EP1516057A2 - Nucleic acid isolated from the 5' non translated region (5'ntr) end of the human fgf-1 gene having ires activity - Google Patents

Abstract

The invention concerns an isolated nucleic acid of the human FGF-1 gene having IRES (internal ribosome entry site) activity, said nucleotide sequence being derived from the 5' non translated region (5'NTR) end of the human FGF-1 gene. The invention also concerns an expression cassette of at least one molecule of interest in a eukaryotic cell comprising at least one of said sequences. The cassette can be inserted directly or in the form of a vector in a host cell, hence the invention further concerns the recombinant vectors and the host cell transfected by said vector or the cassette.

Description

5'NTR ISOLATED NUCLEIC ACID FROM THE HUMAN FGF-1 GENE WITH IRES ACTIVITY

The invention relates to an isolated nucleic acid having IRES (internai ribosome entry site) activity, said nucleotide sequence originating from the 5 'untranslated end (5'NTR) of the human FGF-1 gene. The subject of the invention is also a cassette for the expression of at least one molecule of interest in a eukaryotic cell, said cassette comprising the IRES of FGF-1 and making it possible to coexpress several molecules from the same mRNA. The cassette can be introduced directly or in the form of a vector into a host cell so that the recombinant vectors and the host cells transfected by said vector or the cassette are also part of the invention.

The regulation of gene expression occurs at several levels, transcription, maturation, stability of F mRNA but also translation. Translation of an mRNA is subject to strict regulations and proves to be a key step in the regulation of gene expression. For example, increased translation of the proto-oncogene c-myc is associated with Bloom syndrome (20). This aberrant translational control could contribute to the increased risk that these patients run of developing cancers.

Translation-level regulations involve numerous proteins and are exerted mainly via the 5 'and 3' untranslated regions (NTR) without excluding, however, certain structural elements of the coding part of the mRNAs. The translation initiation stage is also subject to regulations.

The classical initiation of translation is done by recognition of the cap (or CAP) located at the 5 'end of the transcript. This structure allows the setting of the translation initiation factors. In this step, it is the eIF4F complex which plays the role of adapter between the cap and the small subunit of the 40S ribosome. This complex is made up of three factors, respectively eIF4E (“Cap binding protein”), eIF4A (an RNA helicase) and eIF4G which plays a bridge role, between the other translation initiation factors, the mRNA and the small 40S subunit. Once recruited, the latter will “scan” the mRNA until it encounters a translation initiation codon (canonical or not). After which, there will be association of the 60S subunit with 40S and the start of translation of the reading frame to a stop codon, where the ribosome dissociates.

Certain messenger mRNAs lack a cap and therefore cannot be translated by the mechanism described by Marylin Kozak. This is the case with picomaviral RNAs (14), where another translation initiation system has been highlighted, involving an internal ribosome or IRES entry site at the 5 'non-region. translated (5'NTR). This site allows the ribosomes to bind directly inside the mRNA, and this by a mechanism still poorly understood. Since then IRES have been identified in the RNA of other groups of viruses such as flaviviruses, pestiviruses, tobamoviruses and retroviruses (19).

The involvement of cellular proteins for the translation of these viral RNAs motivated the search for this internal entry mechanism at the level of transcripts of cell origin. In 1991, such a site was characterized in the BIP (Immunoglobulin heavy-chain binding protein) (8) and more recently in that of the fibroblastic growth factor FGF-2 (18), of the proto-oncogene c-myc ( 12), PDGF growth factor for “Platelet Derived Growth Factor” (2) and VEGF “Vascular Endothelial Growth Factor” (6).

The FGF-2 mRNA is provided with a long 5'NTR of approximately 500 nucleotides in which alternative initiations of translation have been demonstrated at the level of non-canonical codons CUG (1), which results in the formation of five isoforms of FGF-2. An internal ribosome entry site was also located between nucleotides 154 and 318 (unpublished data).

VEGF mRNA also has a long untranslated sequence in which has been demonstrated, very recently, the presence of an alternative initiation at the level of a non-canonical codon CUG. In this 5'NTR, the presence of two IRES has been demonstrated, one is located upstream of the CUG and the other precedes the classic initiation codon AUG (6).

Surprisingly, these two growth factors have their growth controlled at the translational level by IRES, which have an activity in angiogenesis. These two factors are induced during the neovascularization process of a developing tumor. There is a third major angiogenic factor, FGF-1, the translation of which has not been studied to date.

