NO306623B1 - Method of Preparation of DNA Molecule Containing Parts of a Gene Transcribed by Polymerase III and a DNA Sequence Encoding an Inhibitory RNA Molecule - Google Patents

Abstract

The invention relates to a genetic unit, which is optionally in multicopy form, for the inhibition of RNA. The unit contains the transcription units necessary for transcription by polymerase III, and a DNA which codes for inhibiting RNA and which is arranged within the unit in such a way that the transcribed RNA is part of the polymerase III transcript. These units can be used to achieve increased stability of the inhibiting RNA, which can be in the form of ribozymes or antisense RNAs, while there is no reduction in the activity. The invention furthermore relates to a process for introducing the genetic units into the cell, to the use of these units and to pharmaceutical formulations containing them.

Description

The present invention relates to a method for producing a DNA molecule containing parts of a gene that is transcribed by polymerase III and a DNA sequence that codes for an inhibitory RNA molecule.

The specific inhibition of genes by oligonucleotides, e.g. in the case of a therapeutically effective blocking of deregulated oncogenes or viral genes, depends on the ability of such complementary RNA or DNA sequences, so-called "antisense oligonucleotides", to hybridize with mRNA, processing signals or pre-mRNA, and on thus interrupting the information transfer from genes to proteins.

The application of antisense DNA causes a degradation of complementary target RNA by RNAse H cleavage of the resulting hybrid, the result being the irreversible destruction of the complementary RNA present.

In the case of the use of antisense RNA, a so-called "hybrid arrest" of translation or processing follows, whereby the RNA/RNA hybrids represent the structural obstacle.

It is believed that such hybrids accumulate in the cell; their further fate has not yet been investigated. According to our current knowledge, this mechanism must essentially be a reversible process. The use of antisense RNA molecules has the advantage that these molecules can either be synthesized in vitro and brought into the cells, or that the genes that code for them can be introduced into the cells, so that the inhibitory RNA can be produced inside the cell. However, it has so far not succeeded in bringing such genes into a form that enables the production of an effective amount of antisense RNA in the cell.

Quite recently, a third principle of RNA inhibition was discovered and made available for in vitro application.

This principle is based on the ability of the RNA molecules, the so-called "ribozymes", to recognize specific RNA sequences, bind to these and cleave them. It was deduced from the in vivo observed autocatalytic cleavage reactions for RNA molecules in plant viroids and satellite RNA.

By building on specific structural requirements for the ribozyme-catalyzed RNA cleavage, one can now construct de novo ribozymes that have an in trans endonuclease activity that is aligned to a specific target sequence. Since such ribozymes, of which the most thoroughly explored to date are called "hammerhead" ribozymes due to their structure, can act on numerous different sequences, the corresponding ribozyme can be "tailor-made" for virtually any RNA substrate. This makes the ribozymes interesting, extremely flexible instruments for inhibiting specific genes, and a promising alternative to antisense constructs, which have already been shown to have a potential therapeutic benefit.

One of the now known and previously most thoroughly explored ribozyme models comprises three structural domains; by building on this model, ribozymes against CAT mRNA have already been built with good results. (Haseloff et al., 1988; Uhlenbeck et al., 1987) :

The three domains include:

a) a highly conserved region of nucleotides flanking the cleavage site in the 5' direction. This usually involves the sequence GUC, although a modification also showed an essentially undiminished cleavage activity in GUA or GUU. Cleavage also appeared after the sequences CUC, to a lesser extent also for AUC and UUC (the requirements for efficient cleavage have not yet been fully elucidated). b) the highly conserved sequences contained in naturally occurring cleavage domains in ribozymes, which form a kind of base-paired stem; c) the regions that flank the cleavage site on both sides, and which guarantee the exact alignment of the ribozyme

in relation to the cleavage site and for the connection between substrate and enzyme (in the experiments carried out so far, 8 bases were chosen on each of the two sides).

The RNA enzyme can be constructed according to this model, and has already been shown to be suitable for efficient and specific cleavage of RNA sequences in vitro. (Haseloff et al., 1988).

More recently, additional types of autocatalytic RNA cleavage activity have been discovered that come into consideration for the targeted RNA inhibition. One of these models is the so-called "hairpin" ribozyme, whose active site is derived from the minus strand of the satellite RNA of tobacco ringspot virus (Hampel and Tritz, 1989). Additional self-cleaving RNA activities are associated with hepatitis delta virus (Kuo et al., 1988; Sharmeen et al., 1988; Wu et al., 1989) and with RNAseP (Altmann et al., 1988).

The experiments that preceded the work on the present invention served to compare the effect of antisense RNA, antisense DNA and ribozymes. These experiments were carried out using the snRNP U7-dependent histone pre-mRNA processing reaction, and even in a system in vitro that processes U7-dependent histone pre-mRNA (Mowry et al., 1987 Soldati et al., 1988 ). It was thereby found that antisense RNA represents the strongest inhibitor, whereby the inhibition is reversible. The inhibitory effect of antisense DNA and of ribozymes of the "hammerhead" type, both irreversible, was within the same order of magnitude, whereby complete inhibition sometimes required an approx. 1000-fold excess compared to substrate RNA.

Although the experiments had so far been carried out with ribozymes with naked RNA in protein-free systems, the preliminary experiments for the present invention were the first experiments that had shown that synthetically produced ribozymes that were aligned to a specific sequence also show cleaving action in a protein-containing environment. This fact provided the first hint of a potential in vivo application.

One of the limiting factors in the use of ribozymes for inhibiting the expression of specific genes must lie in the build-up of a ribozyme concentration sufficient to effect efficient switching off of a particular biological reaction; the reasons for this possibly lie, as with the use of antisense RNA, among other things in the much too low stability of RNA.

The present invention was based on the task of providing a system for the use of mRNA as an in vivo inhibitor for mRNA which removes the hitherto existing limitations in the use of RNA, as an effective concentration of inhibitory RNA is made available in the cell.

This task was solved according to the invention by a method for producing a DNA molecule, possibly present in multiple copies, containing parts of a gene that is transcribed by polymerase III and a DNA sequence that codes for an inhibitory RNA molecule characterized by the fact that it contains the necessary transcription units of a tRNA gene for transcription by polymerase III including the sequences that determine the secondary structure of the tRNA, and that the DNA sequence encoding an inhibitory RNA molecule selected from the group ribozyme or antisense RNA is so arranged within the DNA molecule , that the inhibitory RNA molecule is part of the transcript.

The solution to this task was based on the consideration that by introducing a gene that produces the inhibitory RNA, as opposed to the import of RNA as such, there should be a guarantee of a significant amplification of the RNA and thus a sufficient supply of RNA for inhibition of the biological reactions.

The inhibitory RNA can be any ribozyme or other antisense RNA. Fig. 1 shows a schematic tRNA ribozyme model.

In principle, one could imagine carrying out an efficient transport of RNA or the DNA sequence that codes for this, via viruses or viral vectors, e.g. retrovirus.

However, this system has some crucial disadvantages, such as mobilization of endogenous viruses, recombination with endogenous retroviruses, activation of endogenous genes by integration, limitation with regard to host organism and tissue type.

In contrast, carriers are therefore provided according to the invention which do not have these disadvantages.

The genes that are proposed as carrier genes for the RNA genes within the scope of the present invention have the following advantages: They have a compact structure, they can because their small size is more easily transported into the cell, they have a high transcription rate and their expression is not limited to certain tissues, but is expressed ubiquitously, i.e. in almost all cell types.

A further advantage of the polymerase III genes lies in the presence of a very strong termination signal for transcription. Thereby, the likelihood of a nearby cellular gene being activated in an unwanted way is significantly reduced.

The genes that are transcribed by polymerase III have the following distinctive features: These are genes whose promoter is not located upstream of the gene, but within the gene (Geiduschek et al., 1988). These essentially internal control regions for the binding of polymerase III have a discontinuous structure; they consist of two so-called "boxes" (the A-box and the B-box), which are essential for the identification of the transcription factors, as well as an intermediate gene section which is critical with respect to its length (Hofstetter et al., 1981). The length of this sequence in tRNA genes is 31-74 base pairs (Clarkson).

Representatives of these genes transcribed by polymerase III are the tRNA genes, 5S RNA and a number of other small nuclear and cytoplasmic RNA genes: 7SK, 7SL, M, U6 and 4, 5S RNA, as well as the adenovirus genes VAI and VA2 (Geiduschek et al., 1988). Common to these genes is their small size, their compact structure, their high transcription rate and their ubiquitous transcription.

