CA2201302A1 - An attenuated vaccination and gene-transfer virus, a method to make the virus and a pharmaceutical composition comprising the virus - Google Patents

Abstract

RNA polymerase I transcription in vivo in transiently DNA-transfected cells has been used for expression of influenza vRNA molecules coding for chloramphenicol acetyltransferase (CAT) in anti-sense orientation. Influenza virus superinfection served to provide viral RNA polymerase and other proteins for transcriptional conversion of minus-strand vRNA into plus-strand viral mRNA molecules expressing CAT activity. This system has been used for an analysis via nucleotide exchanges as well as deletions and insertions of both terminal segments of the vRNA sequence which cooperatively constitute the vRNA promoter structure. Several mutants with greatly enhanced expression rates over wild-type levels have been constructed, which also can be packaged and serially passaged into progeny virus. The data obtained for the mutations in various promoter elements support a model of consecutive, doublestrand vRNA promoter structures in binding of viral polymerase and initiation of RNA synthesis. Preparations of attenuated influenza virus for vaccination purposes include a single recombinant segment with promoter up mutation(s) for overexpression of an own or foreign gene product, which at the same time because of its over-replication serves to decrease the number of helper virus RNP segments. The same viruses further have been passaged through a step of ribozyme cleavage acting at one of the helper viral segments, which will delete this vital function and structure with high rates from the virus progeny. The resulting attenuated viruses will interact with their target cells in only one round of abortive infection, and are unable to produce viral progeny.

Description

2 2 0 ~ ~ ~ 2 P~.~1~55/03663 AN ATTENUATED VACCINATION AND GENE-TRANSFER VIRUS, A
METEIOD TO MAKE T~E VIRUS AND A P~ARMACEUTICAL COMPO-5 SITION COMPRISING T~IE VIRUS

The object of the present invention was to make a v~crin~tion virus. Thisobjective has been fulfilled with the segmented virus constructed as described hereln.

The genome of infl~n7~ A viruses consists of 8 different single-stranded viral 10 RNA (vRNA) molecules of negative polarity, which have in common 5' and 3' terminal sequences largely complementary to each other. These conserved segments 13 and 12 nucleotides in length are known to form double-stranded RNA
p~nh~ntlle structures (Hsu et al., 1987; Fodor et al., 1993) which have been analysed in more detail recently in vitro using internally deleted model RNAs (Baudin et al., 1994; Tiley et al., 1994). In the virion the panhandle ends of all RNA segments are found in specific binding to viral RNA polymerase complexes, while the rrm~ining internal segments stay single-stranded with viral nucleoprotein (NP) in cooperative binding (Compans et al., 1972; Honda et al., 1988; Martin etal., 1992). Upon infection these viral RNPs initially serve as templates for the20 synthesis of viral mRNAs by a specific cap-~n~trhin~ merh~ni~m (Plotch et al., 1979; Braam et al., 1983), and later on will direct synthesis of full-length complementary RNAs (cRNAs), probably dependent on the absence or presence of newly synth~i7ed NP protein (Shapiro and Krug, 1988). The plus-strand cRNAs are then used as templates for progeny vRNA synthesis.

25 The viral RNA polymerase complex consisting of proteins PB1, PB2, and PA is involved in all three different modes of RNA synthesis during the viral replication cycle, following its specific binding to the terminal p~nh~n~le segments of bothvRNAs and cRNAs. Sequence comparison reveals that the vRNA and cRNA

W O 96110641 ~ 0 2 1 ~ 5~3663 termini have similar, but not identical sequences. For that reason vRNA and cRNA recognition may be ~ ting~ hed because of these structural alterations allowing for asymmetries in initiation of plus and minus strand RNA synthesis, and possibly in viral RNP ~ ging, which has also been suggested to be controlled by the panhandle RNA sequence (Hsu et al., 1987).

Recently, we reported on an in vivo system for the introduction of specific mutations into the genome of influenza viruses: viral cDNA has been inserted in antisense orientation between mouse rDNA promoter and terminator sequences.
This has been derived from in vifro transcription experiments based on nuclear extracts from Ehrlich ascites cells, which resulted in transcripts exactly resembling influenza vRNA. For a series of in vivo studies, the viral coding sequence was replaced by the coding sequence for chlor~mph~nicol-acetyltransferase (CAT), however, with both influenza terminal non-coding sequences being retained exactly on the resulting vRNA transcripts. After transfection of this recombinant DNA template into mouse cells followed by influenza virus infection, CAT
activity was ~etect~hle. Transfer of supernatants to different cells demonstrated that CAT-vRNAs transcribed in vivo by cellular RNA polymerase I were not only transcribed by viral RNA polymerase into plus-strand mRNA and tr~n.~l~t~d into CAT protein, but also were replicated and packaged into infectious progeny virusparticles (Zobel et al., 1993; Nellm~nn et al., 1994).

We have used this system for a stepwise introduction of single and multiple mutations into the conserved panhandle RNA sequence, thereby effectively converting the HA-vRNA promoter sequence into an HA-cRNA promoter sequence and vice versa. For these series of constructs CAT activities have been25 measured both in primarily transfected and infected B82 cells and, after pa~aging of B82 supernatants, in secondarily infected MDCK cells. From the results obtained we propose a model for the terminal RNA sequence as being recognized RNA polymerase in con~ec~ltive steps of different structure when used as a template for initiation of viral mRNA synthesis.

30 The present invention relates to a segmented~RNA virus which comprises one or more segments which have been genetically modified to show improved transcription, replication and/or expression rates.

WO96/10641 2~ 0 ~ PCr/EP95103663 The virus can be one where one or more modifications have been introduced in the noncoding region(s) and/or one or more modifications have been introduced inthe coding region(s). A possibility is that at least one modified segment is derived from an original one by sequence variation(s). It is also possible that at least one 5 modified segment is an artificial addition to the set of original or modified original segments. The virus can be one wherein the modified segment comprises a nucleotide sequence which codes for a protein or peptide which is foreign to theoriginal virus. Preferred is that the foreign protein or peptide constitutes an antigen or antigen-like sequence, a T-cell epitope or related sequence. In such a case it is 10 possible that the segment comprises repetitions of an antigen or epitope or other.
Such an antigen or epitope can be for example derived from HIV, Herpes-Virus, Human Papilloma virus, Rhinovirus, CMV or Hog Cholera Virus (HCV). The virus of the present invention may be a single stranded negative-strand RNA virus as for example one of the Orthomyxoviridae family, the Bunyaviridae family or of15 the Arenaviridae family. The most preferred virus is an influenza virus. The virus of the present invention can also be a double-stranded RNA virus, as for examplea reovirus, a rotavirus or an orbivirus. The viruses of the present invention can be used in gen-therapy.

The present invention also relates to the virus and use of the virus for the 20 preparation of pharmaceuticals.