However, the Applicants have discovered that the initiation of translation of the FGF- mRNA

I involved an internal entry site located in the 5 'untranslated region. In addition, they found that the FGF-1 transcript A in particular unexpectedly exhibited better in vivo IRES activity (mouse muscle) than the other cellular RNA ERES tested to date.

The genomic structure (11) of FGF-1 comprises 3 exons containing the coding sequence (fig. N ° 1). The latter undergo splicing which results in the formation of a unique reading framework. Exon 3 contains the 3'NTR end which is provided with a polyadenylation site.

There are four different promoters for the human FGF-1 gene, which when activated, result in the formation of four different primary transcripts. Following maturation, the only difference between these four transcripts is at their respective 5'NTR (fig. N ° 2). Exon 1 is provided with a splice acceptor site which binds one of the 4 non-coding exons (A, B, C, D) following the activation of their respective promoter.

Transcripts A and B are the majority transcripts which are mainly expressed in the brain, kidney and retina (fig. N ° 2). These two transcripts have respectively 5 'NTR of 435 and 147 nucleotides.

Transcripts C and D are minor transcripts expressed very weakly in vivo (11). However, they are present in lines derived from glioblastoma and can be induced in fibroblasts or smooth muscle cells of vascular origin when subjected to serum deprivation or PMA treatment (phorbol 12-myristate 13 -acetate ). These two transcripts have respectively 5 'NTRs of 149 and 92 nucleotides.

The existence of these four mRNA variants coding for FGF-1 is common to humans, mice (9), and beef (16). The comparison of the sequences of these 5 ′ human NTRs showed good homology with those of the ox and the mouse (9). Indeed, the alignment of the sequences of the 5'NTR of the FGF-1 of these various organisms (9, 11, 16) has shown good conservation of these non-coding sequences, with almost 70% similarity for the 5 'NTR A. Such conservation in the sequences of untranslated zones makes it possible to envisage an important function of these 5'NTRs, regions usually subjected to a relatively low selection pressure and therefore to rapid evolution.

The IRES activity of the 5'NTRs of each of the transcripts A, B, C and D of human and murine FGF-1 is demonstrated in the following description (example 1 B1).

Essentially, the presence of dRES is sought by transfecting Cos-7 cells with CAT-I-NuCAT vectors, containing the 5 'NTRs of FGF-1, then by subjecting the transfected cells to a western blot. The expression of the protein NuCAT is observed in the presence of transcripts A, B, C and D and testifies to the IRES activity of these 4 transcripts (qualitative analysis). Due to the strong homology of the sequences of the human 5 'NTRs with those of the ox and in particular of the mouse (9), the presence of IRES sites is also deduced therefrom in murine and bovine FGF-1.

The in vitro IRES activity of the 5 'NTRs of each of the human FGF-1 transcripts is also studied according to the type and cell state (quantitative analysis) (example 1B1).

To do this, the IRES activity is measured by means of bicistronic vectors coding for reporter genes (in practice the LucF / LucR genes) in different lines. The calculation of the LucF / LucR ratio provides information on the internal input capacity of the 5 'NTR sequence and defines the IRES activity of this sequence. Among the various results, a very good IRES activity of the transcript A is observed in the Hela and Sk-Hep-1 cells, which is greater than that of 1TRES of the FGF-2.

The Applicants have also compared, in vitro, on human cells, the IRES activity of the structure of the human 5 'NTR (transcribed A and B) with that of the murine 5' NTR (example 1 B2). The transfections were carried out in Cos-7, SK-Hep-1 and Saos-2 cells, to make a direct comparison with human sequences. The experiments showed that the IRES activity profiles of the human and murine variants A and B were similar.

The invention consists not only in having demonstrated for the first time the IRES activity of the 5 'NTRs of FGF-1, but also, as will be demonstrated below, in having characterized the IRES sites in the various transcripts A, B , C and D of human FGF-1 (Example 2). More particularly, the Applicant has succeeded in identifying the smallest sequence of each of the transcripts A, B, C, D having IRES activity, both in vitro and in vivo. To do this, different continuous nucleotide sequences originating from the messenger RNA of FGF-1 were subcloned into a bicistronic vector coding for luciferase reporter genes. Different mammalian and human cell lines (in vitro) were transfected and the mouse muscle (in vivo) was transduced by these vectors.

