CA2039359A1 - Cellular binding protein having affinity for the regulatory region of the human papilloma virus type 18 - Google Patents

Abstract

Abstract Cellular binding protein having affinity for the regula-tory region of the human papilloma virus type 18 The invention relates to a DNA sequence which codes for the cellular DNA-binding protein RS1, which specifically binds to the 28 bp E2 subregion of the upstream regula-tory region (URR) of the human papilloma virus (HPV), and to the protein RS1 and proteins which are similar to RS1.

Description

2 [);39359 BEHRINGWERKE AKTIENGESELLSCHAFT HOE 90/B 016 - Ma 825 ANR: 1 000 764 Dr. Lp/Wr Description Cellular binding protein havin~ affinity for the reoula-tory re~ion of the human papilloma yirus tyPe 18 The invention relates to DNA sequences which code for a DNA-binding protein (RSl), and to a novel DNA-binding protein (RSl) which was isolated from primary human keratinocytes with the aid of the recombinant DNA techni-que.

The invention further relates to recombinant DNA mole-cules for use in the preparation of RSl protein or proteins which are similar to and have the biological activity of RS1, and to host organisms such as bacteria, yeasts and mammalian cells which have been transformed with these recombinant DNA molecules.

The invention finally relates to pharmaceutical composi-tions which contain the RSl protein or proteins which are similar to and have the activity of RS1.

Papilloma viruses are small double-stranded DNA viruses and are responsible for benign lesions of the skin and the mucosal 0pithelium. However, in addition to this a connection is also seen between the occurrence of certain types of human papilloma viruses (HPV) and anogenital cancer (The Papovaviridae, Vol. 2, Plenum Publishing Corp., New York, pages 245-263). An important indication of the connection between the development of cancer and the presence of HPV is that in more than 90% of cervical tumors HPV DNA sequences were found after biopsy. Most of these tumors contain HPV type 16 or HPV type 18 DNA. HPV
18 DNA sequences have also been discovered in a number of cells from cell lines which were obtained from cervical tumors. Both in cervical tumors and in cell lines deriv~d X0393S~3 from such tumors the viral DNA is normally integrated in the host genome. It is also known that the viral tran-scription takes place in such cells. The examination of established tumor cell lines suggested that the HPV
sequences are the sequences responsible for the develop-ment of the cancer. There are also strong indications that the expression of the early open reading frames (ORF) of the E6 and E7 genes of the HPV types 1~ and 18, which code for the transforming activity, play a decisive role in initiating and maintaining the transformed cells (Cancer Res. 48: 3780-3786).

Regulatory transcription elements are located in the 80-called URR region (upstream regulatory region) of the HPV
genome which has also been termed the noncoding region or the long-control region. This region extends over 825 base pairs from the end of the late open reading frame (ORF Ll) to the start of the early active gene E6 in the HPV 18 DNA. The ~PV 18 URR region contains a promoter which is active in several tumor lines and has at least 3 different enhancer regions (J. Virol. 62:665-672, 1988).

The object of the present invention follows from the following points:

The regulation of viral genes by cellular proteins plays a key role in tumor development. The deregulation of the expression of the transformed early genes E6 and E7 of HPV 16 and 18 contributes to the process of cancer development in the transformed cells. In this connection the identification, the isolation and preparation of proteins which have a binding affinity for the regulatory region of these early genes are a decisive step in combating the abovementioned tumors. The proteins binding the deregulatory regions therefore possibly represent suitable means for the preparation of an agent for combating cancer.

The present invention therefore had the ob~ect of iden-tifying factors which can influence the regulation and expression of viral DNA in infected human cells, and of providing these in amounts sufficient to be able to use them as agents against carcinoses.

This object has been achieved according to the invention on the one hand by providing a DNA sequence, such as the one depicted in Figure 8 and claim 1, which codes for the cellular DNA-binding protein RSl or a polypeptide which is similar to and has the biological activity of RSl.

On the other hand, a DNA-binding protein (RSl) having the amino acid sequence shown in Figure 8 and claim 11, and the property of binding specifically to the 28 bp E2 region of the HPV 18 upstream regulatory region is provided in order to achieve the ob~ect.

The invention furthermore comprises DNA sequences which are derived from the genome of primary human keratinocytes and code for the cellular DNA-binding protein RS1.

