WO1999006569A1

WO1999006569A1 - Cytotoxic polypeptide (fragment) derived from ribonuclease

Info

Publication number: WO1999006569A1
Application number: PCT/NL1998/000434
Authority: WO
Inventors: Johannes Theodorus Van Oirschot; Christianne Jacqueline Maria Bruschke; Petrus Antonius Van Rijn
Original assignee: Stichting Instituut Voor Dierhouderij En Diergezondheid
Priority date: 1997-07-30
Filing date: 1998-07-29
Publication date: 1999-02-11
Also published as: AU8650898A

Abstract

The invention relates to the field of immune suppression, cytostatic and cytocidal agents and apoptosis by biological immunosuppressive agents. Immunosuppressive therapy is widely applied in the treatment of a varied array of diseases or conditions. The invention provides a polypeptide or fragment thereof, derived from a protein related to the superfamily of ribonucleases, said polypeptide devoid of its specific RNase activity and capable of having an immunosuppressive, cytostatic and/or cytocidal effect. The invention further provides a polypeptide or fragment thereof, capable of having apoptotic effect on lymphocytes, said polypeptide derived from a virus protein.

Description

Title: Cytotoxic polypeptide (fragment) derived from ribonuclease .

The invention relates to the field of cytotoxic agents which are widely applied in the treatment of a varied array of diseases or conditions. Cytotoxic, cytocidal or cytostatic (these terms are interchangeably used throughout this document) agents are widely used with a variety of clinical conditions, such as rheumatism, multiple sclerosis and after (organ) -transplantations . Cytotoxic agents such as azathioprine, cyclosporine, mycophenolic acid, FK-506, also find application in the treatment of cancers, added to or replacing other cytotoxic agents as varied as cyclohexamide and cis-platinum, which have a similar broad action. Above cytotoxic agents act in general on cells of various types by inhibiting or blocking nucleic acid and subsequent protein synthesis, thereby inhibiting cell-division and finally promoting cell-death. These drugs exert major toxic side- effects which can seriously hamper the recovery of a patient, and can also be a serious risk to medical personal, when not handled carefully. Side-effects such as nephrotoxicity, leucopenia, thrombocytopenia, gastro-intestinal problems, and opportunistic infections stem from the in general too broad reactivity of said agents. Attempts to find selective agents with a narrow specificity restricted to a specific organ system or even to specific cells have in general been unsuccessful. Several enzymes which specifically act on nucleic acid, in particular ribonucleases (RNases) , have been reported to have cytotoxic side-effects. These effects, which are thought to be linked to the RNase activity of these enzymes, include anti-fertility, anti-tumor, immunosuppressive, angiogenic, ribosome-inactivating, anti- helmintic, gametophytic self-incompatibility and neurotoxic effects (Youle et al, Crit . Rev. Ther. Drug Carrier Syst . 10, 1-28, 1993) . It has even been demonstrated that adding RNA-se derived activity to a in itself angiogenic protein, can increase the angiogenic activity of said protein ( 091/12322) . RNases having general biological effects are also known as RISBASES (D'Alessio et al , TIBS 16, 104-106, 1991; D'Alessio, Trends in Cell Biology vol. 3, 106-109, 1993) .

The present invention provides a selective cytotoxic (poly) peptide or fragment thereof which is derived from a ribonuclease wherein said cytotoxic activity is not, at least not exclusively, linked to RNase activity. The (poly) peptide or fragment thereof provided by the invention comprises a site related to its cytotoxicity independent of or additional to an RNase catalytic site. Although it has been described that ribonuclease activity may not be essential for general cytotoxicity (Rosenberg, H.T., J. Biol.Chem. 270, 7876-7881, 1995) it has not been recognised that said cytotoxicity can be selective, instead of broadly reactive with a wide range of cells. An example provided by the invention is a polypeptide or fragment thereof which is derived from a viral protein with ribonuclease activity which polypeptide (fragment) comprises selective cytotoxicity, for example for lymphocytes, independent of said ribonuclease activity. A preferred embodiment of the invention is a polypeptide (fragment) derived from a pestivirus E^rns protein which comprises selective cytotoxicity for lymphocytes independent of ribonuclease activity characteristic for a pestivirus E^rπs protein. The invention provides a polypeptide, or fragment thereof, derived from a protein having specific RNase activity and related to the superfamily of ribonucleases, said polypeptide (fragment) having selective cytotoxic effect but partly or wholly lacking said specific RNase activity. The invention also provides a nucleic acid molecule encoding such a polypeptide (or fragment thereof) partly or wholly lacking said specific RNase activity but capable of having a selective cytotoxic effect. As an example, the invention provides a nucleic acid molecule wherein at least one codon encoding a histidine residue being part of a ribonuclease catalytic site has been mutated. As a further example, the invention provides a nucleic acid molecule wherein said codon has been mutated to encode a lysine residue. A vector or an expression system comprising a nucleic acid according to the invention are also provided, as well as a host ^' cell comprising such a nucleic acid or vector according to the invention. The invention provides a polypeptide, or fragment thereof, capable of having selective cytotoxic activity on lymphocytes, thereby eliciting an immunosuppressive effect. As an example of the invention in the experimental part a recombinantly expressed polypeptide derived from a pestivirus E^rns protein capable of having selective cytotoxic activity on lymphocytes is given. The invention provides a RNase-derived polypeptide, or fragments thereof, in which partly or wholly abolishing RNase activity does not or only little affect other biological activity of the protein. The invention provides a manipulated RNase or RNase derived polypeptide or fragment thereof, making it possible to selectively maintain, mitigate or deplete various effects of the protein, thereby making it into a selective agent for therapeutic use without many unwanted side-effects. The invention further provides a polypeptide or derivative or fragment thereof, capable of having specific and selective immunosuppressive effect by its effect on lymphocytes, said polypeptide derived from a (virus) protein but devoid of the specific level of RNase activity of said protein. The invention also provides chimaeric (polypeptides) comprising a (poly) eptide or derivative or fragment thereof according to the invention. In addition, analogs which counteract parts of the biologically activity of a polypeptide (fragment) provided by the invention are also provided. Since routine techniques are available for producing e.g. synthetic peptides, it is within the skill of the art to arrive at analogs or derivatives of the RNase-derived polypeptides or fragments provided by the invention. A (polypeptide) derivative or fragment derived from a ribonuclease as provided by the invention is obtained by various methods, such as synthetic peptide (FMOC) chemistry and synthetic peptide ELISA (PEPSCAN) , whereby amino acid residues can be replaced conventionally, e.g valine replacement by alanine (in synthetic peptides all conventional replacement amino acid residues, be it D- or L-residues, can be used) or whereby a replacing amino acid is selected by replacement net scanning, or by using a combinatorial phage display library, or other methods known in the art can be used to modify a