The IRES activity of human FGF-1 and, by homology, of murine FGF-1, is therefore disclosed for the first time in the present application. This activity has been demonstrated for each transcript A, B, C, D as well as for certain fragments of the 5 'NTR of these transcripts.

In the following description and in the claims, by the expression “homologous sequence” is meant a sequence having at least 60% identity, advantageously 80% preferably 95% with a given sequence, the homology resulting in this case of variability from one species to another, in particular from mice to humans.

Therefore, the invention relates first of all to the use for its IRES activity of an isolated nucleic acid containing a sequence identical or homologous to the sequence SEQ ID 1 located between nucleotides 1 and 435 of the 5 'NTR of transcript A of human FGF-1 mRNA and hereinafter called A. This sequence corresponds to the entire sequence of the 5 'NTR of transcript A.

As already stated, the Applicant has also succeeded in identifying, on transcript A, nucleotide sequences of size smaller than the sequence SEQ ID 1, exhibiting IRES activity, which therefore suggests that the IRES site would be included in these sequences.

Thus, the subject of the invention is also, in particular for use for its IRES activity:

an isolated nucleic acid exhibiting IRES activity comprising a sequence identical to the sequence SEQ O 2, located between nucleotides 1 and 390 of the sequence SEQ ID 1 and hereinafter referred to as A2,

an isolated nucleic acid exhibiting IRES activity comprising a sequence identical or homologous to the sequence SEQ LD 3, located between nucleotides 39 and 435 of the sequence SEQ ID 1 and hereinafter referred to as A3,

- an isolated nucleic acid exhibiting IRES activity comprising a sequence identical or homologous to the sequence SEQ ID 4, located between nucleotides 122 and 435 of the sequence SEQ ID 1 and hereinafter referred to as A4, - an isolated nucleic acid exhibiting activity IRES comprising a sequence identical to or homologous to the sequence SEQ ID 5, located between nucleotides 214 and 435 of the sequence SEQ ID 1 and hereinafter called A5,

- an isolated nucleic acid exhibiting IRES activity comprising a sequence identical or homologous to the sequence SEQ ID 6, located between nucleotides 122 and 390 of the sequence SEQ ID 1 and hereinafter called A6, - an isolated nucleic acid exhibiting activity IRES comprising a sequence identical or homologous to the sequence SEQ LD 7, located between nucleotides 214 and 390 of the sequence SEQ ID 1 and hereinafter referred to as A7,

an isolated nucleic acid exhibiting IRES activity comprising a sequence identical to or homologous to the sequence SEQ ID 8, located between nucleotides 269 and 435 of the sequence SEQ ID 1 and hereinafter referred to as A8,

an isolated nucleic acid exhibiting IRES activity comprising a sequence identical or homologous to the sequence SEQ ID 9, located between nucleotides 269 and 390 of the sequence SEQ LD 1 and hereinafter called A9.

an isolated nucleic acid exhibiting IRES activity comprising a sequence identical or homologous to the sequence SEQ LD 12, located between nucleotides 1 and 366 of the sequence SEQ LD 1 and hereinafter called A12,

- an isolated nucleic acid having an IRES activity comprising a sequence identical or homologous to the sequence SEQ ID 16, located between nucleotides 122 and 366 of the sequence SEQ LD 1 and hereinafter referred to as Al 6, - an isolated nucleic acid having a IRES activity comprising a sequence identical or homologous to the sequence SEQ ID 17, located between nucleotides 214 and 366 of the sequence SEQ ID 1 and hereinafter referred to as A1 7,

an isolated nucleic acid exhibiting IRES activity comprising a sequence identical or homologous to the sequence SEQ ID 18, located between nucleotides 269 and 366 of the sequence SEQ ID 1 and hereinafter referred to as Al 8,

an isolated nucleic acid exhibiting IRES activity comprising a sequence identical or homologous to the sequence SEQ ID 19, located between nucleotides 1 and 344 of the sequence SEQ ID 1 and hereinafter called A 19,

an isolated nucleic acid exhibiting IRES activity comprising a sequence identical or homologous to the sequence SEQ ID 20, located between nucleotides 1 and 330 of the sequence SEQ ID 1 and hereinafter called A20, The nucleic acid corresponding to the sequence (SEQ ID 18 (A18) constitutes the smallest fragment of the transcript A of human FGF-1 exhibiting IRES activity. For a sequence of smaller size, of the order of 100 nucleotides, there is n does not obtain significant IRES activity, at least in human adenocarcinoma cells. Indeed, and as is apparent from the examples below, the IRES activity is not significant for the following fragments:

