WO1987000864A1 - Insertion into animals of genes coding for interferon-induced proteins - Google Patents

Abstract

Interferon-induced proteins having antiviral effects in animals; DNA sequences that code for such proteins and recombinant DNA molecules and transformed hosts that express such proteins; a method of producing an interferon-induced protein in an animal, in the absence of interferon induction or where there is insufficient expression of interferon-induced protein, comprising inserting into an animal a gene coding for such protein; and a method of protecting an animal from viral infection (e.g., influenza) comprising administering to the animal an antiviral-effective amount of an antiviral protein (e.g., Mx protein). The methods may be used to protect an animal against viral infection (e.g., influenza) by inserting a gene coding for an antiviral protein (e.g., Mx protein) into an animal that is susceptible to such infection or by administering a pharmaceutical composition comprising an antiviral-effective amount of such protein to an animal that is susceptible to such infection.

Description

INSERTION INTO ANIMALS OF GENES CODING FOR INTERFERON-INDUCED PROTEINS

The present invention relates to the insertion into an animal of a gene coding for an interferon-induced protein. The present invention also relates to a method of protecting an animal against viral infection comprising inserting into an animal a gene coding for an interferon-induced protein that is capable of protecting said animal from said viral infection. In a preferred, embodiment, the present invention relates to a method of protecting animals, e.g., mammals or birds, against influenza virus.

Orthomyxo virus infections may cause infection in various mammalian species, including humans (influenza), horses (equine influenza and laryngeal influenza) and swine (swine influenza), and in various avian species (Newcastle disease and fowl plague), including turkeys, ducks and chickens (The Influenza Viruses And Influenza, Ed. Kilbourne, Academic Press, New York (1975)). Infection often results in death and/or serious economic loss.

Some 25 years ago it was discovered that the inbred laboratory mouse strain A2G was resistant to doses of influenza virus that were lethal to all other inbred strains tested (Lindenmann, Virology, 16, 203 (1962)), and that this trait was inherited by a single dominant allele named Mx (Lindenmann, Proc. Soc. Exp. Biol. Med., 116, 203, (1964)). Since then, only one other resistant laboratory strain, SL/NiA, was discovered. In vivo and in vitro studies revealed that this resistance to influenza virus was mediated by interferon (either alpha or beta, but not gamma (Haller et al., J. Exp. Med., 149, 601-12 (1979); Haller et al., Nature, 283, 660-62 (1980); Arnheiter & Staeheli (Arch. Vir., 76, 127 (1983); Staeheli et al., Virology, 132, 456-61 (1984)), and that, after stimulation with type I IFNs, Mx⁺ but not Mx- cells produced a protein, designated Mx,

(Horisberger, Staeheli & Haller, Proc. Natl. Acad. Sci. USA, 80, 1910-14 (1983)). As there was no difference in susceptibility of Mx⁺ and Mx- IFN- treated mouse cells to other viruses (Haller et al., Nature, 283, 660-62 (1980)), the effect of the Mx gene is myxovirus-specific. Type I IFNs induce or increase the intracellular level of many proteins in addition to Mx; and until now, it was not possible to determine whether the resistance to influenza virus was due to Mx expression by itself, or whether Mx was required to complement other IFN-induced proteins. Moreover, because of the high frequency of the Mx- gene in both laboratory and wild mice, it was not clear whether Mx⁺ represented the wild or the mutant phenotype.

We have isolated and determined the structure of murine Mx cDNA; we have shown that expression of Mx in mouse cells suffices to impart resistance to influenza virus even in the absence of IFN (interferon) treatment; and we have determined that the

Mx- phenotype results from deletions in the Mx gene.

We have found that rat and human cells treated with IFN-α synthesized proteins that crossreacted with anti-Mx antibodies, indicating that these proteins had structural homologies to murine Mx protein. We have found that human cells synthesize an 80,000 Dalton protein in response to IFN-α, but not IFN-γ, induction and that this protein shares at least one antigenic determinant with murine Mx which is not found on other human or mouse proteins

(Staeheli, P. et al., Molec. Cell. Biol., 5, 2150-53 (1985)). We also found that murine MxcDNA hybridizes to genomic DNA of many mammals. As used herein, and unless indicated otherwise, the term Mx protein shall be understood to include polypeptides that cross-react with anti-Mx antibodies or that are coded for by DNA that hybridizes to murine MxcDNA.

We have found that the resistance of an animal cell that has a defective Mx gene and is susceptible to influenza may be substantially enhanced by inserting an Mx gene into said cell. Although cells that contain an Mx gene as part of their natural complement of genes develop resistance to influenza, they do so only after treatment with interferon. We have found that generating Mx protein within an animal cell by inserting an Mx gene which is permanently expressed into the cell makes the cell permanently resistant to influenza virus. Such a cell, and, in particular, such a cell in an animal, may be more resistant to influenza virus than a cell that has such a gene as part of its natural complement of genes, because the latter must first produce interferon before the antiviral state is established. Similarly, insertion of a gene for an interferoninduced protein, other than Mx protein, into an animal cell may cause that interferon-induced protein to be produced even if such cell has not been stimulated by interferon.

In addition, interferon-induced proteins may provide other beneficial effects. It has been shown, for example, that interferon induces expression of HLA class I proteins and β-microglobulin at the cell surface, and that cancer cells bearing these proteins at the surface are more easily destroyed by the immune system. (Hui et al., Nature, 311, 750-52 (1984); Tanaka et al., Science, 228, 26-30 (1985), Wallich et al., Nature, 315, 301-05 (1985)). Thus, if a tumor cell carrying NLA class I and microglobulin at its surface were to arise in an organism, it would be rapidly destroyed and not give rise to cancer. Our discovery thus provides a method for protecting an animal, including a human, against influenza or other orthomyxoviral infection by providing that animal with a gene that codes for an antiviral interferon-induced protein. This may be done by inserting the appropriate genes into cells of germ line tissues (preferably, into fertilized oocytes or very early embryos) and then facilitating the development of such cells (preferably, fertilized oocytes or very early embryos) into animals. When a gene is inserted into an animal in this fashion rather than into cells of the potential site of infection in that animal, offspring of such an animal can be bred that retain the desired gene permanently and will be resistant to viral infection. Similarly, animals (e.g., mammals, including humans, and birds and fish) may be provided with genes coding for interferon-induced proteins that have beneficial properties other than antiviral properties against orthomyxo virus, e.g., anti-tumor properties or antiviral properties against viral infections such as picorna viral infections, such as foot-and-mouth disease virus, or paramyxovirus infections such as canine distemper or rinderpest (affecting cattle). When a gene is inserted into a cell of germ line tissue, such insertion may be performed with cells that are in an animal or with cells that have been removed from an animal. For example, ova may be flushed from the oviducts of an animal, fertilized in vitro, microinjected with the desired gene, and then surgically returned to the animal or transferred to other animals. It is also possible to microinject embryos, e.g., one-cell or two-cell embryos. These techniques are more fully described in Hammer et al., Nature, 315, 680-83 (1985).

The present invention relates to a method of producing an interferon-induced protein in an animal in the absence of natural interferon induction or where there is insufficient expression of interferon-induced protein, comprising inserting into an animal a gene coding for such protein. In a preferred embodiment of our invention, a recombinant DNA molecule comprising a gene coding for an interferon-induced protein is inserted into an animal cell, such as a one cell embryo, and the development of such cells into animals is facilitated. Animal cells into which such recombinant DNA molecules have been inserted may also be cultured for the purpose of preparing commercially useful amounts of desired proteins.

In a preferred embodiment, the present invention relates to a process of protecting an animal against viral infection, said process comprising inserting into an animal susceptible to such infection a gene coding for an interferon-induced protein that is capable of protecting said animal from said viral infection. Preferably, a recombinant DNA molecule comprising a gene coding for an interferon-induced antiviral protein is inserted into an animal cell, such as a one cell embryo, and the development of that cell into an animal is facilitated. In a more preferred embodiment, the present invention relates to a method of protecting an animal against infection by influenza virus comprising inserting into an animal susceptible to such infection a gene coding for Mx protein. Preferably, the Mx protein will be that Mx protein that is normally found in the species of animal to be protected. Thus, for example, a gene coding for swine Mx protein would be inserted into a swine. In a particularly preferred embodiment, a recombinant DNA molecule comprising a gene coding for Mx protein (e.g., swine Mx protein) is inserted into an animal cell (e.g., a swine cell), such as a one cell embryo, and the development of that cell into an animal is facilitated.

