Phosphopeptide Fingerprints of Nucleoproteins of ... - Semantic Scholar
465
J. gen. Virol. (1985), 66, 465 472. Printed in Great Britain
Key words: influenza virus/nucleoprotein/phosphopeptides
Phosphopeptide Fingerprints of Nucleoproteins of Various Influenza A Virus Strains Grown in Different Host Cells ByO.
KISTNER,
H.
Mi3LLER,
H.
BECHT
AND C. S C H O L T I S S E K *
Institut j~r Virologie, Justus-Liebig-Universitiit Giessen, Frankfurter Strasse 107, D-6300 Giessen, Federal Republic o f Germany (Accepted 1 November 1984)
SUMMARY M D C K , HeLa, L or primary chick embryo cells were infected with different influenza A virus strains and labelled with [32p]orthophosphate. The nucleoprotein was immunoprecipitated and digested by trypsin. The resulting tryptic fingerprints were strain-specific and dependent on the host cell in which the virus strains had been propagated. Virus mutants had different fingerprints. It is suggested that specific cellular protein phosphokinases are involved in virus replication and that these m a y determine host range and cell tropism by site-specific phosphorylation of viral phosphoproteins. INTRODUCTION Three virus-specific phosphoproteins have been isolated from influenza A virus-infected cells: the nucleoprotein (NP), the m e m b r a n e protein (M) and the non-structural protein NS1 (Privalsky & Penhoet, 1977, 1978, 1981 ; Petri & Dimmock, 1981; Petri et al., 1982; Schrom & Caliguiri, 1981 ; A l m o n d & Felsenreich, 1982; Gregoriades et al., 1984). The nucleoprotein of the W S N strain (H 1N 1) contained one major phosphorylation site at a serine residue (Privalsky & Penhoet, 1981). Since the functional specificities of protein phosphokinases vary in different cells (for review, see Cohen, 1982) and phosphorylation of viral proteins varies according to the virus strain, we have examined the hypothesis that these parameters could be at least partly responsible for the host range of influenza viruses. Therefore, we labelled the nucleoproteins of various influenza A virus strains with [32p]orthophosphate in different cells and analysed their tryptic fingerprints after immunoprecipitation. It was found that the fingerprints of phosphopeptides of the nucleoproteins of different influenza A strains were indeed strain-specific, host cell-dependent, and could be influenced by mutation o f the virus. METHODS Virus strains and tissue culture cells. The following influenza A virus strains were investigated: A/fowl plague/Rostock/34 (FPV, H7N1); A/chick/Germany/N/49 (N, H10N7); A/PR/8/34 (PR8, H1N1); A/Hong Kong/l/68 (Ho, H3N2). The temperature-sensitive (ts) mutants ts 19 and ts 81 of FPV were also investigated; these have ts detects in the NP gene (Scholtissek & Bowles, 1975; Scholtissek et al., 1976). Four different host cell lines were used : HeLa, L, MDCK cells and primary chick embryo fibroblasts (CEF) 48 h after seeding. Plaque and haemagglutination tests were performed according to standard techniques. Labelling of the nucleoproteins and their immunoprecipitation. The host ceils were incubated with 2.5 mCi carrierfree [32p]orthophosphate (Amersham) per 2 x l0 Tcells from 2 to 6 h after infection. Thereafter, the cells were covered with 2 ml RIPA buffer [10 mM-phosphate buffer pH 7.2, containing 2 mM-EDTA, 40 mM-NaF, 1~ Triton X-100, 1~osodium deoxycholate, 0.1 ~ SDS, 5 ~ Trasylol (Bayer) (Gilead et al., 1976, as modified by Ziemiecki & Friis, 1980)],left for 1 h on ice, and clarified by centrifugation (60 min, 35000 r.p.m., 4 °C, SW60Ti rotor). Aliquots (300 ~tl)of the supernatant were reacted with 7.5 ~tl anti-NP serum for 2 h on ice. Anti-NP serum was prepared by injecting a rabbit with NP antigen which had been purified from chorioallantoic membranes of FPV-infected eggs by antibody-mediated affinity chromatography (Becht & Malole, 1975). Immune complexes were isolated using 100 ~tl of a 10~o suspension of Staphylococcus aureus (Kessler, 1975). The antigen-antibody complexes were washed three times with wash buffer ( 10 mM-phosphate buffer pH 7, containing 10 mM-EDTA, 40 mM-NaF, 1 mMNaCI, 0.2~ Triton X-100, 5Vo Trasylol) (Ziemiecki & Friis, 1980) and once with water. 0000-6294 © 1985 SGM
466
O. KISTNER AND OTHERS FPV i
Ho
PR8
tt
u
~
N
t f ~ l t
u
N
u
,L
IL...
I
~
w
u
ALL
' v
Z
Fig. 1. Polyacrylamide gel electrophoresis of immune precipitates of 32p-labelled nucleoproteins of various influenza A strains. CEF or MDCK cells were either mock-infected (first two lanes) or infected with influenza strain FPV (H7NI), Ho (H3N2), PR8 (HIN1) or N(H10N7).
Isolation of the labelled nucleoprotein and its digestion by trypsin. The precipitates were dissolved in electrophoresis sample buffer (80 mM-Tris-HCl pH 6.8, containing 2 ~ glycerol, 2 ~ SDS, 2 ~ 2-mercaptoethanol and a trace of bromophenol blue) and subjected to SDS-polyacrylamide gel electrophoresis (PAGE) (Bosch et al., 1979). The gels were washed twice in 25 ~ isopropanol and 10 ~ methanol to remove SDS and urea, and Fuji X-ray films were exposed to the gels for 24 h. The NP bands were excised from the gel, lyophilized and digested in 0.75 ml 0-1 M-ethylmorpholine acetate pH 8-1, containing 1 mg/ml TPCK-trypsin (Worthington) for 15 h at 37 °C. The extracts were lyophilized and resuspended in 15 ~tl water. Two-dimensional separation ofphosphopeptides. Five ~tl aliquots were applied to cellulose thin-layer plates (20 x 20 cm, Polygram CEL300; Machery-Nagel, Dfiren, F.R.G.) for electrophoresis in pyridine:acetic acid:acetone : water (1 : 2 : 8 : 40, by vol.) at 400 V for 2.5 h. After drying the plates were chromatographed in the second dimension with butanol:acetic acid :pyridine:water (15:3:10:12, by vol.) for 7 h (Boege et al., 1980). The phosphopeptides were visualized by autoradiography with intensifying screens (Cronex Lightning Plus, DuPont). The spots were scraped off from the plates, dissolved in scintillation fluid and the radioactivity counted. RESULTS
Immunoprecipitation of 32p-labelled nucleoproteins After labelling of chick embryo cells with 2.5 mCi 32p i per 2 × 107 cells from 2 to 6 h after infection at 37 °C, the cells including the nuclei were disrupted in RIPA buffer and the debris was removed by ultracentrifugation. Only traces of labelled NP were found in the pellets. The NP was localized by comparison with the known PAGE patterns of 35S-labelled extracts of infected cells. The NP was precipitated from the supernatant with monospecific rabbit NP antiserum and the immune complexes were fixed to S. aureus Protein A. The precipitate was dissolved in electrophoresis sample buffer and analysed by P A G E as shown in Fig. 1. No radioactivity co-migrated with the NP protein when mock-infected chick embryo, HeLa or L cells (the latter two not shown) were investigated, but with mock-infected M D C K cells a faint band was sometimes visible in this position. These results made certain that the 32p-labelled band used for preparing the fingerprints represented the NP.
