Journal of General Virology (1992), 73, 17-25. Printed in Great Britain

17

Specific binding of influenza A virus NS1 protein to the virus minus-sense RNA in vitro Eriko Hatada, Takenori Takizawa and Ryuji Fukuda* Department o f Biochemistry, Kanazawa University School o f Medicine, Kanazawa, Ishikawa 920, Japan

The non-structural protein NS1, encoded by genome segment 8 of influenza A virus, was expressed in Escherichia coli from cloned cDNA and purified. The NS1 protein had a specific RNA-binding activity, binding to influenza A virus minus-sense but not plussense RNA synthesized in vitro from cloned DNA using phage RNA polymerase. NS1 bound preferen-

tially to the regions of RNA containing either 5'- or 3'terminal common sequences of the genomic RNA. Binding was inhibited by virion RNA, but not by singlestranded minus-sense cDNA and oligo DNAs having the common sequences. In addition, binding was also inhibited by 28S rRNA but not 18S rRNA prepared from MDCK cells.

Introduction

(Morrongiello & Dales, 1977; Shaw & Compans, 1978; Petri et al., 1982), which contain a mixture of cellular RNA species including 28S ribosomal RNA (Yoshida et al., 1981). However, relatively little is known about the function of NS1 during the virus growth cycle. Studies of several temperature sensitive (ts) mutants of NS1 have suggested that the protein is involved in the synthesis of vRNA or the shutoff of host cell protein synthesis (Wolstenholme et al., 1980; Koennecke et al., 1981). Previously we have reported that synthesis of the late proteins and NS1 by two independent ts mutants, each having a single amino acid substitution in NS1 (Hatada et al., 1990), decreases markedly at a temperature of 40 °C (compared with that at the control temperature of 34 °C). At 40 °C, however, the levels of individual mRNAs, including those for late proteins, are almost the same as those at 34 °C (permissive temperature) and attain wild-type levels later in infection when the synthesis of the late proteins and NS1 decreases greatly. This observation suggests that NS1 is involved in some post-transcriptional process in the synthesis of the late proteins and NS1. To examine the possible role of NS1 in translational control, we carried out experiments to determine whether NS1 binds to viral mRNAs in vitro. We found that NS1 forms RNase-resistant complex(es) with virus minus-sense but not plus-sense RNAs, possibly by interacting with one of the common terminal sequences.

The genome of influenza A virus is composed of eight ssRNA segments of negative polarity (reviewed in Lamb & Choppin, 1983; McCauley & Mahy, 1983). These RNAs (vRNAs) have 13- and 12-nucleotide sequences at the 5' and 3' termini, respectively, that are common to all segments of type A influenza viruses. The common terminal sequences show partial inverted complementarity, and are postulated to have roles as recognition signals for viral transcription, replication or encapsidation (Hsu et al., 1987). Three types of virus-specific RNAs for each segment, mRNA, cRNA (the template for vRNA synthesis) and vRNA, are synthesized in the nuclei of infected cells (Herz et al., 1981). The mRNA has a 5' capped primer oligonucleotide and a 3' poly(A) tail, and is an incomplete transcript of vRNA in that it lacks a copy of the last 17 to 22 nucleotides at the 5' end, whereas the other type of transcript, cRNA, is a full-length copy of vRNA (Krug et al., 1979; Hay et al., 1977). RNA segment 8, the smallest segment, encodes two overlapping non-structural polypeptides, NS1 and NS2 (lnglis et al., 1979; Lamb & Choppin, 1979), which are found in infected cells, but not in the virion. NS1 is encoded by a collinear mRNA transcript, whereas NS2 is encoded by an interrupted mRNA (Inglis et al., 1980; Lamb & Lai, 1980). NS1 is found in the nucleus and nucleolus of infected cells, and is also associated with polysomes (Dimmock, 1969; Lazarowitz et al., 1971; Compans, 1973; Krug & Etkind, 1973). Recently, it has been shown that NS1 contains two independent signals for nuclear localization (Greenspan et al., 1988). In some strains of type A influenza virus NS1 forms electrondense paracrystalline inclusion bodies late in infection 0001-0472 © 1992SGM

Methods Construction of a recombinantplasmidfor expressionof the NS I protein.

The Escherichia coli expressionvectorpKK233-2(Amann& Brosius, 1985)has the tac promoterand the lacZ ribosome-bindingsitefollowed