Preferably, the expression of the Mx gene is under the control of a constitutive promoter, a promoter that is active in the tissue most susceptible to viral infection, for example, the mucosa of the respiratory tract or the intestinal tract, or a promoter that can conveniently be activated by exogenous agents (including interferon). The method of the present invention is useful when an animal does not have the interferon-induced gene or there is insufficient expression of interferon-induced protein. The phrase "insufficient expression of interferon-induced protein," as used herein, means that it is desirable, for one or more reasons, that the expression of an interferon-induced protein be increased and/or made continuous. This may be the case in the following situations:

1) An animal has the interferon-induced gene but the gene only expresses the desired protein upon interferon induction and it is advantageous for expression of the desired protein to be continuous; or

2 ) An animal has the desired interferon induced gene but adding one or more additional such genes to its cells is advantageous. The present invention also relates to an animal cell transformed with a recombinant DNA molecule comprising a gene coding for an interferoninduced protein as well as to an animal comprising such a cell. In a preferred embodiment, the present invention relates to an animal cell, which in its natural state is susceptible to a viral infection, transformed with a recombinant DNA molecule comprising a gene coding for an interferon-induced protein which is capable of protecting said cell against said viral infection and also relates to an animal comprising such a cell. Most preferably, the viral infection is influenza and the gene is a gene coding for Mx protein. Alternatively, a gene coding for a polypeptide that has a similar antiviral effect to that of

Mx protein (for example, a polypeptide that is a fragment of Mx protein or a derivative of Mx protein that has been modified to be more stable, or more effective, or to have a higher therapeutic index in vivo) may be substituted for the gene coding for Mx protein. Thus, as used herein, and unless stated otherwise, the words "interferon-induced protein" should be understood to include not only naturally occurring proteins, such as Mx protein, but also other polypeptides that have biological effects that are similar to those of an interferon-induced protein.

Generally, the methods of the present invention may be applied to animal cells that in their natural state do not have the desired gene for an interferon-induced protein, or do not have it in a constitutively active form, or do not produce the protein in sufficient amounts after IFN stimulation to provide antiviral protection. However, the method of the present invention may also be applied to add one or more additional genes coding for an interferoninduced protein to an animal cell that already has one or more of such genes, such that the interferon induced protein is produced under natural IFN control, constitutively, or at will, in sufficient amounts to produce the desired effect (e.g., resistance to viral infection) in all or some tissues. The present invention also relates to the polypeptide

M D S V N N L C R H Y E E K V R P C I D L I D T L R A L G V E Q D L A L P A I A V I G D Q S S G K S S V L E A L S G V A L P R G S G I V T R C P L V L K L R K L K E G E E W R G K V S Y D D I E V E L S D P S E V E E A I N K G Q N F I A G V G L G I S D K L I S L D V S S P N V P D L T L I D L P G I T R V A V G N Q P A D I G R Q I K R L I K T Y I Q K Q E T I N L V V V P S N V D I A T T E A L S M A Q E V D P E G D R T I G V L T K P D L V D R G A E G K V L D V M R N L VY P L K K G Y M I V K C R G Q Q D I Q E Q L S L T E A F Q K E Q V F F K D H S Y F S I L L E D G K A T VP C L A E R L T E E L T S H I C K S L P L L E D Q I N S S H Q S A S E E L Q K Y G A D I P E D D R T R M S F L V N K I S A F N R N I M N L I Q A Q E T V S E G D S R L F T K L R N E F L A W D D H I E E Y F K K D S P E V Q S K M K E F E N Q Y R G R E L P G F V D Y K A F E S I I K K R V K A L E E S A V N M L R R V T K M V Q T A F V K I L S N D F G D F L N L C C T AK S K I K E I R L N Q E K E A E N L I R L H F Q M E Q I V Y C Q D Q V Y K E T L K T I R E K E A E K E K T K A L I N P A T F Q N N S Q F P Q K G L T T T E M T Q H L K A Y Y Q E C R R N I G R Q I P L I I Q Y F I L K T F G E E I E K M M L Q L L Q D T S K C S W F L E E Q S D T R E K K K F L K R R L L R L D E A R Q K L A K F S D wherein the amino acids represented by the single letter amino acid codes are defined as follows: A = Ala C = Cys D = Asp E = Glu

F = Phe G = Gly H = His I = He K = Lys L = Leu M = Met N = Asn P = Pro Q = Gin R = Arg S = Ser T = Thr V = Val

W = Trp Y = Tyr The present invention also relates to polypeptides that are homologs of murine Mx protein that are synthesized by cells of other animals, and more particularly, to the IFN-inducible 80,000 Dalton human Mx polypeptide that is immunoprecipitable with monoclonal antibody 2C12 (Staeheli P. et al., J. Biol. Chem., 260, 1821-25 (1985)) directed against murine Mx protein. The present invention also relates to pharmaceutical compositions comprising the aforementioned polypeptides and to the use of the polypeptides in treatment (therapeutic or prophylactic) of viral infections in animals (e.g., mammals, including humans, and birds and fish). It will be understood that derivatives and fragments of the aforementioned polypeptides may be prepared that have similar utility in the treatment of viral infections. Such derivatives and fragments are considered to be within the scope of the present invention and a reference to the aforementioned polypeptides (such as, for example, in the discussion of methods of treatment and formulations set forth below), unless otherwise indicated, shall be understood to include such derivatives and fragments.

The present invention also relates to recombinant DNA molecules that are useful in preparing the aforementioned polypeptides. Preferred recombinant DNA molecules are characterized by a DNA sequence selected from the group consisting of

b) DNA sequences that hybridize to the aforementioned DNA sequence and that code on expression for the polypeptide whose amino acid sequence is set forth above or other Mx protein (including human Mx protein); and c) DNA sequences which are degenerate as a result of the genetic code to the aforementioned DNA sequences and that code for the polypeptide whose amino acid sequence is set forth above or other Mx protein (including human Mx protein).

Brief Description of The Drawings

Figure 1 shows an autoradiograph of immunoprecipitated translation products of mRNA from gel fractions 10 to 19, unfractionated mRNA from Mx⁺BALB .A2G-mx cells(+) and unfractionated mRNA from Mx-BALB/c cells(-) analyzed by electrophoresis (see Example 1 for details). Figure 2A depicts a restriction map of the pMx34 cDNA insert.

Figure 2B depicts the nucleotide sequence and the corresponding amino acid sequence of the pMx34 cDNA insert. The numbering of the nucleotides starts at the first nucleotide following the string of Gs.

Figure 2C depicts the amino acid sequence of murine Mx protein. The numbering of the amino acids starts at the first methionine residue. Figure 2D shows the vector pHG 327 into which the Mx-DNA was cloned to yield pMx34. Cloning was into the Sstl site, in an orientation such that the cDNA was transcribed from the SV40 early promoter. Figure 3 shows immunofluorescence with Mx-specific antibodies of G418-resistant NIH 3T3 cells transformed with pSV2-neo (a) or pSV2-neo and pMx34 plasmid DNA (b). Only pMx34-transformed cells contain Mx protein in their nuclei.

Figure 4 shows immunofluorescence with specific antibodies of transfected NIH 3T3 cells infected with influenza virus (a, b) or VSV (c, d). We analyzed the cells for synthesis of recombinant Mx protein (a, c) and for synthesis of influenza virus proteins (b) or VSV G protein (d). The Figure shows that cells which have Mx protein in their nuclei do not show signs of viral replication, while cells with little or no Mx protein in their nuclei have viral proteins in their cytoplasm.

Figure 5 shows Mx-specific transcripts in RNA from IFN-treated Mx⁺ and Mx- cells. We hybridized Northern blots of polysome-bound mRNAs (3 ug) to the 1.0 kb Bam HI restriction fragment of the pMx34 insert. We prepared mRNAs from BALB.A2G-MX (Mx⁺ ) and BALB/c (Mx-) embryo cells which we treated either with serum free medium (C) or with 300 U/ml of mouse IFN-alpha/beta (IFN).

Figure 6 shows Southern transfer analysis of murine genomic sequences related to MxcDNA. We digested mouse liver DNA samples (10 μg) with restriction endocucleases, electrophoresed through 0.8% agarose gels, transferred to a nitrocellulose membrane and hybri .di.zed with (³²P)-radιolabeled fragment of MxcDNA. We hybridized DNA from BALB.A2G- mx (+) or BALB/c (-) mice (A), to a fragment corresponding to nucleotides 658 to 2317 of MxcDNA, or to the 1.0 kb BAM HI fragment of MxcDNA (B). We compared Eco Rl-cleaved DNAs (C) and Hind Ill-cleaved DNAs (D) from different mouse strains by hybridization to the same fragment of MxcDNA corresponding to nucleotides 658 to 2317. Genes coding for interferon-induced proteins other than Mx protein may also be introduced into animal cells, using the methods described below. Such other proteins include 2'-5' oligoadenylate synthetase (Baglioni, Cell, 17, 255 (1979); Baglioni et al., Biochemistry, 18, 1765-70 (1979)), protein kinase (Jarvis et al., Cell, 14, 879-87 (1978)), guanosine binding proteins (Cheng et al., J. Biol. Chem., 258, 7746-50 (1983)), HLA Class I (Basham et al., Proc. Natl. Acad. Sci., 79, 3265-69 (1982); Yoshie et al., J. Biol. Chem., 257, 13169-72 (1982)), β2-microglobulin (Rosa et al., EMBO J., 2, 239 (1983)) and other proteins isolated by Weil et al. (Nature, 301, 437-39 (1983), and Antiviral Res., 3, 303-14 (1983)).

A gene coding for a desired interferoninduced protein will generally be introduced into a cell as part of a recombinant DNA molecule in which the gene is operatively linked to an expression control sequence. Methods of introducing recombinant DNA molecules comprising genes coding for an interferon-induced protein into a host cell include: (a) direct microinjection (see Hammer et al., Nature, 315, 680-83 (1985).