It was surprisingly established that increased stability of the inhibitory RNA can be achieved by means of the genetic units according to the invention, without having to take into account a damage to its activity in terms of activity.

It was already shown by Jennings and Molloy, 1987, that a specific promoter recognized by polymerase III is suitable for directing the synthesis of antisense RNA. In addition, the XbaI/BamH1 fragment in the VAI gene, which contained the promoter, was cloned in the 5' direction in front of the antisense DNA, which corresponds to the usual principle when using polymerase II promoters. An effort was made to obtain a short transcript compared to polymerase II transcripts. The transcript has only a minor modification that allows for base pairing of the ends, to protect the ends from degradation by single-stranded exonucleases.

In contrast to this proposal, where only the promoter sequence of a specific polymerase III gene is used, (the range between Pos.-30 to 73, while the wild-type VAl gene extends to the terminator sequence at pos.+160 - + 200), according to the present invention, the sequences in the polymerase III gene are also purposefully used, which are decisive for the secondary structure of the transcript, in order to utilize these for stabilizing the inhibitory RNA sequences. In contrast to the previously described proposal, the gene sequence for the inhibitory RNA coding end within the gene transcribed by polymerase III is arranged in such a way that, according to the invention, there is a gene cassette. In addition to this, the carrier genes used within the scope of the present invention, in contrast to the viral VAl gene, are not toxic.

The genes that are transcribed by polymerase III can be used flexibly within the scope of the invention. Regarding their function as carrier genes for RNA-inhibiting sequences, the following criteria must be taken into account when selecting or modifying natural genes or when constructing artificial genes:

1) The A and B boxes are highly conserved.

2) A section of 5 to 7 T residues downstream from the B-box is responsible for the termination of transcription. 3) The distance between the A and B box cannot be increased arbitrarily, whereby the maximum is currently assumed at

about 9 0 bp.

4) In some systems, the 5'-flanking sequence has an influence on transcription. 5) There are indications that an intact anticodon stem region is essential for the stability of the transcript.

Particularly suitable as carrier genes for the production of DNA molecules according to the invention are tRNA genes. The RNA encoded by these genes is also, due to its cloverleaf structure, considered particularly suitable for providing stability to the inhibitory RNA. With its help, compact RNA-producing gene units can be produced, as the sequence that codes for

the inhibitory RNA is inserted between the A and B block.

Experiments within the framework of the present invention were carried out with starting methionine tRNA which has an Apa I restriction site in the anticodon stem and loop region located between the A and B blocks (starting methionine tRNA in all higher eukaryotes has this restriction site in the anticodon loop).

In addition to the region between the A and B blocks, other insertion sites are also possible within the carrier DNA sequences (tRNA genes or other polymerase III genes), as long as one ensures that the transcriptional activity of polymerase III and the transcript's stability is preserved.

The DNA molecules produced according to the invention can still in principle be produced with all tRNAs; if necessary, a suitable restriction site can be constructed into which the sequence that codes for the inhibitory RNA is inserted. When inserting, care must only be taken that no base substitution is made in the anticodon stem region, which could affect the gene's transcription rate or the stability of the resulting tRNA (Folk et al., 1983). In addition, account must be taken of the fact that insertions larger than 60 bp can damage the transcription (Ciliberto et al., 1982).

When it comes to choosing suitable carrier genes, it is in principle worth striving to use species-specific tRNA genes or other genes that are transcribed by polymerase III, and which are not toxic.

The experiments within the scope of the present invention were carried out when ribozyme sequences were introduced into normal tRNA genes. When the tRNA ribozyme that is transcribed from this gene is predominantly located in the cytoplasm, while still being desirable for certain applications, e.g. to inhibit core-specific RNA, e.g. RNA from snRNP particles, and locate the tRNA ribozyme and the tRNA ribozymes, resp. tRNA-antisense-RNA in the nucleus, mutants that are predominantly located in the nucleus can be used, for example met in the tRNA gene described by Zasloff et al., 1982, and Zasloff 1983, which has a single base exchange.

The DNA molecules produced according to the invention can be produced e.g. as follows: A copy of the gene transcribed from polymerase III, e.g. a tRNA gene contained on a bacterial plasmid is cleaved with a suitable restriction enzyme at the site intended for insertion, e.g. between the A and B block, and the double-stranded DNA coding for the inhibitory RNA prepared according to standard methods, ligated. Thereby, suitable host organisms are transformed, selected, propagated and the amplified plasmid DNA recovered. The plasmids are examined for the presence of the genetic unit according to the invention; this can be done by restriction cleavage, sequence analysis or by detection of a functional transcript of plasmid DNA in vitro.

The thus produced genetic units can be used both in the form of the circular plasmid and also in the form of the gene unit excised from the plasmid which contains all the necessary information for transcription by polymerase III.

Which form is chosen usually depends on the area of application and on the transport system chosen for the introduction of the gene unit into the cell.

The DNA molecules produced according to the invention can also be present as multiplied copies (thereby it is a matter of tandem structures, in which the genetic units are arranged several times one after the other, whereby the inhibitory inserts can be the same or different).

Transcription of such a tandem gives, due to the promoter and termination signals contained in each individual unit, separate RNA units. Due to this property of the fragments to exist alternately as complete transcription units, the orientation of the individual units to each other is also irrelevant. When producing such tandems, the starting point is plasmids that contain multiple copies of the region that codes for the inhibitory units. Multimeric vectors containing tRNA ribozyme genes can be prepared, for example, starting from that in e.g. 1. described erbBsnit series, the single fragments are ligated using T4 DNA ligase into polymers, and the polymeric genes are re-cloned into a corresponding restriction site in a suitable plasmid vector. The production of such tandems is made possible by the small size of the units produced according to the invention, which preferably amount to 200 to 350 bp. With the help of such a "tandem inhibitor", one can, after producing a heteromeric complex of antisense RNA or ribozymes, test the effectiveness of a batch of inhibitors in a single experiment. For example, a mixture of 5 to 10 ribozymes can be introduced in this way into a cell system. After an inhibition has been achieved by means of the mixture, the individual ribozymes can be examined for their effectiveness. This is an advantage especially due to the fact that the criteria for selection of target RNA sequences for inhibition are not yet all unexplored.

Good target sequences are to be considered e.g. such regions that have no secondary structure, regions close to cleavage signals, regions after the initiation codon, regions without binding sites for specific proteins (such as Sm binding sites in snRNA molecules). With the help of the present invention, especially when used in tandem form, it becomes possible to obtain information about such target sequences, and then to inhibit them effectively.

The use of tandem units is also an advantage when it is necessary for the inactivation of biological processes to cleave RNA, e.g. viral RNA, in multiple locations. A multiple cleavage can be achieved e.g. in that a larger number of different ribozymes are constructed, and a vector containing the genetic units with the relevant DNA sequences that code for them is introduced into the cells. During transcription of these sequences, the ribozyme units will be stored on the corresponding target sequences, whereby the mRNA is cleaved into fragments.

Multimeric copies can be used when greater amounts of a low-molecular-weight ribozyme are required for efficiency in certain applications, for example when using the transferrin-polycation transport system. Plasmids carrying multiple copies of the genetic unit containing the DNA sequence coding for this ribozyme improve the production of the fragment by increasing the yield several times.

Tandem units can also be used advantageously when inhibitory RNA is to be provided at the same time, and which are directed against different RNA types.

When applying the invention to antisense RNA, care must be taken that in the case of tRNA genes, the size of the insert must be limited. When it comes to efficient transcription, the order of magnitude is approx. 60 bp; therefore, when using larger antisense RNA constructs, other polymerase III-transcribed genes, which have a greater capacity for transcription of longer sequences, must be taken into account.

The carrier genes that can be used within the scope of the invention can also be produced synthetically, if they meet the condition that they have the transcription units necessary for transcription, with the help of polymerase III. The use of such synthetic genes can bring with it the following advantages: a) Such genes are not recognized by aminoacyl synthetases and by ribosomes, whereby an intervention in the cell's translation machinery is prevented. b) There is a possibility for, by forming a synthetic construction, to produce inhibitory RNA with larger

stability and higher transcription rate than with et

natural gene.

c) The cloning process can be designed more flexibly by creating synthetic sequences.

Within the scope of the present invention, it was established that the stability of the ribozyme tRNA molecules can be increased by lengthening the anticodon stem region in the carrier tDNA molecule. It could be shown that by lengthening the 5-base-paired anticodon stem region of the wild-type tRNA gene to 9-base pairs, it is possible to obtain a 6-fold amount of transcript compared to the shortened wild-type of the tRNA gene, which can is attributed to an increase in stability and the workability that can be achieved thereby.

