WO2000006775A1

WO2000006775A1 - Exon deletion antisense drug design and therapy

Info

Publication number: WO2000006775A1
Application number: PCT/US1999/016632
Authority: WO
Inventors: Helen Fillmore; William C. Broaddus; George T. Gillies; William S. Conrad
Original assignee: Virginia Commonwealth University
Priority date: 1998-07-27
Filing date: 1999-07-23
Publication date: 2000-02-10
Also published as: WO2000006775A9; AU5124199A

Abstract

The present invention provides a method for designing antisense oligodeoxynucleotides (ODNs) and long antisense expression vectors which produce long antisense RNA sequences for inhibiting the translation of specific isoforms, or families of isoforms, of a protein. The invention also provides methods of use for the ODNs and long antisense expression vectors which produce long antisense RNA sequences.

Description

EXON DELETION ANTISENSE DRUG DESIGN AND THERAPY

DESCRIPTION

BACKGROUND OF THE INVENTION

Field of the Invention The invention relates generally to the inhibition of translation of specific isoforms of proteins. In particular, the invention provides methods for the design of antisense oligodeoxynucleotides (ODNs) or long antisense DNA sequences which inhibit the translation of specific isoforms of proteins by targeting unique exon or splice site junction sequences. In addition, methods of use for the ODNs or long antisense DNA sequences as laboratory or clinical tools and in gene medicine are provided.

Background of the Invention

In the wake of the success of the Human Genome Project, new vistas for molecular medicine are rapidly opening as sequence after sequence of the human genome is revealed. A tantalizing yet challenging use for this new information is that of gene therapy or gene medicine, wherein manipulation of the expression of genes is artificially altered. One particularly promising avenue for carrying out such manipulations is antisense technology.

The techniques of antisense technology are designed to interfere with the process of protein expression. One antisense strategy is to block expression at the level of mRNA translation by introducing into the cell a single-stranded oligodeoxynucleotide (ODN) which is complimentary to a targeted sequence within the mRNA molecule. The ODN binds via Watson-Crick base pairing to the complimentary sequence, and translation is prevented either by physically blocking the translational machinery, or by inducing destruction of the mRNA-

ODN complex by RNAse H. The technique is selective because only mRNA containing the targeted sequence is affected, whereas other mRNAs are translated normally. Therefore, the impact of an antisense ODN on cells which do not contain the target sequence is minimal. For the antisense strategy to be successful, it is necessary to have methods for the rational design of ODNs which are to be used. Many factors must be taken into account. For example, the targeted sequence must be unique to the mRNA which is to be inhibited; the synthetic ODN must exhibit high affinity for the targeted sequence; the ODN sequence should not be likely to form intramolecular secondary structure (e.g. hairpin loops) because this would abrogate its ability to bind to the target sequence; and the nature of the ODN should be such that it can readily cross the cellular membrane and be relatively non-susceptible to intracellular nucleases.

Another powerful antisense technology can be described as long antisense techniques. Also known as "long" antisense techniques, this approach involves the use of expression vectors which encode for a long antisense sequence (often several hundred to several thousand nucleotide bases) designed to target a specific mRNA. Expression vectors are agents or biological constructs designed to efficiently deliver DNA sequences into cells for the purpose of inducing expression of the genetic material encoded by the particular sequence. Examples of expression vectors including modified vectors or plasmids, viral vectors, lipid- or other biomolecule-coated DNA constructs and "naked" DNA in conjunction with agents to induce cellular uptake of the DNA. Using these vector systems, the long antisense DNA sequence can be introduced into cells in such a way as to induce the transcription of new mRNA molecules containing a sequence complimentary to a targeted endogenous (sense) mRNA which would ordinarily be translated by the protein synthetic machinery of the cell to form a particular protein product. Like antisense ODN technology however, the long antisense mRNA can then hybridize with high affinity and specificity to the targeted sense mRNA, leading to its destruction and preventing translation of the sense mRNA into the targeted protein product. This class of techniques has been used to specifically inhibit the expression of a variety of proteins for the purpose of investigating their function, and could be considered for therapeutic applications in diseases characterized by the abnormal expression of specific protein products. Potential examples include tenascin-C in brain tumors, as described below, Alzheimer-specific proteins in that disease state, and many others.