18

E. Hatada,

T. T a k i z a w a

and R. Fukuda

by an ATG initiation codon which is contained within a unique NcoI site. A cDNA copy of genome segment 8 (containing the NS1 gene) of influenza virus A/PR/8/34 (H 1N 1) was cloned in the plasmid pSP64 (a gift from Dr K. Nagata). The recombinant plasmid has two BamHl sites, one between the first and the second codons of the NS 1 gene, and the other in the vector sequences downstream from the NS1 gene. An NS1 cDNA fragment was excised from the recombinant DNA using BamHI and the recessed 3' termini were filled with the Klenow fragment of DNA polymerase I. The fragment was then inserted by blunt-end ligation into the filled NcoI ends of pKK233-2, resulting in the correct NS1 reading frame without any extra amino acid sequences (pKK233-2/NS1). This expression plasmid was introduced into E. coli JM105. Purification o f the NS1 protein. NS1 was purified from E. coli cells (JM105), in which the synthesis of the protein from the expression plasmid pKK233-2/NS1 was induced with 1 mM-IPTG. The purification method was as described by Young et al. (1983). As some contaminating proteins were observed at the final step of this procedure (Fig. 1, lane 3), further purification steps were introduced to remove them. NS1 was precipitated from the solution by raising the MgCI 2 concentration to 17 mM. After mixing for 30 min at 4 °C, the precipitates were collected by centrifugation (12000g, 30 min) and pellets were dissolved in buffer B (40 mi-Tris-HC1 pH 7.5, 2 m i igCl2). NaC1 was then slowly added to a final concentration of 1 M and mixed for 30 min at 4 °C, resulting in the precipitation of NS1, which was collected by centrifugation as described above. The pellet was dissolved in and dialysed against buffer B overnight at 4 °C. Some precipitates appeared during dialysis, were removed by centrifugation and the supernatant was used. Analysis of this material by PAGE and Coomassie blue staining did not reveal any contaminating proteins (Fig. I, lane 2). This protein comigrated on polyacrylamide gels with the authentic NS1 synthesized in infected MDCK cells and was immunoprecipitated by anti-NS 1 antibody (a gift from Dr K. Shimizu). To obtain mock protein, induced E. coli cells carrying the expression vector but not the cDNA were treated according to the NS1 purification procedure of Young et al. (1983) without further purification steps (Fig. 1, lane 4). Construction o f plasmids for in vitro transcription o f virus RNA. To obtain template DNAs for in vitro transcription of virus plus- and minus-sense RNAs, eDNA copies of various genome segments of influenza virus A/Udorn/72 (H3N2) were cloned into plasmids pSP64 and pSP65 (Hatada et al., 1989), or the phagemids pTD-T7 (a gift from Dr T. Date) and pTZ 18U. For genome segments 7 and 8, recombinant phagemids were constructed to produce virus RNA transcripts with the correct 5' and 3' termini but including four extra bases (Fig. 2). EcoRIrestricted eDNA fragments were inserted in the phagemids in both orientations, and the recombinant phagemids were transfected into E. coli CJ236 with dut-1 and ung-1 mutations (Kunkel, 1985). In vitro mutagenesis was performed according to the protocols for the MutaGene phagemid in vitro mutagenesis kit (Bio-Rad), using as primers the oligodeoxynucleotides 5' C C T G C T T T T G C T T A T A G T G A G T C G 3' and 5' G G T C G A C T C T A G A G T A G A A A C A A G G 3' to obtain templates [pTDT7M(+) and pTDT7NS(+)] which provide the 5' and 3' ends of virus plus-sense RNA, respectively, and 5' CCTTGTTTCTACTTATAGTGAGTCG 3' and 5' GGTCGACTCT A G A G C A A A A G C A G G 3' to obtain templates [ p T Z M ( - ) and p T Z N S ( - ) ] which provide the 5' and 3' ends of virus minus-sense RNA, respectively. These recombinant phagemid DNAs, which were linearized by XbaI digestion, provided run-off transcripts having the correct 5' and 3' termini with four extra bases of the XbaI site. In addition, clones which were used as templates for some truncated M and NS transcripts were constructed using appropriate restriction enzymes as indicated in Fig. 2.

For NS1 binding experiments employing RNAs of other genome segments (Fig. 7b), truncated 32p-labelled RNAs were synthesized in vitro as reported previously (Table 1 of Hatada et al., 1989). The plussense and minus-sense RNAs carry the 3'- and Y-terminal common sequences, respectively [indicated in the figure as (3') and (5')], both of which are flanked by the same extra sequence originating from the vector. Preparation o f 3eP-labelled virus RNAs. Template DNAs linearized as mentioned above were transcribed in an in vitro system using T7 or SP6 phage RNA polymerase, and [c~-32p]UTP as described previously (Hatada et al., 1989). RNA binding assay. To study the RNA-binding activity of NS1, we modified the procedures published previously for an RNA mobility shift assay and a ribonuclease protection assay (Garner et al., 1981 ; Zapp & Green, 1989; Heaphy et al., 1990; Marciniak et al., 1990). About 1 ~tg of NS1 (38 pmol) and 100 to 600 fmol RNA labelled with [c~-3~P]UTP (3 x 103 to 5 x 103 c.p.m./pmol) (see above) were incubated in 15 ~1i binding buffer containing 6.6 mM-Tris-HCl pH 7-6, 33 mM-KC1, 1 mM-EDTA and 0-3 mM-DTT for 30 rain at 30 °C. The incubation was continued for 10 to 30 min in the presence or absence of either 0-75 ~tg of RNase A or 1.5 units of RNase T2. After addition of gel loading buffer (5% glycerol, 1 mM-EDTA, 0.05% xylene cyanol and 0.05% bromophenol blue; final concentrations), the mixtures were applied to a non-denaturing 4 ~ polyacrylamide gel (acrylamide :bisacrylamide, 60:1) in 0.5 x TBE buffer (1 × TBE contains 90 mM-Trisborate pH 8.0 and 2.5 mM-EDTA). Electrophoresis was carried out at a constant voltage of 2-5 V/cm in 0-5 x TBE buffer for 14 h at 4 °C. The gels were dried and exposed to Fuji RX X-ray film for autoradiography. For competition experiments NS1 was preincubated with competitor RNA for 30 min at 30 °C prior to addition of the 32p-labelled RNA. Irnrnunoprecipitation assay. Antibody against NS1 was raised in a rabbit, from the serum of which the IgG fraction was precipitated using 3 3 ~ ammonium sulphate. The collected pellets were dissolved and dialysed against PBS, and then applied to a DE52 cellulose column. The flow-through fraction of the column was used as the IgG fraction. Complexes between NS1 and 32p-labelled RNAs in the binding buffer (15 ~tl)were diluted with PBS up to 50 ~tl, and incubated with the anti-NSt IgG fraction (0-5 to 5 ~1) or a non-specific IgG fraction (5 ~tl) in the presence of 2 ~tl of 30~ Protein A-Sepharose for 1 h at room temperature. After washing several times in 100 ~tl of buffer containing 50 mi-Tris-HCl pH 7.5, 150 mM-NaC1 and 5 mM-EDTA, the beads were resuspended in 10 ~tl of 80 ~ formamide, 10 mm-Tris-HCl pH 7.5, 1 mM-EDTA and 1~ SDS, and heated to 95 °C for 3 rain to release the bound RNA. The beads were then removed by centrifugation and the supernatant containing the released RNA was applied to a 6 ~ denaturing polyacrylamide gel containing 7M-urea in 0.5 x TBE buffer. Bands were visualized by autoradiography as described above. Detection o f the NS1 protein in the shifted RNA bands. The radioactive bands on the dried gel corresponding to the shifted RNA bands (complex) were excised and swelled in 200 ~tl of H20 for 3 h at 37 °C. They were then supplemented with 50 ~tl of gel loading buffer containing 250 mM-Tris-HCl pH 6-8, 20 mM-DTT, 2 0 ~ glycerol and 0.05~ bromophenol blue, heated for 2 min at 100 °C, and subjected to 12.5 ~ SDS-PAGE (Laemmli, 1970). Gels were silver-stained (Oakley et al., 1980) to visualize proteins. Materials. Plasmids pKK233-2 and pTZ18U were purchased from Pharmacia and Bio-Rad, respectively. Protein A-Sepharose was from Sigma, proteinase K from Merck and T7 phage RNA polymerase from Takara Syuzo. The oligodeoxynuc!eotides were synthesized using an Applied Biosystems DNA synthesizer (Model 380 B). Sources of other materials have been described previously (Hatada et aL, 1989, 1990).