It was further established that an increase in the stability of the DNA molecules produced according to the invention in the form of tRNA ribozyme genes, or tRNA antisense genes can also be achieved when the DNA sequences coding for the inhibitory RNA function are present as part of introns (in the case ribozymes designated as "ribintrons"). This was based on the assumption that naturally occurring introns do not change the secondary structure of tRNA precursor molecules and that tRNA introns do not show any sequence conservation, whereby the structural change in the resulting tRNA precursor can be minimized and its stability thereby maximized when cloning ribozyme or antisense -sequences as part of introns. With the help of the tyr-tRNA gene used within the scope of the present invention, whereby the splicing of the tRNA precursor molecules only takes place slowly, it was possible to show that the activity of rib-tRNA can be increased with good effect by using tDNA molecules containing "ribintron " sequences. Thereby, it is possible to attack RNA that is localized amplified in the cytoplasm. This system enables the incorporation of suitable structural elements in a simple way to increase the stability of the "ribintrons" compared to degradation by exon nucleases. This can be achieved e.g. by additional base pairings or by larger hairpin regions bordering the ribozyme sequences. In the case of modifications of this type, care must be taken that the structures which determine the splicing of the intron are retained.

The natural intron sequences can be modified, e.g. by inserting suitable restriction sites to enable the cloning of oligonucleotides, or, if they are not found in the naturally occurring gene, by inserting nucleotides that enable base pairing with the anticodon triplet to provide a further stabilizing structural feature.

It could be shown within the framework of the present invention that the expression of ribozymes as a component of introns essentially does not cause any damage to the tRNA secondary structure, whereby the resulting transcript can be accumulated in a higher concentration and processed correctly. If the intron released during the processing does not prove to be sufficiently stable, one can think of suitable structural characteristics which cause an increase in stability against exonuclease degradation.

To bring the DNA molecules produced according to the invention into the cell, several methods can be used: The standard method for introducing DNA into tissue culture cells uses the formation of a coprecipitate between DNA and calcium phosphate (Graham et al., 1973). The precipitate is added to cells that occupy a certain proportion, possibly by a pinocytosis process. A similar method uses a positively charged substance, DEAE-dextran, which facilitates the uptake of DNA into the cell. Methods have also been developed for the introduction of DNA into the cell by electroporation (thereby creating temporary pores with the help of a pulsating electric field)

(Langridge et al., 1987, Fromm et al., 1987). Microinjection techniques for introduction into large cells (Kressmann et al., 1980) and tissue culture cells (Pepperkok et al., 1988) are also applicable. However, these methods are only usable in the laboratory, respectively for in vitro applications. A synthetic cationic peptide (DOTMA) has recently been developed, which spontaneously forms liposomes with DNA, and thus facilitates transport into the cells (Felgner et al., 1987). In principle, as already mentioned, retroviral vectors are also suitable for transporting genetic material into the cells (Stewart et al., 1986, 1987); however, these systems have the disadvantages already stated.

A further transport mechanism relies on using "disarmed" toxins as transport agents. The transport methods used to date are all burdened with the shortcoming that they cannot transport sufficient inhibitory nucleic acid into the cell. With the help of the present invention, it is now possible due to the small size and compactness of the molecules, in terms of the transport capacity to achieve an increase in the number of active inhibitor units. Because of. of the small size and compact structure of the genetic units according to the invention, one can also after insignificant modifications, e.g. conjugation with cholesterol, lipophilic motions or nuclear localization peptides also completely forgo a transport system.

Within the scope of the present invention, it is preferred to use a soluble system for the transport that takes place via receptor-mediated endocytosis. In particular, a transferrin-polycation conjugate that is complexed with the DNA molecule produced according to the invention is therefore preferred; the complex is taken up via the transferrin receptor which is present on virtually all growing cells.

The areas of application for the DNA molecules produced according to the present invention are numerous: For example, transgenic animals can be produced which, due to the presence of the DNA molecule produced according to the invention in their genetic material, have intracellular immunity against viruses, e.g. foot-and-mouth disease virus, Newcastle disease virus, bovine papillomavirus, pseudorabies or infectious gastroenteritis. Similarly, intracellular immunity can also be produced in transgenic plants, e.g. against potato virus PVX.

Furthermore, the DNA molecules produced according to the invention can be introduced into somatic cells to use against pathogenic viruses, such as HIV or related retroviruses, targeted ribozymes or antisense RNA to combat these viral stimulants.

A further area of application lies in gene therapy using RNA construction with complementarity to oncogenes and other key genes, which control the growth and/or differentiation of cells. In such applications, it is important with the high specificity for the RNA inhibition which can be effectively achieved with the help of the present invention, and with whose help it is possible, e.g. to distinguish between proto-oncogene and oncogene transcripts.

Furthermore, the genetic units can be used in plants or animals to prevent the expression of specific genes, in order in this way to bring out desired properties.

The inhibitory effect of RNA can also be used when fighting diseases in such a way that the production of unwanted gene products is prevented, e.g. the production of the main plaque protein ("major plague protein") that occurs in Alzheimer's disease or of proteins that cause autoimmune diseases.

The present invention can also be used in those cases where a regulatory protein, which has an interaction with RNA, is to be switched off by embedding or adaptation of RNA.

With the help of the experiments carried out within the framework of the present invention, transcriptional activity of the tDNA ribozyme gene could be demonstrated. In addition, a tDNA ribozyme gene construct was produced, whereby a DNA sequence encoding a 53 bp long ribozyme directed against the snRNA U7 sequence was inserted into the ApaI restriction site between the A-box and the B-box in the starting methionine tDNA (A- and B- box are the two polymerase recognition sequences; transcription begins 15 bp upstream of the A box and ends in an oligo T sequence downstream of the B box). After microinjection of this gene, transcription was detected; the concentration of the tRNA-ribozyme hybrid was 10-20% of the concentration of tRNA, which was produced by a coinjected wild-type tRNA gene.

RNA molecules, synthesized in vitro from the tRNA ribozyme gene, cleave the target RNA at the expected site. The addition of the tRNA structure to the ribozyme sequence does not block ribozyme activity. tRNA ribozyme molecules, synthesized from genes that have been injected into oocytes, cleave the target RNA similarly at the expected sites and with the same efficiency as in vitro synthesized ribozymes without the additional tRNA structure. It proves that in vivo synthesis and processing of a tRNA ribozyme does not involve modifications that prevent the action of the ribozyme.

Within the scope of the present invention, the in vivo activity of a ribozyme was first demonstrated. In addition, the tDNA/ribozyme gene was injected together with radioactively labeled GTP into the nucleus of oocytes. After 8 hours of incubation, which was required for the ribozyme synthesis, the radioactively labeled substrate RNA (U7 RNA) was injected into the cytoplasm of the oocytes. After a further 2 hours, the nucleic acid from the oocytes was prepared: In the oocytes in which ribozyme synthesis had taken place, no remaining substrate RNA was detected. In contrast, substrate RNA was stable in the oocytes that had not been injected with the tDNA/-ribozyme gene or where the gene had not hit the nucleus.

It could also be shown within the framework of the present invention that the tDNA ribozymes produced according to the invention can inhibit the transforming effect of an oncogene. Using erythroid chicken cells, transformed with the erbB oncogene, the activity of a tRNA ribozyme was demonstrated in erythrocytes via the differentiation that occurred upon inhibition of erbB expression.

The effectiveness of the genetic units can also be checked by considering the resistance of mouse cells to infections, for example polyoma infections, after using genetic units produced according to the invention, which are directed against the virus (possibly against several regions), e.g. against the papilloma virus.

