C. Saxinger, M. S. Reitz, R. E. Gallagher, A. M. Wu, R. C. Gallo, and D. Gillespie. 1974. Systematics of RNA tumor viruses and viruslike particles of human origin. Proc. Natl. Acad. Sci. U.S.A. 71:3177-3181. Pons, M. W., and G. K. Hirst. 1969. Single- and doublestranded RNAs and the proteins of incomplete influ-
VOL. 18, 1976 enza virus. Virology 38:68-72. 12. Roy, P., and D. H. L. Bishop. 1972. Genome homology of vesicular stomatitis virus and defective T particles and evidence for the sequential transcription of the virion ribonucleic acid. J. Virol. 9:946-955. 13. Saxinger, W. C., C. Ponnamperuma, and D. Gillespie. 1972. Nucleic acid hybridization with RNA immobi-
NOTES
369
lized on filter paper. Proc. Natl. Acad. Sci. U.S.A. 69:2975-2978. 14. von Magnus, P. 1954. Incomplete forms of influenza virus. Adv. Virus Res. 2:59-79. 15. Young, Y. P. L., and R. J. Young. 1974. A method for the recovery of nucleic acids from polyacrylamide gels. Anal. Biochem. 58:286-293.
JoURNAL or Vmowoy, Apr. 1976, p. 365-369 Copyright C 1976 American Society for Microbiology
Transcriptase Activity and Genome Composition of Defective Influenza Virus WILLIAM J. BEAN, JR.,' AND ROBERT W. SIMPSON* Waksman Institute of Microbiology, Rutgers University-The State University of New Jersey, New Brunswick, New Jersey 08903
Received for publication 28 August 1975
Eight genome RNA segments are present in both normal and von Magnustype influenza virus preparations and all species are transcribed by the virionassociated polymerase. Although both the RNA polymerase activity and the amount of the three largest RNA segments are reduced in defective influenza virus preparations, these reductions do not appear to be great enough to account for the much greater loss of infectivity. The production of noninfectious particles of influenza virus by passage of undiluted allantoic fluid from infected eggs was first described by von Magnus (14). Although it had been reported the total amount of nucleic acid found in these defective particles was less than that found in complete virions (1), the elucidation of the segmented nature of the influenza genome and the demonstration that the proportion of the largest RNA segment was reduced in incomplete preparations (7, 11) suggested that the noninfectious particles lacked the largest genome segment. However, Pons and Hirst (11) reported that influenza virus subjected to a single undiluted passage showed a 20-fold reduction in infectivity, but no detectable change in the amount of the largest RNA species. This implicated the existence of some other defect that would lead to the loss of the large RNA species after additional high-multiplicity passage. Earlier work in this laboratory (5) suggesting that von Magnus-type virus had a reduced level of RNA polymerase activity and the finding that defective particles of vesicular stomatitis virus lack an active transcriptase (12) suggested that an impaired polymerase might be responsible for the noninfectious nature of the defective particles. This study was initiated to determine if any correlation could be found between the loss of infectivity and the polymerase activity or genome composition of defective influenza populations. Virus strains AO!WSN and A,/WS were propagated in either embryonated chicken eggs or in primary chicken embryo fibroblast cultures as previously described (3). Defective virus preparations were produced by undiluted passage of 1-ml volume of either allantoic fluid or 1 Present address: St. Jude Children's Research Hospital, Memphis, Tenn. 38101.
cell culture medium. Radiolabeled WSN virus was made by the addition of either 20 1ACi of [3H]uridine/ml or 10 /Ci of 32P/ml to the culture medium. Al virus was purified by polyethylene glycol precipitation followed by sucrose equilibrium centrifugation (3). The virion-associated RNA polymerase activity was assayed as described previously (3). The production of defective virus particles after serial passage of undiluted egg or tissue culture fluids with either the WS or WSN strain of influenza virus was monitored by calculating the log of the ratio of infectivity to hemagglutinin titer. These biological characteristics and the corresponding RNA polymerase activities are shown in Table 1. In all cases the amount of RNA polymerase activity of defective preparations was less than that of the corresponding control virus preparation, but these decreases (two- to threefold) were very small when compared with the decreases in infectivity. The greatest decrease in infectivity occurred in the third passage of egg-grown WSN virus. In this case the specific infectivity was reduced by a factor of 10W while the polymerase activity was lowered by a factor of 2.7. Our analysis (data not shown) of the RNA segments of the chicken embryo fibroblastgrown WSN von Magnus preparation using cylindrical polyacrylamide gels by the method of Bishop et al. (4), showed a clear reduction in the amount of the largest RNA size class similar to that shown previously (7, 11). Thus, in the normal virus preparation the large RNA peak represented 48% of the total RNA and this was reduced to 24 and 17%, respectively, in second and third undiluted passage virus. Since a 2/3 reduction in a single RNA species did not appear to be sufficient to explain the much greater loss in infectivity, the possibility was 365