(poly) peptide fragment or derivative provided by the invention. With such techniques it is possible to alter, design or modify an active site, for example by replacing at least one histidine residue being part of an RNase catalytic site with another amino acid residue. An example of such a replacing residue is lysine, however, other amino acid residues, be it D- or L-residues can also be used to modify a (poly) peptide fragment provided by the invention. Other methods are classic recombinant expression of a modified polypeptide (fragment) provided by the invention, whereby changing an appropriate codon allows amino acid modification. The invention also provides a pharmaceutical composition comprising a polypeptide or fragment thereof according to the invention and further comprising a pharmaceutically acceptable substance. The invention provides a polypeptide or fragment thereof or a composition comprising such a polypeptide- (fragment) and its use in immunosupressive, cytostatic or cytocidal therapy. Furthermore, the invention provides the use of a polypeptide, or fragment thereof, derived from a pestivirus E^rns protein, for the manufacture of a medicament for the selective cytotoxic treatment of lymphocytes. A pharmaceutical composition or medicament provided by the invention can specifically be used in immunosuppressive therapy. The invention also provides a nucleic acid encoding a polypeptide or fragment thereof according to the invention, and a vector comprising such a nucleic acid. Also an expression system or host cell comprising such a nucleic acid or vector are provided. The invention provides a pharmaceutical composition comprising a polypeptide or fragment thereof according to the invention. It is within the skills of the artisan to compose a pharmaceutical composition using a pharmaceutically acceptable substance as diluent, carrier or excipient . The invention further provides a pharmaceutical composition not being a vaccine comprising a polypeptide (fragment) derived from a pestivirus E^rns protein. A pharmaceutical composition provided by the invention can be used in therapy, for example in immunosuppressive, cytostatic or cytocidal therapy. A pharmaceutical composition comprising a polypeptide (fragment) derived from a pestivirus E^rns protein can for example be used in immunosuppressive therapy wherein a selective cytotoxic effect on lymphocytes is preferred. The invention also provides the use of a polypeptide or fragment thereof according to the invention for the manufacture of a medicament for immunosuppressive, antiproliverative, cytostatic or cytocidal (cancer) therapy. The invention is further explained in the experimental part which however should not be seen as limiting the invention.

Experimental part.

The superfamily of ribonucleases comprises among others RISBASES from animals (e.g. BS-RNase, angiogenin, eosinophil derived Neurotoxin (liver RNase) , ECP, Rana cataesbiana and Rana pipiens egg RNase) and from plants, fungi and viruses (e.g. RNase of Nicotiana alata, Aspergillus giganteus,

Rhizopus niveus, pestiviruses) . These RISBASES are thought to exert their effects through an RNase catalytic site. A clear cytocidal example is that of S-allele-specific death of male cells in N. alata . BS-R ase is an effective immunosupressive agent, thought to exert its action through its ability to catalyze the cleavage of RΝA of specific cells, with its dimeric structure playing a key role. R . pipiens RNase exerts an antiproliverative (cytotoxic) activity against numerous cancer cell lines. Several studies have shown direct functional relationships between RNase activity and other biological effects of such an enzyme: modifying the RNase active site was found to decrease or even abolish all biological effects (Sorrentino et al . J. Biol . Chem. 267, 14859-14865; Shapiro et al . Biochemistry 28, 7401-7408, 1989; Di-Donato et al . Protein Sci. 4, 1470-1477, 1995). Pestiviruses belong to the family of the Flaviviridae which consists of three genera: flavivirus, pestivirus and hepatitis C virus. Flaviviridae are small, enveloped viruses with a diameter of 40-60 nm. The genome is a single positive stranded RNA molecule of 12.5-16.5 kb . The genomes of flaviviruses and hepatitis C virus encode for two envelope associated glycoproteins : El and E2^7,23. Pestiviruses contain, additionally to El and E2 , an extra envelope glycoprotein, designated E^{rns 4}' ^2S. The viral function of glycoprotein E"¹⁹ , which elicits neutralising antibodies and is secreted in considerable amounts into the extracellular environment¹⁹, is as yet not clear. Its use in a vaccine has been suggested. Recently, E^ns was identified by comparing and aligning with RISBASE amino acid sequences and functionally as having an RNase catalytic site, and thus as an RNase^{10, 20} ; specific for uridine and inhibitable by zinc ions. Classical functions of RNases are the processing of RNA in gene expression and the digestion of dietary RNA. RNases can also exert cytotoxic activities and can be broadly involved in e.g. antitumor and immunomodulatory processes^{1,17, 24}. For pestiviruses, it was recently shown²⁹, that death of virus-infected cells was induced by cytopathic (but not by non-cytopathic) virus via apoptosis. Cells undergoing apoptosis show certain morphological changes (cell shrinkage, pronounced cell surface swelling and chromatin condensation) and DNA fragmentation to oligonucleosomal size, which is taken as one of the biological hallmarks of apoptosis. At least 5-6 serologically different pestivirus types have been found, all with varying E^rns amino acid sequences . The E^rns protein is enzymatically cleaved from one large precursor polyprotein encoded by the genome of pestiviruses. The E^rns protein has formerly also been known under various other names, such as gp44/48, E0 , E2 , the first encoded glycoprotein or a minor glycoprotein of pestiviruses. The various amino-terminal and carboxy terminal and in between amino acid sequences of the E^rns protein of each and every pestivirus isolate can be aligned easily to further identify and characterize the specific sequence corresponding to the specific isolate. Pestivirus infections induce leucopenia and immunosuppression^{3, s}' ^{15, 16, 22}. Pestiviruses of all serotypes can induce a rapid and severe immunosuppression in various animals such as pigs and ruminants. Pestivirus infections have also been reported in humans, comprising human pestivirus isolates. The non-cytopathic classical swine fever virus (CSFV) induces a leucopenia, depresses secondary antibody responses, and significantly lowers lymphocyte responses to T-and B-cell mitogens and induces B-lymphocyte deficiency. In addition, antibodies against CSFV generally appear later than in other viral infections of pigs. At postmortem severe lesions in tissues constituting the immune system are generally observed. Bovine virus diarrhea virus (BVDV) infections in their various cytopathic and non- cytopathic phenotypes are associated with immunosuppression as well. Postnatal infections give rise to leucopenia and can result in a period of immunosuppression and the virus appears to be able to replicate in all of the major lymphocyte subpopulations. Prenatal infections can lead to the birth of immunotolerant healthy calves, which may succumb months later to mucosal disease, which is thought to be initiated by superinfection with a cytopathic (partially) homologous strain of BVDV. Cattle that die from mucosal disease show severe lymphocyte depletion, for example, destruction of the lymphoid tissue in the depletion at postmortem. It is generally thought that these immunosuppressive effects are caused by the pestivirus infecting the cells which relate to the immune system.