Example 1

Construction of tRNA ribozyme genes

a) Construction of pSPT18metl

The methionine initiator 1-tRNA gene from Xenopus. present at a

284 bp EcoRI fragment, which was cloned into pBR322 (Hinfl H-G fragment (Hofstetter et al., 1981, Tellford et al., 1979), was isolated by EcoRI digestion of the pBR322 vector, purified by gel electrophoresis (2% agarose/ TBE) and ligated into the EcoRl site in the bacterial plasmid pSPT18 (Boehringer Mannheim) so that a sense tRNA transcript was obtained by transcription of the plasmid with SP6 polymerase. For this, standard cloning methods described in (Maniatis ).The main advantage of the recloning of the tRNA gene in pSPT18 lies in the presence of SP6 and T7 RNA polymerase promoters which lie opposite each other on the plasmid. Through in vitro transcription, specific RNA transcripts can be obtained which either contain the tRNA ribozyme sequence or the thus complementary to the sequence (Melton et al., 1984). These transcripts are useful to test the cleavage activity of the RNA molecule or to demonstrate by an "RNase Protection Mapping" the presence of tRNA ribozymes in cell extracts are that express the tRNA ribozyme.

b) Construction of tRNA ribozyme genes

The tRNA gene on pSPT18metl was cleaved at the single ApaI site

in the anticodon stem and loop region (see Fig. 2). This figure shows plasmids containing sequences coding for tRNA ribozyme. pSPT18 metl contains the 284 bp EcoRl fragment, which carries the Xenopus laevis initiator tRNA gene. It is therefore about the G-H fragment (Hofstetter et al., 1981), cloned into the EcoRl site of the polylinker in pSPT18. The complementary oligonucleotides encoding ribozymes, which target U7snRNA (CD33 and both sequences in erbB mRNA (ES13, ES53), are shown. The cloning strategy used here led to the removal of the upstream ends in the Apal site in the tRNA gene. The proportion of the inserts coding for the ribozyme and the regions complementary to the target RNA (anti-U7, anti-erbB) are characterized, as well as A- and B-

the box, the section in the 5 T residue (termination signal) and the transcription initiation sites. The plasmid contains the ColEI origin of replication, an ampicillin resistance marker and the promoters for T7 and SP6 RNA polymerase.

To prepare double-stranded synthetic DNA oligonucleotides encoding the viroid cleavage sequence (Haseloff et al., 1988) flanked by the sequences complementary to the target mRNA, single-stranded oligonucleotides were first prepared according to standard methods (Applied Biosystems DNA synthesizer). Complementary oligonucleotides were phosphorylated, annealed and ligated into the ApaI-cleaved pSPT18met1 plasmid using standard methods (Maniatis). The ligation mixture was used to transform E.coli HP101, bacterial clones containing the new plasmid were isolated and the presence of active ribozyme sequences on the bacterial plasmid confirmed in two species: 1) RNA molecules originating from the in vitro SP6 transcription of cloned DNA -plasmids, were incubated with a radioactively labeled RNA containing the target sequence for the ribozyme, and tested for specific cleavage of the target RNA (see Fig. 4). 2) The presence of correctly inserted DNA sequences was confirmed by dideoxy DNA sequencing over the insertion site.

Fig. 4 shows the in vitro ribozyme activity of tRNA ribozymes. Plasmid DNA molecules, which were the B-cut tRNA ribozymes ES13 and ES53, were digested with PvuII and transcribed with SP6 RNA polymerase. This transcription produced 230 nucleotide long RNA molecules containing the tRNA ribozyme sequence and in addition the 5' and 3' flanking sequences derived from the flanking Xenopus sequences and the flanking bacterial plasmid sequences. (See Fig. 2). The ribozyme transcripts were incubated with 20,000 cpm (20 fM) of an RNA molecule with the region of the erbB mRNA, including the initiation codon. The RNA molecule contains the target sequences for both ES13 and ES53 (see Fig. 3). After 2 h incubation of the ribozyme plus target RNA at 37°C in the presence of 10 mM MgCl2, 20 mM TRIS-HCl, pH 7.5 and 150 mM NaCl, EDTA was added to a concentration of 15 mM, the sample was dried, dissolved in 80% formamide/TBE, heated for 30 sec. at 95°C and separated on a 9.5% acrylamide/8.3 M urea/TBE gel. After electrophoresis, the labeled RNA molecules were detected by auto-radiography.

Lane M: Molecular weight marker: pBR322 DNA, digested with Hpall and radioactively labeled with alpha-32p-CTP using the Klenow fragment in DNA polymerase (Maniatis). Immediately before application to the gel, the molecular weight marker was dissolved in 80% formamide/TBE and heated for 3 minutes at 95°C. The molecular weights of some of the fragments (in nucleotides) are indicated on the left.

Lane 1: erbB target mRNA (20,000 cpm, 20 fM) without incubation

Lane 3: erbB target mRNA (20,000 cpm, 20 fM), incubated with MgCl2 at 37°C without ribozymes.

Lane 4: erbB target mRNA (20,000 cpm, 20 fM), incubated with ES13 RNA

(1 dc)

Lane 5: erbB target mRNA (20,000 cpm, 20 fM), incubated with ES53 RNA

The molecular weights (in nucleotides) of the erbB target mRNA and the 5' and 3' cleavage products from both ribozyme cleavage reactions are indicated on the left of the figure. (see also fig. 3).

Fig. 3 shows the complementarity between ribozymes and target RNA: CD33 (A) against U7snRNA (Cotten et al., 1988), ES13 (C) and ES53 (B) against sequences in erbBmRNA (Venustrom et al., 1980).

The tRNA portion of the tRNA ribozymes is not shown for clarity. The cleavage sites for the ribozyme are indicated, as are the initiation codons in the erbB mRNA.

Example 2

tRNA ribozyme transcription in Xenopus oocytes

The transcription of the tRNA ribozyme genes, microinjected into Xenopus oocytes, was carried out according to the method described for tDNA in (Kressman et al., 1978, Hofstetter et al., 1981, Kressmann et al., 1980). In addition, the procedure was as follows: Stage VI oocytes were obtained from adult females of Xenopus laevis stimulated with HCG (Human Chorionic Gonadotropin). The oocytes were briefly centrifuged to bring the nucleus to the rim of the oocytes. Each core was injected with approx. 50 ni of a solution containing 0.3 ixg/fil superspun plasmid DNA (containing the tRNA ribozyme gene according to ex. 1) and 2/iCi//xl 32p-GTP. After 5 to 8 hours of incubation at 20°C, the individual injected oocytes were lysed in 1% SDS, 1 mg/ml proteinase K, 300 mM NaCl, 20 mM Tris, pH 8, 20 mM EDTA (400 µl/oozyte). at 37°C for 45 min, extracted once with phenol and once with phenol/chloroform and precipitated with ethanol. The pooled ethanol precipitates were resolved in 80% formamide/TBE, briefly heated for denaturation at 95°C and separated by electrophoresis on a 10% acrylamide/8.3 M urea/TBE gel and visualized by autoradiography. In all experiments with tRNA ribozyme genes, the injection solutions contained the wild-type metRNA gene at a concentration of 1/6 of the concentration of the tRNA ribozyme gene. TBE buffer (tris, borate, EDTA) was prepared according to the regulations, described in (Maniatis).

The result of these experiments is shown in fig. 5:

Track m: Molecular weight marker: as in fig. 4. The molecular weights of some of the fragments (in nucleotides) are indicated on the left of the figure. Lanes 1, 2, 3: The nucleic acids from single oocytes, injected with the met-tRNA gene and the met-tRNA ribozyme gene metribo 33. On the right of the figure, the positions for met-tRNA (met, 77 nucleotides long) and for tRNA are indicated -the ribozyme (met ribo), 128 nucleotides long.

Example 3

Determination of the ribozyme activity of tRNA ribozyme synthesized by oocytes, compared with in vitro synthesized ribozyme lacking tRNA sequences.

A tRNA ribozyme synthesized in microinjected oocytes and targeting U7 was prepared by separation by electrophoresis, visualized by autoradiography, excised from the polyacrylamide gel, and eluted by overnight incubation in an Eppendorf vibrator in HEP ("Heidelberg Extractionspuffer": 0, 75 M ammonium acetate, 10 mM magnesium acetate, 1% (vol/vol) phenol, 0.1%

(w/v) SDS, 0.1 mM EDTA). The eluted RNA was extracted once with phenol/chloroform and once with chloroform, and precipitated with ethanol in the presence of 10 µg of E.coli tRNA as carrier. The precipitate was collected and quantitatively determined by Cerenkov counting

the 32p labeling using the values for the specific activity (Kressmann et al., 1982). Samples of the tRNA ribozyme were incubated with 32p-labeled RNA containing the U7 sequence (10,000 cpm/sample, 10 fM plus the indicated amounts of unlabeled U7 RNA) for 2 h at 37°C in the presence of 150 mM NaCl, 10 mM MgCl 2 and 10 mM tris-HCl, pH 7.5. The reactions were stopped by adding EDTA up to 15 mM, the samples were dried, dissolved in 80% formamide/TBE, heated at 95°C for 30 sec. and separated on a preheated 9.5% acrylamide/8.3 M urea/TBE gel. The radiolabeled RNA species were visualized by autoradiography at -80°C with a Dr. Goo's "special" enhancer foil.