(b) use of viral vectors (see Munyon et al., J. Virol., 7, 813-20 (1971));

(c) cell-cell fusion, i.e. the fusion to cells of a limited number of chromosomes enveloped in nuclear membranes (see Fournier et al., Proc. Natl. Acad. Sci., 74, 319-23 (1977));

(d) cellular endocytosis of microprecipitates of calcium-DNA complex (see Bachetti and

Graham, Proc. Natl. Acad. Sci., 74, 1590-94 (1977); Maitland and McDougall, Cell, 11, 233-41 (1977); Pellicer et al., Cell, 14, 133-41 (1978); Wigler et al., Cell, 14, 725-31 (1978); and Axel et al., PCT Patent Application Publication No. WO 81/02426, published September 3, 1981);

(e) minicell fusion;

(f) Electroporation (see Neumann et al., EMBO J., 1, 841 (1982)); (g) fusion with liposomes containing DNA (see Froley & Papahadgopoulos, Current Topics in Microbiol. and Immunol., 96, 171-91 (1982)); (h) fusion with bacterial protoplasts containing plasmid DNA (see Schoffner, Proc. Natl. Acad. Sci., 77, 2163-67 (1980); Rassoulzadegan et al., Nature, 295, 257-59 (1982)). Use of any one of the aforementioned techniques will depend upon various factors, such as efficiency of information insertion, selectivity as to the particular nature or information of the DNA, permissible size of the DNA fragment, and the like.

When employing microprecipitates of calcium-DNA complex, a plurality of unrelated genes, including a gene having a selective marker, may be employed and mixtures of DNA which are not covalently linked may be introduced by congression, that is, different fragments of DNA will frequently concurrently enter a susceptible cell. Accordingly, those cells which have the selective marker are also likely to have the genetic capability of the gene that one desires to introduce, e.g., a gene for Mx protein. It is contemplated that in practicing our invention the desired genes will be introduced into one-cell embryos (see, for example, Hammer et al., cited above, describing production of transgenic rabbits, sheep and pigs by microinjection) and the development of embryos into animals will be facilitated (preferably by transfer into other females

(foster mothers)). However, it is also possible to introduce the desired genes into the cells of other tissues, e.g., bone marrow cells, liver cells and the cells of the intestinal mucosa (see U.S. Patent 4,497,796). It is also possible to introduce the desired genes into the cells of the respiratory tract. The usual purpose of introducing genes for interferon-induced proteins into an animal will be to provide the animal with a desired protein, e.g., a protein that has a protective effect or some other beneficial effect. However, the method of the present invention is also applicable to the commercial production of interferon-induced proteins using an appropriate host, for example, animal cells. For the latter purpose, one skilled in the art may select from known methods of transforming an appropriate host with an expression vector having the desired DNA sequence operatively-linked to the expression control sequence of the vector, culturing the host under appropriate conditions of growth and collecting the desired polypeptide from the culture. Preferably, the host cells will be allowed to reach stationary phase before the desired polypeptide is collected.

The polypeptide of the present invention is useful when it is expressed within a cell in protecting that cell against viral infection. However, the polypeptide or another interferon-induced protein may also be applied (e.g., in humans):

(a) as a spray to the respiratory tract (e.g., to the nasal passages and/or to the lungs),

(b) by injection, or

(c) topically, for prophylactic and/or therapeutic use against influenza and/or other viruses against which it may be effective and/or to prevent the propagation of such viral diseases. For such uses, it will be formulated with conventional pharmaceutically acceptable carriers, such as sterile water (for injection or intranasal use) or dermatologically acceptable creams and ointments. Preferably, the polypeptide or other interferon-induced protein will be lyophilized (preferably, using a bulking agent such as glycine) and then reconstituted with a carrier, such as sterile water, before use. The polypeptide of the present invention or other interferon-induced protein may be administered in a single daily dose or in divided doses (e.g., four times per day). The exact dose will be determined by the prescribing physician or other clinician and will depend upon the species of patient, on the age and weight of the patient, on the severity of the patient's condition, on the response of the patient to the prescribed medication, and on any observation of side effects in the patient.

In a preferred embodiment of the present invention, an anti-influenza effective amount of human Mx protein is administered prophylactically or therapeutically to a human in need of such treatment. The preparation of human Mx protein by recombinant DNA methods is described in Example 9. Although the following discussion is intended to be a useful guide to achieving the expression and subsequent isolation of interferon-induced proteins that will be useful in further research or for commercial purposes, one skilled in the art will recognize that many of the techniques described will also be useful when isolation of an interferoninduced protein is not desired but one wishes to achieve expression of an interferon-induced protein for the purpose of causing a desired biological effect within an organism or a cell (e.g., an antiviral effect).

In the cloning and expression of DNA sequences coding for interferon-induced proteins, a wide variety of vectors are useful. These include, for example, vectors consisting of segments of chromosomal, non-chromosomal and synthetic DNA sequences, such as various known derivatives of SV40, known bacterial plasmids, e.g., plasmids from E.coli including col El, pCRl, pBR322, pMB9 and their derivatives, wider host range plasmids, e.g., RP4, phage DNAs, e.g., the numerous derivatives of phage λ, e.g., NM 989, and other DNA phages, e.g.,

M13 and filamentous single stranded DNA phages, yeast plasmids such as the 2μ plasmid or derivatives thereof, and vectors derived from combinations of plasmids and phage DNAs, such as plasmids which have been modified to employ phage DNA or other expression control sequences.

Within each specific cloning or expression vehicle, various sites may be selected for insertion of DNA sequences coding for interferon-induced proteins. These sites are usually designated by the restriction endonuclease which cuts them and are well recognized by those of skill in the art. Various methods for inserting DNA sequences into these sites to form recombinant DNA molecules are also well known. These include, for example, dG-dC or dA-dT tailing, direct ligation, synthetic linkers, exonuclease and polymerase-linked repair reactions followed by ligation, or extension of the DNA strand with DNA polymerase and an appropriate single- stranded template followed by ligation. It is, ofcourse, to be understood that a cloning or expression vehicle useful in this invention need not have a restriction endonuclease site for insertion of the chosen DNA fragment. Instead, the vehicle could be joined to the fragment by alternative means.

For expression of DNA sequences coding for interferon-induced proteins, these DNA sequences are operatively-linked to one or more expression control sequences in the expression vector. Such operative linking, which may be effected before or after the chosen DNA sequence is inserted into a cloning vehicle, enables the expression control sequences to control and promote the expression of the inserted DNA sequence.

Any of the wide variety of expression control sequences (sequences that control the expression of a DNA sequence when operatively linked to it) may be used in these vectors to express the DNA sequence coding for an interferon-induced protein. Such useful expression control sequences, include, for example, the early and late promoters of SV40, the lac system, the trp system, the TAC or TRC system, the major operator and promoter regions of phage λ, the control regions of fd coat protein, the promoter for 3-phos- phoglycerate kinase or other glycolytic enzymes, the promoters of acid phosphatase, e.g., Pho5, the promoters of the yeast α-mating factors, and other sequences known to control the expression of genes of prokaryotic or eukaryotic cells or their viruses, and various combinations thereof. In mammalian cells, it is additionally possible to amplify the expression units by linking the gene to that coding for dehydrofolate reductase and applying a selection to host Chinese hamster ovary cells.

The vector or expression vehicle, and in particular the sites chosen therein for insertion of the selected DNA fragment and the expression control sequence employed in this invention are determined by a variety of factors, e.g., number of sites susceptible to a particular restriction enzyme, size of the protein to be expressed, expression characteristics such as the location of start and stop codons relative to the vector sequences, and other factors recognized by those of skill in the art. The choice of a vector, expression control sequence, and insertion site for a particular protein sequence is determined by a balance of these factors, not all selections being equally effective for a given case. In a preferred embodiment of this invention, when we desire to achieve expression of Mx protein in E.coli, we employ an expression control sequence derived from bacteriophage λ (P_L), which is fused to the Mx cDNA such that the ATG at the end of the prokaryotic sequence abuts the second codon of the Mx sequence.

When we desire to achieve expression of an interferon induced protein, the following different types of promoters may be used:

1) a constitutive promoter;

2) a natural interferon-dependent promoter ( for animals which naturally lack Mx or express it too weakly); 3) a hybrid promoter which has an increased expression rate but is still interferon dependent (for any animal). For a discussion of hybrid promoters see Ryals et al., Cell, 41, 497-507 (1985); or 4) a promoter that is dependent on a substance other than interferon. When we desire to achieve expression of Mx protein in mammalian, including human, or avian cells, we, preferably, employ a metallothionein promoter. The recombinant DNA molecule containing the desired gene operatively linked to an expression control sequence may then be employed to transform a wide variety of appropriate hosts so as to permit such hosts (transformants) to express the gene and to produce the interferon-induced polypeptide for which the hybrid DNA codes. The recombinant DNA molecule may also be employed to transform a host so as to permit that host on replication to produce additional recombinant DNA molecules as a source of genes coding for interferon-induced proteins.

In addition to animal cells (e.g., mammalian, including human, cells or avian cells, as discussed above), a wide variety of hosts are also useful in producing additional recombinant DNA molecules and interferon-induced proteins. These hosts include, for example, bacteria, such as E.coli, Bacillus and Streptomyces, fungi, such as yeasts, and plant cells in tissue culture.