366
NOTES
J. VIROL.
RNA was electrophoretically eluted from each strip and freed of contaminating acrylamide as described by Young and Young (15). Probes for In vithe eight complementary RNA species were Lg tro made by covalently linking the genome then Virus pool polymIU/m HAmlU/ RNA segments to filter disks as described by AU erase Saxinger et al. and Miller et al. (10, 13). Actc Direct hybridization of in vitro-synthesized 500 4.62 6.8 x 108 16,824 WSN/CEF transcriptase product RNA to the filter-immonormal bilized genome RNA segments was complicated 440 2.84 5.6 x 106 8,192 WSN/CEF by the presence of an excess of virion template VMP2 290 2.6 x 104 1.40 1,024 WSN/CEF RNA in the original reaction mixtures. To preVMP3 vent the labeled transcriptase product RNA 520 5.48 6.1 x 108 2,048 WSN/EE from preferentially annealing to the virion normal 190 1.20 6.5 x 104 RNA in solution, rather than to the appropriate 4,096 WSN/EE VMP3 filter-immobilized genome RNA segment, the 5.83 1.1 x 1010 1,820 16,824 WS/EE self-devised continuous flow hybridization sysnormal tem described in Table 2 was used. The anneal790 2.99 8.0 x 106 8,192 WS/EE VMP3 ing of viral complementary RNA, generated by a Normal or defective (von Magnus) preparations were the transcriptase of normal and defective vigrown in embryonated eggs (EE) or chicken embryo fibro- rions, to the individual filter-immobilized geblast (CEF) cultures as described in the text. VMP2 and nome RNA segments is shown in Table 2. The VMP3, second and third von Magnus (undiluted) passage, results indicate that all resolvable RNA species respectively, in the host system indicated; normal, low mul- are transcribed in vitro by both normal and tiplicity passage virus; HAU, hemagglutinin units. b IU, Infectious units per milliliter determined for WSN defective virus preparations. as plaque-forming units on CEF monolayers or for WS as The exact reason for the loss of infectivity of egg-infectious units (mean infective dose of titer per milli- defective influenza particles remains an open liter x 0.699). question. Neither the reductions in the amounts c Picomoles of uridine 5'-monophosphate incorporated per milligram of viral protein after 60-min incubation at of the larger RNA species nor the relatively 32 C. small decrease in the RNA polymerase activity appear to be great enough to account for the considered that the large RNA peak actually significantly higher reduction in infectivity obrepresented a composite of two or more species, served. Although we cannot rule out the possione of which might be preferentially lost in the bility that certain RNA segments may not be defective particles. An effort was therefore fully functional in vivo, all of the eight resolvamade to develop a polyacrylamide gel system ble RNA species were able to initiate transcripthat would provide better resolution of the tion in vitro. An attractive explanation for the noninfeclarge RNA species. Figure 1 shows the separation of 32P-labeled WSN genome segments on a tious nature of von Magnus virus is suggested 2.5% slab gel into eight species including three by the results of Ada and Perry (1) who demonin the high-molecular-weight size class. A strated that the total amount of RNA in von marked reduction in all three of the large Magnus particles was less than that in normal RNAs was evident but there did not appear to virions. The RNA analysis presented in this be a selective loss of any one RNA species in the report and genetic evidence for the existence of von Magnus virus preparations. eight recombination-complementation groups Although we had found that defective influ- (8) suggest that the minimum number of RNA enza virus preparations show only a nominal segments in an infectious particle must be loss of the in vitro RNA polymerase activity eight. It is obvious that if less than this critical associated with infectious virions (Table 1), it number of segments are packaged in any single was of interest to determine if all RNA seg- virion, even if all segments are present in the ments present in normal and defective particles population, these particles by themselves are actually transcribed in vitro. To accomplish would be noninfectious (11, 14). The incorporathis, the genome RNA from 7 mg of purified 32p_ tion of less RNA per particle in defective prepalabeled WSN virus was run on a slab gel as rations would readily explain all of the findings described in Fig. 1, and strips containing the of this investigation. However, one would preeight individual RNA segments were cut from dict that a viral population containing all of the the dried gel by using the autoradiogram as a virion genetic information packaged in subgenreference template. After rehydration, the omic amounts would show multiplicity-dependTABLE 1. Properties of normal and defective influenza virus preparations used in this study
NOTES
VOL. 18, 1976
367
A BC D
40001"%
C) lo C]
-
%4mmop
12 -
t
-.