The ribozyme CD32 was produced by T7 polymerase transcription of a plasmid containing the insert from CD33 (see Fig. 2) and cloned into the HindIII/SalI sites of pSPT19 (Boehringer Mannheim). This transcript contains only the ribozyme sequence plus those sequences complementary to U7, flanked by short segments of the vector sequence. The cleavage activity of this ribozyme was drawn out in comparison with the cleavage activity of the oocyte-synthesized tRNA ribozyme CD33, to assess the influence of the secondary structure of the tRNA as well as the in vivo synthesis and modification on the ribozyme activity. Both ribozymes (CD32 and CD3 3) were found to cleave the RNA containing the U7 sequence (94 nucleotides long) into a 5' cleavage product of 25 nucleotides and a 3' cleavage product of 69 nucleotides.

The result of these experiments is shown in fig. 6:

Track m: Molecular weight marker analogous to fig. 5.

Lane 1: U7 RNA (10,000 cpm, 10 fM), incubated without ribozyme

Lane 2: U7 RNA (10,000 cpm, 100 fM), incubated with 10 fM of the

the oocyte-synthesized tRNA ribozyme CD33

Lane 3: U7 RNA (10,000 cpm, 1 pM), incubated with 10 fM of oocyte-synthesized tRNA ribozyme CD33

Lane 4: U7 RNA (10,000 cpm, 100 fM), incubated with 10 fM of

ribozyme CD32 synthesized in vitro with T7 polymerase. Lane 5: U7 RNA (10,000 cpm, 100 fM), incubated with 1 fM of ribozyme CD32 synthesized in vitro with T7 polymerase.

Example 4:

Determination of ribozyme substrate cleavage in oocytes

A mixture of<32>P-GTP, anti U7-tRNA ribozyme gene and the met-tRNA gene was injected into oocyte nuclei. As described in ex. 2, the injected oocytes were incubated at 20°C for 8 hours to allow transcription to take place. Radiolabeled U7-RNA (50 ni, 100,000 cpm//xl, 100 fM/pil) was then injected into the cytoplasm of the oocytes. The oocytes were then incubated for 2 hours. The preparation of the nucleic acids of single oocytes and their separation by gel electrophoresis was carried out as described above.

The result of these experiments is shown in fig. 7.

Lane m: Molecular weight marker as fig. 5

Lane 1: Nucleic acids from an oocyte injected with met- and

the metribo genes

Lanes 2 and 3: Oocytes injected with met and metribo, followed by

U7 RNA injection

Lanes 4 and 5: Oocytes injected with U7 RNA only

Lane 6: An aliquot of U7 RNA used for injection

Lane 7: U7 RNA (10 fM), incubated with the ribozyme CD32 (10 fM)

during 2 h at 37°C in the presence of 150 mM NaCl, 10 mM MgCl 2 and 20 mM tris-HCl, pH 7.5. Due to the gel conditions, only the 3' cleavage product (69 nucleotides) is shown.

Example 5

Transcriptional activity of tRNA ribozymes in chicken cells.

Plasmid DNA molecules containing the tRNA ribozyme genes were introduced into chicken cells to determine the transcriptional activity of the genes. It was shown that the tRNA ribozyme gene (derived from a Xenopus tRNA gene) is efficiently transcribed in chicken cells. The procedure was as follows: 10<6> primary chicken embryo fibroblast cells (Zenke et al., 1988) were seeded per 10 cm dish according to the standard method, and allowed to grow overnight. In the morning, each dish was transfected with 10 µg of plasmid DNA (containing either gene fragment 13 or gene fragment 53) using the calcium phosphate coprecipitation method (Graham et al., 1973). The cells were exposed to the precipitate overnight, the following morning washed twice with fresh medium and incubated for a further 48 hours in fresh medium. The medium was then removed, the cells taken up in PK/SDS buffer (see above), and the nucleic acid recovered. The nucleic acid was then subjected to RNAse Protection Mapping using 32p-labeled antisense erbBsection 13 or erbBsection 53 RNA probes to detect the presence of tRNA-ribozyme transcripts. Labeled RNA (10,000 cpm/10 fM antisense RNA per sample) was added to the dried ethanol precipitate, the sample was dried again and dissolved in 10 μl of 80% deionized formamide, 400 mM NaCl, 20 mM PIPES, pH 6.5, 10 mM EDTA. The samples were overlaid with sterile paraffin oil, heated for 3 minutes at 95°C and quickly rehydrated in a 45°C water bath, which was allowed to stand overnight. In the morning, 0.3 ml of ice-cold NaCl (300 mM), 30 mM Tris, pH 7.5, 1 mM EDTA, 0.05 mg per ml RNase A and 80 units/ml RNase Tl, under rapid, gentle agitation. The samples were allowed to stand at room temperature for 45 min. Proteinase K and SDS were added at 1 mg/ml and 0.5%, the incubation continued at room temperature and then at 56°C for 20 min. After adding 10 µg of tRNA, the sample was precipitated with ethanol. The collected precipitate was taken up in 80% formamide/TBE and separated on a preheated 9.5% acrylamide/8.3 M urea/TBE gel.

The result of these experiments is shown in fig. 8:

Lane m: Molecular weight marker as in the preceding examples.

Lane 1: Antisense ES 13 probe, hybridized with E. coli tRNA

(10 figs)

Lane 2: Antisense ES 53 probe, hybridized with E. coli tRNA Lanes 3 and 4: Mapping of the nucleic acids from 10,000 and 100,000

cells that had not been transfected with plasmid

DNA, hybridized with the ES 13 probe

Sports 5 and 6: Mapping the nucleic acids from 10,000 and 100,000

cells, transfected with ES 13, hybridized with ES

the 13 probe.

Sports 7 and 8: Mapping the nucleic acids from 10,000 and 100,000

cells, transfected with ES 53, which had been hybridized with the ES 53 probe.

Example 6

Attenuation of the action of the v-erb B oncogene by tDNA ribozyme genes introduced by means of polylysine-transferrin conjugates into v-erb B transformed erythroblasts.

With the help of this example, it could be shown that tDNA ribozymes which are directed against the erb B oncogene can be introduced by means of polylysine-transferrin conjugates into erb B transformed chicken erythroblasts and weaken the transforming effect of the oncogene.

Preliminary attempt 1

Preparation of transferrin-polylysine conjugates

Coupling followed analogously to methods known from the literature (cf. G. Jung, W. Kohnlein and G. Liiders, Biochem. Biophys. Res. Commun. 101 (1981), 599) by introducing disulphide bridges after modification with succinimidylpyridylthio propionate.

Pyridiyldithiopropionate-modified transferrin 1:

6 ml of a Sephadex G-25 gel-filtered solution of 120 mg (1.5 fxmol) transferrin (from hen's egg white, Sigma, Conalbumin type I, iron-free) in 0.1 M sodium phosphate buffer (pH 7.8) was added under good shaking 200 µl of a 15 mM ethanolic solution of succinimidylpyridylthiopropionate (SPDP, Pharmacia) and allowed to react for one hour at room temperature with occasional shaking. Over a gel column (Sephadex G-25, 14 x 180 mm, 0.1 M sodium phosphate buffer pH7.8) low molecular weight reaction products and reagent residues were separated, thereby obtaining 7 ml of the product fraction; the content of pyridyldithiopropionate residues bound to transferrin was determined using an aliquot after reduction with dithiothreitol by photometric determination of the amount of liberated pyridine-2-thione, and was approx. 2.6 fimol.

Mercaptopropionate-modified polylysine 2:

A solution of 18 mg (approx. 1.0/xMol) poly (L) lysine - hydrobromide (Sigma, fluorescein isothiocyanate ( = FITC) - labeled, molecular weight approx. 18.0 00/corresponding to an average degree of polymerization of approx. 90) in 3 ml 0.1 M sodium phosphate (pH 7.8) was filtered over Sephadex G-25. The polylysine solution was further diluted with water

7 ml, with good shaking added 270 µl of a 15 mmol ethanolic solution of SPDP, and allowed to react for 1 hour in the dark at room temperature and with occasional shaking. After the addition of 0.5 ml of 1 M sodium acetate buffer (pH 5.0), low molecular weight substances were filtered over Sephadex G-25 (eluent: 20 mM sodium acetate buffer, pH 5.0). The product fraction (ninhydrin staining, fluorescence) was evaporated in vacuum, brought to pH approx. 7 with buffer, added a solution of 23 mg (150//mol) in dithiothreitol in 200 fil of water, and allowed to stand for one hour at room temperature under argon in the dark. By further gel filtration (Sephadex G-25, 14 x 130 mm column, 10 mM sodium acetate buffer pH 5.0) excess reducing agent was separated and 3.5 ml of product solution of fluorescently labeled polylysine with a content of 3.8 moles of mercapto groups (photometric determination of Ellman's reagent, 5,5'-dithiobis(2-nitrobenzoic acid)).