The selection of an appropriate host for the uses described above is controlled by a number of factors recognized by the art. These include, for example, compatibility with the chosen vector, toxicity of the co-products, ease of recovery of the desired polypeptide, expression characteristics, biosafety and costs. No absolute choice of host may be made for a particular recombinant DNA molecule or polypeptide from any of these factors alone.

Instead, a balance of these factors must be struck with the realization that not all hosts may be equally effective for expression of a particular recombinant DNA molecule. It should be understood that the DNA sequences that are inserted at the selected site of a cloning or expression vehicle may include nucleotides which are not part of the actual gene coding for the desired polypeptide or may include only a fragment of the entire gene for that protein. It is only required that whatever DNA sequence is employed, the transformed host produces the desired interferoninduced polypeptide. For example, the DNA sequences utilized in the method of the present invention may be fused in the same reading frame in an expression vector to a portion of a DNA sequence coding for at least one eukaryotic or prokaryotic carrier protein or a DNA sequence coding for at least one eukaryotic or prokaryotic signal sequence, or combinations thereof. Such constructions may aid in expression of the desired DNA sequence, improve purification or permit secretion and maturation, of the desired polypeptide from the host cell. The DNA sequence may alternatively include an ATG start codon, alone or together with other codons, fused directly to the sequence encoding the first amino acid of a desired polypeptide. Such constructions enable the production of, for example, a methionyl or other peptidyl polypeptide. This N-terminal methionine or peptide may then be cleaved intra- or extra-cellularly by a variety of known processes or the polypeptide used together with the methionine or other fusion attached to it.

The following non-limiting Examples are illustrative of the present invention.

EXAMPLES Example 1

Partial Purification of mRNA Encoding Protein Mx,

Isolation of Corresponding cDNA Clones

And Expression of Murine Mx Protein In E.coli

We prepared mouse embryo cells from BALB/c and congenic BALB.A2G-MX mice (Staeheli et al.,

Journal Of Biological Chemistry, 260, 1821-25 (1985)) as described by Arnheiter & Staeheli, Arch. Virol.,

76, 127 (1983). We cultured the cells in Dulbecco's modified minimal essential medium containing 10% fetal calf serum. We used cells between passages 3 to 5.

We treated confluent cell monolayers (about 10⁷ cells/150-mm dish) for 3 h with serum-free medium containing 300 units/ml of partially purified mouse IFN-alpha/beta (10⁷ units/mg) (Tovey et al., Proc.

Soc. Exp. Biol. Med., 146, 809-15 (1974)). We isolated polysome-bound mRNAs essentially as described by Colonno & Pang (J. Biol. Chem., 257, 9234-37 (1982)), except that poly(A) RNA was further purified by extracting with phenol/chloroform. We translated mRNAs in a rabbit reticulocyte lysate in vitro protein synthesis system from Bethesda Research Laboratories, following the manufacturer's protocol. 10 μg/ml of mRNA were translated in 18 ul reactions using ( ³⁵S)methionine

(1200 Ci/mmol; 2 μCi/μl) as radioactive amino acid.

To partially purify this mRNA, we size-fractionated

60 μg of poly(A)⁺ RNA obtained from Mx⁺ embryo cells treated for 3 h with 1,000 U/ml of mu-IFN-alpha/beta by electrophoresis through methylmercury-hydroxyl agarose as follows:

We precipitated 60 μg of poly(A)⁺ RNA from IFN-treated BALB.A2G-MX embryo cells with ethanol, dissolved in 60 μl 0.5 x electrophoresis buffer (a buffer containing 50 mM boric acid, 5 mM sodium borate, 10 mM sodium sulfate and 1 mM EDTA) containing 20 mM methyl-HgOH, incubated for 5 min at 50°C, loaded into 3 cm x 0.2 cm slots and electrophoresed in the buffer through a 1.2% gel of LGT (low gelling temperature) agarose (FMC Corporation) in electrophoresis buffer containing 6 mM methylHgOH, as described by Bailey and Davidson (Anal. Biochem., 70, 75 (1976)). We then soaked the gel for 10 min in a solution of 0.2 M ammonium acetate and 20 mM 2-mercaptoethanol and then cut it into forty 2 mm slices. We extracted the mRNA from individual gel fractions by melting the individual agarose slices at 65°C for 3 min, extracting with phenol and chloroform and recovering the mRNA of the individual gel fractions by precipitation with ethanol. We translated portions of mRNA of individual gel fractions (10 to 19), or 100 ng of unfractionated mRNA from either IFN-treatedMx⁺ BALB.A2G-mx cells (+) or Mx- BALB/c cells (-) in a reticulocyte lysate in vitro protein synthesis system using ( ³⁵S)-methionine as radioactive amino acid. We diluted 6 μl of each translation product in 43 μl of PBS, we added 1 μl of a mouse hyperimmune serum (Staeheli et al., J. Biol. Chem., 260, 1821-25 (1985)) containing antibodies to Mx protein, and we incubated the mixture for 2 h at 4°C. We recovered antibodies with protein A-Sepharose, we analyzed the immunoprecipitated proteins by electrophoresis through a 8% SDS-polyacrylamide gel (top to bottom) as described by Staeheli et al., J. Biol. Chem., 260, 1821-25 (1985), and we visualized by autoradiography. Gel fractions with mRNA of about 3 kb length contained mRNA encoding protein Mx (Figure 1).

We used 0.5 μg of a poly(A) mRNA fraction enriched 10 to 20 fold in Mx mRNA activity to direct vector-primed cDNA synthesis essentially according to the procedure of Okayama & Berg (Mol. Cell. Biol., 2, 161-70 (1982)), but with the cloning vector pHG327, which was derived from pKCR (O'Hare et al., Proc. Natl. Acad. Sci. USA, 78, 10527-31 (1981)) by inserting a Sst I linker into the unique Bam HI site of pKCR such that Bam HI sites were regenerated flanking the linkers. Carrying out cloning at the Sst I site, permitted convenient excision of the cloned sequence and Bam HI. We used the resulting hybrid plasmids to transfect E.coli strain DH-1 cells by the method of Hanahan (J. Mol. Biol., 166, 557-80 (1983)). One microgram of RNA gave rise to about 2 x 10⁵ transformants.

To identify recombinant plasmids with Mx⁺ - specific cDNA inserts, we took advantage of the availability of the congenic mouse strains BALB/c and BALB.A2G-MX which differ at alleles of the Mx locus (Staeheli et al., Virology, 132, 456 (1984)) and as a consequence differ in ability to synthesize protein Mx (the Balb/c strain is Mx- and does not accumulate Mx protein (Staeheli et al., J. Biol. Chem., 260, 1821-25 (1985); Horisberger, Proc. Natl. Acad. Sci., 80, 1910-14 (1983)). Assuming that the mRNA complexities of IFN-treated cells of Mx-congenic mice were largely identical except for mRNA encoding protein Mx, we screened our cDNA library by differential colony hybridization (St. John, Cell, 16, 443

(1979); Hoeijmakers, Gene, 8, 391 (1980)). As hybridization probes, we used ( ³²P)-labeled cDNA prepared from size-fractionated (about 3 kb long) rather than unfractionated mRNA of either IFN-treated BALB.A2G-MX cells or IFN-treated BALB/c cells. We grew transformed E.coli DH-1 (500 per plate of 9 cm diameter) on LB agar containing 100 μg/ml ampicillin. We prepared filters for colony hybridization by the method of Taub and Thompson, Anal. Bioch., 126, 222-30 (1982). We prepared (³²P)-radiolabeled cDNA

(2 x 10 ⁷ cpm/ug RNA) using oligo dT as primer for reverse transcription of mRNA recovered from selected methyl-HgOH agarose gel fractions. To prepare the

"minus" probe, we used mRNA from IFN-treated BALB/c embryo cells; to prepare the "plus" probe, we used mRNA from IFN-treated congenic BALB.A2G-MX embryo cells. We first hybridized the filters to the "minus" probe. After autoradiography, we washed the hybridized probe from the filters by incubation at room temperature for 5 min in 0.5 M NaOH and neutralization in 0.5 M Tris-HCl (pH 8). We then hybridized the filters to the "plus" probe. We performed hybridization in a solution of 150 mM Tris-HCl (pH 5), 0.75 M NaCl, 5 mM EDTA, 0.2% SDS , 0.02% bovine serum albumin, 0.02% polyvinylpyrrolidine, 0.02% Ficoll, and 0.2 mg/ml salmon sperm DNA at 5 x 10 cpm/ml for 14 h at 65°C. We washed the filters in a solution of 0.2 x SSC (SCC is 0.15 M NaCl, 0.015 sodium citrate), and 0.2% SDS at 65°C and exposed the filters to Fuji RX films at -70°C with intensifier screens. Colonies that hybridized to the "plus" probe but not to the "minus" probe were picked and recloned.

Of 10,000 clones analyzed, we identified 2 clones that strongly hybridized to the cDNA from BALB.A2G-MX cells but not detectably to the cDNA from BALB/c cells. The excised cDNA inserts of these two clones hybridized to each other, indicating that they were probably derived from the same mRNA species. The insert of the larger clone (pMx5) was about 1 kb long. It hybridized to mRNA of IFN- treated BALB.A2G-MX cells but not to mRNA of untreated control cells, as expected for MxcDNA.