Most genes of higher organisms contain both exons and introns. Exons are segments of DNA within a gene that are ultimately translated to form part of a polypeptide gene product. Introns are segments of DNA which are interspersed within the gene but which are not translated and have no known function. During the initial phases of protein synthesis, the entire linear sequence of the gene is transcribed, including both exons and introns, to form a pre- mRNA molecule (Figure 1). The intron sections are then removed from the pre- mRNA and the exon-encoded mRNA segments are joined together via "splicing" mechanisms to form the mature, processed mRNA molecule which is translated. As a result of splicing, successive exons that make up the gene are thus aligned in proper linear sequence prior to ribosomal translation, without the intervening intron sequences. The portion of the mature mRNA which encompasses the point at which two exons have been joined or "spliced" is referred to as a "splice site".

In addition, there are many genes that have alternative or optional exons whose inclusion or omission via "alternative splicing" can result in multiple forms of their respective protein products. When alternative splicing occurs, certain portions of the polypeptide usually remain constant in all forms of the protein product, whereas other portions vary according to which exons are spliced into the mRNA. These different forms of the protein are referred to as isoforms, and they may have different molecular weights, may be expressed at different times in the development of an organism, and may have different biological functions.

Due to such differential expression, protein isoforms provide excellent clues to the origin and progress of cellular development and differentiation. In particular, the investigation of isoforms which are absent in nomal tissue but present in certain disease states can provide valuable clues to the progress of the associated disease. Further, the manipulation of the expression of various isoforms may provide a means for the treatment of such disease states. The uniqueness of the mature mRNA of the individual isoforms suggests that antisense technology could be a viable approach to effecting such manipulation.

Conventional approaches to antisense ODN design involve choosing sequences within the mRNA which often include the translational start site. While this approach is appropriate for targeting all isoforms of a protein, it does not allow for the selective inhibition of isoforms. It would thus be of benefit to have available a method for the design of antisense ODN molecules and long antisense constructs which are adapted specifically for the selective inhibition of the translation of protein isoforms which occur as a result of alternative splicing. The ODNs and long antisense constructs would be useful, for example, in the design of laboratory experiments for investigating such phenomena as cellular development and differentiation; the development of clinical diagnostic methods based on detecting specific isoforms (or families of isoforms) of a protein; and in the design of antisense gene medicine protocols, for example, when the expression of a particular isoform or family of isoforms is associated with the onset and development of a disease. SUMMARY OF THE INVENTION

It is an object of this invention to provide a method of designing and preparing antisense oligodeoxynucleotide (ODN) sequences, and of designing and preparing long antisense constructs that encode sequences, for blocking translation of a specific isoform (or a family of isoforms) of a protein which can be expressed as a plurality of different isoforms. The method includes identifying a nucleotide sequence for a specific region of mRNA which is required to encode the specific isoform or family of isoforms, but which is not required to encode other isoforms of the protein. The specific region may be from, for example, a differentially expressed exon, or from a splice site region of two spliced exons. The method of preparing the ODN or long antisense sequence includes generating a computerized list of candidate sequences which hybridize to the specific region and analyzing the candidate sequences for characteristics which will enhance their efficacy for targeting the protein isoform or family of isoforms. In the case of antisense ODNs, the candidate sequences are analyzed for binding affinity, secondary structure and homodimer formation; the most appropriate final candidates are selected and their specificity is then determined. In the case of long antisense sequences, the candidate sequences are analyzed for their percent homology to targeted isoforms and those isoforms not to be targeted in order to optimize selectivity. The candidate sequences are then searched for appropriate restriction enzyme sites within the sequences in order to allow the planning of cloning techniques that lead to production of an expression vector containing the desired sequence in the desired (antisense) orientation. In a preferred embodiment of the present invention, the protein is Tenascin-C and the specific region is derived from exon 15, or from the splice site region of exons 15 and 16, or exons 14 and 16, or exons 9 and 17. In other preferred embodiments, the proteins are, for example, CD-44 and cyclin Dl, and the specific region is derived from, for example, differentially expressed exons or splice site regions.

In a preferred embodiment of the present invention, the ODNs are SEQ IDs 1- 532.

BRJTEF DESCRIPTION OF THE DRAWINGS

Figure 1. Schematic of mRNA splicing.

Figure 2 A and 2B. Tenascin-C polypeptide isoforms. A: the largest isoform B: the smallest isoform. | = cysteine-rich amino terminal region; H = epidermal growth factor-like repeat region; ^~j = fibronectin type Ill-like repeats, constant;< > = fibronectin type Ill-like repeats, alternatively spliced; ® = fibrinogen-like domain.