Variation of influenza virus NP fingerprints
467
g ×
X
j
HolCEF
VIRUS N/CEF
X
r~
il Qli
x
o ,
FPV/CEF
PRNCEF
$,
0 2
o
ill +
X
11 FPV/MDCK
PR8/MDCK
Chromatography Fig. 2. Phosphopeptidefingerprints of the nucleoproteins of various influenzaA virus strains grown in MDCK cellsor CEF cells. The phosphopeptidesof the PR8 strain grown in CEF are numbered as used in Tables 1 and 2 and in the text.
Phosphopeptide fingerprints of nucleoprotein of various influenza A virus strains The bands with the labelled nucleoproteins were cut out of the gel and digested exhaustively with trypsin. After two-dimensional separation of the phosphopeptides and visualization with X-ray films the following patterns were obtained. There were six phosphopeptides in the nucieoprotein of the PR8 strain (Fig. 2) grown in chick embryo cells. According to the intensity of these six spots, four major and two minor phosphopeptides could be discriminated. The same four major phosphopeptides were found in the NP of FPV (Fig. 2); one of the minor phosphopeptides was regularly missing (no. 1). When the labelled phosphopeptides of the NPs of PR8 and FPV were mixed prior to PAGE, the five common phosphopeptides co-migrated exactly (not shown). Sometimes spot no. 5 could be partially resolved into two spots, indicating that this fraction might consist of two individual phosphopeptides. These qualitative patterns were highly reproducible. Gel slices from lanes of mock-infected cells (Fig. 1) corresponding to the position of the NP protein never gave rise to any radioactivity at the sites of phosphopeptides as shown in Fig. 2, even after an extremely long autoradiographic exposure time. The Hong Kong strain grown in CEF produced one major and one minor phosphopeptide, and for strain N, phosphopeptides no. 2 and 3 were present (Fig. 2). In a FPV-Hong Kong reassortant (81/Ho; Scholtissek et al., 1978), in which RNA segments 3, 4 and 7 were derived from FPV and the other genes including segment 5 coding for the NP from the Hong Kong
468
O.
KISTNER
AND
OTHERS
I
e
~t
e O .
x
I
,.Io II
'
4
I~P/t-'leLo
PRB/HeLo
0 0
+
l
O
. oo
KP/CEF
il /
e qq
PRB/CEF
Chromatography Fig. 3. Phosphopeptide fingerprints of the nucleoproteins of FPV (KP) and PR8 propagated and labelled in CEF or HeLa cells.
parent, the phosphopeptides were the same as those of the Hong Kong parent (not shown). This underlines the strain dependence of the degree of phosphorylation. When M D C K cells were infected with various influenza A virus strains, significant differences in the phosphopeptide fingerprints of the NPs were found when compared with NPs labelled in chick embryo cells. After infection with PR8 there was no or only very little radioactivity at the position of the CEF phosphopeptide no. 6 of Fig. 2. Significant differences can also be recognized in the patterns of NPs of FPV. There was relatively little radioactivity found in phosphopeptide no. 3 in the NP of FPV propagated in M D C K cells. Phosphopeptide no. 6, however, was always labelled. The phosphorylation of the nucleoprotein was also studied in HeLa and L cells. In these cells influenza viruses multiply abortively (Franklin & Breitenfeld, 1959; ter Meulen & Love, 1967). It has been shown by Schrom & Caliguiri (1981) that the NP of the WSN strain is not phosphorylated in HeLa cells. In our hands the NPs of all strains tested, including the WSN strain, were heavily phosphorylated in HeLa as well as in L cells. The fingerprints in Fig. 3 demonstrate that a family of additional phosphopeptides appear in the NPs of FPV and PR8 produced in HeLa cells. The same was found for L cells (not shown). We have no explanation for the discrepancy between the results of Schrom & Caliguiri and our results. One important difference might be that the RIPA buffer used for processing of our labelled cells inhibited the action of phosphatases. The radioactivity in various spots was also quantified by scraping the corresponding material from the thin-layer chromatograms after visualization and counting the samples in a scintillation counter. The results of three independent experiments using three different cell batches are
Variation of influenza virus NP fingerprints
469
Table 1. Relative distribution (% of total) of radioactivity in various phosphopeptides of the
nucleoprotein ofF P V and P R8 grown in the presence of[ 32p]orthophosphate in two different cell types FPV x
Expt.*
~CEF
1
l
0
1
4
2
2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
3 2 11 8 11 48 32 31 5 8 13 19 29 28 17 20 15
2 3 4 4 7 31 11 22 23 24 21 34 45 36 7 14 11
5 5 5 5 7 40 36 37 6 18 7 25 30 25 20 6 19
2 3 7 8 4 23 21 2O 25 24 24 40 44 43 3 1 6
2 3 4 5 6
MDCK
c CEF
PR8 ~
Spot no.
MDCK
* A different cell batch was used for each experiment. summarized in Table 1. Radioactivity of ~< 3 % did not a p p e a r as spots on X-ray films and was regarded as background. The quantitative distribution of radioactivity among the various spots is fairly reproducible. The greatest variations were found in phosphopeptides no. 4 and 6 of the N P of CEF-grown PR8. There was an inverse relationship between their radioactivities: when it was high in one, it was low in the other. In a few experiments spot no. 2 o f the N P of F P V was also more heavily labelled when compared to spot no. 2 in Fig. 2 or Table 1 (see Fig. 3). If two cultures derived from the same cell batch were processed in parallel the results were nearly identical. Thus, the different degrees of phosphorylation in individual cell batches might be due to differences in the metabolic state of the corresponding phosphokinases in these cells.