Mutational analysis of vRNA 3' terminal sequence positions.
Influenza A viral RNA 5' and 3' ends have similar, but not identical sequences with nucleotide mi~m~tclles at positions 3, 5 and 8, and an additional unpaired nucleotide is located at position 10 in the 5' region. Nevertheless, both vRNA
25 termini hybridize into a double stranded panh~n~lle structure made up of twelve and thirteen nucleotides in common for all eight RNA segments, plus in average three additional basepairs specific for each of the vRNA molecules. Due to the deviations mentioned the cRNA or plus strand panhandle structures have to be different from the vRNA structures; however, both are recognized by viral RNA
30 polymerase and are used for initiation of RNA synthesis, i.e. they are constituting a promoter structure. Even if in initial recognition and binding of RNA
polymerase the double stranded RNA p~nh~n~1le structure is known to be the substrate, and is also observed in virion RNPs (Hsu et al., 19~7), for the initiation step of transcription at the 3' ultimate template position this terminal region has to , WO96/10641 ~2 ~ ~ 3 0 2 PCT/~7s~3663 be separated into a partially single stranded, i.e. 'forked' structure (Fodor et al., 1994). RNA polymerase may be predicted to continue its binding interaction with both, the rçm~inin~ double stranded segment: nucleotides 10 to 15 versus 11 to 161, and to the single stranded 3' template segment: nucleotides 1 up to 9, as well 5 as the 5' single stranded end (Tiley et al., 1994). Inkoduction of mutations at specific positions in either strand may hence alterate simultaneously both of these con~ecutive vRNA promoter structures: p~nh~n~le and fork in different ways, and will in addition also result in corresponding variations of the cRNA promoter structure.

10 To investigate the importance of the three mi~m~tc.l~ positions, specific single, double or triple nucleotide exchanges were first introduced into the vRNA 3' endsequence at positions 3, 5 and 8, thereby approa~hing a fully double-stranded vRNA promoter structure, in a step-wise manner. At the same time the vRNA 3' end template sequence will become equivalent to the cRNA 3' end in these 15 positions, but not in regard to the additional nucleotide at position 10. Single nucleotide exchanges according to this scheme (pHL1098, p~1099, pHL1100) abolished the promoter activity, and no CAT activity was observed, as has been reported before with a different method (Luo et al., 1993). Two of the double mutation constructs (pHL1101, pHL1103) also gave negative results.

In contrast, for pHL1102 (G3A, U8C)I a significant CAT activity was detected, distinctly higher than for the corresponding wild-type construct (pHL926;) whichin the conditions applied (8 hr after infection) resulted in rather low levels of CAT
expression. This activity increase is further enhanced for the final construct of this series carrying the triple exchange G3A, C5U and U8C (pHL1104), i. e. trans-fection of pHL1104 DNA followed by influenza virus infection resulted in a very high level of CAT expression, also considerably above the pHL1102 results.

These results have been repeated using various conditions of transfection and infection as well as determining kinetic data during the course of infection. While the pHL1104 variant is always observed far superior over any wild-type construct ~Notations concerning nucleotidesl to 15 refer to positions in the vRNA 3' end, e.g. position 2 designates the penultimate nucleotide; 5' end positions are given in ordinary numbers. The notation G3A describes a mutational change of guanosine to adenosine at position 3.

~ wo9Gno64l 22 ~ ~ 3 ~ ~ PCT/~ 51'~,366~

that ~ ;s~ion ratio may be variable and difficult to quantitate (between around 20 fold and nearly 100 fold). Rather short infective cycles of eight hours as used prevalently appear to put more slowly replicating, i.e. wild-t,vpe molecules at a disadvantage, in particular in F~q~ n~ of packaged pseudo-vRNA molecules-via virus progeny, both is found increased for wildtype and related constructs afterDNA transfection plus ~welve hours of infection (see Ne~lm~nn et al., 1994).
l?~m~inin~ deviations in CAT t;~lt;ssion ratios may be attributed to variations in growth conditions in individual experiments.

Mutational analysis of vRNA ~' terminal sequence positions We also addressed the question whether the unexpectedly high viral mRNA
es~ion rate of p~1104 is the consequence of a stabilized p~nh~ndle double-strand structure or may be directly attributed to the point mutations introduced into the vRNA 3' sequence, and active when being used as a single-stranded template segment, e.g. in the 'forked' structure.

For this purpose we constructed pHL1124, three complementary point mutations introduced at the S' end of the vRNA sequence again in positions 3, 5 and 8 (U3C, GSA, A8G). Together with a sequence wild-type vRNA 3' end these variations again result in a p~nh~nllle structure free of mism~tches and, therefore, pHL 1124 is equivalent in this regard to p~1104, but different in the sequence of its template and non-t~mrl~te single strands. No significant CAT t;~ s~ion was detected for pHL1124. We conclude that the increased CAT activity of pHL1104 is not a consequence of the stabilized panh~n~le structure itself, but at least in part is a consequence of the individual nucleotide exchanges at positions 3, S and 8 at the 3' end of the vRNA sequence, it is also more likely then to originate from other structural intermediates of initiation than a stabilized panhandle.

Mutational analyses of concerted exchanges at both ends of the vRNA
sequence In order to det~rmine in detail the influence of single, double and triple exchanges at the vRNA S' end upon C~AT ~ ession rates we also used the improved vRNA