Influenza virus N S I : RNA-binding specificity

19

RNA Probes for the Gel Shift Assay

Results

35 174 18II 7 9 6 832 Ac~/No }~l ~ A I . . . . . . A£.-. . . . . . . . . . . . = .... n a m e NS cDNA1 l ~ Z ~ a ; , . ~ : . ~ % , ~ L : . 7 > = ~ l ~ 890

Probe

Specific binding of NSI to virus minus-sense RNAs The RNA-binding activity of NS1 was studied using a RNA mobility shift assay and a ribonuclease protection assay. Pure NS1 was prepared from E. coli cells, in which the synthesis of the protein from a cDNA copy of A/PR/8/34 (H 1N 1) segment 8 inserted in the expression vector pKK233-2 had been induced using IPTG (Fig. 1, lane 2 and see Methods). A mock protein preparation (Fig. 1, lane 4) was made from induced E. coli cells carrying the expression vector but not the cDNA insert. Virus plus- or minus-sense RNAs were transcribed in vitro from the virus cDNAs connected to the bacteriophage promoters of SP6, pTZ18U and pTDT7 plasmids (see Methods). These recombinant plasmid DNAs, which were linearized by restriction enzyme digestion, provided run-off transcripts of type A influenza virus [A/Udorn/72 (H3N2)] RNAs with or without the exact common 5'- or Y-terminal sequences (Fig. 2). 32p-Labelled ssRNA transcripts with or without added NS1 were incubated for 30 min at 30 °C, and some were treated with RNase A o r T 2. They were then applied to a 4 ~ non-denaturing polyacrylamide gel in TBE buffer and subjected to electrophoresis at 2-5 V/cm for 14 h at 4 °C. As shown in Fig. 3, long RNAs moved as

NS(-)E 5'4 NS(-) 5'1 NS(-)(189-890) ~'~I27NS(-)(832-890) NS(-)(I 174) 5' IELT_--]3' NS(-)(1-35) 5' I] 3' NS(-)(189-796) 5'~-f] NS(+)E NS(+) NS(+)(I

M(-)E 5'~1 M(-) 5'i M(-)(739-1027) M(-)(1 287) 5'~I--Z-~] M(-)(1 29) M(-)(739-891) M(+)E M(+) M(+)(I 287)

{ ) M[

)(1 29} ~I

35) (+]

1

2

3

174)

3'q] 3'4] 3' - g - - l ~ 5 '

I]~' fl 5' 738 ,~a ,..,, ~

891

f~9 Io27 NSI . . ~ binding

~':~' + ]3' + 5' ~ ! 3 ' + +

~'

s' !1 3' 5' ~ 3'~11 3'R 3'~U---~

3'

+ iI]~ 5' II s' -

s'

5'AGUAGAAACAAGGNNNUtJ[]UUN

Q]

.............

IIAG . . . . .

UACUCCAG

.............

GIIG . . . . .

IJAUCAIJtlAAAUAAG35

5'A(~CAAAAGCAGGN

+ + + + + + +_

I 3'

29 287 Probe /}, ~t name m cDNA I .:-~'~=.~-,r~=~=~ [Z~~m~;%Y-'

N%(

M

I ~ 3' t3' ] 3' s ' , ~ ] 3'

NSI binding

. . . . . . .

b!CCUGCUUUUGCtJ

d'

2~

|{

NAAAAA

N N N CCUUGUULJCUACU

~'

4

• NSI

Fig. 1. Purification of NS1 expressed in E. coli cells. Proteins were fractionated by 17.5~ S D S - P A G E and the gel was stained with Coomassie blue. Lane M, M~ markers; M~s are indicated to the left. Lane 1, 1 gg of bovine trypsinogen as a protein concentration marker ( M r 24 000); lane 2, purified NS 1 fraction (about 1 pg); lane 3, partially purified NSI fraction (about 2 gg); lane 4, mock protein preparation (0-06 A280 units).