Phosphopeptide fingerprints of nucleoproteins of mutants of F P V with ts defects in the nucleoprotein If the phosphorylation pattern of the nucleoprotein is relevant to its function, it should be possible to influence this pattern by mutation which at the same time leads to a defect in the function of the NP. W e have studied two such mutants. M u t a n t ts 81 is unable to synthesize viral R N A at 40 °C (Scholtissek, 1978), and ts 19 has a muturation defect at the non-permissive temperature (Scholtissek & Bowles, 1975). W h e n C E F were infected with these mutants or with wild-type F P V and labelled with 32p i the ts mutants at the permissive temperature exhibited a phosphorylation pattern similar to that of FPV. At 40 °C the overall incorporation of label into the N P of ts 81 was greatly reduced, while that of ts 19 was almost normal. As can be seen in Table 2, the radioactivity in phosphopeptide no. 3 of the N P ofts 19 as well as ts 81, when grown at 40 °C, was greatly reduced. Thus, with ts 19 and ts 81 the phosphorylation pattern of the N P is influenced by ts mutations which also influence the maturation of the virus or R N A synthesis (Scholtissek & Bowles, 1975). The phosphopeptide fingerprints of the N P s of ts ÷ revertants of ts 19 and ts 81 were nearly identical to that of F P V when labelled at 40 °C.
Virus yield after infection of different cells with F P V or PR8 I f the differential phosphorylation of the nucleoprotein represents one of the factors influencing the replication of influenza viruses it has to be expected that the yield of F P V and PR8 in different cells should vary depending on the phosphorylation pattern of their N P s in these cells. As shown in Table 3, this is indeed the case. Even low yields of infectious virus could
470
O. K I S T N E R A N D O T H E R S
Table 2. Radioactivity (% of total) in the various phosphopeptides of F P V and ts mutants thereof grown and labelled in CEF either at 33 or 40 °C FPV* Spot no. 1 2 3 4 5 6
ts
r - - ) ' - - ~
r
19" x
Revertant of t s 81" Revertant of t s 191" • x ~ t s 81"I" 40 °C 33 °C 40 °C 40 °C
•
33 °C
40 °C
33 °C
40 °C
3 18 44 7 16 12
3 13 38 10 19 17
3 22 34 9 19 13
3 13 16 21 24 23
4 8 46 16 17 9
6 15 21 12 23 23
9 14 15 13 30 19
4 6 43 15 23 9
* Average of three independent experiments. t One single experiment.
Table 3. Multiplication of F P V and PR8 on four different host cell types* Yield 9 h after infection &
Host cell CEF MDCK HeLa L
Strain
HA units
FPV PR8 FPV PR8 FPV PR8 FPV PR8
256 256 64 128 8 8 16 32
P.f.u. 1x 1× 8× 1× 6× 8x 1× 2x
108 107 106 105 105 104 105 104
P.f.u./HA 4 x 4 x 1-3 x 8x 7.5 × 1× 6"2 × 6"2 ×
105 104 105 102 104 104 103 102
* Cells were infected with a m.o.i, of about 20 p.f.u./cell. After penetration, the cells were washed twice with cold 0.9 % NaCI, pH 2.5. The yields were determined 9 h after infection after three cycles of freezing and thawing. During the plaque test with PR8, trypsin was added to the overlay medium (Appleyard & Maber, 1974).
be determined by removing the residual virus by washing the infected cells at pH 2.5. There are significant differences not only in the yield (p.f.u.) but also in the quality of the viruses (p.f.u./HA). In cells in which the NPs are overphosphorylated and in which an abortive infection occurs (HeLa and L cells) the yield of infectious virus is more than 100-fold lower than the yield in chick embryo cells. The yield in M D C K cells is intermediate. 'DISCUSSION
Among the three viral proteins which have been reported to carry phosphate groups only the NP appears to be regularly phosphorylated. Not all strains induce the synthesis of a phosphorylated NS1 protein when grown in chick embryo cells (Petri et al., 1982). We have been able to identify a phosphorylated M protein only with the WSN strain (Gregoriades et al., 1984; and our unpublished data). Thus, there are strains which can grow without phosphorylation of the M and/or NS1 proteins. The fact that the nucleoprotein was phosphorylated in all strains tested could indicate that phosphorylation of the NP might be essential for multiplication. Therefore, we have concentrated our analysis on the nucleoprotein of various influenza strains. As can be seen in Fig. 2 and in Table 1, the patterns of the tryptic phosphopeptides of the nucleoprotein are strain-specific. The nucleoprotein of the Hong Kong virus grown in chick embryo cells had one major and one minor phosphopeptide and resembles that of the WSN strain reported by Privalsky & Penhoet (1981). On the other hand, the phosphopeptide fingerprint of the PR8 strain is much more complex, consisting of at least six individual phosphopeptides. This is surprising as all three strains were isolated from man, the WSN and PR8 strains at about the same time. The patterns of the tryptic phosphopeptides are not only virus strain-specific but they are also
Variation of influenza virus NP fingerprints
471
influenced by the host cell in which the virus was propagated. Therefore, we have to assume that cellular protein phosphokinases are involved and not (only) virus-coded enzymes. The virus strains studied here are able to multiply in M D C K as wetl as in chick embryo cells for at least a single cycle. [For multiple cycles all viruses except FPV have to be activated by trypsin (Appleyard & Maber, 1974; Klenk et al., 1975; Lazarowitz & Choppin, 1975).] However, the virus yields in these cells are quite different (Table 3). It is not yet known whether or in what way these differences correlate with the degree of phosphorylation of the NP. There is the striking observation, however, that in abortively infected HeLa (Fig. 4) or L cells, where the virus yield drops by about 2 log10 units (Table 3), the nucleoprotein is overphosphorylated. The phosphorylation pattern can be influenced by mutation in the NP gene, which correlates with a biological defect. It is not yet known whether this change in the phosphorylation pattern is an epiphenomenon or whether it causes the biological defect. Therefore, more mutants with defects in the NP gene and revertants have to be studied in order to show that the changes in the phosphorylation pattern at 40 °C is a constant finding. In one cell type, the host-specific phosphokinases might modify the NP by phosphorylation in such a way that it can cooperate with other viral components like the P or M proteins in all the necessary steps in virus multiplication. In another cell type, the modification of the NP might be different and multiplication of this strain might not be optimal or even not possible. This could be one of the reasons why various influenza virus strains replicate in different host cells with variable efficiencies. Limited replication of a certain virus strain caused by inadequate phosphorylation of its NP might be overcome by replacing it for a functional N P in a reassortant; in this case, phosphorylation of the new NP would mean that the host range of the virus has changed by reassortment, and examples of this are known (Scholtissek et al., 1978, 1979; Vallbracht et al., 1979; Bonin & Scholtissek, 1983). Our results obtained so far do not permit a final conclusion, but they are compatible with the idea that specific host cell protein phosphokinases are involved in cell- and species-specificity of influenza viruses. The research was supported by the Sonderforschungsbereich 47 of the Deutsche Forschungsgemeinschaft. This work is in partial fulfilment of the requirements for the degree Dr. rer. nat. of O. Kistner, Fachbereich Biologie, University of Giessen. REFERENCES ALMOND, J. W. & FELSENREICH, V. (1982). P h o s p h o r y l a t i o n of the nucleoprotein of an avian influenza virus. Journal of General Virology 60, 295-305. APPLEYARD, G. & MABER, H. B. (1974). Plaque formation by influenza viruses in the presence of trypsin. Journal of General Virology 25, 351-357. BECHT, H. & MALOLE, B. (1975). Comparative evaluation of different fixation procedures and different coupling
reagents for demonstration of influenza virus-specific antibodies by the indirect hemagglutination test.