Materials and methods

Abbreviations: BVDV: bovine virus diarrhea virus; cf : control fraction; cpm: counts per minute; CSFV: classical swine fever virus; DME : Dulbecco's minimal essential medium; EBTr: embryonic bovine tracheal; EMEM: Earle's minimal essential medium; MAb: monoclonal antibody; ncpm: netto counts per minute; PBS: phosphate buffered saline; PBMC : peripheral blood mononuclear cells; SK: swine kidney.

E"¹⁸: production and purification

Purified E^rns was prepared as described previously¹⁰. Briefly, Spodoptera frugiperda cells were infected with recombinant baculovirus expressing glycoprotein E^rπs of CSFV strain C². Infected cells were incubated for 96 hours and then lysed in buffer containing 30 mM Tris-HCl, pH 7.5 , 10 mM MgCl₂ and 1% Nonidet-P40. The lysed cultures were centrifuged to remove cell debris and the supernatant was stored at - 20°C. E^rns was purified by immunoaffinity chromatography using MAb C5, directed against E^rns of CSFV strain C. The concentrations of protein in the fractions containing RNase activity were determined by extrapolation of the absorption at 280 nm on a bovine serum albumin standard curve. RNase- specific activity was assayed as described previously¹⁰ and expressed as A₂₆₀ units min^ting^"1. Control fractions were prepared from cells infected with wild-type baculovirus (cfl) and from cells infected with recombinant baculovirus expressing glycoprotein E2 of BVDV strain 4800 (cf2) . The control lysates were chromatographed over similar column material as the E^rns lysates. Hybridoma cells producing MAb C5 were kindly provided by Dr. G. Wensvoort and ascites fluid containing MAb C5 was produced as described by Wensvoort et al.²⁶.

Mutagenesis of the E"¹" gene by PCR.

In general, amino acid residues reported to be necessarily involved in an RNase catalytic site comprise histidine residues. Such histidines being part of a RNase catalytic site can be identified by comparing (aligning) amino acid sequences of one ribonucleases with amino acid sequences of another ribonuclease. Histidines being part of an RNase catalytic site were mutated to generate E^rns protein lacking RNase activity. In a first PCR reaction a part of the E^rns gene was amplified in a 35 cycles reaction with VentTm DNA polymerase (Biolabs) using a baculo transfervector containing wild-type E^rns gene as template. In this reaction, primers in which the histidine codon (CAT) was substituted for a lysine codon (AAA, underlined) were used as forward primers. Primer H>K(1), 35-mer, 5 ' -GG-GTT-AAC-AGA-AGC-TTG- AAA-GGG-ATC-TGGCCG-GGG-3 ' (nt 1245-1279 in the sequence of CSFV strain C, and primer H > K(2) , 24mer, 5 ' -GAA-TGG-AAC- AAA-AAA-GGA-TGG-TGT-3' (nt 1397-1421). A primer with a flanking BamHl site (39 mer, 5'-

ATAGTCGACGGATCCTTAGTACCCTATTTTCGTTGTCAC-3 ') was used as reverse primer. PCR fragments of correct size were isolated from an agarose gel and used in a second PCR reaction to recover the complete E^rns gene. In this reaction the non- coding DNA strand functions as reverse primer, and the wildtype E^rns gene with flanking BamHl sites, isolated from the above mentioned transfervector and re-cloned downstream of the bacteriophage T7 promotor in the BamHl site of pGem4z- bleu, was now used as template. In this reaction a T7 primer was used as forward primer. The crude PCR products were BamHl digested, and the 720 bp mutated E^rns genes were isolated from agarose gel and fused to the signal sequence of glycoprotein gG of pseudorabies virus in the baculo transfervector pAcAS3gX as described. A transfervector with an E^rns gene in which the histidines in both domains were substituted by lysines (H > K(l,2)), was constructed as described above, except that, in the first PCR reaction, the transfervector with the mutation in the second domain (H > K(2)) was used as template and the H > K(l) primer was used as forward primer.

Characterization of mutated ^Erns proteins expressed in insect cells.

Through chemical inactivation Kawate et. al showed that the histidine residues in both ribonuclease domains of RNase T2 , which are homologues to E^rπs, were essential for ribonuclease enzyme activity. To inactivate the RNase activity of E^rn, the histidine residue in the first and/or in the second ribonuclease domain was substituted by a lysine residue. In a two-step PCR reaction, the CAT codon at position 297 (H > K(l)) or the CAT codon at position 346

(H > K(2)) in the amino acid sequence of CSFV strain C was replaced by an AAA codon, or both codons were replaced (H > K(l,2)) . The mutagenesis for H > K(l) and H > K(l,2) is essentially the same, except that different primers or templates were used (see materials and methods) . After mutagenesis the 720 bp long mutated E^rns, genes were analysed by restriction enzyme digestion. Gel analyses showed that the mutated genes lacked the NIalll (CATG) site in the RNase domain that corresponded with the primer used for mutagenesis. The mutated E^rns genes were inserted into the plO locus of baculovirus as described. For all three mutated genes, E^rns expressing recombinant viruses were selected and plaque purified (Bac [H>K(1) ] , Bac[H>K(2)], and Bac [H>K(1, 2) ) virus) . The mutated E^rns proteins and the wt ^Erns protein, were purified by immunoaffinity chromatography from the lysate of Sf21 insect cells infected with these respective recombinant bacoluviruses , and immunologically, biochemically and functionally characterized. Compared to wt E^rns, the mutated proteins H > K(l), H > K(2) , and H > K(l,2) had no detectable RNase activity measured between pH 3.0 and 8.0 at 37 °C, and between 20 °C and 65 °C at pH 4.5. Analyses of the purified proteins on SDS-PAGE, under reducing and non-reducing conditions, showed that the mutated proteins had a similar mobility as wt E^rns and also were efficiently dimerized. They also reacted identical with monoclonal and polyclonal antibodies in a direct ELISA. Moreover, the mutated proteins were able to inhibit infection of SK6 cells with CSFV strain Brescia, and showed no functional differences with the wild type protein in the assays detecting immunosuppression or apoptosis. Substitution of the histidine residues in either of the two catalytic domains by lysine residues inactivated the RNase enzyme activity of the E^rns proteins. As reported for the fungal RNase T2 , the histidines in both domains are essential for ribonuclease activity. Because there is a space of 48 amino acids between both domains, strong interactions between amino acids residues of both domains are probably involved in formation the RNase active site of E^rπs . Moreover, the conformation of E^rns is probably dependent of the formation of this active site. To inactivate the RNase enzyme activity of E^rns, without destroying its conformation, the histidine residues were substituted by positively charged lysine residues .