Fig. 2. The structure of R N A probes used in the R N A - b i n d i n g assay. These probes were transcribed in vitro from linearized phagemid D N A s carrying c D N A copies of genome segments 8 (NS) and 7 (M) (see Methods). E represents 20 to 30 extra bases and numerals in parentheses indicate n u m b e r of nucleotides (numbered from the 5' end of v R N A ) in the c D N A transcribed into R N A . Filled and striped boxes show the c o m m o n terminal sequences of virus minus- and plussense R N A s , respectively. W a v y lines represent extra bases originating from the vector sequence (20 to 30, or four bases); stippled boxes show the respective c D N A s with the positions of restriction enzyme sites at which c D N A s were cleaved to construct the templates for truncated R N A s ; Ac, AI, Ba, Bg, St and Ta represent AccI, AluI, BamHI, BgtlI, StuI and TaqI sites, respectively; the cleavage points are indicated above the sites. At the bottom, the c o m m o n terminal sequences and the sequences of M ( - ) ( 1 29) and N S ( - ) ( 1 - 3 5 ) R N A s are shown. Dashes indicate nucleotides identical to those of minus-sense R N A ( - ) .

a broad smear in this gel (lanes 5, 10, 15 and 20). In the presence of NS1 (molar NS1 :RNA ratio, about 190:1), however, the movement of a portion of the RNAs was decreased (lanes, 1, 6 and 11). Upon digestion by RNase, some radioactivity in the more slowly migrating fraction remained resistant to digestion for minus-sense NS and M RNAs [ N S ( - ) and M ( - ) RNAs, respectively] (lanes 2 and 12), but not that for plus-sense RNAs (lanes 7 and 17). Using a short polynucleotide consisting of the 5'-

20

E. Hatada, T. Takizawa and R. Fukuda 32p-labelled RNA NSlprotein RNase SDS

NS(-) I

+

+

+

-

+

1

2

+ + + 3 4

NS(+) I

5

I

M(-) I

+

+

+

6

+ 7

+ + + 8 9 10

I

+

+

-

+

+

M(+) I

+ + + 11 12 13 14 15

i

M(-X1-29) 1

+

+

+

16

+ + + - + 17 18 19 20

f

+

+

-

+

2l

+

-

I

+ + + 22 23 24

Complex Eomplex

Free (digested)

Free

Fig. 3. Assay of NS 1 binding to influenza virus RNAs. The binding buffer containing 1 ktg of NS 1 (lanes 1 to 3, 6 to 8, 11 to 13, 16 to 18 and 21 to 23) or buffer B (lanes 4, 5, 9, 10, 14, 15, 19, 20 and 24) was mixed with 32p-labelled RNA probes (see Fig. 2): 200 fmol of NS( - ) (lanes 1 to 5), 200 fmol of NS(+) (lanes 6 to 10), 200 fmol of M ( - ) (lanes 11 to 15), 200 fmol of M(+) (lanes 16 to 20) or 600 fmol of M ( - )(1-29) (lanes 21 to 24). The mixtures were incubated for 30 min at 30 °C and continued in the presence (lanes 2 to 4, 7 to 9, 12 to 14, 17 to 19, and 22 to 24) or in the absence (lanes 1, 5, 6, 10, 11, 15, 16, 20 and 21) of 0.75 p.g of RNase A. Some samples were treated with 1.25~ SDS (lanes 3, 8, 13, 18 and 23) for 2 min at 100 °C. The position of the NS1-RNA complex is indicated.

t e r m i n a l 29 n u c l e o t i d e s o f t h e m i n u s - s e n s e M R N A [designated M(-)(1-29)], half the amount of the p o l y n u c l e o t i d e m i g r a t e d s l o w l y (lane 21) a n d a b o u t 3 0 ~ o f t h e r a d i o a c t i v i t y in t h i s f r a c t i o n r e m a i n e d R N a s e r e s i s t a n t (lane 22) in t h e p r e s e n c e o f N S 1 in a 60-fold m o l a r excess. U p o n a d d i t i o n o f 1"25~o S D S , t h e s e R N a s e - r e s i s t a n t b a n d s d i s a p p e a r e d (lanes 3, 13 a n d 23). To characterize these RNase-resistant bands further, d o s e - r e s p o n s e e x p e r i m e n t s w e r e p e r f o r m e d u s i n g 32p_ labelled M(-)(1-29) RNA. Increasing the amount of R N A a d d e d to 38 p m o l N S 1 , a s a t u r a t i o n l e v e l w a s a t t a i n e d w i t h 600 f m o l (Fig. 4, l a n e 3). U s i n g 600 f m o l R N A , t h e a m o u n t o f b i n d i n g c o m p l e x i n c r e a s e d in p r o p o r t i o n to a d d e d N S 1 (Fig. 4, l a n e s 5 to 8). A s s h o w n in l a n e 10, t h e m o c k p r o t e i n p r e p a r a t i o n d i d n o t g i v e t h e RNase-resistant complex. No RNase-resistant band was o b s e r v e d w h e n N S 1 w a s h e a t e d for 2 m i n at 100 °C (lane 12), or t r e a t e d w i t h 500 rtg/ml p r o t e i n a s e K (lane 13) before mixing with 32p-labelled RNA. In the latter experiment, proteinase K digested RNase and thus digestion of 32p-labelled RNA was incomplete. T o d e m o n s t r a t e f u r t h e r t h e b i n d i n g o f N S 1 to R N A , t h e R N A [ 3 2 p - l a b e l l e d M ( - )(1-29)] w a s i n c u b a t e d w i t h either NS1 or the mock protein preparation. These mixtures were then treated with either the anti-NS 1 IgG

1

Complex ~ ~

2

3 4

5

6

7

8

9

10 11 1213

JJ6 4 Free

Free (digested)

.~ Free (digested)

Fig. 4. Characterization of the RNA-binding activity of NS1. Increasing amounts of M(-)(1-29) labelled RNA (200, 400, 600 and 800fmol; lanes 1 to 4, respectively) were added to 1 ~tg NS1 or increasing amounts of NS1 (0.25, 0.5, l-0 and 2.0~g; lanes 5 to 8, respectively) were added to 600 fmol of M(-)(1-29) labelled RNA. Using 200 fmol of labelled NS(--) RNA, the binding complex was observed with 1 ~tg of NS1 (lane 9), but not with 0.06 Az8o units of a mock protein preparation (lane 10). Prior to mixing with NS(-)(832890) labelled RNA (600 fmol), 1 ~tg of NS1 was not treated (lane 11) or pretreated by heating for 2 min at 100 °C (lane 12), or digested with 0.5 mg/ml or proteinase Kfor 30 rriin at 37 ~C (lane 13). Note that 32p. labelled RNA digestion was incomplete because proteinase K inactivated the RNase (lane 13).