Transferrin-polylysine conjugates 3:

The solution of modified transferrin 1 obtained as described above (7 ml in 0.1 M sodium phosphate buffer pH 7.8, about 1.5/imole transferrin with about 2.6/j.mol pyridyldithiopropionate residues) was flushed with argon; 2.0 ml of the above-described solution of mercaptomodified polylysine 2 was added (in 10 mM sodium acetate buffer the pH is 5.0, it corresponds to approx. 0.6/imole polylysine with approx. 2.2/xmol mercapto groups) , flushed with argon, shaken, and allowed to react for 18 hours at room temperature in the dark and under argon. The reaction mixture was diluted with water to 14 ml and separated by ion exchange chromatography (Pharmacia Mono S column HR10/10, gradient elution, buffer A: 50 mM HEPES pH 7.9, buffer B: A + 3 M sodium chloride, 0.5 ml/min) . Unconjugated transferrin was eluted initially, the product fractions at ca. 0.66 - 1.5 M sodium chloride. The conjugated products (ninhydrin staining, in UV at 280 nm protein absorption, and fluorescence) were collected in 6 fractions with a content in each of approx. 10 mg transferrin. The fractions were then dialyzed against a 100 mM iron(III) citrate solution (brought up to pH 7.8 with sodium bicarbonate), and then a further two times against 1 mM HEPES buffer (pH 7.5).

Sodium dodecyl sulfate gel electrophoresis (10% SDS, 8% polyacrylamide) showed, upon pretreatment with 2-mercaptoethanol, in all 6 fractions an approximately equal content of transferrin, while in non-reduced samples no band was visible for free transferrin, but only conjugates which walked shorter.

Preliminary attempt 2

Transport of transferrin-polylysine conjugates in living cells.

In order to demonstrate that the transferrin-polylysine conjugates described in preliminary experiment 1 are taken up efficiently in living erythroblasts, these conjugates were labeled with FITC. It is known (Schmidt et al., 1986) that after a few hours of incubation with erythroblasts from which transferrin had been previously removed, FITC-labelled transferrin could be detected in vesicles inside the cell (fluorescence microscope examination).

In the present example, erythroblasts (transformed by anEGF receptor retrovirus, Kahazaie et al., 1988) were incubated for 18 hours in transferrin-free differentiation medium (composition in Zenke et al., 1988) at 37°C (cell concentration 1.5 x 10<6 >/ml). After addition of the various transferrin-polylysine conjugates (or as a control the corresponding amount of sterile aq bidest) the cells were incubated at 37°C in the presence of 10 ng/ml EGF (to maintain the transformed state). After 24 and 48 hours approx. 5 x 10<5> cells, washed once in phosphate-buffered saline (PBS; pH 7.2), fixed with the 50-fold volume of a mixture of 3.7% formaldehyde and 0.02% glutaraldehyde in PBS (10 min, 40°C), washed 1 time in PBS, embedded in elvanol and examined in a fluorescence microscope (Zeiss Axiophot, Narrow band FITC and TRITC impulse). At the same time, the cells' growth rate was determined in other aliquots of the different approaches. For this, 100 µl of cell suspension was taken and the incorporation of <3>H-thymidine (8£tCi/ml, 2 hours) determined, as described in Leutz et al., 1984. From fig. 10, it appears that the erythroblasts incubated with transferrin-polylysine after 24 hours display 2 to 10 strongly fluorescent vesicles, which cannot be detected in the controls. Table A shows that with the exception of fraction 6, all the conjugates have been taken up by practically all the cells.

The fact that the cells in all samples grow equally rapidly (as measured by incorporation of tritiated thymidine (<3>H TdR) ; Table A) proves that the cells are not damaged by the polylysine constructs and that a nonspecific uptake (e.g. through cell membranes that have become permeable) can therefore be ruled out.

Preliminary attempt 3

Polylysine-transferrin constructs can functionally replace the natural transferrin-iron complex in the in vitro-induced maturation of chicken erythroblasts into erythrocytes.

The purpose of this experiment was to show that the transferrin-polylysine conjugates used here are used by the cell as natural transferrin, i.e. that they go through the normal transferrin cycle with similar efficiency. As a test system for this, erythroblasts which can be induced to mature into normal erythrocytes by "disconnecting" the transforming oncogene are particularly suitable (Beug et al., 1982). From the literature it appears that for normal maturation such cells require high concentrations of transferrin-iron complex (100 - 200 ug/ml), 3 times lower concentrations prevent the maturation of the cells and lead to cell death after several days (Kowenz et al., 1986) . It could also be shown (Schmidt et al., 1986) that "recycling", i.e. reuse of transferrin receptors and thus a transferrin cycle that takes place at optimal speed, is necessary for normal in vitro differentiation.

Erythroblasts (transformed by the EGF receptor retrovirus) were induced to differentiate by withdrawal of EGF and addition of an optimal amount of partially purified chicken erythropoietin (Kowens et al., 1986, free of transferrin). The incubation took place at a cell concentration of 1 x 10<6>/ml in transferrin-free differentiation medium at 42°C and 5% C02. At the beginning of the incubation, either natural transferrin-iron complex (Sigma, 100 µg/ml) or transferrin-polylysine conjugates saturated with iron (concentration also 100 µg/ml) were added. The growth and maturation state of the cells was analyzed in the following way after 24 and 48 hours: 1. Determination of the cell number (in Coulter Counter, Modell ZM, Beug et al., 1984) 2. Recording of cell size distribution (in Coulter-Charmelyzer

Model 256) and

3. Photometric measurement of the hemoglobin content in the cell (Kowens et al., 1986)

Also, after 72 hours, aliquots of the preparations were centrifuged in a cytocentrifuge (Shandon) on glass slides and subjected to a histochemical detection for hemoglobin (staining with neutral benzidine plus Diff-Quick quick staining for blood cells, Beug et al., 1982).

The results in Table B clearly show that cells which, in the presence of polylysine-transferrin conjugates, were induced to differentiate fractions 1 to 5, will mature as efficiently and at the same rate as those incubated with native transferrin-iron. The cells in the transferrin-free control, on the other hand, showed a greatly reduced cell growth, and only accumulated small amounts of hemoglobin. Examination of the cell phenotype on stained cytospin preparations showed that the cells incubated with polylysine-transferrin conjugates had matured in the same way into late reticulocytes (Beug et al., 1982) as those treated with native transferrin, while those cells incubated without transferrin were a mixture of disintegrated and immature, erythroblast-like cells (Schmidt et al., 1986). Only the cells treated with transferrin-polylysine fraction 6 had a lower hemoglobin content and a higher percentage of immature cells (Table B). This shows that fraction 6, which was conjugated with a particularly large amount of polylysine, works less well in the transferrin cycle. At the same time, this result refers to the sensitivity of the test method.

Preliminary attempt 4

Polylysine-transferrin conjugates enable uptake of DNA in chicken erythroblasts.

In the present experiment, it was to be investigated whether DNA can be efficiently transported by transferrin-polylysine conjugates into the interior of the cell in a size that corresponds to the size of tDNA ribozymes (see example 1). In the present example, tDNA was used with an insert with the sequence.