Cloning and expressing E^røs protein or mutants or parts thereof .

Generation of primers.

The sequences encoding the boundaries of the E^ws gene are highly conserved among the various serotypes of the pestiviruses. This is demonstrated for the amino acid sequences in the alignment shown in figure 4. (Sequences are obtained from the EMBL GenBank) . Based on this high conservation, forward and reverse primers can be generated matching and hybridizing to these regions. Alternatively, a forward primer can be generated from published sequences of the 5 ' untranslated region (UTR) , located upstream from the

E^ms encoding region. A reverse primer can be generated from published sequences of the E2 gene, located downstream from the E^ms encoding region. After RT-PCR and sequencing, primers as described above can be generated from the determined nucleotide sequence.

Amplification of E™^s genes.

By use of these synthesized primers in a reverse transcriptase reaction followed by polymerase chain reaction (RT-PCR) , E™^s genes of pestiviruses can be amplified. Synthesized primers can be supplied by several companies, and one-tube RT-PCR protocols and kits are commercially available .

Cloning of E™^s genes. Subsequently, amplified E^ws fragments/genes can be cloned. Alternatively, using synthesized primers containing appropriate restriction sites, the amplified fragment can be directly cloned in expression plasmids, including prokaryotic, eukaryotic expression plasmids as well as yeast and baculovirus transfector plasmids. An example of the latter has been published by Hulst et al . , 1994.

Purification of E^*³ protein.

For recombinant baculovirus expressing E^ms, this protein can be purified by immuno-affinity chromatography in large amounts. Alternatively, the E^ gene can be fused to a "tag, for example six codons for Histidin. Using commercially available techniques, such fusion proteins can be purified and are fully active.

Deletions and site-directed mutagenesis of E™^s.

To map an active E¹^ protein (RNase activity, immunosuppressivity or other activities) , the E^ws protein can be mutated on different levels, including deletions and changes of specific amino acid residues. Defined deletions are introduced by use of common recombinant DNA technology in combination with PCR. Changes of specific amino acid residues are carried out by mutation of codons of these amino acids by several techniques, including site-directed mutagenisis kits and the PCR technology.

Assays for the activity of E™^s .

In vitro

Influence on the protein synthesis in different types of cells, (cell-specificity).

Influence on the proliferation of different types of lymphocytes (species, but also subsets of lymphocytes). Influence on apoptosis . In vivo - Influence on immune system Influence on other cells

Isolation of lymphocytes and the lymphocyte proliferation assay Blood of specific-pathogen-free animals housed in isolation units and of humans was used for isolation of lymphocytes. From each species the same animal or person was blood sampled for each successive experiment. Blood (10 ml) was collected in heparin coated tubes, diluted 1:1 in PBS, layered onto 5 ml Ficoll Paque (Pharmacia Biotech) and centrifuged for 20 minutes at 20°C. Human and porcine blood were centrifuged at 500*g, ovine blood at 900*g and bovine blood at 1400*g. PBMC at the interface were collected, resuspended in 10 ml PBS and washed twice. They were counted with a microcell counter (Sysmex, F-800) and suspended at a concentration of 4 x 10⁶ cells per ml in RPMI containing 10% fetal calf serum, 0.5 % antibiotic cocktail and 5 x 10^"s M 2- mercapto-ethanol . The fetal calf serum was BVDV and BVDV- antibody negative. The antibiotic cocktail consisted of penicillin 10.000 IU/ml , streptomycin 11250 μg/ml, kanamycin 10.000 μg/ml and nystatin 5000 IU/ml. Each culture of the lymphocyte proliferation assays was set up in triplicate and assays were performed at least twice. To each well of round bottom micro titre plates, containing 25 μl of cell suspension (10⁵ PBMC) , 25 μl of a dilution of E^rns, a similar dilution of cfl, cf2 or cell culture medium was added. _.Plates were incubated for one hour at 37°C in a 5% C0₂, humidified atmosphere. Then 50 μl concanavalin A (type IV-S, Sigma) in RPMI was added to a final concentration of 5 μg per ml. Plates were cultured 72 hours at 37°C in a 5% C0₂ humidified atmosphere and then pulsed with 0.4 μCi of [³H] -- thymidine in 10 μl RPMI per well. After 16 additional hours, the cells were harvested with an automatic 96-wells harvester (Tomtec) onto glass fiber filters. Filters were dried for 1 hour at 60 °C and radioactivity was measured in a liquid scintillation counter (Wallac; 1450, micro beta plus). The results are expressed as mean ncpm of the different assays and ncpm = cpm of the triplicate stimulated cultures minus the mean cpm of the non-stimulated cultures. To confirm the specificity of the inhibition by E^rns, a blocking assay with neutralizing MAb C5 was performed. E^rn and MAb C5 were incubated for 1 hour at 37°C in a humidified atmosphere before use in the lymphocyte proliferation assay. An anti-bovine herpesvirus 1 MAb, MAb 41, was used as control .

Cell viability

Cell viability after the concanavalin A induced lymphocyte proliferation was determined with nigrosine staining. Aliquots of cells were mixed with an equal volume of 0.2% nigrosine in PBS. Tests were performed in triplicate and the percentage of viable cells was determined microscopically . Protein synthesis assay

Inhibition of protein synthesis was used to assay the direct cytotoxic action of E^{rns 12}.

PBMC were isolated as described above, washed twice and resuspended in leucine-free RPMI, without fetal calf serum and antibiotic cocktail, at a concentration of 2 x 10⁶ cells per ml. To each well of a round bottom micro titre plate, containing 50 μl cell suspension (10^s PBMC) , 50 μl of dilutions of E^rns similar dilutions of cf2 or leucine free RPMI were added. Plates were incubated in 5% C0₂ humidified atmosphere for 1-24 hours and then 10 μl RPMI containing 0.4 μCi [³H] -leucine was added. Incubation was continued for 1 hour. Cells were harvested on DEAE filters using an automatic 96 wells harvester (Tomtec) . Filters were dried for 1 hour at 60°C and radioactivity was measured in a liquid scintillation counter (Wallac, 1450, microbeta plus). Tests were performed in duplicate. The inhibition of the protein synthesis of EBTr and SK6¹³ by E^mB was also studied. These semi -continuous adherent cell types were cultured in 165 cm² cell culture flasks (Costar) , in EMEM, 10% normal calf serum and 0,5% antibiotic cocktail. The normal calf serum was collected from specific pathogen free calves and was BVDV and BVDV antibody free. For the protein synthesis assay the adherent cells were trypsinized, washed in cell culture medium, resuspended and 5 x 10⁶ cells per well were dispersed in 96-wells flat bottom micro-titre plates. After 24 hours at 37°C, 5% C0₂ in regular cell culture medium, the attached cells were washed twice with leucine-free EMEM. To each well 100 μl of 0.4 mM of E^rns, a similar dilution of cf2 or leucine-free EMEM was added. Plates were incubated at 37°C in a 5% C0₂ humidified atmosphere. After a defined incubation time 10 μl PBS containing 0.4 μCi [³H] -leucine was added and the incubation was continued for 2 hours. Then the cells were washed, trypsinized and harvested on DEAE filters. Filters were dried for one hour at 60 °C and counted in a liquid scintillation counter (Wallac, 1450, microbeta plus). The protein synthesis after incubation of the cells with Eⁿ³ or the control fractions is expressed as a percentage of the protein synthesis of the cells in leucine free cell culture medium.