Figure 3. Tenascin-C pre-mRNA containing all 27 exons. Black boxes (Exons 10-16) represent exons which code for the seven fibronectin type Ill-like repeats.

Figure 4A-E. Computer generated list of candidate sequences for the splice junction of exons 14 and 16.

Figure 5A-E. Computer generated list of candidate sequences for the splice junction of exons 15 and 16.

Figure 6. Percent invasion of U87 cells treated with or without 10 μM Tenascin

C antisense phosphothioate oligonucleotides.

Figure 7. Percent invasion of U373 cells treated with or without 10 uM Tenascin C antisense phosphothioate oligonucleotides.

ft Figure 8. Immunoblot of total protein of Tenascin-C.

DETAILED DESCRD7TION OF THE PREFERRED EMBODIMENTS OF THE INVENTION

The present invention provides a method for rationally and systematically designing oligodeoxynucleotides (ODNs) or long antisense expression vectors which produce long antisense RNA sequences to selectively block the translation of mRNA containing targeted isoform sequences. The present invention also provides methods of use of the designed ODNs and long antisense constructs, including clinical diagnostic uses whereby the ODNs are used to detect the mRNA containing targeted isoform sequences of a protein, and gene medicine uses whereby the ODNs and long antisense constructs are used to inhibit the translation of mRNA encoding targeted isoform sequences of a protein. By "targeted isoform sequences" we mean sequences of mRNA which are unique to a particular isoform of a protein, or to a particular family of isoforms of a protein, the isoforms being generated by differential or alternative splicing of the pre-mRNA which encodes the protein.

In a preferred embodiment of the present invention, the specific isoforms are those of the protein Tenascin-C. Tenascin-C is an example of a protein which exhibits various alternatively spliced isoforms with differing biological functions. Tenascin C is a oligomeric extracellular matrix (ECM) protein composed of six similar subunits joined together at their amino termini by disulfide bonds (Borsi 1994). Each of the subunits consists of a cysteine-rich amino-terminal domain which is involved in oligomerization, a series of epidermal growth factor (EGF)-like repeats, followed by a series of fibronectin type m-like repeats (termed A1-A4, B, C, and D), and a fibrinogen-like globular domain (Figure 2, A and B). The number of fibronectin type Ill-like domains varies among isoforms, thus accounting for the eight different isoforms which have so far been identified (Wilson, 1996). Two of the isoforms predominate. The larger of the two contains seven alternatively spliced fibronectin type Ill-like domains (Figure 2A) while the smaller isoform contains no fibronectin type IE- like domains (Figure 2B). The six other isoforms contain from one to six of the fibronectin type Hi-like domains. Figure 3 is a schematic representation of Tenascin-C unspliced pre-mRNA containing all 27 exons. The black boxes correspond to Exons 10-16, which code for the seven fibronectin type Ill-like domains which are differentially spliced.

During normal embryogenesis, Tenascin C is present within the developing central nervous system (CNS), in the extracellular matrix which lines the pathways of migratory cells, in mesenchyme at sites of mesenchymal- epidermal interaction, and in developing connective tissue (Chiquet-Ehrismann,

1995). However, in adult tissue, the expression of Tenascin-C is attenuated, and only low amounts of the smaller isoform are found in normal tissue (Leprini 1994). The larger isoform is found in adult tissue at only two locations: in the ECM of healing wounds and in tumors (Higuchi 1993). For example, the cells of glioblastomas (highly malignant primary neoplasms of the brain) make only the

"long isoform" of the Tenascin-C protein, in contrast to normal adult tissue which makes only the "short isoform". This difference is thought to be partly responsible for the ability of glioblastoma cells to invade normal brain and to divide rapidly, features that make these tumors highly malignant. This may be due to properties of the long isoform of Tenascin-C which are believed to promote tumor growth. For example, the so-called "anti-adhesive" effects reduce the adhesion of tumor cells to other matrix proteins, promoting their mobility and invasiveness. Tenescin-C also causes the secretion of matrix metalloproteinases (MMPs) which degrade the ECM, making it more susceptible to tumor invasion (Mackie 1997). The protein may also play a role in angiogenesis, thus providing proliferating tumors with new blood supplies.