Medical Microbiology and Immunology 162, 43-53. BOEGE, U., WENGLER, G., WENGLER, G. & WITTMANN-LIEBOLD, B. (1980). Partial amino acid sequences of Sindbis and Semliki Forest virus-specific core proteins. Virology 103, 178-190. BONIN, J. & SCHOLTISSEK, C. (1983). Mouse neurotropic recombinants of influenza A viruses. Archit'esof Virology 75, 255 268. BOSCH, F. X., ORLICH,M., KLENK, H.-D. & ROTT. R. (1979). The structure of the hemagglutinin, a determinant for the pathogenicity of influenza viruses. Virology 95, 197 207. COHEN, P. (1982). The role of protein phosphorylation in neural and hormonal control of cellular activity. Nature, London 296, 613 620. FRANKLIN, R. M. & BREITENFELD, P. M. (1959). The abortive infection of Earle's L-cells by fowl plague virus. Virology 8, 293-307. GILEAD~ Z., JENG, Y. H., WOLD, W. S. M., SUGAWARA, K., RHO, H. M., HARTER, M. L. & GREEN, M.(1976). Immunological identification of two adenovirus 2-induced early proteins possibly involved in cell transformation. Nature, London 264, 263-266. GREGORIADES, A., CHRISTIE, T. & MARKARIAN, K. (1984). The membrane (M1) protein of influenza virus occurs in two forms and is a phosphoproteim Journal of Virology 49, 229-235. KESSLER, S. W. (1975). Rapid isolation o f a n t i g e n s from cells with a staphylococcal Protein A-antibody adsorbent: parameters of the interaction of antibody antigen complexes with protein A. Journal of lmmunology 115, 1617 1624.
472
o. KISTNER AND OTHERS
KLENK, H.-D., ROTT, R., ORLICH, M. & BL(JDORN, J. (1975). Activation of influenza viruses by trypsin treatment.
Virology 68, 426-439. LAZAROWITZ, S. G. & CHOPPIN, P. W. (1975). E n h a n c e m e n t of the infectivity of influenza A and B viruses by proteolytic cleavage of the hemagglutinin polypeptide. Virology 68, 440-454. PETRI, T. & DIMMOCK, N. J. (1981). Phosphorylation of influenza virus nucleoprotein in vivo. Journal of General Virology 57, 185-190. PETRI, T., PATTERSON, S. & DIMMOCK, N. J. (1982). Polymorphism of the NSI proteins of type A influenza virus. Journal of General Virology 61, 217-231. PRIVALSKY, M. L. & PENHOET, E. E. (1977). Phosphorylated protein component present in influenza virions. Journal of Virology 24, 401-405. PRIVALSKY, M. L. & PENHOET, E. E. (1978). Influenza virus proteins: identity, synthesis, and modification analyzed by two-dimensional gel electrophoresis. Proceedings of the National Academy of Sciences, U.S.A. 75, 36253629. PRIVALSKY,M. L. & PENHOET,E. E. (1981 ). The structure and synthesis of influenza virus phosphoproteins. Journal of Biological Chemistry 256, 5368-5376. SCHOLTISSEK,C. (1978). The genome of the influenza virus. Current Topics in Microbiology and Immunology 80, 139169. SCHOLTISSEK, C. & BOWLES, A. L. (1975). Isolation and characterization of temperature-sensitive mutants of fowl plague virus. Virology 67, 576-587. SCHOLTISSEK, C., HARMS, E., ROHDE, W., ORLICH, M. & ROTT, R. (1976). Correlation between R N A fragments of fowl plague virus and their corresponding gene functions. Virology 74, 332 344. SCHOLTISSEK, C., KOENNECKE, I. & ROTT, R. (1978). Host range recombinants of fowl plague (influenza A) virus. Virology 91, 79 85. SCHOLTISSEK, C., VALLBRACHT, A., FLEHMIG, B. & ROTT, R. (1979). Correlation of pathogenicity and gene constellation of influenza A viruses. 1I. Highly neurovirulent recombinants derived from non-neurovirulent or weakly neurovirulent parent virus strains. Virology 95, 492 500. SCHROM, M. & CALIGUIRI, L. A. (1981). Heterogeneity of influenza virus polypeptides during productive and abortive infections. In The Replication ofNegatit, e Strand Viruses, pp. 285-290. Edited by D. H. L. Bishop & R. W. Compans. New York: Elsevier/North-Holland. TER MEULEN,V. & LOVE, R. (1967). Virological, immunochemical and cytochemical studies of four HeLa-cell lines infected with two strains of influenza virus. Journal of Virology 1,626-639. VALLBRACHT, A., FLEHM1G, B. & GERTH, H.-J. (1979). Influenza virus: appearance of high mouse-neurovirulent recombinants, lntervirology 11, 16-22. ZIEbIIECKI, A. & FRIIS, R. R. (1980). Simultaneous injection of newborn rabbits with the Schrnidt-Ruppin and Prague strains of Rous sarcoma virus induces antibodies which recognize the pp60 src of both strains. Journal of General Virology 50, 211-216. (Receit~ed 24 July 1984)
J. gen. Virol. (1985), 66, 465 472. Printed in Great Britain
Key words: influenza virus/nucleoprotein/phosphopeptides
Phosphopeptide Fingerprints of Nucleoproteins of Various Influenza A Virus Strains Grown in Different Host Cells ByO.