Detection of apoptosis

Apoptosis of bovine lymphocytes was determined with the cell death detection ELISA (Boehringer Mannheim, cat.no 1544 675) , which is based on the specific detection of mono- and oligo-nucleosomes in the cytoplasmic fraction of apoptotic cells. The apoptosis in 100 μl cell suspension (10^s PBMC) was determined after 8 hours incubation with 0.4 μM E^rπs, a similar dilution of the recombinant baculovirus control fraction or RPMI 1640. Similarly, apoptosis of EBTr cells was measured, except that these cells were trypsinized, because they were attached to flat bottomed micro titre plates. The results are expressed as the enrichment factor which represents the enrichment of mono- and oligo-nucleosomes in the apoptotic cells and is the quotient of the average absorbance of the test sample and the average absorbance of an untreated sample.

Studies of E¹"⁸ as immuno-suppressive agent at the level of lymphocytes, endothelial cells and enterocytes, and its role in haemorrhagic aspects, and vasculitis.

Test immunosuppressive agent: Experiments Mice

1. Indication dose range: 5 mice/group, 5 fold dose difference.

Start 10, 50, 100, 200 microgram/day, 3 days, blood samples lymphopenia day 5, organ weight (thymus, spleen) Increase doses until animals are depleted of lymphocytes, organ weight or die. 2. Survival LD50

For each dose 10 animals; of all moribund animals section, pathology: all lymphoid tissues (include spleen, bone marrow, lymph nodes (including mesenterial lymph nodes) , GI-tract mucosa, lung mucosa, endothelial sites

At day 1, 2, 4, 6, 9, 12 after application: section of 4 animals for each dose Look for endothelial damage and lymphocyte depletion, follicle formation.

3. Reduction of antibody responses in simple model: Use doses with established lymphopenia Ficoll-TNP, KLH-, HSA- , CGG- or Ova-TNP (titre, cytochemistry T (4/8 thymus) /B/macrophages , cytokines, endothelial cell markers) DTH to SRBC (according to Vissinga et al . , 1987, use C57B1 mice). Skin graft (black to cba, black to balb/c) .

Demonstration of immuno-suppressive potential of E^TOS

Demonstration of efficacy in vivo

1. Immune suppression measured as reduction of circulating lymphocytes. Adult (8-10 wk old ) male and female mice (Balb/c, C57B1/, CBA) were treated with increasing doses of E^rns during three consecutive days. The number of leucocytes in peripheral blood was determined at various days after administration. To determine the effect of E^rns on T- and B- lymphocytes the frequency of CD4 , CD8 and IgM surface marker positive cells was determined in peripheral blood leucocytes using FACS-analysis . In addition organ weight was determined for thymus and spleen being indicative for cell-loss from the leucocyte compartment of the immune system, indicating a clear dose-response relationship.

2. To demonstrate that the immunosuppressive effect of E^r dependent on the doses administered is not compatible with life LD50 survival experiments were performed. Adult (8-10 wk old ) male and female mice (Balb/c, C57B1) were treated with selected doses and survival was determined. Histopathological analysis of the animals at the level of primary and secondary lymphoid organs revealed that no detrimental effects were seen.

3. Immunosuppressive effects of E^rns subsequently were demonstrated in functional immuno-assays relevant for potential application fields.

A. Specific antibody responses to antigens administered simultaneous or following administration of E^rns allows to conclude at which level of effector cells E^rns is most effective. Antibody responses can be distinguished in two types: T cell independent (TI) antibody responses for which polysaccharides are good models and T-cell dependent (TD) antibody responses for which proteins are useful models. For T-independent responses B-cells and antigen presenting cells are mainly responsible. In T-dependent immune responses T- cells are involved as regulator cells while B-cells are effector cells.

Mice (adult, 8-10 wk old, male and female Balb/c and C57B1 mice) were immunised with Ficoll-TNP (TI) or Ova-TNP (TD) simultaneous with or without application of the E^rns immunosuppression protocol. The doses of antigen were chosen such that suboptimal responses were expected to allow evaluation of immuno-modulation effects. In the mice antibody responses were determined at days 7 and 14 after administration of antigen. In the peripheral blood was drawn at various days after application of the immunosuppression protocol to establish the reduction in frequencies of CD4 , CD8 and IgM cells as well as the specific antibody titre. Antibody titres were determined in ELISA. B. Delayed type hvpersensitivitv is regarded as a measure for T-cell activity which regulates leucocyte effector cells. This type of immune response is thought to be to a large extent responsible for graft rejection. As a consequence suppression of DTH activity is a model to prospectively evaluate the capacity of an immunosuppressive agent in vivo. DTH responses can be performed with for protein antigen or complex corpuscular antigens. As the protein antigens are already evaluated in antibody responses we preferred the mouse response to sheep red blood cells (SRBC) to determine the effect of E^rns on DTH responsiveness (Vissinga et al . , 1987) .

C. Immunosuppression is best demonstrated in a transplantation model for which immunosuppressive agents are mostly wanted. A direct and well accepted transplantation model is skin transplantation. Well defined skin transplants are performed between two mouse strains which fully differ in major histocompatibility antigens (H2) such as C57B1 (H2b) to Balb/c (H2d) and vice versa. Skin sections are transplanted from the donor (H2b) to recipient (H2d) . Simultaneous in the same recipient a sham transplantation is performed with the host skin. The take of the graft is monitored after transplantation. Mice treated with an immunosuppression protocol will allow grafting whereas in non-suppressed mice a rejection is monitored. The delay in rejection is a measure for immunosuppression.