While the Tenascin-C isoforms are targeted in preferred embodiments of the present invention, those of skill in the art will recognize that the methods of the present invention may be practiced with a wide variety of proteins whose sequences are encoded by exons which are spliced, for example CD44 and cyclin Dl.

In preferred embodiments of the present invention, the unique sequences which are targeted may be derived from an exon which is differentially expressed, or from a region of the mRNA that encompasses the splice site between two exons. In both cases, the targeted sequences may be differentially expressed in, for example, certain cell types, or at certain times in the cell cycle, or at certain times during the development of the organism, or during particular states such as a disease state, pregnancy, etc. However, it is not necessary that the targeted isoform sequences be expressed at a time different from that of other isoform sequences. The invention also encompasses the targeting of a specific isoform or family of isoforms which are expressed concomitantly with other isoforms. Those of skill in the art will recognize that the ODNs and long antisense constructs of the present invention may be used to detect or inhibit a specific isoform or family of isoforms within a mixture of many isoforms or families of isoforms. The present invention provides methods for the design and use of antisense ODNs or long antisense constructs encoding antisense RNAs which hybridize to any mRNA sequence which codes uniquely for a specific isoform of a protein, or a family of isoforms of a protein, regardless of whether or not the isoform is uniquely expressed in a temporal sense.

The ODNs and long antisense RNA sequences of the present invention may be designed to inhibit the translation of a single isoform of a protein, or to inhibit the translation of more than one isoform of a protein, (i.e. a "family of isoforms") so long as the mRNA of all the isoforms to be inhibited contains the targeted isoform sequence.

The present invention provides a method for the design of ODNs which are complimentary to targeted isoform sequences. According to the method of ODN design of the present invention, regions of mature mRNA which are unique for a given isoform of a protein must be identified. For example, the unique region may be an exon which is differentially expressed, or a splice site region between two exons. In a preferred embodiment, a list of all sequences of varying lengths which are complimentary to the sense strand of the region of interest was generated using the Excel program. Excel was programmed to generate a set of potential antisense ODN sequences for each given length which differ in register by one nucleotide. Examples of such sets for Tenascin-C are given in Figures 4A-4E (splice site junction of exons 14 and 16) and Figures 5A- 5E (splice site junction of exons 15 and 16). In a preferred embodiment of the present invention, the software program is Primer Premier v4.04, by Premier Biosoft International. However, those of skill in the art will recognize that several equivalent primer design software packages exist and may be used in the practice of the present invention. Potential ODN sequences are then analyzed by a second software program to determine their binding affinity to the targeted mRNA sequence, and potential secondary structure formation (e.g. hair pin loops) or potential homodimer formation, which would interfere with the binding reaction. In a preferred embodiment of the present invention, the software utilized for this analysis was Oligotech, by Oligo Therapeutics, Inc. However, those of skill in the art will recognize that several equivalent software programs exist and may be used in the practice of the present invention. The best sequences for use as antisense ODNs are selected based on the above parameters and the desired ODN length. For example, longer sequences may have higher binding affinities, but they may be more costly to synthesize. The final choice of which sequences to utilize is up to the discretion of the trained investigator. Those of skill in the art will recognize the salient factors which are taken into account when making such a decision. Finally, the specificity of the sequences is analyzed using a search engine capable of accessing an extensive database of appropriate DNA sequences. The specificity is analyzed in order to ensure that the selected ODN sequences are indeed specific for the isoform of interest and do not hybridize to other known sequences within the genome of the species of interest. In a preferred embodiment of the present invention, the BLAST search engine which is maintained by the National Center for Biotechnology Information (NCBI) was utilized. However, those of skill in the art will recognize that other appropriate search engines are also available. Any suitable software programs and search engines may be used in the practice of the present invention, so long as the resulting ODNs are able to specifically and effectively bind to and inhibit the expression of, or to effect the detection of, the targeted isoform sequences.

In embodiments of the present invention, the ODNs are synthetic, single- stranded oligodeoxynucleotides. In a preferred embodiment of the present invention, the ODNs are phosphothioate ODNs. However, one of skill in the art will recognize that many types of ODNs of varying chemistries or modifications are known and suitable in the practice of the present invention. Any type of single-stranded ODN which has the ability to bind to the targeted mRNA sequence and inhibit its translation is encompassed by the methods of the present invention.