In addition, the invention comprises DNA sequences which hybridize with one of the abovementioned DNA sequences.
These DNA sequences can be of natural, semisynthetic or synthetic origin and are related to one of the above-mentioned DNA sequences by mutation such as nucleotide substitution, nucleotide deletion, nucleotide insertion or inversion of nucleotide sections. Furthermore, these last-mentioned DNA sequences characteristically code for a polypeptide which is similar to and has the biological activity of RSl.

One biological activity of RSl i8, for example, the specific affinity (high affinity binding) of RSl for the E2 subregion with a length of 28 bp of the upstream regulatory region of HPV types 16 and 18. Proteins similar to RS1 are proteins which, like the protein RSl, 2~ 3359 have an affinity for the DNA E2 subregion of the URR of the HPV types 16-18.

The invention further comprises recombinant DNA molecules for cloning in vectors, the DNA sequence being selected from the group of the abovemsntioned DNA sequences which code for the cellular DNA-binding protein RSl or proteins which are similar to RSl.

These recombinant DNA molecules are operatively connected to an expression control sequence.

Recombinant DNA molecules with an expression control sequence selected from an E.coli promoter system, such as the E.coli lac system, the E.coli ~-lactamase system, the E.coli trp system, the E.coli lipoprotein promoter, or with a yeast expression control sequence or another eukaryotic expression control sequence are preferred.

The invention furthermore comprises host organisms which are transformed with at least one of these recombinant DNA molecules.

Preferably host organisms selected from the following group are used: E.coli, a different bacterium, yeast, a different fungus, an animal or human cell.

Particular preference is given to the use of a recom-binant lambda gt 11. The DNA-binding protein RSl according to the invention has the property of specifi-cally binding to the 28 bp E2 region of the URR of HPVtype 18 and 16. DNA-binding proteins according to the invention also comprise proteins having the biological properties of RSl, ie. all proteins which are similar to RSl and likewise have the property of binding to the E2 region of the URR of HPV type 18 and 16, or else they have the action of RS1, namely of influencing the HPV by specific interaction with the E2 region of the URR or overlapping regions, such as regulating the early genes 5 - '~03935~
E6 and E7.

The proteins according to the invention are encoded by one of the abovementioned DNA sequences according to the invention and are produced with the aid of the recom-binant DNA technology in a manner known per se using ahost organism according to the invention.

In these processes for preparing RSl or polypeptides which are similar to and have the biological activity of RSl, a host organism according to the invention is transformed with a recombinant DNA molecule according to the invention. The transformed organism is cultivated, and the RS1 or the pol~peptide which is similar to RSl is isolated from the culture medium after expression.

Proteins which can be obtained by these processes are also proteins according to the invention.

Because of the biological activity of the proteins according to the invention which make it possible to influence the requlation of cells transformed with HPV
with respect to a viral expression, these proteins are suitable for the treatment of cancer.

The present invention therefore also comprises pharmaceu-tical compositions which contain the protein RSl or polypeptides which are similar to RSl. These agents can be employed in the treatment of cancer or tumors.

In the attempt to identify and isolate URR-binding proteins, a ~trategy was used which makes it po~sible to i~olate the factors encoded by complementary DNA (cDNAs) on the basis of their property of sequence-specific binding to DNA fragments. This method is based on a high expression rate of the DNA-binding region encoding the DNA-binding factor in bacteriophage lambda gt 11. The interaction of this factor with double-stranded, radio-actively labeled DNA leads to a recognition signal. The - 6 - ~ ~39359 corresponding recombinant in the lambda gt 11 expression library can then be isolated.

Brief descri~tion of the fi~ures:

Fig. 1:

Diagrammatic representation of the regulatory region URR
(upstream regulatory region) with human papilloma virus HPV 18. CAAT box and TATA box are indicated by an ellipse or by a rectangle, respectively. The E2 6ubregion which contains the two palindromic repeats is shown underneath.
L0 The palindromic repeats are indicated by arrows. They are the recognition region for the viral E2 proteins. The core sequence which is common to all ~he nucleotides (Fiq. 2) and i8 bound by the DNA-binding protein RSl has been boxed in.