Results

Immunosuppressive effect of E^ras

We determined the immunosuppressive effect of pestivirus envelope glycoprotein E^rns. Immunoaffinity purified E^rns was found to have a strong inhibitory (cytotoxic) effect on concanavalin A induced proliferation of bovine lymphocytes. Incubation of the lymphocytes with a concentration of 0.4 mM of E^rns completely inhibited the ³H-thymidine incorporation and even 0.01 mM E^rns resulted in a slight but significant inhibition. The specificity of the immunosuppressive effect on the lymphocyte proliferation was confirmed in two different ways. To exclude the possibility that contaminants in the purified E^rns fraction caused the inhibitory effect, cfl was tested in similar dilutions as E^rns in the lymphocyte proliferation assays. Cfl did not inhibit the lymphocyte proliferation (figure 1) . Additionally, blocking assays were performed with MAb C5 , which is neutralizing and directed against E^rns of CSFV strain C. The immunosuppressive effect of E^rns could be blocked partially by MAb C5. Incubation of 0,2 mM E^rns with MAb C5 (0,4 mg/ml) before adding to the lymphocytes, resulted in an increase of the proliferation from 1% to 35% of the maximum proliferation. Incubation of E^rπs with anti-bovine herpesvirus 1 MAb 41 (0,5 mg/ml) did not result in an increase of the lymphocyte proliferation. These results of the lymphocyte proliferation assays show that E^rns is immunosuppressive for bovine lymphocytes. The immunosuppressive effect of E^rns was then studied in proliferation assays with human, porcine and ovine lymphocytes. Proliferation of lymphocytes of these species was also completely inhibited by 0.4 mM E^rns. From these results we conclude that E^rns is cytotoxic for lymphocytes of species in which pestiviruses normally replicate as well as for lymphocytes of species in which pestiviruses normally do not replicate.

Cytotoxic action of E^rns

The viability of bovine lymphocytes during the proliferation assay was studied by nigrosine staining of aliquots of cells. At the start of the assay all wells contained more than 90% viable lymphocytes. After 4 hours incubation the control wells contained 93% and the wells incubated with 0.2 and 0.4 mM E^rns contained 85% viable lymphocytes. After 3 days incubation the control wells still contained 93% and the wells incubated with 0.2 mM E^rns 71% viable lymphocytes. None of the lymphocytes in the wells incubated with 0.4 mM of E^rns were viable at that time (figure 2) . The nigrosine staining showed death of lymphocytes after E^rns incubation, which again demonstrates cytotoxic effect of E^rns. The kinetic of the protein synthesis of lymphocytes was used to quantify the cytotoxic effect. Firstly, the effect of E^rns on the protein synthesis of bovine lymphocytes was determined. A dose dependent inhibition of the protein synthesis was observed, that became apparent as soon as after 4 hours incubation. Incubation with 0.4 mM E^rns resulted after 8 hours in a decrease of the protein synthesis of 41% and after 24 hours of 63%. Concentrations of 0.1 and 0.025 mM also resulted in a decrease of the protein synthesis at all measured time points. A concentration of 0.006 mM E^rns significantly enhanced instead of inhibited the protein synthesis at all time points (figure 3) . When comparing in a separate experiment, performed in duplicate, the protein synthesis inhibitory activity of E^rns, the RNase mutant

E^rnsHis, and a control fraction, it was found that both the wild-type and the mutant could induce comparable cytotoxicity, percentages found were: E^rns 40, 31; E^rnsHis 50, 39; control 96, 108. This E^rns induced inhibition of the protein synthesis was not specific for bovine lymphocytes.

The protein synthesis of human, porcine and ovine lymphocytes was measured after 8 hours incubation with 0.4 mM E^rns and these results were similar to the results with the bovine lymphocytes. After investigation of the species specificity we also studied the cell type specificity of the cytotoxic effect of E^rns. We selected 2 epithelial cell lines in which pestiviruses normally replicate and measured the protein synthesis after 8 hours incubation with 0.4 mM E^rns. In contrast to the results obtained with the lymphocytes we did not detect an inhibition of the protein synthesis in these cell types. In EBTr cells a decrease of 5% and in SK6 cells an increase of 8% was measured. Incubation with cf2 did not affect the protein synthesis. These results indicate that E^rns is selectively cytotoxic for lymphocytes of different species and not for epithelial cells. Apoptosis

Virus induced apoptosis of infected cells is described for several virus species, including pestiviruses²⁹. We showed that purified envelope glycoprotein E^rns induced apoptosis in lymphocytes. Lymphocytes were incubated for 8 hours with E^rns and then tested in the cell death detection ELISA. After incubation with E^rns the amount of mono- and oligo-nucleosomes in the lymphocytes, which indicates apoptosis of the cells, increased. The average enrichment factor after E^rns incubation compared to incubation in regular cell culture medium was 3.9 ± 1.2. The control fraction cf2 did not induce apoptosis as shown by an enrichment factor of 1.0 + 0.1. When comparing in a separate experiment, performed in triplicate, the apoptotic activity of E^rπs, the RN-ase mutant E^rnsHis, and a control fraction, it was found that both the wild-type and the mutant could induce comparable apotosis, enrichment factor values found were: E^rns 4.3, 4.8, 2.4; E^rnaHis 4.3, 6.4, 4.2; control 0.9, 1.1, 0.8. Although it is described that cythopathic effect of pestiviruses is mediated by apoptosis, we did not detect apoptosis of EBTr cells after E^rns incubation. The earlier described cytophatic effect of pestivruses can thus not be contributed to the action of the E^rns protein observed in vi tro . The enrichment factor for mono- and oligo-nucleosomes in these cells was 1.1 ± 0.02. In this study we investigated the immunomodulatory properties E^rns . E^rns inhibited concanavalin A induced proliferation of bovine, porcine, ovine and human lymphocytes very efficiently. The reduction of the ³H-thymidine incorporation was dependent on the concentration of E^rns in the lymphocyte cultures and could be partially blocked by a neutralizing MAb directed against E^rns. The specific inhibition was not dependend on the RNA-se activity of the protein, a RNA-se negative mutant protein also inhibited the lymfocyte proliferation. Nigrosine staining of aliquots of stimulated lymphocytes, which indicates damage of the cell membrane, showed cell death after E^rπs incubation. The percentage of dead lymphocytes was dependent on the concentration of E^rns and on the incubation time. To quantify the direct cytotoxic effect of E^rns on the lymphocytes we performed the protein synthesis assay. After 4 hours incubation the inhibition of the protein synthesis was significant for all studied concentrations of E^rns, except for the lowest concentration tested. This concentration had a stimulatory effect. After 4 hours incubation the nigrosine staining did not yet show a significant increase in the percentage of dead cells. Thus, E^rns induces a significant inhibition of the protein synthesis of lymphocytes before the cell membrane becomes permeable for nigrosine. Therefore, the cytotoxic effect of E^rns is probably not caused by direct damage of the cell membrane. The hypothesis that E^rns does not damage the cell membrane directly, is further supported by the fact that a concentration of 0.006 mM E^rns enhanced instead of inhibited the protein synthesis.