The antisense ODNs of the present invention are single-stranded ODNs which are approximately 10 to 25 nucleotides in length. However, the exact length of the ODN is not critical to the methods of the present invention. The choice of the length of the designed ODN will vary depending on the nature of the targeted isoform sequence, the exact sequence and composition of the ODN itself, the affinity of the ODN for the target sequence, the cost of synthesis of the ODN, and various other factors which are well-known to those of skill in the art. Those of skill in the art will recognize that ODNs of many different lengths are comprehended by the methods of the present invention.

The present invention also encompasses a method of design of long (or "full-length") antisense expression vectors which encode long (or "full-length") antisense RNA sequences. The first steps of the design process are similar to those for antisense ODN design as described above. Regions of mature mRNA which are unique for a given isoform of a protein must be identified and candidate antisense sequences may be generated by computer as described for antisense ODNs. However, the candidate sequences in this case will be much longer than those of the antisense ODNs. For example, in long antisense, the sequence of the target gene may be several hundred to several thousand thousand base pairs in length. The candidate sequences are analyzed for their percent homology to targeted isoforms and those isoforms not to be targeted in order to optimize selectivity. Appropriate antisense sequences will display relatively high sequence homology to targeted isoforms and relatively low sequence homology to non-targeted isoforms. Techniques for ascertaining sequence homology and reasonable criteria for assessing homology are well known to those of skill in the art. Appropriate sequences are selected by the investigator and the selected sequences are then searched for appropriate restriction enzyme sites within the sequences in order to plan for cloning of the selected sequences into appropriate expression vectors. The cDNA of a given selected sequence is then obtained either cloning or by utilizing polymerase chain reaction (PCR) techniques and the cDNA is subcloned in reverse orientation into an appropriate expression vector, inducible vector, or viral vector, forming a long antisense expression vector. The DNA must be cloned so that when it is expressed within the cell, the RNA which is transcribed is the compliment

(antisense) of the targeted mRNA which is to be inhibited. Examples of appropriate expression vectors are various viral vectors or constituitive or inducible expression vectors. Many such expression vectors and methods of cloning DNA into them are well known to those of skill in the art. The long antisense expression vector is introduced into the cell which contains the targeted mRNA by one of several techniques which are well known to those of skill in the art, for example by transfection or transduction. Other carrier * molecules, for example lipids, may also be employed to effect introduction of the long antisense expression vector into the cell. The methods involved in the cloning and expression of long ("full-length") antisense RNA and well known to those of skill in the art. For an example, see Kadono et al., 1998, and references therein.

Those of skill in the art will recognize that the exact sequences of ODNs and long antisense constructs designed according to the method of the present invention will vary from case to case according to the mRNA sequence which is to be inhibited. Also, those of skill in the art will recognize that the sequence of the ODNs and long antisense constructs will vary according to other factors, for example, the GC content of the ODN (GC base pairing is more stable than AT base pairing), the desired affinity of the the antisense RNA for the target sequence, the elimination of potential intramolecular secondary structure within the ODN itself, and the like. The sequence of the antisense RNA may be precisely complimentary to that of the targeted isoform sequence, or may be modified according to the above-mentioned parameters, and other parameters which are well-known to those of skill in the art.

In a preferred embodiment of the present invention, the targeted sequences are from a particular isoform (the "long" isoform) of the Tenascin-C protein. In particular, the targeted sequences are derived from Exon 15, (fibronectin type Ill-like repeat "C"), which is the first exon spliced out of the Tenascin-C pre-mRNA during the formation of intermediate isoforms (Chiquet-

Ehrismann, 1995; Wilson, 1996; Gulcher, 1997). Exon 15 is thus unique to the "long" isoform which contains all known exons. The two ODNs which are complimentary to targeted sequences within Exon 15 are referred to as "ten-1" and "ten-2" (SEQ IDs #1 and #2 and Table 1). In another preferred embodiment of the present invention, an ODN was designed to specifically target the shortest isoform of Tenascin-C. The ODN

(SEQ ID #3, or "ten-3" of Table 1) is complementary to the splice site region between exons 9 and 17. In yet another preferred embodiment of the present invention, a Tenascin-C isoform of intermediate length was targeted by designing an ODN complimentary to the splice site region between Exon 14(B) and Exon 16(D). The sequence is given in SEQ ID #4 (also referred to as "ten- 4" in Table 1).