- sequence identical or homologous to the sequence SEQ ID 10, located between nucleotides 303 and 435 of the sequence SEQ ID 1 and hereinafter called A 10, - sequence identical or homologous to the sequence SEQ LD 11, located between nucleotides 303 and 390 of the sequence SEQ ID 1 and hereinafter called Ail,

- sequence identical or homologous to the sequence SEQ LD 13, located between nucleotides 282 and 435 of the sequence SEQ ID 1 and hereinafter referred to as Al 3,

- sequence identical or homologous to the sequence SEQ ID 14, located between nucleotides 282 and 390 of the sequence SEQ ID 1 and hereinafter called A14,

- sequence identical or homologous to the sequence SEQ ID 15, located between nucleotides 282 and 366 of the sequence SEQ ID 1 and hereinafter referred to as A15.

As already stated, the Applicant has demonstrated that the 5 ′ NTR end of the B transcript of human FGF-1 exhibits significant IRES activity, in particular in myoblasts.

Consequently, the invention also relates to an isolated nucleic acid exhibiting an IRES activity which is characterized in that it contains a sequence identical or homologous to the sequence SEQ ID 30 located between nucleotides 1 and 147 of the 5 'NTR of transcript B of human FGF-1 mRNA and hereinafter referred to as B. In other words, the sequence SEQ ID 30 corresponds to the entire sequence of the 5 'NTR of transcript B.

The Applicant has also succeeded in identifying, on transcript B, nucleotide sequences of size smaller than the sequence SEQ ID 30, exhibiting an activity IRES, which therefore suggests that the IRES site would be included in these sequences.

Thus, the subject of the invention is also:

- an isolated nucleic acid exhibiting IRES activity comprising a sequence identical to the sequence SEQ ID 31, located between nucleotides 1 and 93 of the sequence

SEQ ID 30 and hereinafter referred to as B2,

The nucleic acid corresponding to the sequence SEQ ID 31 (B2) constitutes the smallest fragment of the transcript B of human FGF-1 exhibiting IRES activity. For a sequence from which the first 39 nt are deleted, no significant IRES activity is obtained, at least in human adenocarcinoma cells. Indeed, and as appears from the examples below (FIG. 8B), the IRES activity is not significant for the following fragment:

- sequence identical or homologous to the sequence SEQ ID 32, located between nucleotides 39 and 147 of the sequence SEQ ID 30 and hereinafter called B3,

- sequence identical or homologous to the sequence SEQ ID 33, located between nucleotides 39 and 93 of the sequence SEQ ID 30 and hereinafter called B4.

The Applicant has also demonstrated that the 5 ′ NTR end of the C transcript of human FGF-1 and one of its subfragments exhibits significant IRES activity.

Consequently, the subject of the invention is also an isolated nucleic acid exhibiting an IRES activity which is characterized in that it contains a sequence identical or homologous to the sequence SEQ ID 36 located between nucleotides 1 and 149 of the 5 'NTR of the transcript C of human FGF-1 mRNA and hereinafter referred to as C. This sequence corresponds to the entire sequence of the 5 'NTR of transcript C.

The IRES site having been identified by the Applicant between nucleotides 1 and 103 of the sequence SEQ ID 36, the invention also relates to an isolated nucleic acid exhibiting an IRES activity which is characterized in that it contains a sequence identical or homologous to the sequence SEQ ID 37 located between nucleotides 1 and 103 of the sequence SEQ ID 36 hereinafter called C2.

For a sequence from which the first 59 nt are deleted, no significant IRES activity is obtained, at least in human adenocarcinoma cells. Indeed, and as appears from the examples below (FIG 8B), the IRES activity is not significant for the following fragment:

- an isolated nucleic acid having an IRES activity comprising a sequence identical to the sequence SEQ ID 38, located between nucleotides 59 and 149 of the sequence SEQ ID 36 and hereinafter referred to as C3, - an isolated nucleic acid having an IRES activity comprising a sequence identical to the sequence SEQ ID 39, located between nucleotides 59 and 103 of the sequence SEQ ID 36 and hereinafter called C4.