Influenza virus N S I : RNA-binding specificity (a)

P

1

2

3

4

5

6

7

1

2

3

4

5

6

7

8

21 9

Q

M(-)

(1-29)

(b)

C C

P

1

2

C

C'

3

4

5

6

M

NS1

k~

Fig. 5. Detection of NS1 in the complex (C). (a)Immunoprecipitation of M(-)(1-29)-NS1 complex with anti-NS1 antibody. Binding reactions were performed using 600 fmol of 32p-labelled M(-)(1-29) and 1 ~tg of NS1 (lanes 1 to 5) or 0.06 A280 units of mock protein preparation (lanes 6 and 7), and treated with an increasing amount of polyclonal anti-NS1 IgG (lanes 1 to 4, and 6; 0.5, 1, 2, 5 and 5 ~tl, respectively) or 5 ~tl of non-specific IgG (lanes 5 and 7) in the presence of Protein A Sepharose. 32p-labelled RNA was released from the immunoprecipitates (see Methods) and subjected to electrophoresis on a 6 ~ polyacrylamide-7 M-urea gel in 0.5 × TBE buffer, followed by autoradiography. Lane P, 50fmol of 3zP-labelled M ( - ) ( I 29). (b) Presence of NS1 in shifted RNA bands. Lanes P, 1 and 2 show an autoradiogram of an RNA-binding assay; for the binding reaction, 1 ~tg of NS1 was incubated with 600 fmol of 32p-labelled M(-)(1-29) (lane 1) or 200 fmol of 32p_labelled N S ( - ) (lane 2); lane P, 32P-labelled M(-)(1 29) alone (150 fmol). The radioactive bands on the dried gel corresponding to complexes (as indicated in the figure) were cut out, treated as described in Methods and subjected to 12.5~ SDS-PAGE (second gel electrophoresis). The gel was silver-stained (lanes C, C', 3 to 6 and M). The following specimens were subjected to a second gel electrophoresis (all the gel pieces contained pieces of 3MM filter paper on which the gel had been mounted): a piece of blank gel (lanes C and C'); pieces of gel containing complexes from lanes 1 and 2 (lanes 3 and 4, respectively); a piece of gel containing free NS1 (lane 5); N S1 (lane 6); the same Mr markers as in Fig. 1 (lane M). f r a c t i o n or a n o n - s p e c i f i c I g G f r a c t i o n (Fig. 5a). I n t h e p r e s e n c e o f b o t h NS1 a n d a n t i - N S 1 I g G 32p-labelled R N A was r e c o v e r e d in t h e i m m u n o p r e c i p i t a t e s (lanes 1 to 4), b u t n o t w h e n n o n - s p e c i f i c I g G or the m o c k p r e p a r a t i o n were e m p l o y e d (lanes 5, 6 a n d 7). T h e

Fig. 6. Competition among various RNAs for NS1 binding. Prior to addition of 32p-labelled N S ( - ) (200 fmol), 1 ~tg of NS1 was preincubated with an excess of various RNAs; no competitor (lanes 1 and 8), 2 pmol of M(+) RNA (lane 2), 2 pmol of M ( - ) RNA (lane 3), 60 pmol of 18S rRNA and 26 pmol of 28S rRNA prepared from MDCK cells (Scherrer & Darnell, 1962) (lanes 4 and 5, respectively), 1 ktg of vRNA prepared from WSN virions (lane 6), 80 pmol ofE. colitRNA (lane 7), 2 pmol of homologous N S ( - ) RNA (lane 9). p r e s e n c e o f N S 1 in the R N a s e - r e s i s t a n t b a n d s o f the first gel was d e m o n s t r a t e d d i r e c t l y b y e x c i s i n g the b a n d s , s u b j e c t i n g t h e m to a s e c o n d S D S P A G E a n d s t a i n i n g w i t h silver (Fig. 5 b ; R N A w a s n o t s t a i n e d ) . A s the gel p i e c e s f r o m the first gel also c o n t a i n e d p i e c e s o f filter p a p e r w h i c h c o u l d n o t be s t r i p p e d off t h e gel, silvers t a i n i n g b a n d s o r i g i n a t i n g f r o m t h e p a p e r were seen n e a r the t o p o f the gel (lanes C a n d C'). E x c l u d i n g these b a n d s , t h e o n l y v i s i b l e b a n d c o m i g r a t e d w i t h NS1 ( c o m p a r e l a n e s 3 a n d 4 w i t h l a n e 5; see legend). T h e s e results s h o w e d t h a t the R N a s e - r e s i s t a n t b a n d s were N S 1 - R N A c o m p l e x e s , i n d i c a t i n g specific b i n d i n g o f NS1 to viral m i n u s - s e n s e R N A .

Competition o f R N A binding o f NS1 by various R N A s T o e x a m i n e t h e specificity o f c o m p l e x f o r m a t i o n , NS1 was p r e i n c u b a t e d w i t h a n excess o f v a r i o u s u n l a b e l l e d R N A s (Fig. 6). B i n d i n g to l a b e l l e d N S ( - ) R N A was

E. Hatada, T. Takizawa and R. Fukuda

22

(a) l

2

3

4 5 6

7

8

P

9

10 11 12 13 14

O

Complex

Free (digestedl

(b) 1 2 3

4

5 6

7

8

9 10 11 12

1314

Complex •

Free

(digested)

inhibited almost completely with a 10-fold molar excess of the homologous N S ( - ) R N A (lane 9). W i t h similar quantities of M ( - ) R N A , binding decreased to 4 0 ~ o f that in the absence o f competition (lane 3), but remained u n c h a n g e d with M ( + ) R N A (lane 2). Total virus genomic R N A (16-fold excess with regard to the n u m b e r o f R N A molecules) inhibited binding almost completely (lane 6). N o appreciable inhibition was observed with 18S ribosomal R N A (300-fold molar excess, lane 4) or t R N A (400-fold molar excess, lane 7). In contrast, 28S ribosomal R N A (130-fold molar excess) greatly inhibited binding (lane 5), in agreement with a previous report showing that NS1 binds to 28S ribosomal R N A in vivo (Yoshida et al., 1981). N o competition was observed with single-stranded viral c D N A s of either sense (data not shown).