3' end sequences of pHL1104 and pHL1102 as starting points rather than the corresponding wild-type sequence. From the series of experiments related to p~1104 and from the equivalent series related to pHL 1102 it can be concluded that ret~ining a G residue in position 5 is the most important single feature in -WO96/10641 2~ ~ ~ 3 0 2 P~ ;15S/03663 these S' end variations. A single exchange into an A residue at position S as inp~1185 will render the promoter entirely inactive, while single exchanges in positions 3 or 8, as well as a 3 plus 8 double exchange will retain promoter activity even if reduced from the level observed for pHL1104, but still above S wild-type expression rates. While the GSA nucleotide substitution opposite nucleotide CS in the 3' terminus results in losing one basepair (in the p~nh~n~le context) within the p~l104 series, a basepair is indeed gained by exactly the same GSA exchange within the p~llO2 series, i.e. opposite the US residue as present in the pE~1102 vRNA 3' end. Since again the GSA exchange results in 10 loss of promoter function inspite of gaining one basepair we conclude that the guanosine at position S may be important for RNA polymerase binding within the S' non-template single strand rather than being part of the panhandle double-stranded structure in this region. The importance of a G residue at this position has been shown earlier in a single-step mutational analysis (Li and Palese, 1992), 15 while non-t~mpl~te strand binding of RNA polymerase has been studied recentlyin vitro (Tiley et al., 1994). Different from the deleterious effect of an exchange at position S exchanges at positions 3 and in particular 8 are of minor importance.The series of S' nucleotide exchanges has also been repeated for the p~llO2 version of the vRNA 3' end yielding exactly the same pattern of results, albeit at 20 the somewhat reduced levels characteristic for p~l102. The only result in both series not quite in agreement with a uniquely important role for a G-residue in position 5 is the triple exchange of pHL1126 which retains low promoter activityinspite of an A residue in that position. Due to altogether six concerted exchanges in positions 3, S and 8 as well as 3 , S and 8 from the S' and 3' end of the 25 vRNA sequence the pHLl 126 vRNA panhandle structure is indeed nearly equivalent to a wild-type cRNA panh~n~lle, with the exception of an unpaired adenosine being present in position 10 of pHL1126 while an unpaired uridine at position 10 is part of the wild-type cRNA structure. This correlation may indicate a correct structure in pHL1126 for several other residues of (minor) importance 30 which, therefore, appa,c;llLly allows to compensate for the mis~ing G residue in position S, even if at a clearly reduced level of activity. - In the parallel pHL1102 series the corresponding triple exchange clone pHLl 125 does not show any promoter activity; however, because of its deviation at position S it does not completely resemble the cRNA panhandle structure.
--WO96110641 2~ PCT/~ 5~`~3663 Mutational analysis of the panhandle bulge structure around nucleotide 10 An extra, unpaired residue in position 10 at the 5' end is a specific feature of the influenza viral RNA p~nh~ntlle structure. It is causing or at least enforcing a major bulge of the structure, together with unpaired residues at position 9, and might be 5 part of a specific recognition element of that structure by viral RNA polymerase.
In order to investigate the importance of that particular structural feature, a further series of plasmid constructs has been initiated, again based on pEILl 104 and its 3' termin~l sequence as a reference. A perfectly matched RNA double-strand without any bulge has been achieved either by inserting an additional U residue in the 3' end sequence opposite A10 (p~1140) or by deleting the A10 residue from the 5' sequence (pHL1152). Finally, a bulge of opposite direction was created in the panhandle structure of p~ll64 with an extra U residue in position rO of the 3' end, and position 10 deleted from the 5' end sequence. While the latter two constructs proved inactive in the CAT assay, p~ll40 did show some promoter 15 activity, albeit at a reduced level. We conclude from this result that a bulge in this region may not be recognized directly by viral RNA polymerase but may serve as a flexible joint between two more rigid structural elements that are involved inrnmediate contact with viral polymerase. The necessary RNA bending may aiso, but less efficiently be achieved in an A U-basepaired structure like pHL1140, 20 while the other two structures would not permit such type of interaction with RNA
polymerase. This inl~ elalion has also been substantiated in a further series ofvariations in this region .

Serial r ~s;-ging of influenza virus carrying promoter mutants All previous experiments consisted of a first measurement of viral mRNA
- 25 synthesis in DNA-transfected and infected B82 cells, followed by a second measurement of viral mRNA synthesis in infected MDCK cells, after p~sagin~ of progeny virus cont~ining supernatants. CAT expression in infected cells upon viral p~c~ging requires paçk~ging of pseudo-viral vRNAs, in addition to new rounds of viral mRNA synthesis in those cells leading to CAT expression again. All viral 30 promoter mutants analysed and found active in transfected and helper-infected B82 cells also resulted in CAT t;x~.es~ion after transfer, and consistently in equivalent ratios of activity. PaçL ~in~ therefore, cannot be correlated with any specific element in the vRNA promoter structure so far, and does not appear to be a Iimiting factor in constructing influenza virus mutants in this system. While CAT
3S expression after p~SaginSg in general appeared to be increased over the levels WO96110641 ~ i 3 0 2 PCT/~;l9sl~3663 before p~.cs~gin~ this might have been simply the result of different cells being used for the first and second step of CAT analysis, with MDCK being superior to B82 cells in influenza mRNA synthesis and also in progeny yields. Therefore, several experiments of serial passage have been pelrolllled using pHL1104 derived S influenza supernatants and others, in MDCK cells. In these serial passages, always done using aliquots of supernatants harvested eight hours after infection for further transfer, a stepwise increase of CAT ~ ession is observed (Fig. 2). Appalelllly the superior performance of viral RNA promoters carrying sequence deviations according to pE~1104 is not only true for viral mRNA synthesis, but also for viral 10 RNA replication.

Therefore, mutant viral RNAs of this character become accumulated and effectively selected in further p~ ginf~, while p~ck~Ein~ may be a neutral eventin this regard, at least for the variants analysed here.

Serial p~s~in~ extended 15 During further pa.~s~in~ of supernatants the CAT containing influenza segmentcarrying the mutationally altered viral promotor sequences became accumulated ina stepwise manner in the population of progeny viruses. In order to demonstrate this effect on the level of individual viruses being transferred we isolated in three indepentent experiments 50 to 85 plaques each after a third round of passage on 20 MDCK cells. Each cell Iysate obtained for the individual plaques was assayed for CAT activity according to the standard protocol. While in two of the experimentsthe fraction of CAT positive plaques was in the range of 4 to 8% (I out of 50, 4out of 40 plaques) in one of these series this fraction amounted to 47% (19 of 40 plaques). Both of these results demonstrate a substantial increase over the initial 25 fraction of CAT-segment cont~ining virus, which may be calculated to be in the range of 10-5 or at most 104, and slight variations in the conditions of growth during three steps of transfer may precipitate to result in the observed differences of CAT positive plaques. While every CAT positive plaque demonstrates the amplification of nine (not eight) viral RNA segments present in the initially 30 infected cell, this may have resulted from a single virus carrying nine or more RNA segments or from coinfection by two defective viruses able to complement each other.

- WO 96/10641 ~ 3 0 2 Pcr/k~s,~3663 .

Neces~nly, ~ccllmul~tion of a pseudo-viral segment not contributiong to viral growth will, in further steps, become lethal to viral growth, even if a majority of virions may contain an average of eleven rather than eight RNA segments (Hsu et.al. 1987). Par~ ing of viral RNA-segments based on a general p~e~ing signal 5 identical for all eight segments and realized via a specific interaction chain: vRNA
p~nh~nrile structure - viral RNA polymerase - viral NP protein - viral Ml protein will reflect the pools of the various vRNA segments in infected cells, and therefore may be biased towards an RNA segment superior in replication and ovel-~pl~;sented in that pool. Biased replication and pacL-~ing will, however, lead }0 to ~cllm~ tion of lethal viral particles due to an imbalance between the eight (or nine) viral RNA segments. This prediction is borne out in continl~ing the viral passage of pHLl 104 derived influenza supernatants beyond step three as exemplified in Fig. 2. While CAT expression based on transcription of the pHL1104 derived pseudoviral RNA segments is increased further up to the fifth 15 passage the number of viable viruses reaches a maximum already after the second step of viral p~ ginP thus demonstrating the continuous accumulation of an over-replicated foreign segment, based on a superior panhandle sequence.

At a stage repres~nting the third or fourth passage as displayed in Fig. 2 a virus preparation obtained in this way can be regarded as the equivalent of an attenll~ted 20 viral strain. While the concentration of ~tt~nll~ted virus particles that can be achieved in this way may appear to be limited a stage equivalent to passage 4 inFig. 2 may be delayed upon coinfection with wildtype helper virus during first or second steps of transfer, and considerably increased concentrations of ~ttenl-~ted virus preparations might be achievable in this way.