The cDNA was further identified by the following hybridization-translation assay: mMx5 plasmid DNA or vector DNA was attached to nitrocellulose filters and hybridized to poly(A)⁺ RNA of IFNtreated BALB.A2G-MX cells. Immobilized pMx5 DNA, but not vector DNA, efficiently bound mRNA encoding Mx protein as demonstrated by translating in vitro the RNA released from the filter under denaturing conditions and assaying the product by immunoprecipitation and gel electrophoresis. We concluded that pMx5 probably contained cDNA complementary to Mx RNA encoding Mx protein. Definitive proof for this conclusion was obtained by the transformation experiments described below.

We performed a preliminary Northern transfer analysis as follows: We electrophoresed 3 μg of poly(A) RNA through methyl-HgOH agarose gels as described above. After electrophoresis, we soaked the gel for 15 min in 0.2 M ammonium acetate, 20 mM 2-mercaptoethanol, rinsed for 10 min in water and transferred the RNA to a nitrocellulose membrane using 20 x SSC (Thomas, Proc. Natl. Acad. Sci., 77, 5201-05 (1980)). We baked the membrane under reduced pressure for 2 h at 80°C and prehybridized for 6 h at 42°C in a solution of 20 mM Pipes (pH 6.4), 5 x SSC, 50% formamide, 0.2 mg/ml salmon sperm DNA, 0.1% SDS, 0.04% bovine serum albumine, 0.04% polyvinyl pyrrolidine, and 0.04% Ficoll. We hybridized for 18 h at 42°C in the same buffer at 2 x 10⁵ cpm/ml of a restriction fragment of MxcDNA radiolabeled by nicktranslation to a specific activity of 2 x 10⁸ cpm/ug. We then washed the membrane in 1 x SSC, 0.1% SDS, at 50°C, and autoradiographed at -70°C. This analysis showed that full length Mx in RNA was about 3.5 kb long, far longer than the cDNA insert of pMx5. We therefore undertook the isolation of full length Mx cDNA.

To isolate full-length MxcDNA clones, we amplified our cDNA library (four batches with a total of approximately 100,000 independent transformants) in liquid culture. Transformed E.coli DH-1 (2.5 x 10 4 per 1 liter of medium) were grown to stationary phase in LB medium containing 100 μg/ml ampici lin.

Plasmids prepared from these cultures were enriched for supercoiled forms by pH 12.4 treatment for 10 min at room temperature, followed by renaturation with Tris-NCl (pH 7.5), phenol extraction and ethanol precipitation. The DNA was then electrophoresed through 0.8% gels of LGT agarose and plasmids with inserts larger than about 2,000 bp and having up to about 5,000 bp (as estimated from migration relative to an appropriate reference) were isolated and used to transform E.coli DH-1. Filters for colony hybridization were prepared (Taub & Thompson, Anal. Bioch., 126, 222-30 (1982)) and hybridized to nick translated, ( ³²P)-radiolabeled (108 cpm/ug) insert of pMx5. Hybridization and washing conditions were as above. Positive clones were picked and recloned. We analyzed about 200,000 ampicillinresistant colonies by hybridization with a nicktranslated insert of pMx5. About 1% of the 200,000 clones gave a positive hybridization signal. We further characterized forty-eight randomly chosen clones. The inserts of all clones contained a single Bam HI site located about 1 kb upstream from poly A. The insert of one clone was 3.3 kb long (pMx34), 12 clones (including pMx41) had an insert of 2.5 to 2.8 kb, and the remaining clones had an insert shorter than 2.5 kb. pMx34 and pMx41 were characterized in detail as described in Example 2.

Moreover, we carried out a perfect construction for the expression of murine Mx protein in E.coli DS-10. We first fused the Trp promoter to the Mx coding sequence so that there was a flush end at the AUG codon. We transformed E.coli DS-10 with the resulting recombinant DNA molecule and grew the transformants at 30°C. We then disrupted the transformants with lysozyme and freeze thawing. After centrifuging the disrupted bacteria, we applied the supernatant to an SDS-polyacrylamide gel electrophoresis, then transferred the putative Mx protein to a membrane filter. We next incubated the protein-containing membrane with mouse antibodies against Mx. Using goat anti-mouse antibodies linked to S-peroxidase enzyme, we detected the antigen-antigen complex and thus revealed a 72,000 to 75,000 Dalton band positively identified as murine Mx protein.

Example 2 MxcDNA Sequence And Deduced Sequence Of Protein Mx

Restriction fragments obtained as described m Example 1 were 5' ( ³²P)-labeled with polynucleotide kinase or 3' (³²P)-labeled with Klenow DNA polymerase or with terminal transferase using radiolabeled dideoxy-ATP, and sequenced according to the method of Maxam and Gilbert (Proc. Natl. Acad. Sci. USA,

74, 560-64 (1977)). A restriction map and the nucleotide sequence of the cDNA insert of pMx34 are shown in Fig. 2. The heteropolymeric sequence comprised 3218 nucleotides and was preceded by 12 G residues and followed by about 80 A residues. Numbering starts with the first nucleotide following the string of Gs. An open reading frame extending from the first ATG at nucleotide 214 to a TAA stop codon at position 2107 encodes a protein with 631 amino acids. The sequence upstream of the first ATG codon contains translational stop signals in all 3 reading frames indicating that pMx34 contains the complete coding region of the MxcDNA and that translation most likely initiates at the ATG at position 214 because the next ATG codon is located at position 802. The MxcDNA coding sequence is followed by a 3'-nontranslated region of 1,108 bp which contains the consensus poly A addition signal AATAAA at position 3199. An E.coli containing plasmid pMx34 (designated E.coli DH-1/pMx34) was deposited with the American Type Culture Collection on July 30, 1985 and was assigned ATCC number 53207.

To minimize reverse transcriptase errors and other cloning artifacts, we determined the nucleotide sequence of pMx41, which arose independent of pMx34. The 2650 bp insert of pMx41 was an incomplete copy of the Mx⁺-specific mRNA. The sequence of pMx41 corresponded to nucleotides 658 to 3218 of pMx34, without a single nucleotide difference.

The predicted amino acid sequence of protein Mx is shown in Fig. 2B. Numbering starts with the first methionine of the sequence. The primary Mx translation product consists of 631 amino acid residues and has a molecular weight of 72,037.

Mx protein, as deduced from the nucleotide sequence of the cDNA clone, comprises 631 amino acids. The calculated molecular weight of Mx protein, 72,037 may be compared with experimental values of 72,500 (Horisberger et al., Proc. Natl. Acad. Sci., 80, 1910-14 (1983)), 75,000 (Staeheli et al., J. Biol. Chem., 260, 1821-25 (1985)), and 78,000 (Horisberger et al., J. Biol. Chem., 260, 1730-33 (1985)) estimated by SDS-polyaery1amide gel electrophoresis of natural Mx protein.

It is striking that Mx protein contains several domains which are very rich in charged amino acids; for example, the segment from position 76 to 89 has 3 negatively and 6 positively charged amino acids, the segment from position 93 to 107 has 8 negatively charged residues, and the stretch from position 511 to 522, with the exception of one amino acid, consists of alternating basic and acidic residues. The 40 carboxy terminal residues comprise 26 hydrophilic residues, of which 20 are charged; the segment from position 606 to 614 consists of 7 basic amino acids. The positively charged domains might interact with negatively charged cell components like nucleic acids. Some of the hydrophillic stretches consist of a dense array of acidic residues, others of alternatingly basic and acidic, and others of predominantly basic residues, in particular the sequence Arg-Glu- Lys-Lys-Lys-Phe-Leu-Lys-Arg-Arg near the carboxy terminus; a stretch of (carboxy proximal) basic amino acids Pro-Lys-Lys-Lys-Arg-Lys-Val is held responsible for the nuclear location of SV40 large T antigen (Kalderon et al., Cell, 39, 499-509 (1984)).

In a computer-assisted comparison of the predicted sequence of the Mx protein with more than 4,000 published protein sequences, we detected no significant homology between influenza virus proteins and the Mx protein. Example 3

Expression Of MxcDNA In 3T3 Mouse Cells

Relative positioning to the SV40 early promoter and the orientation of MxcDNA in plasmid pMx34 allowed us to study its expression in eucaryotic cells. We grew 5 x 10⁵ NIH 3T3 mouse cells in Dulbecco's modified minimal essential medium with 10% fetal calf serum in a 90 mm-dish for 18 h. We then added 10 ml fresh medium and 4 h later added 20 μg of calcium phosphate-precipitated DNA.