Fig. 7. Binding of NSI to truncated virus RNAs. (a) NS1 binding to various truncated N S ( - ) and M ( - ) RNAs. To the binding buffer containing 1 ~tg NS1 (odd-numbered lanes) or without NS1 (even-numbered lanes), 32p-labelled in vitro transcripts were added: 200 fmol N S ( - ) (lanes 1 and 2), 400 fmol of NS(-)(189-890) (lanes 3 and 4), 400 fmol of NS(-)(189-769) (lanes 5 and 6), 400 fmol of N S ( - ) ( l - 3 5 )

(lanes 7 and 8), 400 fmol of M(-)(739-1027) (lanes 9 and 10), 400 fmol of M(- )(739 891) (lanes 11 and 12), 400 fmol of M(-)(1 29) (lanes 13 and 14). After incubation, the mixtures were digested with RNase A and subjected to gel electrophoresis. (b) NSI binding to truncated plus- or minussense RNA of other virus genome segments. These 32p. labelled RNAs were synthesized in vitro as described in Methods. The plus-sense (+) and the minus-sense ( - ) RNAs carry the 3'- and Y-terminal common sequences, respectively. In lanes 1 to 6, NSI binding to NP RNAs was examined. To the binding buffer containing 1 ~tgNSI (lanes 1, 2, 4 and 5), or no NS1 (lanes 3 and 6), 400 fmol of 32p. labelled NP(-)(5') (lanes 1 to 3) or NP(+)(3') (lanes 4 to 6) was added. RNase digestion was performed in lanes 2, 3, 5 and 6, but not in lanes 1 and 4. In lanes 7 to 14, 1 ~tg of NS1 was incubated with 400 fmol of 3zp-labelled NA(-)(5'), N A( + )(3'), HA( - )(5'), HA(+)(33, PA(-)(5'), PA( + )(3"), PB 1( - )(5') and PB1(+)(3'), respectively. RNA in lanes 7 to 14 was digested with RNase.

Requirement of terminal sequences of viral minus-sense RNA for NS1 binding To examine the sequences required for binding o f NS 1 to viral R N A , various deletions were generated in M or N S R N A s as shown in Fig. 2. In Fig. 7 (a), binding patterns are shown only for deletion mutations of N S ( - ) and M ( - ) R N A s (other results are summarized in Fig. 2). Full-length N S ( - ) R N A was the best substrate examined for NS1 binding. In the presence o f either terminal sequence considerable binding activity was retained, as shown with N S ( - ) ( 1 8 9 - 8 9 0 ) and N S ( - ) ( 1 - 3 5 ) R N A s which had the intact c o m m o n 3'- and 5'-terminal sequence o f genomic R N A , respectively (although the former R N A had four extra bases) (lanes 3 and 7). In contrast, the binding activity decreased appreciably with

Influenza virus N S I : RNA-binding specificity

RNA with deletions in both terminal regions (lane 5). The binding activity of full-length M ( - ) RNA was moderate (Fig. 3, compare lanes 12 and 2) and increased with truncated RNAs retaining either terminal region. A 5'-terminal region as short as 29 nucleotides [M(-)(129)] exhibited good binding activity, as did M(-)(7391027), which has the Y-terminal region (lanes 13 and 9). Removal of both terminal regions eliminated binding activity (lane 11). The binding activity of plus-sense RNAs was very weak compared to that of their minus-sense counterparts, even if they had intact common terminal sequences. Preferential binding to viral minus-sense RNAs was also observed for NP, NA, HA, PA and PBI RNAs, as shown in Fig. 7(b). The presence of extra sequences at either end of viral RNA had little if any effect on binding activity. The common sequences exhibiting NS 1-binding activity are both common terminal sequences of viral genomic RNA. NS1 may discriminate between and bind to either of these sequences. Deoxyoligonucleotides of these common sequences neither competed with the NS1binding activity of these RNAs nor formed complexes with NS1 (data not shown), indicating that NS1 is an RNA-binding protein.

Discussion In this report we have demonstrated that the NS1 polypeptide binds preferentially to minus-sense RNAs of influenza virus segments 7 and 8 having either 5'- or 3'terminal common sequence (Fig. 3 and 7a). This preferential binding has also been observed with minussense RNAs of segments 2, 3, 4, 5 and 6 having the 5'terminal sequence (Fig. 7b). These observations strongly suggest that NS1 recognizes both the 5'- and Y-terminal common sequences of virus minus-sense RNAs. If it is assumed that the interaction is sequence-dependent, it is important to elucidate how each of the two common sequences participates in the binding, whether independently or cooperatively as a hybrid. NS1 is an RNAbinding protein because it bound neither to oligodeoxynucleotides of these common terminal sequences nor to ssDNA copies of virus minus-sense RNAs (data not shown). Further studies, including footprinting experiments, are now in progress to analyse the binding properties and the exact binding sequences in the viral RNAs. Thirteen and 12 nucleotides at the 5' and 3' termini, respectively, of influenza type A vRNAs are common to all RNA segments, except for a variation in the fourth base from the 3' end. These common sequences show partial inverted complementarity and form a hybrid in the virion and the infected cell (Hsu et