Molecular weight approx. 3 00.00, end-labeled with gamma<32>P ATP (Maniatis) . About 0.3 µg of this DNA, dissolved in 20 µl TE buffer, was either mixed with 10 µg native transferrin, with 10 µg transferrin-polylysine conjugate fraction 3, each time dissolved in 50 µl ag bidest plus 400 /ig/ml bovine serum albumin (Beug et al., 1982) or with 50 /xl of this solvent without transferrin. The DNA-protein mixtures were added to 2 ml of transferrin-free differentiation medium, 4 x 10<6> chicken erythroblasts, which had been transformed with an EGF receptor retrovirus and were pre-incubated for 18 hours in transferrin-free medium in the presence of EGF, (Kahazaie et al., 1988) added, and the mixtures incubated for 8 hours at 37°C and 5% C02. The cells were then centrifuged, the supernatant removed and the cells washed 3 times in transferrin-free medium. The cell pellet and culture medium were taken up in 1% SDS, 1 mg/ml proteinase K, 300 mM NaCl, 10 mM Tris pH 8.0, 10 mM EDTA (PK/SDS buffer), incubated for 30 min at 37° C, extracted with phenol/chloroform, and DNA isolated by ethanol precipitation. Isolated DNA with a total radioactivity of 2,000 cpm was separated on a non-denaturing 3.5% acrylamide gel (TBE, Maniatis) and DNA detected by autoradiography. The figure shows fluorescence recording of chicken erythroblasts that were incubated for 24 hours without (A) or with FITC-labelled transferrin-polylysine conjugates (B,C). Upon stimulation with blue light (B, for the detection of FITC), several fluorescent vesicles can be clearly observed in each cell. The specificity of this fluorescence is shown by the fact that the vesicle fluorescence does not appear (C) upon stimulation with green light (whereby a similar non-specific fluorescence can be seen in the cells as in A). This figure shows that in the cell sample that has been treated with transferrin-polylysine, approximately 5-10 times more DNA has been taken up from the cells than in the control samples with natural transferrin or without transferrin.

Two tRNA-ribozyme genes, targeting the translation initiation region in erbB were constructed (compare fig. 2 and 3, ex. 1). About. 100 µg of each plasmid containing the gene was EcoRI digested to release the tRNA ribozyme gene on a 225 bp fragment. Cleavage product one was end-labeled using Klenow fragment and purified by gel electrophoresis through a 2% agarose/TBE gel. The vector fragment and tRNA ribozyme gene fragments were located by ethidium bromide staining, excised and recovered by electroelution, phenol/chloroform and chloroform extraction and ethanol precipitation. The purified radiolabeled DNA fragments were then used using the transferrin-polylysine transport system to determine the uptake and inhibition of erbB RNA. The vector pSPT 18 was used as control DNA.

A chicken erythroblast cell line transformed by a temperature-sensitive mutant (ts 34, Graf et al., 1978) from the avian erythroblastosis virus AEV (Beug et al., 1982) was chosen as the test cell system

b). The erb A oncogene, which is also expressed in these cells, can be inhibited by a specific protein kinase inhibitor (H7). It was

established that the v-erbA oncogene is phosphorylated in vivo and in vitro (ie as bacterially expressed protein) in 2 sites, namely Ser 28 and Ser 29, by protein kinase C resp. by cAMP-dependent protein kinase. Mutation of these serines to alanines prevents the phosphorylation and destroys the v-erbA oncogene activity. H7 is a specific inhibitor of both of these kinases, and is able, in v-erbA-v-erbB containing erythroblasts, to selectively abrogate the changes caused by v-erbA (e.g. blocking the differentiation .

It is known that erythroblasts whose erbB oncogene is activated - e.g. by raising the temperature in the case of a temperature-sensitive erbB mutant - is induced to mature into erythrocytes. One of the first signs of this development is the induction of hemoglobin synthesis, which can be detected by a sensitive detection (acid benzidine staining, Orkin et al., 1975, Graf et al., 1978) at the single cell level. As a phenotypic effect of a ribozyme directed against erbB in this test system, one should therefore expect a specific increase in the number of benzidine-positive cells.

The series of experiments on which this example is based was carried out in the following way: The different DNA preparations (see above and Table C), dissolved in 30 μl of TE buffer, were all mixed with 10 μl of natural transferrin-iron complex or transferrin- polylysine conjugate (dissolved in 50 µg. bidest), and incubated for 30 minutes at 37°C.

In the case of vector DNA (10 fig), correspondingly more (100 fig) of the transferrin preparations were used. The DNA-transferrin-DNA mixtures were all added to 1 ml of transferrin-free differentiation medium (Zenke et al., 1988). The test cells (per batch 3 x 10<6>) were incubated for 60 min before the experiment. in transferrin-free differentiation medium at 42°C (thus enhancing the transferrin uptake) and added to the DNA-transferrin-containing preparations. After 6 hours, 18 hours and 68 hours (for treatment of the cells see below) samples were taken, as described, from the supernatant and the cell sediment separated, taken up in PK-SDS buffer and the DNA analyzed.

After the end of incubation (6 hours), the cells were centrifuged and incubated for a further 72 hours in transferrin-containing differentiation medium with erythropoietin and insulin (Kowens et al., 1986, Zenke et al., 1988), 2 ml per approach and at 37°C, i.e. in the presence of an active v-erb B protein.

The following results were obtained:

1. As in preliminary experiment 4, an enhanced uptake of

DNA in the size of hereditary DNAs in the cell samples that had been treated with transferrin polylysine (approximately 5-fold).

2. By transfection of chicken fibroblasts with genetic DNA

it could be shown that hereditary ribozyme tDNA is expressed in chicken cells (see ex. 5). 3. Table C shows that in all cases in which erb-cut ribozyme tDNA was introduced into erb-B-transformed erythroblasts using the polylysine-transferrin construct, the percentage of benzidine-positive cells was significantly increased (to about twofold) (as reference served those samples that were treated with vector DNA and in which the use of polylysine-transferrin conjugates did not, as expected, lead to an increase in the number of benzidine-positive cells).

Example 7

Extending the anticodon stem region increases the yield of rib-tRNA.

The human met-tRNA gene (cloned as a BamHI/Rsal fragment, completed by EcoRI linker, in the EcoRI cut site of the bluescript vector) was cleaved at its individual ApaI cut sites in the anticodon stem region. The single-stranded overhangs were removed by T4 DNA polymerase treatment, and oligonucleotides containing the ribozyme sequences inserted. The ribozyme insertion used in the previous examples resulted in a rib-tRNA molecule containing a 3-base stem (Fig. 11 shows wild-type tRNAmet and tRNArib with the shortened anticodon stem). In the second construction, an oligonucleotide was used which differed from the first by the sequence GGTTAT complementary to the wild-type sequence at the 5' end. Thereby, the wild-type strain was produced again and extended by 4 additional base pairs (Fig. 12 shows the construction of tRNArib with the amplified anticodon strain. The ribozyme sequence is produced in Fig. 13; at the left (5') end the differences in the nucleotide sequence between the region that codes for the shortened strain and the amplified strain (Fig. 13 also shows the sequence for the met-tRNA gene). Both ligation products were transformed into E. coli HB 101 according to the standard method described in Maniatis, suitable clones were selected, plasmid DNA isolated and sequenced to determine the correct structure. Both of the resulting rib-tRNA genes were examined in the presence of<32>P-GTP on the rib-tRNA molecules' transcriptional activity and accumulation by microinjection of the cloned DNA into Xenopus oocytes, carried out as described in ex. 2 or 3. The Vi 11type tRNA gene was co-injected at a 10-fold lower concentration.

The injected oocytes were incubated for 7 h at 20°C, the resulting RNA was harvested (cf. Ex. 2) and electrophoretically separated (10% acrylamide/8.3 M urea/TBE gel) and the RNA molecules visualized ( fig. 14) by autoradiography (exposed for 2 days at -70°C). The shortened rib-tRNA gene delivers approx. 1/10 of the amount of RNA that is transcribed from the wild-type gene; compared thus, the gene that codes for the rib-tRNA molecule with the extended stem delivers the 6-fold amount of RNA.

Example 8

Expression of ribozyme genes as part of introns.

A Xenopus laevis tRNA<tyr>C gene (oocyte type) was chosen as the starting gene, which contains an intron comprising 13 nucleotides. The natural intron sequence was modified as follows: First, a suitable restriction site (Apal; GGGCCC) was inserted, to enable a subsequent cloning of oligonucleotides, then complementary nucleotides to the anticodon triplet to dispose of a further stabilizing structural feature when extending the intron sequence. The size of the intron in the modified gene is increased from 13 nucleotides to 15 nucleotides (Fig. 15).

The modification of the intron sequence was carried out by means of the polymerase chain reaction (PCR; Ho et al., 1989). Therefore, 4 primers were synthesized, two of which contained the changed intron sequence (complementary to each other), and two were directed to the gene's 5'- or 3'-end to introduce an EcoRI resp. Sali the restriction cut site. The wild-type gene was present as Hhal fragment (258 bp), cloned in pBR327 (Stutz et al., 1989). The resulting PCR product was purified over agarose gel, cut with the aforementioned restriction enzymes and ligated with the vector pAALM (=pSP64+ T7 promoter; Vieira and Messing, 1982). The construct was transformed into E.coli HB 101, and clones containing the desired insert identified by sequence analysis.