25 p~l 104-mediated high-rate expression of foreign proteins can also be used (after two or more steps of amplification via serial p~s~ging on ~CK cells) for high rate synthesis of foreign proteins in embryonated chicken eggs, following a general method of pl~pal~Lion of viral stocks as used for influenza and other viruses, i. e. injection of virus suspensions into the yolk sac. Protein preparations 30 isolated from those embryonated and infected cells will be glykosylated and modified in other ways according to their origination from eukaryotic cells.

A second method of influenza virus ~ ml~tion has been achieved via cleavage of either one of the influenza viral RNAs, preferrably the M or NP gene (segments 7 WO 96/10641 2 ~ Q 2 P~ 5~J~3663 or 5), via ribozyme hydrolysis in a spe~ ed mode of action. The ribozyme RNAs which may be covalently inserted into the pSV2neo early mRNA, located between the neomycin resistance gene and the small t intron sequence origin7~tin~;
from SV40 viral DNA, or expressed from similar t;~ ssion c~et~e.~, are directed 5 against the 5' end sequence of segment 7 (or another of the influenza vRNA
segments).

During initiation of mRNA synthesis the 5' terminal sequence which is involved in formation of the p~nh~n~le structure is at first covered by viral RNA polymerasein association with that double stranded promoter region. It will, however, become 10 single-stranded and free of protein, since the polymerase molecule will starttranscription at the 3' end and move along the 3' template sequence while synthe7.i.~in~ a perfectly hybridizing 5' RNA d~ughter strand, superior in that regard to the parental panhandle 5' segment. Ribozyme RNAs which may be inhibited in their activity either by RNA substrates involved in double strand 15 formation or if RNA substrates are covered by protein, have been directed with a 3' complementary sequence towards that protein-free 5' sequence of the substratevRNA molecule for initiation of hybridization, which then will be extended across the entire complementary region of approximately 100 nucleotides i. e. well intothe vRNA sequence initially covered by NP protein.

20 A second feature of the ribozyme RNAs as applied for inactivation of influenza vRNA molecules is their double-unit hammerhead character, directed against not one, but two close GUY cleavage sites, e. g. GWl6 and GW36 in segment 7 or GUC30 and GUC48 in segment 5, which are also known to be invariable in sequence comparisons of influenza isolates.

25 Both features of anti-influenza ribozymes as pointed out contribute to a reduction of typically two logs (up to three logs) in production of viral progeny in template ribozyme DNA transfected cells as compared to infection of mock transfected cells, both at moi 1 and 20 h after Lipofect~min-DNA treatment. Ribozyme treatment can be applied after two or three rounds in MDCK cells of pHL1104-30 promoted amplification of a pseudo-viral RNA segment origin~ting from RNA-Polymerase I transcription, in the presence of helper virus as used in initial superinfection.

WO 96/10641 ~ r~,l/hr5~/03663 In a simple version ribozyme treatment as described above is employed as a selection technique. Here, its application is ~,pl~ iate if the pseudoviral RNA is indeed a (foreign) influenza segment carrying particular mutations but capable in principle to act as a functional substitute for the helper viral segment destroyed b~
5 ribozyme cleavage. For that purpose the substitute viral segment to be selected in that procedure has to be ml1t~ni7ed in advance at the two cleavage site~
indicated above in order to become resistant against ribozyme inLelrelellce. In another application ribozyme cleavage of helpervirus vRNA can be used for ?~ttpnl~tion of recombinant influenza virus prepal~ions. Here, the pseudo-viral 10 RNA segment may be designed in a way which renders it incapable to substitutefor a helpervirus gene. Therefore, viral p~ ing into ribozyme template DNA
transfected cells would lead to an abortive infection only, because of ribozyme mediated destruction of an important viral gene, if its gene product would not be added for complementation via expression from a cDNA construct which is alsc 15 DNA-transfected into the cell together with ribozyme-expressing DNA 20 h before viral infection. In this way viral progeny is obtained that is ~ n~ted because ot ribozyme cleavage of one of the vRNA segments, and effectively that segment is mi~ing in the virions because it can no longer be packaged. Viral pl~pal~lions obtained in this way are capable of only one round of infection because of their20 inherent Ml + M2 protein complementation, and therefore are suited for vaccination purposes. Animal infection with progeny virus as isolated after the ribozyme attenuation step results in abortive infection, but viral proteins synthesized in infected cells are able to induce B-cell and T-cell responses in such ~nlm~

2~ In influenza viral RNA synthesis parental negative-strand vRNA is copied intoplus-strand cRNA, which again is copied into progeny vRNA, from the first to thelast nucleotide. This amplification of viral RNAs, however, proceeds in an inherently asymmetric way, since vRNA molecules are synthesized in excess over cRNA molecules. This result is con~i~t~.nt with the idea that cRNA carries a 30 promoter structure more active in binding viral RNA polymerase and in initiation of RNA synthesis, i.e. 'stronger' than does vRNA. While at first simply the two 3' ends of single-stranded vRNA and cRNA templates have been implicated as promoter sequences, the detection of double-stranded panhandle structures involving both ends of the vRNA sequence in virions (Hsu et al., 1987) suggested35 more complicated substrates for RNA polymerase binding and initation of WO96/106~1 ~2 ~ ~ 3 0 2 ~ li~55J~3~63 ght~r-strand synthesis. A slightly different panhandle structure has also been observed with model vRNA molecules in the absence of viral proteins in vifro (Baudin et al., 1994), possibly calling for a structural change upon viral RNA
polymerase binding, i.e. a bulge may be shifted from position 4 to position 10 in 5 that reaction (see Fig. 1). While originally several of the RNA polymerase / vRNA
binding experiments in vi~ro appeared to show recognition only of 3' end oligonucleotides, this has since been shown to be an artifact after pure, recombinant viral polymerase free of residual RNA became available, instead of enzyme plt;p~lions from virions. Under these conditions RNA polymerase 10 binding to viral RNA as well as endonucleolytic cleavage of cellular mRNAs by subunit PB2 was observed to depend on vRNA 5' plus 3' terminal sequence binding, with even higher affinity for the 5' non-template segment (Hagen et al., 1994; Tiley et al., 1994).

Different from the employment of both vRNA and cRNA promoter structures in 15 replication physiologically only vRNA promoters will also serve in initiation of viral mRNA synthesis according to the cap-.cn~tçhing mech~ni~m (Plotch et al., 1979; Braam et al., 1983). While it has been claimed that cRNA promoters would not have the capacity to act according to this scheme (Tiley et al., 1994), the failure to observe viral antisense mRNA molecules may simply reflect the 20 inavailability of cRNA molecules early in infection, i.e. in the absence of surplus viral NP protein, and small amounts of such molecules might even have gone undetçcted In this invention we describe a mutagenizational analysis of the vRNApromoter structure in vivo which in approaching the structure of the cRNA
promoter via three nucleotide exchanges shows considerably improved activity in 25 viral mRNA synthesis over vRNA promoter wild-type levels. Co~tinl~ing increase of viral CAT mRNA ~pression during consecutive steps of viral p~s~ging suggests that the same vRNA promoter ~ also show increased activity in cRNA synthesis, both in accordance with the idea that the cRNA promoter structure might be 'stronger' than the vRNA promoter, also in initiation of viral 30 mRNA synthesis.