The present invention also provides methods for the use of the ODNs and long antisense RNA sequences which are prepared in the above-described manner. One such use is the detection of specific isoforms of a protein. For example, the ODNs and long antisense RNA sequences may be useful for investigative purposes in laboratory settings, or for clinical diagnostics. In a preferred embodiment of the present invention, SEQ ID #1 and SEQ ID # 2 can be used to detect the presence of the long form of Tenascin-C in brain tissue biopsies which are suspected of containing cancerous glioblastoma cells.

In a preferred embodiment of the present invention, the ODNs and long antisense constructs designed by the methods of the present invention may be used in gene medicine applications. For example, ODNs and long antisense constructs which specifically bind to and inhibit the long form of Tenascin-C may be used to treat malignant glioblastomas and other diseases.

EXAMPLES

Methods Matrigel Invasion Asays: Oligonucleotides were purchased from Oligotech (OR) with phosphothioate modifications. Cells (100 cells/μl) in

DMEM containing 2% fetal bovine serum (FBS) +/- 10 μM antisense ODNs were plated on Transwell inserts (Costar, Cambridge, MA; 12 mm, 8 um pore size) coated with 100 μl of a Matrigel solution (Collaborative Research, Bedford MA; 1 mg/ml in DMEM). The lower chamber of the Transwell was filled with 500μl of DMEM containing 2% FBS. Following a 48 hour time point inserts were fixed in methanol and stained with Eosin Y and methylene blue. To obtain percent invasion, cells (bottom and upper of insert filter) were counted, then the cells on the upper side of insert were removed with a cotton swab and the number of cells that had migrated to the lower side were counted. For each insert (three in each experimental condition), eight random fields were selected, and cells were counted on an inverted microscope using a 20X objective. Cells that had invaded to the bottom side were divided by the total number of cells (%

Invasion).

EXAMPLE 1. Design of ODNs Specific for Various Tenascin-C Isoforms A. Design of ODNs Specific for Exon 15 of Tenascin-C

Tenascin-C mRNA was analyzed to determine features which were unique to each isoform. Since the longest isoform contains all of the alternative splicing domains, any antisense ODN targeted to any of those domains should inhibit the long isoform. However, in order to eliminate inhibition of intermediate-sized isoforms which contain some of the domains, Exon 15 was targeted. Exon 15 (fibronectin type Ill-like repeat C) is the first exon spliced out of the mRNA to form the smaller, intermediate isoforms (Chiquet-Ehrismann,

1995; Wilson, 1996; Gulcher, 1997). It is therefore present almost exclusively in the longest isoform.

Potential antisense ODNs complimentary to sequences within Exon 15 were designed using Primer Premier v4.04 (Premier Biosoft International). The potential ODN sequences were then analyzed by the Oligotech program (Oligo

Therapeutics, Inc.) to determine their binding affinity for Exon 15, possible intrastrand secondary structure, and potential homodimer formation. Two ODNs, ten-1 and ten-2, were chosen. The sequences of these 16-mers are given in Table 1.

To determine the specificity of these ODNs, a BLAST search was ft conducted at the internet site: (htt : //www, ncbi. nlm. gov) of the National Center for Biotechnology Information (NCBI). This database search determines whether the sequence of interest matches any known human genome sequences.

ODNs ten-1 and ten-2 matched only human Tenascin-C mRNA. These ODNs are therefore likely to be specific for the longest isoform of Tenascin-C.

Another ODN complimentary to portions of the fibronectin type Ill-like repeats coded by Exon 8 and Exon 18 was also synthesized. After design and optimization using the Primer Premier and Oligotech software, the control ODN was analyzed using BLAST and, as would be predicted, did not match any known sequences.

B. Design of ODNs Specific for the Smallest Isoform of Tenascin-C

The smallest isoform of Tenascin-C contains no fibronectin type Ill-like domains. To target this isoform, an ODN was designed which was complementary to the splice site junction region between exons 9 and 17. Primer Premier and Oligotech software were used to optimize the sequence and the resulting ODN sequence (ten-3) is depicted in Table 1. A BLAST search of this sequence returned no matches. This is consistent with this sequence being unique for the smallest isoform, since BLAST searches only the full-length unspliced pre-mRNA, and this particular sequence is generated by the splicing of exons 9 and 17.