Fig. 2:

Oligonucleotides which were used in this invention:
oligonucleotides A1 to A3 contain sequences which are derived from the E2 subregion described in Figure 1. ~hey have a core sequence with a length of 28 base pairs (boxed in and doubled in A3) in common.

Oligonucleotides ~ to F and G are monomers or tetramers which are derived from other regions of the URR of the HPV 18 or HPV 16.

Fig. 3:

Sequence of the RSl cDNA and the RSl protein. The amino acid sequence is specified using the 3 letter code.

Isolation of the cDNA clone cRS1.

In order to isolate the cellul~r DNA-binding protein RS1 which binds to the regulatory region URR of HPV 18 - 7 - ~0393~9 (Fig. l), a lambda gt 11 expression library which con-tains cDNA of primary epidermal keratinocytes, which are the natural host cells of HPV, was tested. In this test, the method originally applied by Singh et al., Cell 52:415-423, 1988 and modified by Vinson et al., (Genes Dev. 1988 2:801-806) was used. The probe used to find binding proteins was the oligonucleotide A3 (Fig. 2).
This oligonucleotide contains a duplication of 2 tandem repeats which include the recognition region for the viral E2 transactivator/ transrepressor (Nature 325:70-73, 198~). In a first approach 500,000 pfu of the lambda gt ll expression library were tested. Poly-dIdC, which does not hold any biological information, was used to prevent nonspecific binding. In a further approach, a further 500,000 pfu were tested, denatured and sonicated calf thymus DNA being used in this approach to prevent nonspecific binding. Both rounds of screening led to only one positive signal which, additionally, occurred at the same site on the original and the filter copy. Restric-tion analysis and partial sequencing of the cDNA frag-ments (inserts) showed that both clones are identical and probably are copies of the same phage. The phage which was identified (referred to as RS1 phage) wa~ purified in 4 round~ of plating and then analyzed in detail.

S~ecificity of binding:

The binding specificity was analyzed with the aid of filter binding tests, 3 different oligonucleotide~ (A1, A2 and A3) all of which contained the E2 subregion with a length of 28 base pairs being used. As comparison several nucleotides which are not related to the E2 subregion but are derived from the regulatory region URR
of HPV 18 and HPV 16 were used as negative controls (Fig.
2). Concentrated samples of phage (5 x 107 pfu) were sprayed onto freshly poured E.coli Y lO90 lawn. The DNA
binding properties of RS1 and of a product of a negative control clone which was selected from the lambda gt 11 expression library at random were analyzed with the aid of filter binding tests. In all cases, the product of the negative control clone did not interact with any of the tested oligonucleotides. In contrast to this, RSl bound specifically to the oligonucleotides A1, A2 and A3. An interaction between RSl and other control oligonucleo-tides was not observed.

It was also found that RS1 has a high and specific binding activity to single-stranded DNA, namely to the noncoding (anti-sense) strand (Fig. 2) of the oligonucle-otides Al, A2 and ~3. An affinity of RSl to the coding(sense) strand of the oligonucleotides A1, A2 and A3 or to another single strand of the control oligonucleotides was not observed. The negative control phage exhibited no factor which has an activity toward single-stranded DNA.

RSl also binds to DNA at high salt concentration~ (400 mM
NaCl), which indicates a high binding specificity.

Even under high salt concentrations, exclusive binding of RSl to oligonucleotides which contained the core sequence with a length of 28 base pairs was observed. No interaction with other double-stranded or single-stranded DNA fragments was observed.

Analysis of the ~-galactosidase fusion protein encoded by the RS1 phage:

Lysogenic RS1 phages and negative control phages were isolated and induced to produce large amounts of their corresponding ~-galactosidase fusion protein. Western blot proteins of induced and noninduced lysogenic phages were treated with anti-~-galactosidase mouse antibodies and then coated with an anti-mouse antibody, which was covalently bonded to alkaline phosphatase. IPTG induced a fusion protein of approximately 150 R in the lysogenic RSl phage.