KISTNER,
H.
Mi3LLER,
H.
BECHT
AND C. S C H O L T I S S E K *
Institut j~r Virologie, Justus-Liebig-Universitiit Giessen, Frankfurter Strasse 107, D-6300 Giessen, Federal Republic o f Germany (Accepted 1 November 1984)
SUMMARY M D C K , HeLa, L or primary chick embryo cells were infected with different influenza A virus strains and labelled with [32p]orthophosphate. The nucleoprotein was immunoprecipitated and digested by trypsin. The resulting tryptic fingerprints were strain-specific and dependent on the host cell in which the virus strains had been propagated. Virus mutants had different fingerprints. It is suggested that specific cellular protein phosphokinases are involved in virus replication and that these m a y determine host range and cell tropism by site-specific phosphorylation of viral phosphoproteins. INTRODUCTION Three virus-specific phosphoproteins have been isolated from influenza A virus-infected cells: the nucleoprotein (NP), the m e m b r a n e protein (M) and the non-structural protein NS1 (Privalsky & Penhoet, 1977, 1978, 1981 ; Petri & Dimmock, 1981; Petri et al., 1982; Schrom & Caliguiri, 1981 ; A l m o n d & Felsenreich, 1982; Gregoriades et al., 1984). The nucleoprotein of the W S N strain (H 1N 1) contained one major phosphorylation site at a serine residue (Privalsky & Penhoet, 1981). Since the functional specificities of protein phosphokinases vary in different cells (for review, see Cohen, 1982) and phosphorylation of viral proteins varies according to the virus strain, we have examined the hypothesis that these parameters could be at least partly responsible for the host range of influenza viruses. Therefore, we labelled the nucleoproteins of various influenza A virus strains with [32p]orthophosphate in different cells and analysed their tryptic fingerprints after immunoprecipitation. It was found that the fingerprints of phosphopeptides of the nucleoproteins of different influenza A strains were indeed strain-specific, host cell-dependent, and could be influenced by mutation o f the virus. METHODS Virus strains and tissue culture cells. The following influenza A virus strains were investigated: A/fowl plague/Rostock/34 (FPV, H7N1); A/chick/Germany/N/49 (N, H10N7); A/PR/8/34 (PR8, H1N1); A/Hong Kong/l/68 (Ho, H3N2). The temperature-sensitive (ts) mutants ts 19 and ts 81 of FPV were also investigated; these have ts detects in the NP gene (Scholtissek & Bowles, 1975; Scholtissek et al., 1976). Four different host cell lines were used : HeLa, L, MDCK cells and primary chick embryo fibroblasts (CEF) 48 h after seeding. Plaque and haemagglutination tests were performed according to standard techniques. Labelling of the nucleoproteins and their immunoprecipitation. The host ceils were incubated with 2.5 mCi carrierfree [32p]orthophosphate (Amersham) per 2 x l0 Tcells from 2 to 6 h after infection. Thereafter, the cells were covered with 2 ml RIPA buffer [10 mM-phosphate buffer pH 7.2, containing 2 mM-EDTA, 40 mM-NaF, 1~ Triton X-100, 1~osodium deoxycholate, 0.1 ~ SDS, 5 ~ Trasylol (Bayer) (Gilead et al., 1976, as modified by Ziemiecki & Friis, 1980)],left for 1 h on ice, and clarified by centrifugation (60 min, 35000 r.p.m., 4 °C, SW60Ti rotor). Aliquots (300 ~tl)of the supernatant were reacted with 7.5 ~tl anti-NP serum for 2 h on ice. Anti-NP serum was prepared by injecting a rabbit with NP antigen which had been purified from chorioallantoic membranes of FPV-infected eggs by antibody-mediated affinity chromatography (Becht & Malole, 1975). Immune complexes were isolated using 100 ~tl of a 10~o suspension of Staphylococcus aureus (Kessler, 1975). The antigen-antibody complexes were washed three times with wash buffer ( 10 mM-phosphate buffer pH 7, containing 10 mM-EDTA, 40 mM-NaF, 1 mMNaCI, 0.2~ Triton X-100, 5Vo Trasylol) (Ziemiecki & Friis, 1980) and once with water. 0000-6294 © 1985 SGM
466
O. KISTNER AND OTHERS FPV i
Ho
PR8
tt
u
~
N
t f ~ l t
u
N
u
,L
IL...
I
~
w
u
ALL
' v
Z
Fig. 1. Polyacrylamide gel electrophoresis of immune precipitates of 32p-labelled nucleoproteins of various influenza A strains. CEF or MDCK cells were either mock-infected (first two lanes) or infected with influenza strain FPV (H7NI), Ho (H3N2), PR8 (HIN1) or N(H10N7).
Isolation of the labelled nucleoprotein and its digestion by trypsin. The precipitates were dissolved in electrophoresis sample buffer (80 mM-Tris-HCl pH 6.8, containing 2 ~ glycerol, 2 ~ SDS, 2 ~ 2-mercaptoethanol and a trace of bromophenol blue) and subjected to SDS-polyacrylamide gel electrophoresis (PAGE) (Bosch et al., 1979). The gels were washed twice in 25 ~ isopropanol and 10 ~ methanol to remove SDS and urea, and Fuji X-ray films were exposed to the gels for 24 h. The NP bands were excised from the gel, lyophilized and digested in 0.75 ml 0-1 M-ethylmorpholine acetate pH 8-1, containing 1 mg/ml TPCK-trypsin (Worthington) for 15 h at 37 °C. The extracts were lyophilized and resuspended in 15 ~tl water. Two-dimensional separation ofphosphopeptides. Five ~tl aliquots were applied to cellulose thin-layer plates (20 x 20 cm, Polygram CEL300; Machery-Nagel, Dfiren, F.R.G.) for electrophoresis in pyridine:acetic acid:acetone : water (1 : 2 : 8 : 40, by vol.) at 400 V for 2.5 h. After drying the plates were chromatographed in the second dimension with butanol:acetic acid :pyridine:water (15:3:10:12, by vol.) for 7 h (Boege et al., 1980). The phosphopeptides were visualized by autoradiography with intensifying screens (Cronex Lightning Plus, DuPont). The spots were scraped off from the plates, dissolved in scintillation fluid and the radioactivity counted. RESULTS
Immunoprecipitation of 32p-labelled nucleoproteins After labelling of chick embryo cells with 2.5 mCi 32p i per 2 × 107 cells from 2 to 6 h after infection at 37 °C, the cells including the nuclei were disrupted in RIPA buffer and the debris was removed by ultracentrifugation. Only traces of labelled NP were found in the pellets. The NP was localized by comparison with the known PAGE patterns of 35S-labelled extracts of infected cells. The NP was precipitated from the supernatant with monospecific rabbit NP antiserum and the immune complexes were fixed to S. aureus Protein A. The precipitate was dissolved in electrophoresis sample buffer and analysed by P A G E as shown in Fig. 1. No radioactivity co-migrated with the NP protein when mock-infected chick embryo, HeLa or L cells (the latter two not shown) were investigated, but with mock-infected M D C K cells a faint band was sometimes visible in this position. These results made certain that the 32p-labelled band used for preparing the fingerprints represented the NP.