C. Design of ODNs Specific for an Intermediate Isoform of Tenascin-C.

An ODN was designed to target the largest intermediate isoform of Tenascin-C, which is missing Exon 15(C). Recall that Exon 15 is usually the first exon to be spliced out of the pre-mRNA. Thus, this ODN is designed to hybridize to the splice site junction region between Exon 14 and Exon 16. The sequence of this ODN, ten-4, which is depicted in Table 1, was also analyzed and optimized using the Primer Premier and Oligotech software, and searched using BLAST.

TABLE 1. Sequences of Tenascin-C derived Antisense ODNs and their Location of Hybridization.

EXAMPLE 2. Effect of ODNs on Invasiveness of Glioblastoma Cells

A well characterized matrigel invasion assay was used to study the effects of the antisense phosphothioate oligonucleotides listed in Table 1 on the invasiveness of tumor cells. Two glioma tumor cell lines were treated with or without antisense phosphothioate oligonucleotides directed to different regions of the Tenascin C gene as described in the Methods section. The results for glioma tumor cell lines U87 and U373 are given in Figures 6 and 7, respectively.

As can be seen, for the U87 cells, there was a significant decrease in the percent of cells that invaded the matrigel coated membrane inserts in cells treated with tenl, ten3, ten4, and ten5 compared to cells treated with no oligo or with 10% serum. However, U87 cells treated with ten2 had a large standard deviation and it is therefore not known if the effect seen with ten2 is significant. For the U373 cells, there was a significant decrease in the percent of cells that invaded the matrigel coated membrane inserts in cells treated with Ten4 compared to control and cells treated with Tenl and Ten2. The effect of the ODNs on invasiveness thus appears to vary with tumor cell type.

Example 3. Immunoblot of Total Protein using anti-Tenascin-C Antibody An immunoblot of total protein from U87 cells treated with or without the ODNs of Table 1 was probed with an antibody for Tenascin-C as described in Methods. The results are shown in Figure 8. As can be seen, lower amounts of Tenascin-C, especially the high molecular weight (220 kilodalton) form are produced in cells exposed to ODNs ten-2 and ten-4, when compared to control cells, or cells treated with other ODNs.

While the invention has been described in terms of its preferred embodiments, those skilled in the art will recognize that the invention can be practiced with modification within the spirit and scope of the appended claims. Accordingly, the present invention should not be limited to the embodiments as described above, but should further include all modifications and equivalents thereof within the spirit and scope of the description provided herein.

REFERENCES Borsi, L., Balza, E., Castellani, P. et al.; "Cell Cycle Dependant Alternative Splicing of the Tenascin Primary Transcript"; Cell Adhesion and Communication. 1994.

Chiquet-Ehrismann, R.; "Tenascins, a growing family of extracellular matrix proteins": Exoerientisi: 51: 853-62; 1995.

Gulcher, J. R., Nies, D. E., Alexakos, M. J., Ravikant, N. A. et al.; "Structure of the human hexabrachion (tenascin) gene"; Proceedings of the National Academy of Science US A: 88: 9438-9442; 1991.

Higuchi, M., Ohnishi, N., Arita, S. et al. ; "Expression of tenascin in human gliomas: its relation to histological malignancy, tumor dedifferentiation and angiogenesis: Acta Neuropatholoaicd 1993.

Kadono, et al. "Membrane type 1 -matrix metalloproteinase is involved in the formation of hepatocyte growth factor/scatter factor-induced branching tubules in madin-darby canine kidney epithelial cells. Biochemical and Biophvical Research Communications 251: 681-687, 1998.

Leprini, A., Querze, G, and Zardi, L.; "Tenascin Isoforms: Possible Targets for

Diagnosis and Therapy of Cancer and Mechanisms Regulating Their Expression" ; Perspectives on Developmental Neurobiologv. 1994.

Mackie, E. J.; "Molecules in Focus: Tenascin-C"; International Journal of Biochemistry and Cell Biology: 29: 1133-37; 1997.

Wilson, K. E., Langdon, S. P., Lessells, A. M., and Miller, W. R.; "Expression of the extracellular matix protein tenascin in malignant and benign ovarian tumors: British Journal of Cancer: 74: 999-1004; 1996.