No species specificity was detected for the cytotoxic effect of E^rns against lymphocytes. The inhibition of the lymphocyte proliferation and of the protein synthesis were similar for all tested species, including humans. It was rather unexpected that this viral protein had such a strong immunosuppressive effect on human lymphocytes because pestiviruses normally do not often infect humans. Viruses that induce immunosuppression in humans carry a protein with a similar function as E^rns . Pestiviruses preferably replicate in lymphoid tissue but also replicate in epithelial cells. We used SK6 and EBTr cells to investigate possible cytotoxic properties of E^rπs against epithelial cells. No cytotoxic action was detected against the epithelial cells. Hence, the cytotoxic action of CSFV E^rns in vitro is specifically directed against lymphocytes and is not restricted to cells of the natural host. Pestivirus envelope glycoprotein E^rns is defined as a RNase and the catalytically important amino acid residues are highly conserved within the pestivirus genus^10,20. The function of E^rns, which is secreted from infected cells¹⁹, in the viral life cycle is as yet not clear. From other RNases is known that they can have, additionally to the ribonucleolytic activity, different non-catalytic biological actions. They have been shown to be involved in anti-tumour, neurotoxic, helmintotoxic and immunomodulatory processes^{1, 14,} ¹⁷²⁴. A specific cytotoxic action of E^rns against lymphocytes is consistent with the pathogenesis of pestivirus infections. Acute virulent and persistent CSFV infections are characterized by severe leukopenia and immunosuppression^{16, 22}. In other pestivirus infections immunosuppression is also a common phenomenon^{3, 5}' ¹⁵. Immune deficiency caused by apoptosis of infected lymphocytes is described previously for a number of viruses such as human immunodeficiency virus⁷, influenza viruses ⁸ and feline leukemia virus¹⁸. Our results of the protein synthesis assay and the nigrosine staining however, show the induction of apoptosis in non- infected lymphocytes by glycoprotein E^rns of CSFV. E™⁸ can be in monomeric or multimeric form, be it unglycosylated or partly or fully attached with carbohydrate chains having similar activity. In the cell death detection ELISA was clearly shown that E^rns induced apoptosis in lymphocytes. To our knowledge this is the first time that selective apoptosis of lymphocytes induced by a purified viral glycoprotein is described. In our experiments E^rns induced apoptosis in lymphocytes and not in epithelial cells. E^rns is involved in virus-cell interaction of several epithelial cells. It interacts with the cell surface of PK15, SK6 , fetal bovine epithelium, SF21 and CL626 cells, probably by binding to a common cell surface receptor. Since E^rns only slightly inhibited the protein synthesis of EBTr and even slightly enhanced the protein synthesis of SK6 cells we conclude that binding of E^rns to a non-immune system cell surface does not give a cytotoxic effect or induces apoptosis of such cells. After binding of an RNase to a cell surface several processes can occur. It can remain attached to the cell surface receptor without any effect on the cell . It can damage the cell membrane directly and give cell necrosis²⁸. After binding and activation of the cell surface receptor²⁷ the RNase could also be internalized by membrane penetration or endocytosis. Then, degradation of RNA could induce apoptosis. Schneider et . al . (1994) showed a pronounced substrate specificity of E^rns for uridine and reduced activities for RNA substrates with high amounts of secondary structure. This suggests a RNA substrate specificity of E^rns . For other RNases a substrate specificity is also described⁶. Selective internalization of E^rns by lymphoreticular cells or even by particular lymphocyte subsets might also explain selective induction of apoptosis, this apoptosis, however, can not be explained by the RNase activity of the E^rns, since RNase deficient mutants also induce apoptosis and reduce protein synthesis. Here we found that, contrary to what would have been suspected, partly or wholly abolishing the RNase activity of E^rns did not effect other biological activity of the protein. Manipulating an RNase thus makes it possible to selectively maintain or deplete various effects of the protein, thereby making it into a potent and selective agent for therapeutic use.

Legends to the figures

figure 1:

The effect of immuno-affinity purified E^rns on the concanavalin induced proliferation of bovine lymphocytes. Of the control fraction similar dilutions are used. Results are expressed as netto counts per minute (NCPM) : counts per minute of the stimulated cultures - counts per minute of the unstimulated control cultures. Values represent the average of triplicate samples and have a standard error of less than 3%.

figure 2 :

Viability of lymphocytes as determined by the nigrosine staining of stimulated lymphocytes after incubation with dilutions of E^rns .

figure 3 :

The protein synthesis of bovine lymphocytes in presence of specific dilutions of E^rns . Average results of triplicate samples (standard deviation < 7%) are expressed as counts per minute (CPM) .

figure 4 : Multiple comparison of amino acid sequences of the E^rns protein of several pestivirus strains belonging to various serotypes. Cleaving/processing positions of the N-terminal and C-terminal end of the protein is located between the amino acid residues indicated by *. Pestivirus strains ALD, GPE, C (Chinese vaccine strain) bre (Brescia) and alf

(Alfort) are CSFV strains. NAD (NADL) sdl , osl (Osloss) , and bvd27 are BVDV type 1 strains, bd78, BVD2 , (178003) and 5250 are BVDV type 2 strains, wis (Wisman) is a sofar uncharacterized strain. REFERENCES

1. D' Alessio, G.. 1993. New and cryptic biological messages from RNases. Trends Cell. Biol . 3:106-109

2. Bognar, K. and J. Meszaros. 1963. Experiences with lapinised hog cholera virus strain of decreased virulence. Acta Vet. Acad. Sci . Hung., 13: 429-438.

3. Bolin, S.R., A.W. McClurkin, and M.F. Coria. 1985. Effects of bovine viral diarrhea virus on the percentages of circulating B and T lymphocytes in cattle. Am. J. Vet. Res. 46 (4) : 884-886 4. Collet, M.S., R. Larson, S.K. Belzer and E. Petzel. 1988. Proteins encoded by bovine viral diarrhea virus : the genomic organisation of a pestivirus. Virology 165: 200-208.

5. Ellis, J.A., W.C. Davis, E.L. Belden and D.L. Pratt. 1988.

Flow Cytofluorimetric Analysis of Lymphocyte Subset Alterations in Cattle Infected with Bovine Viral Diarrhea Virus. Vet. Path. 25: 231-236

6. Endo, Y., P.W. Hubert and I.G. Wool. 1983. The ribonuclease activity of the cytotoxin a-sarcin. J.Biol.Chem. 258, 2662-2667. 7.Gougeon M. and L. Montagnier. 1993. Apoptosis in Aids.

Science, 260: 1269-1270.