The bindi~g properties of the ~-gal-RSl fusion protein 2C)393~;~
were tested with the aid of South-Western analysis of the entire protein extract of induced and noninduced lysoge-nic cultures. After the transfer, the immobilized proteins were denatured with 7 ~ guanidine hydrochloride for 60 min and then renatured for 24 h. Science (1988) 241:577-580. The denaturing/renaturing cycle improved the recognition ~ignals of the proteins in the South-Western analysis and the total protein of induced lysogenic phages of a negative control and RSl was analyzed with the aid of the radioactively labeled oligonucleotide A3b probe. A single band which represents a protein with a molecular mass of approximately 150 R was specifically detected in the lysogenic phage RSl. The specificity of DNA binding by this 150 K fusion protein of the South-Western analysis additionally confirmed the binding specificity of RSl as demonstrated in the filter binding tests. The 150 K protein specifically binds to the oligonucleotides Al, A2 and A3 (Fig. 2) and to their corresponding noncoding (anti-sense) strands, and does not bind to the control oligonucleotides.

Structure of the cDNA present in phage RSl:

The phage RSl contains a cDNA fragment with a length of 1338 nucleotides. One of the two strands starts at its 5'-end with an open reading frame tORF) of 906 nucleo-tides followed by a 3'-nontranslated region of 432 nucleotides. The ORF with a length of 906 nucleotides, which corresponds to 302 amino acids, is bonded in frame to the ~-galactosidase gene at the 3'-end. The ~-galac-tosidase-RSl fusion protein has a molecular mass of approximately 150 K. The ~-galactosidase portion of this fusion protein has a molecular mass of approximately 120 kD so that the cDNA-encoded portion of the fusion protein must therefore have a molecular mass of approximately 30 kD.

These results are consistent with the molecular weight of 33.6 ~, which can be calculated for an oligopeptide with - lo - ~0393S9 302 amino acids, which are encoded by the ORF, with a length of 906 nucleotides, of the RSl phage cDNA. In the ORF of 906 nucleotides a potential translation-initiation codon i8 present at base 192. However, the surrounding bases do not follow the Kozak consensus sequence for eukaryotic translation/initiation sites (Cell 44: 283-292). For this reason it may be possible that the AUG
codon in position 192 is not the original translation/initiation codon for the RS1 gene. Two polyadenylation consensus sequences (AATAAA) are present in the 3~-region of the RSl cDNA and start at nucleotides 1167 and 1221 respectively. However, they are not fol-lowed by a poly(A) tail.

Comparison of the RS1 cDNA sequence or the protein sequence deduced therefrom with several sequence data banks (Gene bank, Swiss Prot, Dayhoff) showed no sig-nificant nucleotide or amino acid sequence homology of RSl with known proteins. The amino acid sequence of RS1 shows several interesting characteristics. The protein is predominantly rich in arginine (34 of 302 amino acid6) and proline (35 of 302 amino acids). Proline is in part present in the form of clusters (e.g. amino acids 278-296). As is known, these proline clusters participate in the activation of transcription by several transcription factors. A region of 51 amino acids (33-84) i8 rich in proline (20%), glutamic acid (10%), serine (10%) and threonine (18%). Analogous regions are present in the amino acid sequence of a group of proteins with a very short in vivo half life ("PEST" proteins). Residues 10 to 51 have a high content of the basic amino acid arginine (20%). RS1 also does not contain a region containing zinc fingers, homeoboxes or leucine zippers, which repre~ent domains, such as are present in several other DNA-binding factors (Science 1989, 245: 371-378).

11 21~39359 Example 1 A cDNA expression library A cDNA library of human epidermal keratinocytes, which is cloned in the expression vector lambda gt 11, was ob-tained from Clontech Laboratories Inc. The librarycontained approximately 1.7 x 106 independent recombinant clones with an averaqe fragment (insert) size of 1.1 kb.

The lambda gt 11 recombinants were screened with the aid of the host organism E.coli Y 1090. Lysogenic phages were generated in E.coli Y 1089 or Y 1090 (in: DNA-Cloning -A Practical Approach, Vol. 1, D.M. Glover ed. (Oxford:
IRL Press), 49-78).

Example 2 DNA probes and ~est for DNA-binding proteins Oligonucleotides were prepared with an Applied Biosystems 380A synthesizer with the aid of the standard phosphori-mide technique. The products were purified by means of gel electrophoresis. Subsequently they were labeled with (32p) dATP with the aid of polynucleotide kinase (supplied by Boehringer Mannheim) under standard conditions.
5Onicated calf thymus DNA (supplied by Pharmacia) and poly-dIdC (supplied by Pharmacia) were u~ed to avoid non-specific binding.