Variation of influenza virus NP fingerprints
467
g ×
X
j
HolCEF
VIRUS N/CEF
X
r~
il Qli
x
o ,
FPV/CEF
PRNCEF
$,
0 2
o
ill +
X
11 FPV/MDCK
PR8/MDCK
Chromatography Fig. 2. Phosphopeptidefingerprints of the nucleoproteins of various influenzaA virus strains grown in MDCK cellsor CEF cells. The phosphopeptidesof the PR8 strain grown in CEF are numbered as used in Tables 1 and 2 and in the text.
Phosphopeptide fingerprints of nucleoprotein of various influenza A virus strains The bands with the labelled nucleoproteins were cut out of the gel and digested exhaustively with trypsin. After two-dimensional separation of the phosphopeptides and visualization with X-ray films the following patterns were obtained. There were six phosphopeptides in the nucieoprotein of the PR8 strain (Fig. 2) grown in chick embryo cells. According to the intensity of these six spots, four major and two minor phosphopeptides could be discriminated. The same four major phosphopeptides were found in the NP of FPV (Fig. 2); one of the minor phosphopeptides was regularly missing (no. 1). When the labelled phosphopeptides of the NPs of PR8 and FPV were mixed prior to PAGE, the five common phosphopeptides co-migrated exactly (not shown). Sometimes spot no. 5 could be partially resolved into two spots, indicating that this fraction might consist of two individual phosphopeptides. These qualitative patterns were highly reproducible. Gel slices from lanes of mock-infected cells (Fig. 1) corresponding to the position of the NP protein never gave rise to any radioactivity at the sites of phosphopeptides as shown in Fig. 2, even after an extremely long autoradiographic exposure time. The Hong Kong strain grown in CEF produced one major and one minor phosphopeptide, and for strain N, phosphopeptides no. 2 and 3 were present (Fig. 2). In a FPV-Hong Kong reassortant (81/Ho; Scholtissek et al., 1978), in which RNA segments 3, 4 and 7 were derived from FPV and the other genes including segment 5 coding for the NP from the Hong Kong
468
O.
KISTNER
AND
OTHERS
I
e
~t
e O .
x
I
,.Io II
'
4
I~P/t-'leLo
PRB/HeLo
0 0
+
l
O
. oo
KP/CEF
il /
e qq
PRB/CEF
Chromatography Fig. 3. Phosphopeptide fingerprints of the nucleoproteins of FPV (KP) and PR8 propagated and labelled in CEF or HeLa cells.
parent, the phosphopeptides were the same as those of the Hong Kong parent (not shown). This underlines the strain dependence of the degree of phosphorylation. When M D C K cells were infected with various influenza A virus strains, significant differences in the phosphopeptide fingerprints of the NPs were found when compared with NPs labelled in chick embryo cells. After infection with PR8 there was no or only very little radioactivity at the position of the CEF phosphopeptide no. 6 of Fig. 2. Significant differences can also be recognized in the patterns of NPs of FPV. There was relatively little radioactivity found in phosphopeptide no. 3 in the NP of FPV propagated in M D C K cells. Phosphopeptide no. 6, however, was always labelled. The phosphorylation of the nucleoprotein was also studied in HeLa and L cells. In these cells influenza viruses multiply abortively (Franklin & Breitenfeld, 1959; ter Meulen & Love, 1967). It has been shown by Schrom & Caliguiri (1981) that the NP of the WSN strain is not phosphorylated in HeLa cells. In our hands the NPs of all strains tested, including the WSN strain, were heavily phosphorylated in HeLa as well as in L cells. The fingerprints in Fig. 3 demonstrate that a family of additional phosphopeptides appear in the NPs of FPV and PR8 produced in HeLa cells. The same was found for L cells (not shown). We have no explanation for the discrepancy between the results of Schrom & Caliguiri and our results. One important difference might be that the RIPA buffer used for processing of our labelled cells inhibited the action of phosphatases. The radioactivity in various spots was also quantified by scraping the corresponding material from the thin-layer chromatograms after visualization and counting the samples in a scintillation counter. The results of three independent experiments using three different cell batches are
Variation of influenza virus NP fingerprints
469
Table 1. Relative distribution (% of total) of radioactivity in various phosphopeptides of the
nucleoprotein ofF P V and P R8 grown in the presence of[ 32p]orthophosphate in two different cell types FPV x
Expt.*
~CEF
1
l
0
1
4
2
2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
3 2 11 8 11 48 32 31 5 8 13 19 29 28 17 20 15
2 3 4 4 7 31 11 22 23 24 21 34 45 36 7 14 11
5 5 5 5 7 40 36 37 6 18 7 25 30 25 20 6 19
2 3 7 8 4 23 21 2O 25 24 24 40 44 43 3 1 6
2 3 4 5 6
MDCK
c CEF
PR8 ~
Spot no.
MDCK
* A different cell batch was used for each experiment. summarized in Table 1. Radioactivity of ~< 3 % did not a p p e a r as spots on X-ray films and was regarded as background. The quantitative distribution of radioactivity among the various spots is fairly reproducible. The greatest variations were found in phosphopeptides no. 4 and 6 of the N P of CEF-grown PR8. There was an inverse relationship between their radioactivities: when it was high in one, it was low in the other. In a few experiments spot no. 2 o f the N P of F P V was also more heavily labelled when compared to spot no. 2 in Fig. 2 or Table 1 (see Fig. 3). If two cultures derived from the same cell batch were processed in parallel the results were nearly identical. Thus, the different degrees of phosphorylation in individual cell batches might be due to differences in the metabolic state of the corresponding phosphokinases in these cells.