%'
4mmu'2
.3
A4 ~~~~
(9 14-
z
16-
.*-' 7 4*
8
(I) 18FIG. 1. Autoradiogram of 32P-labeled WSN RNA genome segments, resolved by polyacrylamide gel electrophoresis. Virus pools were grown in the presence of 32P and purified as described in the text. Viral RNAs were resolved on polyacrylamide slab gels (0.5 by 10 by 18 cm) containing 2.5% acrylamide, 0.125% methylene bis acrylamide, 0.1 % ammonium persulfate, and 0.1% NNN',N'-tetramethylethylenediamine. Acrylamide and bis acrylamide (Eastman) were recrystallized from chloroform and acetone, respectively. The gel buffer and electrode buffer contained 0.4 M Tris (pH 7.4), 0.02 M sodium acetate, 0.002 M EDTA, and 0.1% sodium dodecyl sulfate (SDS) (4). Slabs were cast and run in a vertical glass cell by using a modified Buchler apparatus (model 3-1071) equipped with a pump to continuously circulate the electrode buffer between the upper and lower reservoirs. Purified 32P-labeled virus was lysed with SDS (2%) and electrophoresed for 14 1/2 h with a constant current of 50 mA. The gels were then dried with heat and vacuum and autoradiographed on Kodak NS-2 X-ray film. (A) Normal WSN, log plaque-forming unitslhemagglutinin units (log PFUIHA = 4.94); (B) WSN VMP2 (log PFU/HA = 2.61) (C) WSN VMP1 (log PFU/HA = 4.17); (D) normal WSN (log PFU/HA = 4.97).
368
NOTES
J. VIROL.
TABLE 2. Annealing of normal and von Magnus transciptase product RNA (cRNA) to individual genome RNA segments of WSN virus immobilized on filters Counts/min annealed for genome segment:
Virus used as source of cRNA
None
1b
2
3
4
5
6
Normal WSN
283 (3.4)r
666 (8.0)
815 (9.8)
822 (9.9)
833
(10.0)
716 (8.6)
158 (1.8)
463 (5.3)
752 (8.7)
772 (8.9)
759 (8.7)
652 (7.5)
718 (8.3)
(11.5)
WSN VMP1
7
8
744
905
(8.9)
(10.9)
630 (7.6)
998
891 (10.3)
WSN VMP2
208 609 931 847 827 623 1021 1158 1228 (2.3) (6.6) (10.1) (9.2) (9.0) (6.8) (11.1) (12.6) (13.3) a WSN genome RNA segments were separated by polyacrylamide gel electrophoresis by using the conditions described in Fig. 1. The segments were eluted, purified, and attached to phosphocellulose filters (0.15 Ag/filter) as described in the text. Transcriptase product RNA was prepared in 625 Al of polymerase reaction mixtures containing 500 ,ig of purified virus (Lowry determination) and five times the normal amount of 3H-uridine 5'-triphosphate. Reactions were run for 2 h and the product RNA was purified as described by Bishop et al. (4). Aliquots of product RNA synthesized by normal, VMP1, and VMP2 WSN (about 5,000 counts/min) were dissolved in 300 Ml of 3 x SSC, 50% formamide (13). The solution was continuously cycled through a length of silicone tubing which was maintained at 95 C to melt any templateproduct complexes which formed in solution, and then through a chamber containing the individual filterimmobilized genome RNA segments which were maintained at annealing temperature (37 C). A detailed description of this hybridization system will be published separately (Bean and Simpson, manuscript in preparation). After 16 h the filters were removed and the amount of product RNA retained by each filter was determined. The RNA retained on each set of filters represents 77, 71, and 81% ofthe input radioactivity, for normal, VMP1, and VMP2, respectively. b Genome RNA segment on filter. c Percentage of total input counts/min.