The test for DNA-binding proteins was carried out in the following manner:

Lambda gt 11 recombinants were distributed in a Nunc bioassay dish in 0.7% agarose at a maximum of 1.5 x 105pfu/600 cm3. Nitrocellulose ~ilters (PA 85, ~chleicher and Schuell~ saturated with 10 ~M isopropyl-~-D-thioglucopyranoside were, as a covering layer, placed on top at 42C for 3 h. After incubating at 37C for a - 12 - 2~39359 further 6 hours, the filters were removed and another filter was placed on top at 37C for 3 h. The filters were dried at room temperature for 10 min and then incubated in buffer A (50 mM NaCl, 0.5 mM dithiothreitol (DTT), 25 mM HEPES (pH 7.9) and 6 M guanidine hydro-chloride). After slight shaking for 10 min, the solution was replaced with the same buffer. After a further 10 min, the solution was diluted in four 5 min steps (Genes Dev.: 1988 2:801-806) with the same volume of buffer B
(50 mM NaCl, 0.5 mM DDT and 25 mM HEPES (pH 7.9)).

Finally the filters were washed twice for 5 minutes with buffer B and then incubated at 4C for 30 min in buffer C (5~D Carnation nonfat dry milk, 50 mM NaCl, 0.5 mM DTT
and 25 mM HEPES (pH 7~9)). The solution was replaced by buffer D (0.25% Carnation nonfat dry milk, 50 mM NaCl, 0.5 mM DTT and 25 mM HEP~S (ph 7.9)) and incubated for 1 min. The actual binding reaction was carried out in buffer D which additionally contained 3 x 106 cpm/ml (1 x 108 cpm/~g) labeled double-stranded oligonucleotide and 10 ~g/ml nonspecific DNA (denatured sonicated calf thymus DNA, average length: 3000 bp, or poly-dIdC, average length: 8260 bp). After 60 min at 4C, the filters were washed 3 times for 5 min with buffer D, dried with the aid of 3 MN paper, and then a Rodak x-omat AR film was exposed to the filters overnight at -70C
with the aid of an intensifying diaphragm.

Example 3:

Analysis of the ~-galactosidase fusion protein Lysogenic phages were prepared by the method of Huynh (DNA Cloning - A Practical Approach, Vol. 1, Oxford: IRL
Press, 49-78). The protein extracts of lysogenic phages were prepared (Cell 1988 52:415-423) and stored at 70DC
in aliquots. A Western blot analysis was carried out with the aid of an anti-~-galactosidase mouse monoclonal antibody (supplied by Promega) and a phosphatase-based - 13 - ~ ~39~5~
protoblot Western blot AP ~ystem tPromega) under the conditions recommended by the supplier. In order to carry out South~Western analyses, the protein extracts (50 ~g of crude extract) were boil~d for 5 min in sample buffer (Proc. Natl. Acad. Sci. USA 82: 6741-6744) and then applied to a 10~ SDS PAGE gel. After the electrophoresi~, the proteins were transferred onto nitrocellulose with the aid of electroblotting (semi-dry system, model SDl, CTI GmbH). The filters were then treated as described in Science 19~8 241:577-580.

Example 4:

Sequencing:

The fragments (inserts) of the recombinant phages were prepared with the aid of the lambda phage adsorption system (Promega). The inserts were subcloned in M13mpl8 and then sequenced with the aid of the dideoxy method (SEQUENASE; United States Biochemical Inc.). Both strands were sequenced completely.

Example 5:

Northern and Southern blot hybridization:

DNA was isolated from HeLa, CG13 (tumorigenic HeLa-fibroblast hybrids (Science 1982 215:252-259)), 444 (nontumorigenic HeLa-fibroblast hybrids), primary human fibroblasts, SV 80 (SV 40-transformed fibroblafits), HaCat (spontaneously immortalized keratinocytes (Journal of Cell Biology, 1988, 106:761-771)), Caski (cervical carcinoma), SW 480 (colon carcinoma), Wilms, Hep G2 (hepatoma) and human TR (osteosarcoma) cell lines, namely by the method which is described in Molecular Cloning: A
Laboratory Manual: Cold Spring Harbor Laboratory. Res-triction digests were fractionated on a 1% agarose gel and then transferred onto GeneScreen Plus membranes (Du Pont). The filters were prehybridized and hybridized with - 14 - ~ ~3935 the 32P-labeled RS1 cDNA under stringent conditions.