Phosphopeptide fingerprints of nucleoproteins of mutants of F P V with ts defects in the nucleoprotein If the phosphorylation pattern of the nucleoprotein is relevant to its function, it should be possible to influence this pattern by mutation which at the same time leads to a defect in the function of the NP. W e have studied two such mutants. M u t a n t ts 81 is unable to synthesize viral R N A at 40 °C (Scholtissek, 1978), and ts 19 has a muturation defect at the non-permissive temperature (Scholtissek & Bowles, 1975). W h e n C E F were infected with these mutants or with wild-type F P V and labelled with 32p i the ts mutants at the permissive temperature exhibited a phosphorylation pattern similar to that of FPV. At 40 °C the overall incorporation of label into the N P of ts 81 was greatly reduced, while that of ts 19 was almost normal. As can be seen in Table 2, the radioactivity in phosphopeptide no. 3 of the N P ofts 19 as well as ts 81, when grown at 40 °C, was greatly reduced. Thus, with ts 19 and ts 81 the phosphorylation pattern of the N P is influenced by ts mutations which also influence the maturation of the virus or R N A synthesis (Scholtissek & Bowles, 1975). The phosphopeptide fingerprints of the N P s of ts ÷ revertants of ts 19 and ts 81 were nearly identical to that of F P V when labelled at 40 °C.
Virus yield after infection of different cells with F P V or PR8 I f the differential phosphorylation of the nucleoprotein represents one of the factors influencing the replication of influenza viruses it has to be expected that the yield of F P V and PR8 in different cells should vary depending on the phosphorylation pattern of their N P s in these cells. As shown in Table 3, this is indeed the case. Even low yields of infectious virus could
470
O. K I S T N E R A N D O T H E R S
Table 2. Radioactivity (% of total) in the various phosphopeptides of F P V and ts mutants thereof grown and labelled in CEF either at 33 or 40 °C FPV* Spot no. 1 2 3 4 5 6
ts
r - - ) ' - - ~
r
19" x
Revertant of t s 81" Revertant of t s 191" • x ~ t s 81"I" 40 °C 33 °C 40 °C 40 °C
•
33 °C
40 °C
33 °C
40 °C
3 18 44 7 16 12
3 13 38 10 19 17
3 22 34 9 19 13
3 13 16 21 24 23
4 8 46 16 17 9
6 15 21 12 23 23
9 14 15 13 30 19
4 6 43 15 23 9
* Average of three independent experiments. t One single experiment.
Table 3. Multiplication of F P V and PR8 on four different host cell types* Yield 9 h after infection &
Host cell CEF MDCK HeLa L
Strain
HA units
FPV PR8 FPV PR8 FPV PR8 FPV PR8
256 256 64 128 8 8 16 32
P.f.u. 1x 1× 8× 1× 6× 8x 1× 2x
108 107 106 105 105 104 105 104
P.f.u./HA 4 x 4 x 1-3 x 8x 7.5 × 1× 6"2 × 6"2 ×
105 104 105 102 104 104 103 102
* Cells were infected with a m.o.i, of about 20 p.f.u./cell. After penetration, the cells were washed twice with cold 0.9 % NaCI, pH 2.5. The yields were determined 9 h after infection after three cycles of freezing and thawing. During the plaque test with PR8, trypsin was added to the overlay medium (Appleyard & Maber, 1974).
be determined by removing the residual virus by washing the infected cells at pH 2.5. There are significant differences not only in the yield (p.f.u.) but also in the quality of the viruses (p.f.u./HA). In cells in which the NPs are overphosphorylated and in which an abortive infection occurs (HeLa and L cells) the yield of infectious virus is more than 100-fold lower than the yield in chick embryo cells. The yield in M D C K cells is intermediate. 'DISCUSSION
Among the three viral proteins which have been reported to carry phosphate groups only the NP appears to be regularly phosphorylated. Not all strains induce the synthesis of a phosphorylated NS1 protein when grown in chick embryo cells (Petri et al., 1982). We have been able to identify a phosphorylated M protein only with the WSN strain (Gregoriades et al., 1984; and our unpublished data). Thus, there are strains which can grow without phosphorylation of the M and/or NS1 proteins. The fact that the nucleoprotein was phosphorylated in all strains tested could indicate that phosphorylation of the NP might be essential for multiplication. Therefore, we have concentrated our analysis on the nucleoprotein of various influenza strains. As can be seen in Fig. 2 and in Table 1, the patterns of the tryptic phosphopeptides of the nucleoprotein are strain-specific. The nucleoprotein of the Hong Kong virus grown in chick embryo cells had one major and one minor phosphopeptide and resembles that of the WSN strain reported by Privalsky & Penhoet (1981). On the other hand, the phosphopeptide fingerprint of the PR8 strain is much more complex, consisting of at least six individual phosphopeptides. This is surprising as all three strains were isolated from man, the WSN and PR8 strains at about the same time. The patterns of the tryptic phosphopeptides are not only virus strain-specific but they are also
Variation of influenza virus NP fingerprints
471
influenced by the host cell in which the virus was propagated. Therefore, we have to assume that cellular protein phosphokinases are involved and not (only) virus-coded enzymes. The virus strains studied here are able to multiply in M D C K as wetl as in chick embryo cells for at least a single cycle. [For multiple cycles all viruses except FPV have to be activated by trypsin (Appleyard & Maber, 1974; Klenk et al., 1975; Lazarowitz & Choppin, 1975).] However, the virus yields in these cells are quite different (Table 3). It is not yet known whether or in what way these differences correlate with the degree of phosphorylation of the NP. There is the striking observation, however, that in abortively infected HeLa (Fig. 4) or L cells, where the virus yield drops by about 2 log10 units (Table 3), the nucleoprotein is overphosphorylated. The phosphorylation pattern can be influenced by mutation in the NP gene, which correlates with a biological defect. It is not yet known whether this change in the phosphorylation pattern is an epiphenomenon or whether it causes the biological defect. Therefore, more mutants with defects in the NP gene and revertants have to be studied in order to show that the changes in the phosphorylation pattern at 40 °C is a constant finding. In one cell type, the host-specific phosphokinases might modify the NP by phosphorylation in such a way that it can cooperate with other viral components like the P or M proteins in all the necessary steps in virus multiplication. In another cell type, the modification of the NP might be different and multiplication of this strain might not be optimal or even not possible. This could be one of the reasons why various influenza virus strains replicate in different host cells with variable efficiencies. Limited replication of a certain virus strain caused by inadequate phosphorylation of its NP might be overcome by replacing it for a functional N P in a reassortant; in this case, phosphorylation of the new NP would mean that the host range of the virus has changed by reassortment, and examples of this are known (Scholtissek et al., 1978, 1979; Vallbracht et al., 1979; Bonin & Scholtissek, 1983). Our results obtained so far do not permit a final conclusion, but they are compatible with the idea that specific host cell protein phosphokinases are involved in cell- and species-specificity of influenza viruses. The research was supported by the Sonderforschungsbereich 47 of the Deutsche Forschungsgemeinschaft. This work is in partial fulfilment of the requirements for the degree Dr. rer. nat. of O. Kistner, Fachbereich Biologie, University of Giessen. REFERENCES ALMOND, J. W. & FELSENREICH, V. (1982). P h o s p h o r y l a t i o n of the nucleoprotein of an avian influenza virus. Journal of General Virology 60, 295-305. APPLEYARD, G. & MABER, H. B. (1974). Plaque formation by influenza viruses in the presence of trypsin. Journal of General Virology 25, 351-357. BECHT, H. & MALOLE, B. (1975). Comparative evaluation of different fixation procedures and different coupling
reagents for demonstration of influenza virus-specific antibodies by the indirect hemagglutination test.