Furthermore, the Applicant has demonstrated that the 5 'NTR end of the D transcript of human FGF-1 exhibits significant IRES activity, in particular on myoblasts.

Consequently, the invention also relates to an isolated or recombinant nucleic acid exhibiting an IRES activity which is characterized in that it contains a sequence identical or homologous to the sequence SEQ ID 42 located between nucleotides 1 and 92 of the 5 'NTR of the transcript D of human FGF-1 mRNA and hereinafter referred to as D. The sequence SEQ ID 42 corresponds to the entire sequence of the 5 'NTR of transcript D.

The Applicants have moreover demonstrated that the IRES activity of certain fragments can be increased by adding a certain number of nucleotides 3 'to the IRES.

In particular, the invention covers:

- an isolated nucleic acid having an IRES activity containing a sequence identical or homologous to the sequence SEQ ID 26 (A26) - an isolated nucleic acid having an IRES activity containing a sequence identical or homologous to the sequence SEQ ID 21 (A21) - an isolated nucleic acid exhibiting IRES activity containing a sequence identical or homologous to the sequence SEQ ID 34 (B10)

- an isolated nucleic acid having an IRES activity containing a sequence identical or homologous to the sequence SEQ LD 40 (C10) - an isolated nucleic acid having an IRES activity containing a sequence identical or homologous to the sequence SEQ ID 41 (Ci l)

The IRES activity of human and murine FGF-1 makes them potential candidates for the production of cassettes making it possible to express one or more genes of interest in eukaryotic cells, from the same mRNA under the control of a sole promoter.

In the following description and in the claims, the expression “gene of interest” denotes nucleic acids coding for polypeptides such as proteins of therapeutic interest, antigens, anti-angiogens, pro-angiogens, suppressors. tumors or genes intended to compensate for a genetic anomaly of the mutation type, etc.

Consequently and according to another aspect, the invention relates to a cassette for the expression of at least one gene of interest in a eukaryotic cell, said cassette being characterized in that it contains at least one nucleic acid as described above. .

As will be seen below, cassettes comprising a single gene can be used either for the production of molecules of interest in vitro, or in gene therapy.

It is also possible to create a bicistronic cassette resulting from the insertion of one of the abovementioned nucleic acid sequences between two genes of interest, the expression of the second gene being coordinated with that of the first gene. More precisely, the first gene (first cistron) provides the level of dependent translation while the second gene (second cistron) is expressed in proportion to the internal input activity of the ribosomes at the level of different 5 'NTR from FGF-1

This principle can extend to the coexpression of 3, 4 or more molecules insofar as these are separated by a TRES sequence as defined above, leading to the development of multicistronic cassettes. Of course, the multicistronic cassette of the invention comprises IRES of FGF-1 as they are characterized in the present application, associated or not with other IRES such as, for example, those identified on BIP (Immunoglobulin heavy -chain binding protein), the fibroblastic growth factor FGF-2, the proto-oncogene c-myc, the growth factor PDGF "Platelet Derived Growth Factor" and the VEGF "Vascular Endothelial Growth Factor".

Therefore and in a particular embodiment, the expression cassette contains at least two genes of interest separated by a nucleic acid as previously described, the first gene (first cistron) providing the level of dependent translation while the second gene (second cistron) is expressed in proportion to the activity of the IRES site present in said nucleic acid.

In known manner, the cassette also contains a promoter making it possible to control the expression of the gene (s).

The main application of multicistronic expression cassettes is gene therapy insofar as such cassettes make it possible to express the product of the nucleotide or nucleotides or the genes which they contain, from the same mRNA in vivo. These cassettes can also make it possible to express different types of molecules of industrial interest such as receptors, enzymes, etc. Gene therapy can be envisaged according to two distinct techniques, perfectly known to those skilled in the art.

The first technique consists in introducing, in vivo, directly into the host organism (human or animal), a pharmaceutical composition containing the naked cassette contained in double stranded DNA, said pharmaceutical composition forming part of the present invention.

In the case of the use of the naked cassette, this is coupled, in the pharmaceutical composition, to any compound capable of promoting penetration of the cassette into cells, in particular PEI and its derivatives, liposomes and derivatives ...

The recombinant vectors contained in the pharmaceutical composition, into which the expression cassettes are inserted, are also part of the invention. Thus, the expression cassettes can be inserted into plasmid or viral vectors such as retrovirus, lentivirus, adenovirus and AAV.