We used 1 μg of pSV2-neo (Southern and Berg, J. Mol. Appl. Gen., 1, 327-41 (1982)) plasmid DNA and 20μg of high molecular weight carrier DNA from BALB/c liver, or 1 μg of pSV2-neo and 20 μg of Sal I-linearized pMx34 plasmid DNA to prepare DNA precipitates according to the method of Wigler et al. (Proc. Natl. Acad. Sci. USA, 77, 3567-70 (1979)). After maintaining the mixture for 20 h at 37°C, we replaced the DNA-containing medium by fresh medium and we allowed the cells to grow for 24 h. We then trypsinized them and seeded them into new plates at a splitting ratio of 1 to 8. We replaced the medium 24 h later by a medium containing 1 mg/ml G418 (Geneticin, Gibco), changing the medium every 2 to 3 days. Resistant clones (50 to 100 per plate) appeared after 2 weeks. We co-transfected these cells with 1 part of pSV2-neo plasmid DNA (conferring to transfected cells resistance against the drug G418) and either 20 parts of linearized pMx34 plasmid DNA or 20 parts of carrier DNA. These cells were kept in medium supplemented with 1 mg/ml of the neomycin analog G418 (Gibco).

We selected permanently transfected cells expressing G418 resistance, and rather than isolating single clones, we pooled the G418-resistant cells of each plate. Using immunofluorescence, we analyzed the ability of pools representing the progeny of about 100 individual G418-resistant transfectants to synthesize protein Mx.

G418-resistant cells were grown for 20 h on glass cover slips, washed with PBS, fixed at 25°C for 10 min with 3% paraformaldehyde and permeabilized for 5 min with 0.5% Triton X-100. To detect Mx protein, fixed and permeabilized cells were incubated for 15 min at 25°C with 0.4% of mouse hyperimmune serum with antibodies to Mx protein (Staeheli et al., J. Biol. Chem., 260, 1821-25 (1985)) in PBS containing 5% normal goat serum. To reveal bound antibody, the cover slips were incubated for 15 min at 25°C with rhodamine-conjugated goat anti-mouse IgG (Nordic), diluted 1 to 50 in PBS containing 5% normal goat serum, washed with PBS and mounted in 50 mM Tris-HCl (pH 8.6) and 50% glycerol. A Reichert-Jung Polyvar microscope was used for ultraviolet incident light fluorescence microscopy.

Parental Mx 3T3 mouse cells and 3T3 cells which had been transfected with pSV2-neo plasmid but not with MxcDNA were unable to synthesize protein Mx whether or not they were treated with IFN (Fig. 3A and Fig. 4). In contrast, a high proportion of 3T3 cells transfected with MxcDNA synthesized Mx protein. Since MxcDNA was expressed under control of the SV40 early promoter, recombinant Mx protein was synthesized constituitively in transfected cells and did not require induction by IFN. Like the natural protein in IFN-treated Mx cells, recombinant protein Mx accumulated in the nuclei of transfected 3T3 cells (Fig. 3B). About 30% of the transfected cells originating from plate 34/6 synthesized Mx protein (Fig. 3B).

Treatment with IFN did not detectably influence Mx protein expression in transfected cells. Because in the transformed cells Mx protein is constitutively transcribed under the direction of the SV40 early promoter this is the expected result inasmuch as expression of the Mx gene is solely under transcriptional control. The level of Mx protein expression in individual cells was variable. A minority of transfected cells contained as much Mx protein (as determined by immunofluorescence staining) as fully induced Mx⁺ embryo cells treated with 1,000 U/ml of IFN- alpha/beta for 18 h, whereas the majority of the Mx protein-expressing 3T3 cells contained low concentrations of Mx protein. Recombinant and natural Mx protein were indistinguishable in reactivity with three distinct, specific monoclonal antibodies, and, on Western blots, they had the same apparent molecular weights. We grew cells of pool 34/6 for more than 3 months in G418 containing medium. Over this time period, the number of protein Mx-expressing cells as well as the relative expression levels (determined by immunofluorescence analysis) stayed relatively constant, indicating that protein Mx did not interfere with cell growth. By limiting dilution, we cloned protein Mx-producing 3T3 cells. After two cycles of cloning and amplification, we obtained cultures with about 95% protein Mx-producing cells, but these cultures still contained non-producing cells, and, as in culture 34/6, protein Mx expression levels of individual cells were not uniform, suggesting a complex regulation of Mx protein synthesis in transfected cells.

Example 4

Mx Protein Inhibits Influenza Virus Replication

We infected cultures of transfected 34/6 cells on glass cover slips with either strain FPV-B of influenza A virus or, in control experiments, with the Indiana strain of vesicular stomatitis virus (VSV). We prepared each of the viruses from plaque purified virus grown on NIH 3T3 cells to about 10⁷ plaque-forming units (PFU) per ml. We infected the cells with either virus for 3 h at room temperature in medium containing 2% FCS (fetal calf serum) at a multiplicity of 10 PFU per cell. We then washed the cells twice with medium and incubated for 3 to 4 h at 37°C in medium containing 2% FCS. Under these conditions, 98% of individual cells of a NIH 3T3 control cell culture were positive when assayed by immunofluorescence with specific antiviral antibodies. We fixed cells at room temperature for 10 min with 3% paraformaldehyde and permeabilized them for 5 min with 0.1% Triton X-100. To simultanously monitor Mx protein and virus proteins (i.e., influenza virus proteins or VSV G protein, depending on which virus was used for infection), we treated, fixed and permeabilized cells with a mixture of mouse antibodies directed against Mx protein and of rabbit antibodies against virus proteins (influenza or VSV). We used a mouse hyperimmune serum containing antibodies to Mx protein (Staeheli et al., J. Biol. Chem., 260, 1821-25 (1985)) at 0.4%, a rabbit antiserum against influenza virus at 0.2%, and a rabbit antiserum against the VSV G protein (Arnheiter et al., Cell, 39, 99-109 (1984)) at 2.5%. We diluted antisera in PBS containing 5% normal goat serum. We incubated the mixtures with these antibodies 15 min at 25°C. To visualize antibody bound to Mx protein, we used rhodamine-conjugated goat anti-mouse antibodies (Nordic). To visualize antibody bound to virus proteins, we used fluoresceine-conjugated goat antirabbit antibodies (Nordic). We reacted the cultures with these labeled antibodies for 10 min at room temperature at 2% in PBS containing 5% normal goat serum. We washed the cover slips with PBS, mounted them in 50 mM Tris-HCl (pH 8.6), 50% glycol, and analyzed individual cells for both rhodamine- and fluorescein-specific fluorescence by uv incident light fluorescence microscopy. Thus, using immunofluorescence techniques, we were able to assay individual cells for ability to synthesize protein Mx and at the same time for synthesis of influenza virus proteins. As shown in Fig. 4, cells that produced Mx, as evidenced by a fluorescing nucleus (Fig. 4a), did not contain influenza-specific proteins (Fig. 4b); cells producing influenza-specific proteins are Mx negative.

Thus, transfected cells containing recombinant Mx protein did not allow synthesis of influenza virus proteins, while cells lacking Mx protein were susceptible to the influenza A virus FPV-B (Fig. 4a and 4b). Mx protein-expressing cells were selectively protected against influenza viruses. Both cells lacking Mx protein and cells containing Mx protein were fully susceptible to the rhabdovirus VSV (Fig. 4c and 4d). The degree of resistance against influenza viruses of individual cells was variable. Cells with high concentrations of Mx protein were fully protected, while cells containing only low concentrations of Mx protein were protected to a lesser extent. We recorded the relative amounts of Mx protein and of influenza virus proteins of 746 individual 34/6 cells infected with the influenza A virus FPV-B at a multiplicity of 10 plaque forming units per cell. After 3 h at 37°C, we classified, by immunofluorescence analysis, each cell as a high producer, a low producer or a non-producer of Mx protein and of influenza virus proteins, respectively. Table 1 shows that 100% of the cells that synthesized high amounts of Mx protein were resistant to influenza virus.

About 50% of the cells that produced little Mx protein were stainable with specific antibodies to influenza virus, while 98% of the cells that did not contain detectable amounts of Mx protein were influenza virus positive.

By showing in one and the same culture that cells expressing high levels of Mx were resistant to influenza virus, while cells devoid of Mx protein were fully susceptible, we proved that the cloned sequence encoded the Mx function. The question as to whether Mx protein could exert its antiviral activity in the absence of added IFN was answered in the affirmative: MxcDNA transformed 3T3 cells that expressed Mx protein at a level similar to that of IFN-induced Mx cells were resistant to influenza virus infection even without addition of IFN.

Thus, unexpectedly, the presence of Mx, in the absence of IFN treatment, suffices to protect cells against influenza virus infection. Natural Mx cells, however, only develop resistance to influenza virus after treatment with IFN. Example 5

Mx-Specific mRNA Synthesis Is Stringently Controlled By IFN

We treated cells of BALB/c or congenic BALB.A2G-MX mice for 3 h with either serum-free medium or medium with 300 U/ml of IFN-alpha/beta, and we prepared polysome bound poly(A) RNAs. We electrophoresed 3 μg of mRNA of each preparation through a methylmercuryhydroxyl-agarose gel (Bailey and Davidson, Anal. Biochem., 70, 75 (1976)) transferred the mRNA to a nitrocellulose membrane and then hybridized the mRNA to a radiolabeled MxcDNA probe (Thomas, Proc. Natl. Acad. Sci., 77, 5201-05 (1980)). IFN-treated Mx⁺ cells contained an about 3.5 kb long mRNA which strongly hybridized to MxcDNA. mRNAs from non-treated control cells did not hybridize to MxcDNA (Fig. 5). We obtained identical results with a probe derived from the coding region of MxcDNA and with a probe derived from the 3' non-coding region of MxcDNA which contains a repetitive sequence.