Additional variations of the 5' terminal sequence clearly indicate the major importance of a G residue in position 5, irrespective of complementarity or not to position 5 at the 3' end. The unique role of this G residue has been observed before in a serial mutagenizational analysis (Li and Palese, 1992). According to WO96110641 ~ 3 ~ 5~3663 both data guanosine residue 5 may be involved in single-strand binding of RNA
polymerase as has indeed been observed for the non-template strand terminal segment (Tiley et al., 1994). While p~nh~ndle double-strand structures are likely to constitute the initial RNA polymerase binding substrate a partial separation of template and non-template strands is expected to take place consecutively resulting in a 'forked structure' such as proposed by Fodor et al. (1994). Specific and tight binding of RNA polymerase in this structure may predominantly be oriented towards sequence elements in the non-template strand, since the growing point ofRNA synthesis will have to move along the entire template strand following its initiation. It is, therefore, possible that such a binding interaction survives most or all of an individual round of mRNA synthesis as has been proposed (Tiley et al.,1994).

The triple nucleotide exchanges as introduced in vRNA molecules derived from pHLI 104 templates will create three additional basepairs able to stabilize the p~nh~nclle structure in general, but more specifically they will favor a bulged ~t?no.~ine 10 over the bulged adenosine 4 conformation as observed for the wild-type sequence in vi~ro (Baudin et al., 1994). Since the changes introduced here lead to a considerable enhancement of promoter activity we propose that a bulged10 conformation may be the structure underlying the vRNA / polymerase binding reaction, which otherwise would have to be achieved only as a result of that interaction. A bulged 10 adenosine residue may constitute a kind of flexible joint or angular kink which in turn suggests two major, structurally stable binding sites to the left and right of this element. One of these sites has to be the double-stranded sequence element of (in average) six basepairs extending from positions11 to 16 and 10 to 15, respectively. While the distal three basepairs are known to be variable for the various R~A segments, basepair 13/12 has been shown to beexch~nge~ble experimentally, and also the number of basepairs has been reduced to four without complete loss of function (Luo et al., 1991). With all of these data it seems clear that the main recognition element in this region is an RNA double-strand of certain stability, while it remains possible that residue 12 guanosine and potentially others are also recognized individually within that structure. A major second binding element for RNA polymerase on the other side relative to position10 is less evident, but may be located in a distance of nearly one helical turn in the de-bulged region around position 4, since direct contacts are suggested by that initial conformational interaction, and also by the specific requirement of a WO96/10641 2~ Q ~ 3 0 2 1~,1i~5~/03663 guanosine residue in position 5, which is likely to interact not only during, but also before partial strand separation in that region, i.e. in the p~nh~n~le as well as the forked structure. While an extra adenosine residue in position 10 may be optimal for creating a correctly shaped bulge in this region of RNA, structural variants are possible in this regard (see pHL 1140) which excludes direct interactions between RNA polymerase and residues constituting that bulge.

In sllmm~ry we are proposing a model (see Fig. 1) of consecutive steps of interaction between a vRNA or cRNA promoter structure and viral RNA
polymerase:
bulged 4 p~nh~n~le ~ bulged 10 panh~nclle / polymerase ~ forked RNA /
polymerase (bound to 5G and ds element 11-16) ~ initiation of RNA synthesis (recognition of 3' end of template).

,~tteml~tion of influenza viruses for plt;pal~Lion of a live nasal vaccine relies on two mec.h~ni~m~: 1) preferential amplification of a recombinant viral segment carrying the pHL1104 promoter mutation, which will increase its rate in packagedviral RNP particles and indirectly decrease that of the eight helper virus RNP
particles. This competition results in an increase of defective viruses from which one or more of the regular viral gene segments are mis~ing 2) Sequence specific ribozyme cleavage of one or more helper virus RNA segments, if compensated through gene product expression for functional complementation. This dual interaction will result in virus progeny, which is capable of only one round of infection, abortive because of the mi~ing viral protein(s) that are required for their propagation. - Ribozyme cleavage of one out of two sister viral gene segments, sensitive and (artificially) insensitive for its hydrolysis may also be used (repeatedly) for selection purposes, including selection for viral gene constructs expressed via RNA polymerase I transcription.

MATERIALS AND MET}IODS

Plasmid constructions Plasmids with mutated vRNA and/or mutated cRNA promoter sequences are derivatives of pE~926 (Zobel et al., 1993; Nellm~nn et al., 1994). In pE~926 a hybrid CAT cDNA with flanking non-coding sequences derived from influenza vRNA segments has been precisely inserted in antisense orientation between WO96/10641 ~ 5SJ'~,3663 mouse rDNA promoter and terminator sequences. The CAT reporter gene in this way has been introduced by exactly replacing the coding sequence for h~rn~g~lutinin, ret~ining the untr~n~l~ted viral 5' and 3' sequences of segment 4.

vRNA 5' end mutations were created by PCR, using a general primer hybridizing to a position in the fl~nkin~ rDNA promoter sequence, and a specific primer carrying the desired nucleotide substitution to be introduced in the viral termin~l sequence. The polymerase chain reaction products were first digested by the restriction enzymes BglII and SpeI, inserted into the left boundary position by exch~n~in~ the segment between these appropriate restricton sites in pHL926, andfinally confirmed in their constitution by DNA Sanger sequencing.

Generation of vRNA 3' end mutations followed the same general scheme at the right boundary. PCR products were obtained by using a general primer complementary to a CAT gene internal sequence position, and a specific primer with appropriate nucleotide exchanges inserted into its sequence. Following digestion with restriction enzymes NcoI and ScaI, the PCR products were cloned into NcoI- and ScaI(partially)-digested plasmid pHL926. Any PCR derived sequences were investigated by DNA sequencing.

For constructs with both 5' end and 3' end mutations in combination, 5' variation containing fragments were obtained by BglII and SpeI restriction and inserted into the appropriate 3' terminal variation plasmids.

Cells and viruses Influenza A/FP~/Bratislava viruses were grown in NIH3T3 cells. For transfection and p~e.~ging experiments B82 cells (a mouse L cell line) and MDCK cells were used.