8. Hinshaw, V.S., C.W. Olsen, N. Dybdahl-Sissiko and D. Evans. 1994. Apoptosis: a mechanism of killing by influenza A and B viruses. J. Virol. 68(6): 3667-3673. 9. Howard C.R. 1992. Hepatitis C virus: A comparison with Pestiviruses and other Flaviviruses . In Proceedings of the second symposium on pestiviruses, 1992. European Society for Veterinary Virology. 10. Hulst, M.M., G. Himes, E. Newbigin and R.J.M. Moormann. 1994. Glycoprotein E2 of Classical Swine Fever Virus:

Expression in Insect Cells and Identification as a Ribonuclease. Virology 200: 558-565.

11. Hulst, M.M. manuscript in preparation

12. Johnson, G.V., D. Wilson, L. Greenfield and R.J. Youle. 1988. The Role of the Diphteria Toxin Receptor in Cytosol Translation. J.Biol.Chem. 263 (3): 1295-1300.

13. Kasza, L., J.A. Shadduck and G.J. Christofinis . 1992. Establishment, viral susceptibility and biological characteristics of a swine kidney cell line SK-6. Res. Vet. Sci. 13: 46-51. 14. Maήcheno, J.M., M. Gasset, M. Onaderra, J.G. Gavilanes and G. D'Alessio. 1994. Bovine seminal ribonuclease destabilizes negatively charged membranes. Biochemical and biophysical research communications 199 (1) 119-124 15. Marshall, D.J., G.A. Perry, CA. Kuszynski, K.M. Eskridge and C.L. Kelling. 1994. Flow Cytometric Analysis of

Lymphocyte Subsets in Peripheral Blood and Lymphocyte Tissues of Gnotobiotic Calves during Primary Acute Postnatal Infections of Bovine Viral Diarrhea Virus. Viral Immunology 7(3): 141-149 16. Van der Molen, E.J. and J.T. van Oirschot. 1981. Congenital persistent swine fever (Hog Cholera) . I . Pathomorphological lesions in lymphoid tissues, kidney and adrenal. Zbl . Vet. Med. B, 28: 89-101.

17. Peterson, C.G.B., V. Skoog and P. Venge. 1986. Human eosinophil Cationic Proteins (ECP and EPX) and Their

Suppressive Effects on Lymphocyte Proliferation. Immunobiol . 171: 1-13.

18. Rojko, J.L., R.M. Fulton, L.J. Rezanka, L.L. Williams, E. Copelan, CM. Cheney, G.S. Reichel, J.C. Neil, L.E. Mathes, T.G. Fisher and M.W. Cloyd.1992. Lymphocytotoxic strains of feline leukemia virus induced apoptosis in feline T4-thymic lymphoma cells. J. Clinic. Invest. 66: 418-426

19. Rύ enapf, T., R. Stark, G. Meyers and H-J. Thiel . 1993. Structural proteins of hog cholera virus expressed by vaccinia virus: Further characterization and induction of protective immunity. J. Virol. 67: 589-597.

20. Schneider, R., G. Unger, R. Stark, E. Schneider-Scherzer and H.J. Thiel. 1993. Identification of a Structural Glycoprotein of an RNA Virus as a Ribonuclease. Science 261: 1169-1171.

21. Smith G.H., J.K. Collins, J. Carman and H.C Minocha. 1988. Detection of cytopathic and non-cytopathic bovine viral diarrhea virus with an immunoperoxidase test. J. Virol. Meth. 19: 319-324.

22. Susa, M., M. Kδnig, A. Saalmύller, M.J. Reddehase and H- J. hiel. 1992. Pathogenesis of Classical Swine Fever: B- lymphocyte Deficiency Caused by Hog Cholera Virus. J. Virol.66 (2): 1171-1175. 23. Takamizawa, A., C Mori, I. Fuke, S. Manabe, S.

Murakami, J. Fujita, E. Onishi, T. Andoh, I. Yoshida and H. Okayama. 1991. Structure and Organistion of the Hepatitis C Virus Genome Isolated from Human Carriers. J. Virol. 65(3): 1105-1113. 24. Tamburrini, M. , G. Scala, C Verde, M.R. Ruocco, S.

Parente and G. D'Alessio. 1990. Immunosuppressive activity of bovine seminal RNase on T-cell proliferation. Eur. J. Biochem. 190: 145-148.

25. Thiel, H-J., R. Stark, E. Weiland, T. Rύmenapf and G. Meyers. 1991. Hog cholera virus: Molecular composition of virions of a pestivirus. J.Virol. 65: 4705-4712.

26. Wensvoort, G., E.P. de luyver, J. Boonstra and C Terpstra. 1989. Production and characterization of monoclonal antibodies against the "Chinese" strain of hog cholera vaccine virus. Ph.D. thesis, University of Utrecht, The

Netherlands .

27. Wu,Y.N., S.M. Mikulski, W. Ardelt, S.M. Rybak and R.J.

Youle. 1993. A cytotoxic ribonuclease, study of the mechanism of onconase cytotoxicity. J.Biol.Chem. 268, 10686-10693 28. Young, J.D., CG.B Peterson, P. Venge and Z.A. Co n.

1986. Mechanism of membrane damage mediated by human eosinophil cationic protein. Nature, 321, 613-616.

29. Zhang, G., S. Aldridge, M.C Clarke and J.W. McCauly.

1996. Cell death induced by bovine viral diarrhoea virus is mediated by apoptosis. J. Gen. Virol. 77, 1677-1681.

Claims

I. A selective cytotoxic polypeptide or fragment thereof, derived from a protein having ribonuclease activity, said polypeptide (fragment) partly or wholly lacking said ribonuclease activity.

2. A polypeptide or fragment thereof according to claim 1, derived from a viral protein.

3. A polypeptide or fragment thereof according to claim 2, derived from a pestivirus E^rns protein.

4. A nucleic acid molecule encoding a polypeptide or fragment thereof according to any of claims 1 to 3.

5. A vector comprising a nucleic acid according to claim 4.

6. An expression system comprising a nucleic acid according to claim 4 or a vector according to claim 5.

7. A host cell comprising a nucleic acid according to claims 4 or a vector according to claim 5.

8. A pharmaceutical composition comprising a polypeptide or fragment thereof according to any of claim 1, 2 or 3 and further comprising a pharmaceutically acceptable substance.

9. A pharmaceutical composition not being a vaccine comprising a polypeptide (fragment) derived from a pestivirus E^rns protein and further comprising a pharmaceutically acceptable substance.

10. A polypeptide or fragment thereof according to any of claims 1, 2 or 3 or a composition according to claim 8 or 9 for use in therapy.

II. Use of a polypeptide or fragment thereof derived from a pestivirus E^rns protein for the manufacture of a medicament with a selective cytotoxic effect on lymphocytes.