Example 6

Mx- Cells Synthesize A Truncated Mx-Specific mRNA

We found that in response to IFN-alpha/beta, Mx- embryo cells synthesized an mRNA which hybridized to MxcDNA probes. This mRNA was about 200 to 500 nucleotides shorter than the Mx-specific mRNA of Mx⁺ cells. Polysomal poly(A) RNA preparations of Mx- cells contained very low concentrations of this mRNA, just enough to be detected on Northern blots (Fig. 5). We failed to detect the Mx-specific mRNAs in untreated control cells. Synthesis of the Mx- specific mRNAs of Mx⁺ and of Mx- cells would thus seem to be under similar control. We also performed analytical gel electrophoresis by loading 3 μg of poly(A) RNA into 0.5 cm x 0.2 cm slots and electrophoresing through 1.2% gels of Sigma type II agarose. After electrophoresis, we soaked the gels in a solution of 0.2 M ammonium acetate and 20 mM 2-mercaptoethanol for 10 min, and stained with ethidium bromide or used the gels for Northern transfer analysis.

Example 7 Mx Gene Deletions In Mx- Mice

To learn about the nature of the defect of Mx- mice, we analyzed Southern blots of chromosomal DNA from BALB/c mice using the radiolabeled 1.65 kb Bam HI fragment of pMx41 as probe. As shown in Figure 6A, the restriction patterns of congenic BALB.A2G-MX and BALB/c mice were not identical. The bands at 5 kb and at 7.5 kb of Eco RI- cleaved Mx ± DNA were absent in BALB/c DNA, but instead, a new band at about 4 kb appeared. The weak band at 2.3 kb and the strong band at 6.5 kb of Bam Hl-cleaved DNA were absent in Mx- BALB/c DNA, instead, a new band at 7 kb was detected. Similarly, a 3 kb and a 10 kb Hind III fragment of Mx ± DNA were replaced by a 9 kb fragment in Mx- DNA, and a 5 kb Pst I fragment was replaced by a 2.5 kb fragment. Using the radiolabeled 1.0 kb Ba HI fragment of MxcDNA as Southern probe, we obtained a smear and a few superimposed bands with DNA from Mx- mice as with DNA from Mx⁺ mice (Fig. 6B). These results are compatible with a deletion in the Mx gene of the BALB/c genome. This deletion most likely includes both exons and introns and seems to be located upstream of the Bam HI site at position 2317 of the MxcDNA.

We compared the restriction pattern of eight different mouse strains (Fig. 6C and 6D). The two inbred strains BALB.A2G-MX and SL/NiA as well as a heterozygous backcross animal (T9xCBA)FluR were phenotypically Mx⁺, while the inbred strains BALB/cJ, A/J, 129/J and C57BL/6J gave identical EcoRI restriction patterns, suggesting that these strains may carry identical deletions of the Mx gene. The restriction patterns of strain CBA were clearly different from the other Mx- strains and in the Hind Hl-cleaved DNA clear differences also became visible between CBA DNA and Mx+ DNA. The strong band at 10 kb obtained with Mx⁺ DNA was not detected in CBA mice, but instead, a band at 5 kb was found. Thus, at least two different Mx- alleles occur among inbred strains with deletions differing in length and/or location. It is not yet known whether these two types of deletions are independent or whether one is derived from the other. These results clearly establish that despite the frequency of Mx-, the wild phenotype is Mx⁺ and Mx- is derived from it by deletion. The preponderance of Mx- in laboratory strains may be due to a founder effect, however the high frequency of the Mx- allele found by Haller et al. in wild strains remains to be explained.

Example 8

Immunoprecipitation Assay Of Other Mammalian Proteins With Homology To Murine Mx

To determine if there are homologous proteins to murine Mx in other mammals, we analyzed the cross-reactivity of murine Mx specific antibodies to interferon-induced proteins from other species.

We used three different monoclonal antibodies (5D11, 6D4 and 2C12) (Staeheli et al., J. Biol. Chem., 260, 1821-25 (1985)) to screen human and rat cells for the presence of proteins homologous to murine Mx. We metabolically labeled IFN-treated and non-treated (control) cells with [ ³⁵S] methionine and used cytoplasmic extracts of the cells for immunoprecipitation assays. We prepared monoclonal antibodies 5D11, 6D4 and 2C12 from BALB/c (Mx-) mice which were immunized with cell extracts from IFN- alpha-beta-treated congenic BALB .A2G-Mx(Mx⁺) cells (Staeheli et al., supra).

In immunoprecipitation assays of cytoplasmic extracts of murine cells, all three monoclonal antibodies showed high specificities for murine Mx protein and precipitated a single 75,000 Dalton protein, as did polyclonal anti-Mx antisera (Staeheli et al., J. Biol. Chem., 260, 1821-25 (1985)). Upon immunoprecipitation assays of cytoplasmic extracts of human cells, we found a 2C12- cross-reacting human protein synthesized, after IFN-α induction, by such diverse cells as fresh pulmonary blood lymphocytes (PBLs), cultured human fetal lung cells (HFLCs), and cultured skin fibroblasts from newborn and adult donors. We incubated human PBLs from healthy donors with E . coli-produced IFN-α₂ (10⁸ U/mg) or natural IFN-α (10⁶ U/mg). We left other PBLs untreated as controls. After maintaining the cells at 37°C for 2 h, we washed the cells and metabolically labeled the cell proteins with [³⁵S]methιonine. We next prepared cell extracts and assayed portions according to the procedure of

Staeheli et al. (J. Biol. Chem., 260, 1821-25 (1985)) for proteins immunoprecipitable with antibodies to murine Mx protein.

IFN-treated PBLs of all human donors (11 of 11) synthesized an 80,000 Dalton protein that was immunoprecipitable with monoclonal antibody 2C12. We did not detect this protein in untreated control cells. E . coli-produced IFN-α₂ and natural IFN-α were indistinguishable in their ability to induce in PBLs the synthesis of this protein. The two other monoclonal antibodies to murine Mx protein (6D4 and 5D11) did not react with the human protein; and a polyclonal antiserum (Staeheli et al., supra) had a low but significant titer of cross-reactivity. We next assayed HFLCs for synthesis of proteins homologous to murine Mx. Again, monoclonal antibody 2C12, but not 6D4 or 5D11, precipitated an IFN-α₂-induced, 80,000 Dalton protein. We failed to detect this protein in untreated HFLCs. Concentrations of 5 to 25 U of IFN-α₂ per ml clearly induced the synthesis of the cross-reactive human protein. Induction was more pronounced with 125.U/ml and reached a maximum with 625 or 3,125 U/ml. HFLCs synthesized the cross-reactive protein at a relatively low rate 2 h after the onset of IFN treatment. The rate of synthesis increased with time and reached a maximum between 6 and 12 h.

In HFLC extracts, monoclonal 2C12 reacted with a second IFN-α₂-induced protein (molecular weight approximately 75,000 Daltons) which was present at low concentrations. This protein may represent a degradation product or an unmodified precursor of the 80,000 Dalton protein, or it may be the product of a unique mRNA.

We next treated extracts of HFLCs for various times with high concentrations (10⁴ U/ml) of E.coli-produced IFN-γ (5 x 10⁷ U/mg), but failed to detect significant amounts of the 80,000 Dalton protein. This preparation of IFN-γ is otherwise active on HFLCs, as was evidenced by its ability to induce synthesis of 67k GBP, a guanylate-binding protein which is induced in fibroblast cells by different types of IFNs (Cheng et al., J. Biol. Chem., 258, 7746-50 (1983)), and by its ability to inhibit virus replication in HFLCs.

In additon, using similar procedures we found that IFN-treated, but not untreated, rat embryo cells synthesized proteins that cross-reacted with anti-Mx antibodies. Monoclonal antibodies 5D11 and

6D4 immunoprecipitated a single IFN-induced 72,000

Dalton protein while 2C12 immunoprecipitated large quantities of two 80,000 Dalton proteins and small quantities of a 72,000 and a 65,000 Dalton protein. As in the human and mouse, all rat proteins that cross-reacted with anti-Mx antibodies were absent in untreated control cells and strongly induced in cells treated with IFN-α, but not in cells treated with IFN-γ. Moreover, antibody 2C12 exhibited a high degree of specificity: it failed to recognize any human proteins present in normal, untreated cells; it also failed to specifically react with any murine proteins other than protein Mx (Dreiding et al., Virology, 140, 192-96 (1984); Staeheli et al., J. Biol. Chem., 260, 1821-25 (1985)).

The foregoing results indicate that 2C12- cross-reacting protein shares at least one unique antigenie determinant (not found on other human or mouse proteins) with murine Mx and that synthesis of both the human and mouse proteins described in this Example is induced by IFN-α but not by IFN-γ and thus seem to be under similar control. Thus, the

2C12-cross-reacting protein described above appears to be the human homolog of murine Mx protein.