23

al., 1987). Thus, it is possible that NS1 interacts with the hybrid region of vRNAs which is thought to be associated with viral polymerase proteins and NP in the infected cell, and regulates either viral transcription, replication or encapsidation, for which the terminal common sequences are postulated to have roles as recognition signals. Previous studies of two ts mutants of segment 8 have suggested an involvement of the protein in the synthesis of vRNA (Wolstenholme et al., 1980). However, there is no evidence from studies of NS1 mutants that NS1 participates in viral mRNA synthesis or in vitro viral transcription. In a previous paper we reported that the synthesis of viral late proteins and of NS1 by two independent ts mutants decreased severely without affecting the level of viral mRNA synthesis (Hatada et al., 1990). The observations suggest that NS 1, most of which is localized in the nucleus at least during the first few hours of infection, is involved in some post-transcriptional process in the synthesis of the late proteins and NS1. There are several possible models for this mechanism. One possibility is that NS 1 regulates the nuclear-cytoplasmic transport of mRNAs encoding these proteins, which is not the case for M mRNA (Hatada et al., 1990). Another is the gene gating hypothesis proposed by Blobel (1985). NS1 may locate vRNA (the template to be transcribed) via specific interactions with the terminal sequences, in a nuclear subcompartment favourable for translation factor assembly, processing factor and other factors necessary for post-transcriptional control. The RNA transcribed in this subcompartment would be efficiently transported to and translated in the cytoplasm. As indicated by competition experiments (Fig. 6), NS 1 interacted with 28S rRNA. This observation is in agreement with that of Yoshida et al. (1981), who reported that cytoplasmic paracrystalline inclusions are composed of NS1 in association with RNA, including 28S rRNA (in BHK cells), but not poly(A)-containing RNA. Computer searches of rRNA sequences were performed to find those resembling the 5'- and Y-terminal common sequences of the minus-strand. As data were not available for canine rRNAs, the sequences of human and mouse rRNA were employed, both of which gave the same results. Six matching sequences of six bases each are present in 28S rRNA and the Y-terminal common sequence (Fig. 8, sequences a to f), and there are one seven-base and four six-base matches between 28S rRNA and the Y-terminal common sequence. On the other hand, one seven-base (Fig. 8, sequence g) and one six-base homology (sequence b) are present in 18S rRNA and the 5'-terminal common sequence, whereas no significant homology is found between 18S rRNA and the Y-terminal common sequence. From competition

24

E. Hatada, T. Takizawa and R. Fukuda

28S rRNA (human and mouse) 5' consensus of minus strand 5' 3r AGUAGAAACAAGG

3' consensus of minus strand 5~ 3' CCUGCUUUUGCU

bc

L__I Six six-base homologous sequences

One seven-base and four six-base homologous sequences

18S rRNA (human and mouse) 5' consensus of minus strand

3' consensus of minus strand

5' 3' AGUAGAAACAAGG

5' 3' CCUGCUUUUGCU

9 One seven-base and one six-base homologous sequence

No homologous sequence

Fig. 8. Identification of regions of homology rRNA sequences and the common terminal sequences of the viral minus-sense RNA. For the analysis, DNASIS-DBREF50 (Ver. 7.0) (Hitachi SK software) was employed.

experiments (Fig. 6) we can deduce that sequence a, c or d (Fig. 8) of 28S rRNA is involved in the NS1 interaction. Of these, sequence d is present in the sequence of the NS plus strand and can be excluded. It is thus suggested that either sequence a or sequence c is the most probable part of the 5' terminus which is recognized by NS1. Detailed analysis of the N S I - r R N A interaction in vivo and in vitro is in progress. When this manuscript was in preparation, Skorko et al. (1991) reported that NSI binds non-specifically to all ssRNAs and has greater RNA-binding activity than was shown in this study. In the absence of a detailed description of their binding conditions, we cannot explain this discrepancy, although they did not perform an RNase protection assay. We clearly demonstrated that NS1 binding to minus-sense R N A was greater, even in the absence of RNase digestion (Fig. 7b, compare lanes 1 and 4). We thank Dr K. Nagata for the NS cDNA (A/PR8/34), Dr K. Shimizu for the anti-NS1 antibody that was used to identify NS1 and Dr T. Date for the pTDT7 plasmid. We are very grateful to Dr M. Enami for comments on this manuscript.

References AMANN, E.& BROSIUS,J. (1985). 'ATG vectors' for regulated high-level expression of cloned genes in Escherichia coli. Gene 40, 183-190. BLOBEL, G. (1985). Gene gating: a hypothesis. Proceedings of the National Academy of Sciences, U.S.A. 82, 8527-8529.