Medical Microbiology and Immunology 162, 43-53. BOEGE, U., WENGLER, G., WENGLER, G. & WITTMANN-LIEBOLD, B. (1980). Partial amino acid sequences of Sindbis and Semliki Forest virus-specific core proteins. Virology 103, 178-190. BONIN, J. & SCHOLTISSEK, C. (1983). Mouse neurotropic recombinants of influenza A viruses. Archit'esof Virology 75, 255 268. BOSCH, F. X., ORLICH,M., KLENK, H.-D. & ROTT. R. (1979). The structure of the hemagglutinin, a determinant for the pathogenicity of influenza viruses. Virology 95, 197 207. COHEN, P. (1982). The role of protein phosphorylation in neural and hormonal control of cellular activity. Nature, London 296, 613 620. FRANKLIN, R. M. & BREITENFELD, P. M. (1959). The abortive infection of Earle's L-cells by fowl plague virus. Virology 8, 293-307. GILEAD~ Z., JENG, Y. H., WOLD, W. S. M., SUGAWARA, K., RHO, H. M., HARTER, M. L. & GREEN, M.(1976). Immunological identification of two adenovirus 2-induced early proteins possibly involved in cell transformation. Nature, London 264, 263-266. GREGORIADES, A., CHRISTIE, T. & MARKARIAN, K. (1984). The membrane (M1) protein of influenza virus occurs in two forms and is a phosphoproteim Journal of Virology 49, 229-235. KESSLER, S. W. (1975). Rapid isolation o f a n t i g e n s from cells with a staphylococcal Protein A-antibody adsorbent: parameters of the interaction of antibody antigen complexes with protein A. Journal of lmmunology 115, 1617 1624.
472
o. KISTNER AND OTHERS
KLENK, H.-D., ROTT, R., ORLICH, M. & BL(JDORN, J. (1975). Activation of influenza viruses by trypsin treatment.
Virology 68, 426-439. LAZAROWITZ, S. G. & CHOPPIN, P. W. (1975). E n h a n c e m e n t of the infectivity of influenza A and B viruses by proteolytic cleavage of the hemagglutinin polypeptide. Virology 68, 440-454. PETRI, T. & DIMMOCK, N. J. (1981). Phosphorylation of influenza virus nucleoprotein in vivo. Journal of General Virology 57, 185-190. PETRI, T., PATTERSON, S. & DIMMOCK, N. J. (1982). Polymorphism of the NSI proteins of type A influenza virus. Journal of General Virology 61, 217-231. PRIVALSKY, M. L. & PENHOET, E. E. (1977). Phosphorylated protein component present in influenza virions. Journal of Virology 24, 401-405. PRIVALSKY, M. L. & PENHOET, E. E. (1978). Influenza virus proteins: identity, synthesis, and modification analyzed by two-dimensional gel electrophoresis. Proceedings of the National Academy of Sciences, U.S.A. 75, 36253629. PRIVALSKY,M. L. & PENHOET,E. E. (1981 ). The structure and synthesis of influenza virus phosphoproteins. Journal of Biological Chemistry 256, 5368-5376. SCHOLTISSEK,C. (1978). The genome of the influenza virus. Current Topics in Microbiology and Immunology 80, 139169. SCHOLTISSEK, C. & BOWLES, A. L. (1975). Isolation and characterization of temperature-sensitive mutants of fowl plague virus. Virology 67, 576-587. SCHOLTISSEK, C., HARMS, E., ROHDE, W., ORLICH, M. & ROTT, R. (1976). Correlation between R N A fragments of fowl plague virus and their corresponding gene functions. Virology 74, 332 344. SCHOLTISSEK, C., KOENNECKE, I. & ROTT, R. (1978). Host range recombinants of fowl plague (influenza A) virus. Virology 91, 79 85. SCHOLTISSEK, C., VALLBRACHT, A., FLEHMIG, B. & ROTT, R. (1979). Correlation of pathogenicity and gene constellation of influenza A viruses. 1I. Highly neurovirulent recombinants derived from non-neurovirulent or weakly neurovirulent parent virus strains. Virology 95, 492 500. SCHROM, M. & CALIGUIRI, L. A. (1981). Heterogeneity of influenza virus polypeptides during productive and abortive infections. In The Replication ofNegatit, e Strand Viruses, pp. 285-290. Edited by D. H. L. Bishop & R. W. Compans. New York: Elsevier/North-Holland. TER MEULEN,V. & LOVE, R. (1967). Virological, immunochemical and cytochemical studies of four HeLa-cell lines infected with two strains of influenza virus. Journal of Virology 1,626-639. VALLBRACHT, A., FLEHM1G, B. & GERTH, H.-J. (1979). Influenza virus: appearance of high mouse-neurovirulent recombinants, lntervirology 11, 16-22. ZIEbIIECKI, A. & FRIIS, R. R. (1980). Simultaneous injection of newborn rabbits with the Schrnidt-Ruppin and Prague strains of Rous sarcoma virus induces antibodies which recognize the pp60 src of both strains. Journal of General Virology 50, 211-216. (Receit~ed 24 July 1984)