Cytoplasmatic RNA was isolated by the method described in Cancer Research 1988, 48: 3780-3786. Northern blot analysis was carried out with approximately 10 ~g of cytoplasmatic RNA which had been fractionated on a 1 agarose gel. The gel-running buffer was composed of 0.2 M morpholinopropanesulfonic acid pH 7.0; 50 mM ~odium acetate and 1 mM EDTA pH 8Ø The fractionated cytoplas-matic RNA was then transferred onto GeneScreen Plus filters (Du Pont). The filters were hybridized and washed under stringent conditions with randomly primed 32p_ labeled probes in a manner Xnown per se.

Example 6:

Filter binding test:

Concentrated (1 x 101 pfu/ml) phage solution of plaque lysates was sprayed directly onto freshly poured E.coli Y 1090 host cell lawn. This culture was incubated at 42C
for 2 hours and then covered with nitrocellulose filters which were saturated with 10 ~M IPTG. The incubation was then continued at 37C for 2 h. Subsequently, the filters were collected and treated as described in Example 2.

Claims (18)

1. A DNA sequence which codes for the cellular DNA-binding protein RS1 or a polypeptide which is similar to and has the biological activity of RS1.

2. A DNA sequence which is derived from the genome of primary human keratinocytes and which codes for the cellular binding protein RS1 or a polypeptide which is similar to and has the biological activity of RS1.

3. A DNA sequence which hybridizes with a DNA sequence as claimed in claims 1 and 2, which is of natural, semi-synthetic or synthetic origin, which is related to a DNA
sequence as claimed in claim 1 or 2 by mutation such as nucleotide substitution, nucleotide deletion, nucleotide insertion or inversion of nucleotide sections, and which codes for a polypeptide which is similar to and has the biological activity of RS1.

4. A recombinant DNA molecule for cloning, wherein the DNA sequence as claimed in one of claims l to 3 is selected.

5. A recombinant DNA molecule as claimed in claim 4, wherein the DNA sequence is operatively connected to an expression control sequence.

6. A recombinant DNA molecule as claimed in claim 5, wherein the expression control sequence is selected from an E.coli promoter system, the E.coli lac system, the E.coli .beta.-lactamase system, the E.coli trp system, the E.coli lipoprotein promoter, a yeast expression control sequence or another eukaryotic expression control se-quence.

7. A host organism which contains a DNA sequence coding for the DNA-binding protein RS1.

8. A host organism as claimed in claim 7, which is transformed with at least one of the recombinant DNA
molecules claimed in one of claims 4 to 6.

9. A host organism as claimed in claim 8, which is selected from the group comprising E.coli, a different bacterium, yeast, a different fungus, an animal or human cell.

10. A host organism as claimed in claim 9, which is lambda gt 11.

11. The DNA-binding protein RS1 having the following amino acid sequences

12. A DNA-binding protein with the property of binding specifically to the 28 bp E2 subregion of the upstream regulatory region of the human papilloma virus HPV 18, which subregion has the base sequence ACCGAAAACGGTCGGGACCGAAAACGGT.

13. A DNA-binding protein having the biological activity of RS1, which is encoded by a DNA sequence as claimed in one of claims 1 to 3.

14. A DNA-binding protein which is prepared with the aid of the recombinant DNA technology using a host organism of claims 7 to 10.

15. A process for preparing RS1 or polypeptides which are similar to and have the biological activity of RS1, which comprises transforming a host organism with a recombinant DNA molecule as claimed in claim 5 or 6, cultivating the transformed organism and isolating the RS1 or the polypeptide which is similar to RS1 after expression.

16. A polypeptide having the biological activity of RS1, which can be obtained by a process as claimed in claim 15.

17. A pharmaceutical composition containing the polypep-tide RS1 or polypeptides which are similar to and have the biological activity of RS1, as claimed in claims 11 to 14, or 16, as an agent in the treatment of cancer or tumors.

18. The DNA sequence as claimed in claim 1 and substantially as described herein.