Claims

We claim: ft 1. A method for preparing an antisense oligodeoxynucleotide sequence for blocking translations of a specific isoform of a protein which can be expressed as a plurality of different isoforms of said protein, comprising: identifying a nucleotide sequence for a specific region of mRNA which is required to encode for said specific isoform of said protein, but which is not required to code for other isoforms of said plurality of different isoforms of said protein; and producing at least one hybridizing nucleotide sequence which hybridizes to said nucleotide sequence identified in said identifying step.

2. The method of claim 1 wherein said nucleotide sequence identified in said identifying step is specific for an exon required only for producing said specific isoform of said protein.

3. The method of claim 1 wherein said nucleotide sequence identified in said identifying step is specific for a splice region between two exons.

4. The method of claim 3 wherein said two exons are adjacent to one another in an mRNA which codes for a longest isoform of said plurality of isoforms of said protein.

5. The method of claim 3 wherein said two exons are separate from one another in an mRNA which codes for a longest isoform of said plurality of isoforms of said protein.

6. The method of claim 1 wherein said protein is selected from the group consisting of Tenascin C and CD44.

7. The method of claim 1 wherein said hybridizing sequence produced in said producing step is 10 to 25 nucleotides in length.

8. The method of claim 1 wherein said hybridizing sequence produced in said producing step includes a sequence which hybridizes to a splice region between two exons.

9. The method of claim 8 wherein said two exons are adjacent to one another in an mRNA which codes for a longest isoform of said plurality of isoforms of said protein.

10. The method of claim 8 wherein said two exons separate from one another in an mRNA which codes for a longest isoform of said plurality of isoforms of said protein.

11. The method of claim 1 wherein said hybridizing nucleotide sequence is specific for said specific region of mRNA.

12. The method of claim 1 wherein said step of producing said hybridizing nucleotide sequence includes the steps of: generating a computerized listing of candidate sequences which hybridize to said specific region of mRNA; analyzing said candidate sequences for binding affinity, internal secondary structure and homodimer formation; and determining specificity of said candidate sequences.

13. A method for preparing an antisense oligodeoxynucleotide sequence for blocking translations of a specific isoform of Tenascin-C protein, comprising: identifying a nucleotide sequence for a specific region of mRNA which is required to encode for said specific isoform of Tenascin-C, but which is not required to code for other isoforms of Tenascin-C; and producing at least one hybridizing nucleotide sequence which hybridizes to said nucleotide sequence identified in said identifying step.

14. The method of claim 13 wherein said nucleotide sequence identified in said identifying step is specific for exon 15 of Tenascin-C.

15. The method of claim 13 wherein said nucleotide sequence identified in said identifying step is specific for a splice region between two exons of Tenascin-C.

16. The method of claim 15 wherein said two exons are exon 15 and exon 16.

17. The method of claim 15 wherein said two exons are exon 14 and exon 16.

18. The method of claim 15 wherein said two exons are exon 9 and exon 17.

19. A nucleotide sequence as set forth in Sequence ID No. 1.

20. A nucleotide sequence as set forth in Sequence ID No. 2.

21. A nucleotide sequence as set forth in Sequence ID No. 3.

22. A nucleotide sequence as set forth in Sequence ID No. 4.

23. A nucleotide sequence selected from the group consisting of Sequence ID Nos. 5 to 532.

24. A method for preparing an antisense oligodeoxynucleotide sequence for blocking translations of a specific family of isoforms of a protein which can be expressed as a plurality of different isoforms of said protein, comprising: identifying a nucleotide sequence for a specific region of mRNA which is required to encode for said specific family of isoforms of said protein, but which is not required to code for other isoforms of said plurality of different isoforms of said protein; and producing at least one hybridizing nucleotide sequence which hybridizes to said nucleotide sequence identified in said identifying step.

25. A method for preparing a long antisense expression vector which encodes a long antisense RNA sequence for blocking translations of a specific isoform of a protein which can be expressed as a plurality of different isoforms of said protein, comprising: identifying a nucleotide sequence for a specific region of mRNA which is required to encode for said specific isoform of said protein, but which is not required to code for other isoforms of said plurality of different isoforms of said protein; and producing at least one long antisense expression vector which encodes a long antisense RNA sequence which hybridizes to said nucleotide sequence identified in said identifying step.

26. The method of claim 25 wherein said step of producing said long antisense expression vector includes the steps of: generating a computerized listing of candidate sequences which hybridize to said specific region of mRNA; analyzing said candidate sequences for homology to said specific region A; locating restriction enzyme sites for cloning, and cloning the long antisense sequence into a vector.