The activity of the modified gene (tRNA<tyr>M) was compared with the activity of the wild-type gene by microinjection into Xenopus oocytes, whereby as an internal standard a 5S-RNA gene (50-fold lower concentration) present on the plasmid pUC-10- 5S (Carroll and Brown, 1976) in the presence of<32>P-GTP. The injected oocytes were incubated for 20 hours at 20°C, RNA was separated on an 8% acrylamide/8.3 M urea/TBE gel and autoradiographed (Fig. 16). The primary tRNA<tyr->transcript became visible at 100 (102) nt next to the 5S-RNA at 120 nt, the 5'- and 3'-processed precursor form at 90 (92) nt, as well as the fully processed tyrosine tRNA at 76 nt. The splicing of the precursor form of 90 nt seems to be the limiting factor, so that the majority of the formed transcript (approx. 80%) exists in this form. As expected, the biological activity was not reduced by the modification compared to the wild-type gene.

In a further experiment, the performance capability of the system was tested. Two ribozyme sequences containing oligodeoxyribonucleotides were synthesized, which already contain Apal ends and could thereby be cloned directly into the intron sequence in the modified tRNAtyr gene. In the case of an oligonucleotide, 12 nt were inserted at both ends to form stable "hairpins" in ribintrons and thereby counteract exonuclease degradation. The overall size of the resulting intron was 89 nt (ribozyme HP) compared to 65 nt in the case of the unprotected ribozyme sequence (ribozyme C). The sequences of the ribozymes and the cloning core are shown in fig. 17.

For the microinjections in Xeno<p>us oocytes described above, a third, which contains the ribozyme HP in dimeric form and thus increases the intron size to 163 nt (ribozyme D), was used alongside both the described constructions. The concentration of the coinjected 5S standard was 1 : 20 in construction HP and D, 1 : 1 in construction C (Fig. 18). The experiment shows that the constructions HP and C, despite significantly enlarged introns, transcribe very actively and are processed with the same efficiency as the wild-type tRNAtyr gene.

In the case of construct D, only a minimal amount of transcript is detectable, because the necessary secondary structure for the formation of a Pol-III transcription complex was obviously destroyed by that intronic sequence.

It could be shown that the expression of ribozymes as introns of tRNA does not lead to any significant damage to the tRNA secondary structure, through which the resulting transcript can be accumulated in high concentration and processed correctly.

Literature:

Altmann et al. (1988), Structure and Function of Major

and Minor Small Nuclear Ribonucleoprotein Particles,

Publisher Birnstiel, Springer Verlag, Berlin, 183-195 Beug et al. (1982a), J. Cell Physiol. Suppl.l, 195-207

Beug et al. (1982b), Cell 28, 907-919

Beug et al. (1984), Cell 36, 963-972

Carroll and Brown, (1976), Cell 7, 477-486

Ciliberto et al. (1982), Proe.Nati.Acad.Sei. 79, 1921-1925 Clarkson, Eukaryotic Genes, Their Structure, Activity and Regulation, Publisher: McLean et al., Butterworth (1983),

239-261

Cotten et al. (1988), EMBO J., 7, 801-808

Felgner et al. (1987), Proe.Nati.Acad.Sei. 84, 7413-7417 People et al. (1983), Cell 33, 585-593

Fromm et al. (1987) Methods in Enzymol. 153, 351-366 Geiduschek et al. (1988), Ann.Rev.Biochem. 57, 873-914

Graf et al. (1978) Nature 275, 496-501

Graham et al (1973), Virology 52, 456-467

Hampel and Tritz (1989), Biochemistry 28, 4929-4933

Haseloff et al. (1988), Nature 334, 585-591

Ho et al. (1989), Gene 77, 51-59

Hofstetter et al (1981), Cell 24, 573-585

Jennings et al. (1987), EMBO J. Vol. 6, 10, 3043-3047

Jung et al. (1981), Biochem. Res. Commun. 101, 599

Kahazaie et al. (1988) EMBO J. 10, 3061-3071

Kowenz et al (1986), Mod.Trends in Human Leukemia VII,

Springer Verlag, 199-209

Kressmann et al. (1978), Proe.Nati.Acad.Sei. 75, 1176-1180 Kuo et al. (1988), J. Virol. 62, 4439-4444

Langridge et al. (1987), Methods Enzymol. 153, 336-351

Leutz et al. (1984), EMBO J. 3, 3191-3197

Maniatis et al. (1982), Molecular Cloning, Cold Spring Harbor Melton et al. (1984), Nucleic Acids Res. 12, 7035-7056

Mowry et al (1987), Science 238, 1682-1687

Orkin et al. (1975), Proe.Nati.Acad.Sei. 72, 98-102

Pepperkok et al. (1988), Proe.Nati.Acad.Sei.85, 6748-6752 Schmidt et al (1986), Cell 46, 41-51

Sharmeen et al. (1988), J. Virol. 62, 2674-2679

Soldati et al. (1988), Mol. Cell. Biol. 8, 1518-1524

Stewart et al. (1987), EMBO J. 6, 383-388

Stutz et al. (1989), Genes & Development Vol. 3, 8, 1190-1198 Tellford et al. (1979), Proe.Acad.Sei. 76, 2590-2594 Uhlenbeck et al. (1987), Nature 328, 596-600

Vennstrom et al. (1980), J. Virol. 36, 575-585

Vieira and Messing, (1982), Gene 19, 259-268

Wu et al. (1989), Proe.Nati.Acad.Sei. 86, 1831-1835

Zasloff et al. (1982), Nature 300, 81-84

Zasloff et al. (1983), Proe.Acad.Sei. 80, 6426-6440

Zenke et al. (1988), Cell 52, 107-119.

Claims (15)

1. Method for producing a DNA molecule, possibly present in multiple copies, containing parts of a gene that is transcribed by polymerase III and a DNA sequence that codes for an inhibitory RNA molecule, characterized in that it contains the necessary transcription units for a tRNA gene for transcription of polymerase III including the sequences that determine the secondary structure of the tRNA, and that the DNA sequence that codes for an inhibitory RNA molecule selected from the group of ribozyme or antisense RNA is arranged within the DNA molecule in such a way that the inhibitory RNA -molecule is part of the transcript. 2. Method for producing a DNA molecule according to claim 1, characterized in that said ribozyme is of the "hammerhead" type. 3. Method for producing a DNA molecule according to claim 1, characterized in that it contains the transcription units of an initiation Met-tDNA. 4. Method for producing a DNA molecule according to claim 1, characterized in that it contains the transcription units of a bull tDNA. 5. Method for producing a DNA molecule according to one of claims 1-4, characterized in that it contains the DNA sequence that codes for the inhibitory RNA molecule as an insert between the A block and the B block of the tRNA gene. 6. Method for producing a DNA molecule according to claim 3 or 5, characterized in that it contains the DNA sequence that codes for the inhibitory RNA molecule as an insert in the natural ApaI restriction cutting site between the A and B block. 7. Method for producing a DNA molecule according to claim 4, characterized in that it contains the DNA sequence that codes for the inhibitory RNA molecule as an insertion of the intron. 8. Method for producing a DNA molecule according to one of the preceding claims, characterized in that it contains an oligonucleotide sequence such that the anticodon stem region of the tRNA transcript is extended in relation to that of the wild-type tRNA transcript. 9. Method for producing a DNA molecule according to claim 7 or 8, characterized in that it contains the DNA sequence that codes for the inhibitory RNA molecule as an insert in an artificially introduced restriction site. 10. Method for producing a DNA molecule according to one of claims 7 to 9, characterized in that the intron is modified in such a way that the secondary structure of the output tRNA is stabilized, during which the necessary structures for splicing are maintained. 11. Method for producing a DNA molecule according to one of claims 1-10, characterized in that the inhibitory RNA molecule is directed against viral RNA. 12. Method for producing a DNA molecule according to one of claims 1-10, characterized in that the inhibitory RNA molecule is directed against oncogenes or other key genes that control the growth and/or differentiation of cells. 13. Method for producing a DNA molecule according to one of claims 1-12, characterized in that it is present as multicopy, whereby identical or different sub-units containing DNA sequences coding for an inhibitory RNA molecule are present as complete transcription units. 14. Method for introducing DNA molecules according to one of claims 1-13 into higher eukaryotic cells in vitro, characterized in that the DNA molecules are modified so that their uptake into the cell through receptor-mediated endocytosis takes place, and that the cells are brought into contact with the DNA molecules. 15. Method according to claim 14, characterized by complexing the DNA molecules with a transferrin-polycation conjugate and bringing the cells into contact with this complex.