COMPANS, R. W. (1973). Influenza virus proteins. II. Association with components of the cytoplasm. Virology 51, 56-70. DIMMOCK, N. J. (1969). New virus-specific antigens in cells infected with influenza virus. Virology 39, 224-234. GARNER, M. M. & REVZIN, A. (1981). A gel electrophoresis method for quantifying the binding of proteins to specific DNA regions: application to components of the Escherichia coli lactose operon regulatory system. Nucleic Acids Research 9, 3047 3060. GREENSPAN, D., PALESE, P. & KRYSTAL, M. (1988). Two nuclear location signals in the influenza virus NS1 nonstructural protein. Journal of Virology 62, 3020-3026. HATADA, E., HASEGAWA,M., MUKAIGAWA,J., SHIMIZU, K. & FUKUDA, R. (1989). Control of influenza virus gene expression: quantitative analysis of each viral RNA species in infected cells. Journal of Biochemistry 105, 537-546. HATADA, E., HASEGAWA,M., SHIMIZU, K., HATANAKA,M. & FUKUDA, R. (1990). Analysis of influenza A virus temperature-sensitive mutants with mutations in RNA segment 8. Journal of General Virology 71, 1283-1292. HAY, A. J., ABRAHAM,G., SKEHEL, J. J., SMITH, J. C. & FELLNER, P. (1977). Influenza virus messenger RNAs are incomplete transcripts of the genome RNAs. Nucleic Acids Research 4, 4197-4209. HEAPHY, S., DINGWALL, C., ERNBER.G, I., GAIT, M. J., GREEN, S. M., KARN, J., LOWE, A. D., SINGH, M. & SKINNER, M. A. (1990). HIV-1 regulator of virion expression (rev) protein binds to an RNA stemloop structure located within the rev response element region. Cell60, 685 693. HERZ, C., STAVNEZER, E., KRUG, R. M. & GURNEY, T., JR (1981). Influenza virus, an RNA virus, synthesizes its messenger RNA in the nucleus of infected cells. Cell 26, 391-400. Hsu, M. T., PARVIN, J. D., GUPTA, S., KRYSTAL, i . & PALESE, P. (1987). Genomic RNAs of influenza viruses are held in a circular conformation in virions and in infected cells by a terminal panhandle. Proceedings of the National Academy of Sciences, U.S.A. 84, 8140-8144. INGLIS, S. C., BARRETT, T., BROWN, C. M. & ALMOND, J. W. (1979). The smallest genome RNA segment of influenza virus contains two genes that may overlap. Proceedings of the National Academy of Sciences, U.S.A. 76, 3790-3794. INGLIS, S. C., GETHING, M.-J. & BROWN, C. M. (1980). Relationship between the messenger RNAs transcribed from two overlapping genes of influenza virus. Nucleic Acids Research 8, 3575 3589. KOENNECKE, I., BOSCHEK, C. B. & SCHOLTISSEK, C. (1981). Isolation and properties of a temperature-sensitive mutant (ts412) of an influenza A virus recombinant with a ts lesion in the gene coding for the nonstructural protein. Virology 110, 16-25. KRUG, R. M. & ETKIND, P. R. (1973). Cytoplasmic and nuclear virusspecific proteins in influenza virus-infected MDCK ceils. Virology 56, 334-348. KRUG, R. M., BRONI, B. B. & BOULOY, M. (1979). Are the 5' ends of influenza viral mRNAs synthesized in vivo donated by host mRNAs ? Cell 18, 329-334. KUNKEL, T. A. (1985). Rapid and efficient site-specific mutagenesis without phenotypic selection. Proceedings of the National Academy of Sciences, U.S.A. 82, 488-492. LAEMMLI, U. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature, London 227, 680685. LAMB, R. A. & CHOPPIN, P. W. (1979). Segment 8 of the influenza virus genome is unique in coding for two polypeptides. Proceedings of the National Academy of Sciences, U.S.A. 76, 4908-4912. LAMB, R. A. & CHOPPIN, P. W. (1983). Gene structure and replication of influenza virus. Annual Review of Biochemistry 52, 467-506. LAMB, R. A. & LAI, C.-J. (1980). Sequence of interrupted and uninterrupted mRNAs and cloned DNA coding for the two overlapping nonstructural proteins of influenza virus. Cell 21, 475485. LAZAROWlTZ, S. G., COMPANS, R. W. & CHOPPIN, P. W. (1971). Influenza virus structural and nonstructural proteins in infected cells and their plasma membranes. Virology 46, 830 843. MCCAULEY, J. W. & MAHY, B. W. J. (1983). Structure and function of the influenza virus genome (a review). Biochemical Journal 211, 281294.

Influenza virus N S 1 : R N A - b i n d i n g specificity

MARCINIAK, R. A., CALNAN, B. J., FRANKEL, A. D. & SHARP, P. A. (1990). HIV-1 tat protein trans-activates transcription in vitro. Cell 63, 791 802. MORRONGIELLO, M. P. & DALES, S. (1977). Characterization of cytoplasmic inclusions formed during influenza/WSN virus infection of chick embryo fibroblasts. Intervirology 8, 281-293. OAKLEY, B. R., KIRSCH, D. R. & MORRIS, N. R. (1980). A simplified ultrasensitive silver stain for detecting proteins in polyacrylamide gels. Analytical Biochemistry 105, 361-363. 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. SCHERRER, K. & DARNELL, J. E. (1962). Sedimentation characteristics of rapidly labelled RNA from HeLa cells. Biochemical and Biophysical Research Communications 7, 486~90. SHAW, M. W. & COMPANS,R. W. (1978). Isolation and characterization of cytoplasmic inclusions from influenca A virus-infected cells. Journal of Virology' 25, 605 615. SKORKO, R., SUMMERS, D. F. & GALARZA, J. M. (1991). Influenza A virus in vitro transcription: roles of NSI and NP proteins in regulating RNA synthesis. Virology 180, 668-677.

25

WOLSTENttOLME, A. J., BARRETT, T., NICHOL, S. T. & MAHY, W. J. (1980). Influenza virus-specific RNA and protein synthesis in cells infected with temperature-sensitive mutants defective in the genome segment encoding non-structural proteins. Journal of Virology 35, 1-7. YOSHIDA, T., SHAW, M. W., YOUNG, J. F. & COMPANS, R. W. (1981). Characterization of the RNA associated with influenza A cytoplasmic inclusions and the interaction of NS1 protein with RNA. Virology 110, 87-97. YOUNG, J. F., DESSELBERGER, U., PALESE, P., FERGUSON, B., SHATZMAN, A. R. & ROSENBERG, M. (1983). Efficient expression of influenza virus NSI nonstructural proteins in Escherichia coll. Proceedings of the National Academy of Sciences, U.S.A. 80, 61056109. ZAPP, M. L. & GREEN, M. R. (1989). Sequence-specific RNA binding by the HIV-I rev protein. Nature, London 342, 714-716.

(Received 4 July 1991 ; Accepted 5 September 1991)