ent reactivation, a phenomenon which has not been demonstrated with von Magnus virus (2). In this respect, it should be noted that Hirst and Pons (9) have presented evidence that a large fraction of virions in normal WSN populations apparently contain less than a complete set of viral RNA segments. When the particles of such populations were allowed to aggregate the infectivity was significantly increased. Thus, these workers have provided evidence that several incomplete particles in "normal" myxovirus populations can initiate a complete infectious cycle when introduced into the same cell at the same site. The possibility that von Magnus particles might be reactivated under these conditions deserves consideration. This study was supported by funds from Public Health Service research grant AI-09124 from the National Institute of Allergy and Infectious Diseases. W. J. Bean, Jr. was supported as a Predoctoral Research Intern by funds from grant AI-09124 and as a Postdoctoral Fellow by the Charles and Johanna Busch Foundation. The technical assistance of Sheila Mazar, Rebecca M. Shamy, and Kathleen Strzelec is gratefully acknowledged. LITERATURE CITED 1. Ada, G. L., and B. T. Perry. 1956. Influenza viral nucleic acid: relationship between biological characteristics of the virus particle and properties of the nucleic acid. J. Gen. Microbiol. 14:623-633. 2. Barry, R. D. 1961. The multiplication of influenza vi-
rus. 1. The formation of incomplete virus.
3. 4.
5.
6.
7.
8.
9.
10.
11.
Virology
14:389-397. Bean, W. J., and R. W. Simpson. 1975. Virion associated transcriptase activity of influenza recombinant and mutant strains. J. Virol. 16:516-524. Bishop, D. H. L., J. F. Obijeski, and R. W. Simpson. 1971. Transcription of the influenza ribonucleic acid genome by a vision polymerase. II. Nature of the in vitro polymerase product. J. Virol. 8:74-80. Chow, N., and R. W. Simpson. 1971. RNA-dependent RNA polymerase activity associated with virions and subviral particles of myxoviruses. Proc. Natl. Acad. Sci. U.S.A. 68:752-756. Company, R. W., N. J. Dimmock, and H. Meier-Ewert. 1970. An electron microscopic study of the influenza virus-infected cell, p. 87-108. In R. D. Barry and B. W. J. Mahy (ed.), The biology of large RNA viruses Academic Press Inc., New York. Duesberg, P. H. 1968. The RNA's of influenza virus. Proc. Natl. Acad. Sci. U.S.A. 59:930-937. Hirst, G. K. 1973. Mechanism of influenza recombination. I. Factors influencing recombination rates between temperature sensitive mutants of strain WSN and the classification of mutants into complementation-recombination groups. Virology 55:81-93. Hirst, G. K., and M. W. Pons. 1973. Mechanism of influenza recombination. II. Virus aggregation and its effect on plaque formation by so-called noninfectious virus. Virology 56:620-631. Miller, N. R., W. C. Saxinger, M. S. Reitz, R. E. Gallagher, A. M. Wu, R. C. Gallo, and D. Gillespie. 1974. Systematics of RNA tumor viruses and viruslike particles of human origin. Proc. Natl. Acad. Sci. U.S.A. 71:3177-3181. Pons, M. W., and G. K. Hirst. 1969. Single- and doublestranded RNAs and the proteins of incomplete influ-
VOL. 18, 1976 enza virus. Virology 38:68-72. 12. Roy, P., and D. H. L. Bishop. 1972. Genome homology of vesicular stomatitis virus and defective T particles and evidence for the sequential transcription of the virion ribonucleic acid. J. Virol. 9:946-955. 13. Saxinger, W. C., C. Ponnamperuma, and D. Gillespie. 1972. Nucleic acid hybridization with RNA immobi-
NOTES
369
lized on filter paper. Proc. Natl. Acad. Sci. U.S.A. 69:2975-2978. 14. von Magnus, P. 1954. Incomplete forms of influenza virus. Adv. Virus Res. 2:59-79. 15. Young, Y. P. L., and R. J. Young. 1974. A method for the recovery of nucleic acids from polyacrylamide gels. Anal. Biochem. 58:286-293.