Lipofect~-nin DNA transfection and influenza virus helper infection For DNA transfection 107 B82 cells were used. 5 ~lg of plasmid DNA were mixed with 60 ~lg of Lipofect~min (Lipofe~ ninTM, GIBCO/BRL) in serum-free medium and incubated at room temperature for 10-15 min. This mixture was added to the cells washed twice with serum-free medium, and the incubation with 30 Lipofect~min/DNA was continued for l hr. After further incubation with DMEM
medium for 1 hr the transfected B82 cells were infected with influenza WO96/10641 ~ 3 0 2 P~ ~gSI~3663 A/FPV/Bratislava at a multiplicity of infection of 0.01 to 1 for another 30-60 min.
Further incubation was performed with DMEM medium.

P~cs~ of virus containing supernatants Under standard conditions 8 hr after influenza infection (at moi 0.1 to 1) cellswere harvested for CAT assays, and supernatants were collected and spun down at 1200 rpm for 5 min for removal of cell debris.

Aliquots of virus cont~ining cleared supernatants were used for plaque tests, and another aliquot was adsorbed to 107 MDCK cells for 30-60 min for further p~ gin~ Again 8 hr after infection the CPE was verified, and cells and supernatants were collected and treated as before.

CAT assay Cell extracts were prepared as desribed by Gorman et al. (1982). CAT assays weredone with [l4C]chloramphenicol or fluorescent-labeled chloramphenicol (borondipyrromethane difluoride fluorophore; FLASH CAT Kit, Stratagene) as substrates.

For [l4C]chloramphenicol the assay mixture contained: 0.1 ~lCi [l4C]chloramphenicol, 20 ~1 4 mM Acetyl-CoA, 25 1ll 1 M Tris-HCI (pH 7.5) and 50 1ll of cell Iysate in a total volume of 150 ~11. The assay mixture for the fluorescent-labeled substrate contained (in a final volume of 80 ,ul): 10 ~1 0.25 M
Tris-HCI (pH 7.5), 1 0 !11 4 M Acetyl-CoA, I 0 1ll fluorescent-labeled chloramphenicol, and 50 ~LI of cell Iysate. After an incubation time of 16 hr the reaction products were separated by chrom~tc)graphy and either autoradiographed or ViS~ i7~d by UV illumination and photography.

REFERENCES
2S Baudin,F., Bach,C., Cusack,S. and Ruigrok,R.W.H. (1994) The EMBO J., 13, 3158-3165.
Braam,J., Uhmanen,I. and Krug,R. (1983) Cell, 34, 609-618.
Compans,R.W., Content,J. and Duesberg,P.H. (1972) J. Virol., 10, 795-800.
Fodor,E., Seong,B.L. and Brownlee,G.G. (1993) J. Gen. Virol., 74, 1327-1333.
Fodor,E., Pritlove,D.C. and Brownlee,G.G. (1994) J. Virol., 68, 4092-4096.
Gorman,M., Moffat,L. and Howard,B. (1982) Mol. Cell Biol., 2, 1044-1057.

WO96/10641 a~ ~ ~ 3 0 ~ P~li~:r951f~3663 n,l\~, Chung,T.D.Y., Butcher,J.A. andKrystal,M. (1994).~. Virol., 68, lS09-1515.
Honda,A., Ueda,K., Nagata,K. and T~hih~nna,A. (1987) J. Bioc~em., 102, 1241-Hsu,M., Parvin,J.D., Gupta,S., Krystal,M. and Palese,P. (19g7) proG Nafl. Acad.
Sci. USA, 84, 8140-8144.
Li,X. and Palese,P. (1992).J. Virol., 66, 4331-4338.
Luo,G., Luytjes,W., F.n~mi,M and Palese,P. (1991) J. Virol., 65, 2861-2867.
Martin,J., Albo,C., Ortin,J., Melero,J.A. and Portela,A. (1992) ~. Gen. Virol., 73, 1855-18~9.
Ne~lm~nn,G., Zobel,A. and Hobom,G. (1994) Virology, 202, 477-479.
Plotch,S., Bouloy,M. and Krug,R.M. (1979) Proc. Natl. Acad. Sci. USA, 76, 1618-1622.
Seong,B.L. and Brownlee7G.G. (1992) J. Virol., 73, 3115-3124.
Shapiro,G. and Krug,R. (1988) J. Virol., 62, 2285-2290.
Tiley,L.S., Hagen,M., Matthews,J.T. and Krystal,M. (1994) J. Virol., 68, 5108-5116.
Y~m~n~k~,~., Ogasawara,N., Yoshikawa,H., Tehih~m~,~. and Nagata,K. (1991) Proc. Natl. Acad. Sci. USA, 88: 5369-5373.
Zobel,A., Nel-m~nn,G. and Hobom,G. (1993) Nucl. Acids Res., 21, 3607-3614.

LEGENDS TO FIGURES

Fig. 1. Proposed scheme of consecutive conformational steps occurring prior to initiation of viral mRNA synthesis in influenza vRNA, in wildtype and pHL1104 derived mutant sequences. Positions of triple mutation in pHL1104 vRNA are 25 indicated in bold and larger size letters.

(A) Free R~A panhandle structure, bulged at position 4 (wildtype vRNA; Baudin et al, 1994) or at position 10 (mutant vRNA). (B) Bulged 10 p~nh~ndle structuresafter binding of viral RNA polymerase; proposed protein binding positions markedby underlignments. (C) Forked structures of partial strand separation. (D) Initiation 30 of viral mRNA synthesis via hybridization of capped primer oligonucleotide.

Fig. 2. Serial p~.e~sging of p~ 1104-derived progeny viruses.

WO96/10641 ~ 0 1 3Q 2 P~ 9S~3663 107 B82 cells were transfected with 5~ of pHL 1104 DNA (in 60 }~g Lipofect~min) and infected two hours later with influenza A/FPV/Bratislava (m.o.i:l). 8 hr post-infection the cells were assayed for plaque forming units (dark column) and for CAT activity (hatched column) After sedimentation an other S aliquot of the supernatant was adsorbed to 107 MDCK cells. Further rounds of p~.qq~ging were done equivalently by harvesting the cells 8 hr after infection for assaying CAT activities, whereas an aliquot of the supernatant was always adsorbed to fresh MDCK cells. Numbers of serial passages are indicated at the bottom. Plaque forming units per ml refer to the left ordinate, CAT ~ ession 10 rates (relative to primary imfection levels) to the ordinate on the right.