Example 9 Isolation and Expression of Human Mx cDNA We first determine optimal conditions for cross-hybridization of human gene with mouse cDNA. For this determination, we first prepare a Southern blot of the entire human DNA using three restriction enzymes (for example, Bam HI, Eco II and Bgl II) and a mouse cDNA control. After hybridizing the human

DNA with the mouse cDNA under permissive conditions, we then wash at different stringency conditions to determine optimal conditions for cross-hybridization. We next determine the human cell line or tissue with appropriate Mx mRNA content and the optimum IFN induction conditions. For this determination, we first treat human leukocytes and human cell lines (e.g., HeLa, HEL60, WISH, monocyte, lymphoblastoid and Daudi) with human interferon alpha at different concentrations (0, 30, 300, 3000 U/ml) for 0, 2, 4 and 10 hours. We then purify RNA, and, if necessary, poly(A) RNA, from each sample as well as from positive control cells ( IFN-induced murine Mx⁺ cells). We then dot and hybridize the RNA under the optimal conditions as determined above, using nick-translated mouse Mx cDNA. We next quantitate hybridization by radioactivity analysis and thus determine the human cell line or tissue which has the appropriate Mx mRNA content, as well as the optimum IFN induction conditions.

To prepare and characterize human Mx mRNA-containing poly(A)⁺ RNA, we induce with interferon alpha, under the conditions determined above, cells found to be the best source of human Mx mRNA We then purify poly(A)⁺ RNA by standard procedures and perform a Northern blot analysis to determine the length of the Mx mRNA.

We next prepare and screen a cDNA bank. We use the poly(A)⁺ RNA, characterized by the Northern blot analysis, as a template for preparing cDNA by the Gubler-Hofmann method. - We treat the double stranded cDNA with EcoRI methylase, then we link the double stranded cDNA to EcoRI linkers using DNA ligase. After fractionating on an agarose gel, we ligate the cDNA fractions, corresponding to the length of the Mx mRNa ± 200 nucleotides, to the lateral fragments (arms) of lambda gt10, and then package the resulting recombinant phage. We then use the phage to infect E.coli. We plate the infected bacteria as usual (T. Maniatis et al., Molecular Cloning, Cold Spring Harbor Laboratory, New York (1982)). We then screen the resulting bank with nick-translated murine cDNA under optimum crosshybridization conditions as determined above.

To characterize the cloned Mx cDNA, we first subject the putative Mx cDNA to restriction mapping by the Smith-Birnstiel (Nucleic Acid Res., 3, 2387 (1976)) procedure. We then sequence the Mx cDNA by the Maxam-Gilbert method (Proc. Natl. Acad. Sci. USA, 74, 560-64 (1977)) and compare the sequence to the sequence of murine Mx cDNA to determine the degree of homology.

To test the functional competence of the human Mx cDNA, we first clone the coding sequence determined as discussed above into the mammalian expression vector pβG under the control of the (constitutive) SV40 early promoter. We then introduce the vector into Mx murine L cells as well as into human HeLa cells and Cos cells. We then test expression of Mx in situ in transiently and permanently transformed cells using fluorescent antibodies against murine Mx (which have been shown to cross-react with the human protein). After challenging transformed cells with influenza virus, we assess viral replication in situ with the use of fluorescent anti-influenza antibodies, as described previously (see Staeheli et al., Cell, 44, 147-58 (1986)).

To produce the human Mx protein, we fuse the coding sequence of human Mx to a variety of promoters (e.g., Trp, lambda, P_L or tac) and different

3' non-coding regions, for expression in E.coli. Alternatively, we join the Mx coding sequence to the IFN signal sequence for expression in a DHFR-plasmid, using the methotrexate amplification technique. We then measure expression by an immunological method (see, e.g., Staeheli P. et al., J. Biol. Chem., 260, 1821-25 (1985)). We purify Mx by procedures involving immunoaffinity chromatography. To assess the biological activity of murine and human Mx protein in cell free systems and in tissue culture, we use the following approaches : (i) microinjection into cells and testing for virus resistance as above; (ii) use of a cell-free system (see, e.g., Krug et al., Virology, 56, 201-06 (1985)) for influenza virus replication (it is believed that Mx acts by inhibiting influenza virus RNA synthesis, and one can assay for this activity); (iii) introduction of Mx into cells using Mx-loaded erythrocytes and cell fusion, or fusion of Mx-producing E.coli with L cells, or fusion of Mx-loaded liposomes with murine L929 cells, or simple treatment of murine L929 cells with solutions of Mx, and assay for influenza virus resistance as above.

To assess the biological activity of Mx protein in animals, we intravenously inject Mx- mice with human or murine protein solutions or liposome- packaged Mx. After challenging with influenza virus, we score the surviving mice.

While we have hereinbefore described a number of embodiments of this invention, it is apparent that the embodiments herein described can be altered to provide other embodiments which utilize the processes and compositions of this invention.

Therefore, it will be appreciated that the scope of this invention is to be defined by the claims appended hereto rather than by the specific embodiments which have been presented hereinbefore by way of example.

Claims

CLAIMS :

1. A method of protecting an animal against viral infection comprising inserting into an animal a gene coding for an interferon-induced protein which is capable of protecting said animal against said viral infection.

2. A method according to claim 1, wherein said viral infection is caused by influenza virus and said gene codes for Mx protein.

3. A method according to claim 1 or 2, wherein said gene is operatively linked to an expression control sequence in said recombinant DNA molecule.

4. A host transformed with a recombinant DNA molecule comprising a gene coding for an interferon-induced protein.

5. A host according to claim 4, wherein said host is an animal cell which in its natural state is susceptible to a viral infection and said recombinant DNA molecule comprises a gene coding for an interferon-induced protein which is capable of protecting said cell against said viral infection.

6. A host according to claim 5, wherein the viral infection is influenza and the gene is a gene coding for Mx protein.

7. A host according to any one of claims 4-6, wherein said gene is operatively linked to an expression control sequence in the recombinant DNA molecule.

9. A recombinant DNA molecule comprising a DNA sequence selected from the group consisting of (a) the DNA sequence of claim 8,

(b) DNA sequences that hybridize to the DNA sequence of claim 8 and that code on expression for Mx protein, and

(c) DNA sequences that are degenerate as a result of the genetic code to the aforementioned DNA sequences and that code on expression for Mx protein.

10. A recombinant DNA molecule according to claim 9, wherein said DNA sequence is operatively linked to an expression control sequence in the recombinant DNA molecule.

11. A recombinant DNA molecule according to claim 9, said molecule comprising a promoter selected from the group consisting of (a) a constitutive promoter,

(b) a natural interferon-dependent promoter,

(c) a hybrid promoter that is interferon- dependent but provides higher expression than that of a natural interferon-dependent promoter, and

(d) a promoter that is dependent on a substance other than interferon.

12. A polypeptide selected from the group consisting of

(a) a first polypeptide comprising the amino acid sequence M D S V N N L C R H Y E E K V R P C I D L I D T L R A L G V E Q D L A L P A I A V I G D Q S S G K S S V L E A L S G V A L P R G S G I V T R C P L V L K L R K L K E G E E W R G K V S Y D D I E V E L S D P S E V E E A I N K G Q N F I A G V G L G I S D K L I S L D V S S P N V P D L T L I D L P G I T R V A V G N Q P A D I G R Q I K R L I K T Y I Q K Q E T I N L V V V P S N V D I A T T E A L S M A Q E V D P E G D R T I G V L T K P D L V D R G A E G K V L D V M R N L V Y P L K K G Y M I V K C R G Q Q D I Q E Q L S L T E A F Q K E Q V F F K D H S Y F S I L L E D G K A T V P C L A E R L T E E L T S H I C K S L P L L E D Q I N S S H Q S A S E E L Q K Y G A D I P E D D R T R M S F L V N K I S A F N R N I M N L I Q A Q E T V S E G D S R L F T K L R N E F L A W D D H I E E Y F K K D S P E V Q S K M K E F E N Q Y R G R E L P G F V D Y K A F E S I I K K R V K A L E E S A V N M L R R V T K M V Q T A F V K I L S N D F G D F L N L C C T A K S K I K E I R L N Q E K E A E N L I R L H F Q M E Q I V Y C Q D Q V Y K E T L K T I R E K E A E K E K T K A L I N P A T F Q N N S Q F P Q K G L T T T E M T Q H L K A Y Y Q E C R R N I G R Q I P L I I Q Y F I L K T F G E E I E K M M L Q L L Q D T S K C S W F L E E Q S D T R E K K K F L K R R L L R L D E A R Q K L A K F S D, and

(b) a second polypeptide comprising at least one segment of said first polypeptide, said second polypeptide having a protective effect in an animal against influenza.

13. A polypeptide selected from the group consisting of

(a) a protein having a molecular weight of about 80,000 Daltons that is immunoprecipitable with monoclonal antibody 2C12 and that is inducible in human peripheral blood lymphocytes, human fetal lung cells and human fibroblasts by interferon-α or interferon-β but not by interferon-γ, and

(b) a polypeptide comprising at least one segment of said protein, said polypeptide having a protective effect in an animal against influenza.

14. An antiviral pharmaceutical composition comprising an antiviral effective amount of a polypeptide according to claim 12 or 13 and a pharmaceutically acceptable carrier.

15. A method of treating or preventing viral infections in animals comprising administering to an animal susceptible to such infection an antiviral effective amount of a polypeptide according to claim 12 or 13.