WO 96/10641 ~ PCr/~;r75J~i3663 SEQUENCE LISTING
( 1 ) ~R~ ~T- INFORMATION:
(i) APPLICANT:
(A) NAME: Bayer AG
(B) STREET: Bayerwerk (C) CITY: Leverkusen ~ (E) COUNTRY: Deutschland (F) POSTAL CODE (ZIP): 51368 (G) TELEPHONE: (0)214-3061455 (H) TELEFAX: (0)214-303482 (ii) TITLE OF lNv~NllON: Vaccination virus, method of making it and pharmaceutical composition comprising that virus (iii) NUMBER OF SEQUENCES: 8 (iv) COMPUTER READABLE FORM:
(A) MEDIUM TYPE: Floppy disk _ (B) COMPUTER: IBM PC compatible (C) OPERATING SYSTEM: PC-DOStMS-DOS
(D) SOFTWARE: PatentIn Release #l.o, Version #1.30B (EPO) 20 (2) INFORMATION FOR SEQ ID NO: 1:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 5241 ~ase pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: double (D) TOPOLOGY: circular (ii) MOLECULE TYPE: DNA (genomic) (iii) HYPOTHETICAL: NO
(iv) ANTI-SENSE: NO
(vi) ORIGINAL SOURCE:
(A) ORGANISM: Influenza virus, RNA sequence (C) INDIVIDUAL ISOLATE: pHL926 (xi) SEQUENCE DESCRIPTION: SEQ ID NO: 1:

W 096/10641 ~ ~ ~ 1 3 0 2 PCT/~rSSJ~3663 ATTTTGCCTT C~-l~lllllG CTCACCCAGA AACGCTGGTG AAAGTAAAAG ATGCTGAAGA 540 CGTAGAAAAG ATCAAAGGAT CTTCTTGAGA TC~ l CTGCGCGTAA TCTGCTGCTT 1500 WO96/10641 ~ ~ 7 3 Q p~ ,5,~3663 25 TTACGTTGAC ACCATCGAAT GGTGCAAAAC CTTTCGCGGT ATGGCATGAT AGCGCCCGGA. 3360 W 096/10641 ~2 ~130 2 pCTi~75~3663 GACCTGGAGA TAGGTAGTAG AAACAAGGGT ~-ll"l''ll'AAAT ACTAGTACAT TACGCCCCGC 4260 lS CACAGACGGC ATGATGAACC TGAATCGCCA GCGGCATCAG CACCTTGTCG CCTTGCGTAT 4380 TT~Lllllll TTTTTCCCCC GATGCTGGAG GTCGACCAGA TGTCCGAAAG TGTCCCCCCC 5040 CCCCCCCCCC CCCGGCGCGG AACGGCGGGG CCACTCTGGA CT~lllllll 'l''l''l''l''l''l''l''L'l''l' 5100 ~ WO96110641 ~2 Q ~ 3 0 ~ PCT/EP9S/03663 'lllllllllG GGGATCCTCT AGAGTCGACC TGCAGCCCAA GCTAGCGGCC GCTAGCTTCT S160 (2) INFORMATION FOR SEQ ID NO: 2:
- 5 (i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 5241 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: double (D) TOPOLOGY: circular 0 (ii) MOLECULE TYPE: DNA (genomic) (iii~ HYPOTHETICAL: NO
(iv) ANTI-SENSE: NO
(vi) ORIGINAL SOURCE:
(A) ORGANISM: Influenza virus, RNA sequence (c) INDIVIDUAL ISOLATE: pHL1104 (xi) SEQUENCE DESCRIPTION: SEQ ID NO: 2:

= ATAAAACGAA AGGCTCAGTC GAAAGACTGG GCCTTTCGTT TTATCTGTTG TTTGTCGGTG 180 W 096/10641 22 ~ ~ ~ 0 2 P~ gS~3663 AGTAAGAGAA TTATGCAGTG CTGCCATAAC CATGAGTGAT AACACTGCGG CCAACTTACT a40 CGTAGAAAAG ATCAAAGGAT CTTCTTGAGA TC~11L1111 CTGCGCGTAA TCTGCTGCTT 1500 W V96/10641 ~ ~ ~ 1 3 Q ~ 1951'~3663 TTACATTAAT TGCGTTGCGC TCACTGCCCG CTTTCCAGTC GGGAAACCTG TCGTGCCAGC 26~0 ~1L1L1CTTT TCACCAGTGA GACGGGCAAC AGCTGATTGC CCTTCACCGC CTGGCCCTGA 2760 CTTACACATC CCAGCCCTGA AAAAGGGCAT CAAAATAAAC CACACCTATG GTGTATGCAT 3 ~ 00 TGCCGGTGTC TCTTATCAGA CCGTTTCCCG CGTGGTGAAC CAGGCCAGCC ACGTTTCTGC 34 ~ 0 WO 96/10641 2 ~ 2 P~ ,03663 GACCTGGAGA TAGGTAGTAG AAACAAGGGT ~11111AAAT ACTAGTACAT TACGCCCCGC 4260 ~ TTTTTCCCCC GATGCTGGAG GTCGACCAGA TGTCCGAAAG TGTCCCCCCC 5040 CCCCCCCCCC CCCGGCGCGG AACGGCGGGG CCACTCTGGA CT~L111111 ~L11L1111L~ 5100 1111111$1G GGGATCCTCT AGAGTCGACC TGCAGCCCAA GCTAGCGGCC GCTAGCTTCT 5l60 (2) INFORMATION FOR SEQ ID NO: 3:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 13 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: ~ingle (D) TOPOLOGY: linear (ii) MOLECULE TYPE: RNA (genomic) (iii) HYPOTHETICAL: NO

~ WO96110641 ~ ~ 1 3 Q 2 Pcr~P9S/03663 (iv) ANTI-SENSE: NO
(vi) ORIGINAL SOURCE:
(A) ORGANISM: Influenza virus, 3' RNA sequence (C) INDIVIDUAL ISOLATE: Wild Type vRNA Promotor Element (xi) SEQUENCE DESCRIPTION: SEQ ID NO: 3:

(2) INFORMATION FOR SEQ ID NO: 4:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 13 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: RNA (geno~ic) (iii) HYPOTHETICAL: NO
1~ (iv) ANTI-SENSE: NO
(vi) ORIGINAL SOURCE:
(A) ORGANISM: Influenza virus, vRNA 5' sequence (C) INDIVIDUAL ISOLATE: pHLl104 vRNA Promoter Element (xi) SEQUENCE DESCRIPTION: SEQ ID NO: 4:

(2) INFORMATION FOR SEQ ID NO: 5:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 14 base pairs (B) TYPE: nucleic acid 2~ (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: RNA (genomic) (iii) HYPOTHETICAL: NO
(iv) ANTI-SENSE: NO
(vi) ORIGINAL SOURCE:
(A) ORGANISM: In~luenza virus, cRNA 3' sequence (C) INDIVIDUAL ISOLATE: cRNA Promoter element =

WO96/10641 ~2 ~ 2 PCT/EP9S/03663 (xi) SEQUENCE DESCRIPTION: SEQ ID NO: 5:

(2) INFORMATION FOR SEQ ID NO: 6:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 6802 base pairs (B) TYPE; nucleic acid (C) STRANDEDNESS: double (D) TOPOLOGY: circular (ii) MOLECULE TYPE: DNA (genomic) (iii) HYPOTHETICAL: NO
(iv) ANTI-SENSE: NO
(vi) ORIGINAL SOURCE:
(A) ORGANISM: Influenza virus, RNA sequence (C) INDIVIDUAL ISOLATE: pHLll91 (xi) SEQUENCE DESCRIPTION: SEQ ID NO: 6:

-WO96110641 ~ 3 o ~ PCT/~5J~3663 TATTGTGGAT TCTTGATCGT ~~ CA AATGCATTTA CCGTCGCTTT AAATACGGAC goo GGGAAACTGA ATAGGGTAAT CGAGAAGACG AACGAGAAAT TCCATCA~AT CGAAAAGGAA 1320 CGCCCATTCT CCGCCCCATG GCTGACTAAT LlllllLATT TATGCAGAGG CCGAGGCCGC 2400 -W 096/10641 ~ S~3663 lS TCGCCTTCTT GACGAGTTCT TCTGAGCGGG ACTCTGGGGT TCGAAATGAC CGACCAAGCG 3360 AGCATCACAA ATTTCACAAA TAAAGCATTT TTTTCACTGC ATTCTAGTTG TG~lll~lCC 3600 ~ WO 96/10641 ~ 103663 AGCAGAGCGA GGTATGTAGG CGGTGCTACA GA~ll~lLGA AGTGGTGGCC TAACTACGGC 4500 AGAGTTGGTA GCTCTTGATC CGGCAAACAA ACCACCGCTG GTAGCGGTGG lllllll~ll 4620 0 TGCAAGCAGC AGATTACGCG CAGAAAAAAA GGATCTCAAG AAGATCCTTT GAT~-LL~ l 4680 ~5 CTCTCAAGGA TCTTACCGCT GTTGAGATCC AGTTCGATGT AACCCACTCG TGCACCCAAC 5580 W 096/10641 ~2 ~13~ 2 P~ 5~'~3663 GGTTTGACTC ACGGGGATTT CCAAGTCTCC ACCCCATTGA CGTCAATGGG A~~ ll 6540 ( 2) INFORMATION FOR SEQ ID NO: 7:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 5825 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: double (D) TOPOLOGY: circular (ii) MOLECULE TYPE: DNA (genomic) (iii) HYPOTHETICAL: NO
(iv) ANTI-SENSE: NO
(vi) ORIGINAL SOURCE:
(A) ORGANISM: Influenza virus, RNA sequence =

WO96/10~41 ~ 2 Q 1 3 ~ ~ PCT~;r~S~'~,3663 .

(C~ INDIVIDUAL ISOLATL: pHL1489 (Xi) SEQUENCE DESCRIPTION: SEQ ID NO: 7:

AATCGTTTTC CGGGACGCCG GcTGGATGAT CCTCCAGCGC GGGGATCTCA TGCTGGAGTT 240 ATGAGGCTAC TGCTGACTCT CAACATTCTA CTCCTCCAAA AAA~A~ AAGGTAGAAG 720 ACCCCAAGGA CTTTCCTTCA GAATTGCTAA ~llllllGAG TCATGCTGTG TTTAGTAATA 780 AATTGTGTAC CTTTAGCTTT TTAATTTGTA AAGGGGTTAA T~AGGAATAT TTGATGTATA 1020 WO 96/10641 2 ~ Q 2 P~ r5SI'~,3663 W 096/10641 2 ~ ~ 7 ~ ~ ~ r~ll~r~S/03663 GTAGAAAAGA TCAAAGGATC TTCTTGAGAT C~llllllC TGCGCGTAAT CTGCTGCTTG 3300 CAAACAAAAA AACCACCGCT ACCAGCGGTG ~lll~lllGC CGGATCAAGA GCTACCAACT 3360 ~ lllCCGA AGGTAACTGG CTTCAGCAGA GCGCAGATAC CAAATACTGT CCTTCTAGTG 3420 TAA~ TATTTATGCA GAGGCCGAGG CCGCCTCGGC CTCTGAGCTA TTCCAGAAGT 4620 AGTGAGGAGG ~llllllGGA GGCCTAGGCT TTTGCAAAAA GCTTCACGCT GCCGCAAGCA 4680 WO 96/10641 2 ~ 0 2 PCTIEP9S/03663 AAAGAGAAAG CAGGTAGCTT GCAGTGGGCT TACATGGCGA TAGCTAGACT GGGCGGTTTT 4860 .

CGGCTGCTCT GATGCCGCCG TGTTCCGGCT GTCAGCGCAG GGGCGCCCGG TT~LlLLL~I 5160 (2) INFORMATION FOR SEQ ID NO: 8:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 4023 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: double (D) TOPOLOGY: circular (ii) MOLECULE TYPE: DNA (genomic) (iii) HYPOTHETICAL: NO
(iv) ANTI-SENSE: NO

- ~ WO96/1~ Q ~ PCT/EP95/03663 (vi) ORIGINAL SOURCE: .
(A) O~GANISM: Influenza virus, RNA sequence ~C) INDIVIDUAL ISOLATE: pHL1490 (xi) SEQUENCE DESCRIPTION: SEQ ID NO: 8:

TGACTAAGGG GATTTTAGGA TTTGTGTTCA CGCTCACCGT GCCCAGTGAG CGAGGACTGC 2~0 ~lllllllCA AATGCATTTA CCGTCGCTTT AAATACGGAC TGAAAGGAGG GCCTTCTACG 900 W 096/10641 ~ ~ ~130 2 ~CT/~55~3663 CTCTTGATCC GGCAAACAAA CCACCGCTGG TAGCGGTGGT ~ ll GCAAGCAGCA 2220 WO96/10641 ~ 5 ~ 9SJ~3663 S AAAGGGAATA AGGGCGACAC GGA~ATGTTG AATACTCATA CTCTTCCTTT TTCAATATTA 3300 0 AACAACAGAT A~AACGAAAG GCCCAGTCTT TCGACTGAGC CTTTCGTTTT ATTTGATGCC 3600

Claims (10)

CLAIMS:

1. A segmented RNA virus characterized in that it comprises one or more segments which have been genetically modified to show improved transcription, replication and/or expression rates and which may contain non viral genes to be transferred to the host.

2. The virus of claim 1, wherein one or more modifications have been introduced in the noncoding region(s) and/or in the coding region(s).

3. The virus of any of claims 1 or 2, wherein at least one modified segment is derived from an original one by sequence variation(s).

4. The virus of any of claims 1 or 2, wherein at least one modified segment is an artificial addition to the set of original segments.

5. The virus of claim 4, wherein the modified segment comprises a nucleotide sequence which codes for a protein or peptide which is foreign to the original virus.

6. The virus of claim 5, wherein the foreign protein or peptide constitutes an antigen or antigen-like sequence, a T-cell epitope or related sequence.

7. The virus of any of claims 5 or 6, wherein the segment comprises repetitions of an antigen or epitope or other peptide or protein.

8. The virus of any of claims 7 to 9, wherein the antigen or epitope is derived from HIV, Herpes-Virus, Rhinovirus, CMV, papilloma viruses, Hepatitis viruses and other human viruses or Hog Cholera Virus.

9. A pharmaceutical preparation of a vaccine comprising the virus of any of claims 1 to 8.

10. Use of the virus of any of claims 1 to 8 for the preparation of pharma-ceuticals.