The second gene therapy technique consists of removing a cell from the host organism, then in vitro, introducing the naked or inserted cassette into a vector in the cell, the modified cell then being reintroduced into the host organism. In this way, the product of the cassette can be expressed in the patient's body.

As already said, when the expression cassette contains only one molecule of interest, this cassette can be used for the in vitro production of said molecule from cell cultures. As previously, the cassette can be in naked form or inserted into a vector. The cassette or the vector are then introduced into a host cell by transfection technique, well known to those skilled in the art. Such techniques can for example be physical techniques (electroporation), chemical techniques, synthetic vectors, calcium phosphate precipitation or other methods such as cell fusion, conjugation, etc.

Consequently, the invention also relates to a recombinant host cell, which is characterized in that it comprises the expression cassette or a recombinant vector as previously described.

The recombinant host cells of the invention are in particular, but not limiting:

911 human embryonic retinoblast cells, no ATCC number Gen line _t hon

C2 / 7 murine myoblast cells; no ATCC number Gen line _t hon

CHQ5B human embryonic myoblast cells

C2C12 Line derived from murine myoblast 91031101

Cos-7 CV-1 transformed by the SV-40 CRL-1654 T antigen

Hela CCL-2 Human Uterine Carcinoma Epithelial Cells

L929 Connective tissue of immunocompromised mice CCL-1

NIH-3T3 Murine embryonic fibroblatic cells CRL1658

Saos-2 Human osteosarcoma epithelial cells (p53 - / -). HTB-85

SK-Hep-1 Human liver adenocarcinoma. HTB-52

SK-N-AS Human neuroblastoma cells 94092302

SK-N-BE Human neuroblastoma cells 95011815

In practice, the production of the peptide is obtained by transfection, in vitro, of the host cell cultured by the naked cassette or inserted into a vector.

Analysis of the expression of the peptide is carried out from cell extracts prepared from the transfected cells. Depending on the nature of the peptide of interest, its level of expression will be analyzed by different detection techniques (enzymatic assay, immuno logical detection, etc.). For the purpose of industrial use of the molecules of interest, these can be purified. For other uses, the cells genetically modified by the expression cassette will be used to produce the molecules of interest (ex vivo gene therapy, pharmacological screening).

The expression cassette produced from the IRES sequences which are the subject of the present invention can also be inserted into the cells (egg) of an animal to create a transgenic animal.

Therefore, the invention also relates to a transgenic animal which is characterized in what its cells include the naked cassette or inserted into a vector, as previously described. In practice, the animal is obtained by introducing the cassette or the vector into the cells at the embryonic stage. This type of technology is perfectly known to those skilled in the art.

The invention is clear from the examples below, in support of the appended figures.

Figure 1: The fibroblast growth factor 1 gene: Organization of the human FGF-1 gene. Exons 1, 2 and 3 constitute the 3 coding exons; splicing non-coding exons -1 A, 1B, 1C, 1D at exon 1 will generate mRNAs A, B, C, D respectively. The transcription starting point for transcript A is preceded by the canonical sequences CCAAT and TATA.

Figure 2: Representation of the 4 transcripts A, B, C, D encoding FGF-1: transcripts A and B are predominant and expressed in the kidney, brain and retina. Transcripts C and D are in the minority and expressed in fibroblasts and Vascular Smooth Muscle Cells (VML) after induction with serum or PMA. They are also expressed in cells derived from prostate cancer (PC3), lungs, or cells derived from glioblastomas

Figure 3: IRES study strategy: Principle of bicistronic vectors: The first cistron of bicistronic mRNA provides the level of dependent translation, the second is expressed in proportion to the internal input activity at the 5 'level FGF-1 NTR. The calculation of the protein Y / protein X ratio defines the IRES activity of the sequence located between the two cistrons.

Figure 4: Consecutive Western blotting of CAT-I-NuCAT vector transfections in Cos-7 cells: The intensity of the bands corresponding to CAT provides the level of cap-dependent translation. The intensity of the NuCAT bands corresponds to

1 'IRES activity of the 5' NTR considered. NT: extract from non-transfected cells; FGF-2: vector containing the 5 'NTR of FGF-2. A, B, C, D,: vectors respectively containing the 5 'NTR A, B, C, D of human FGF-1; arrow 1: high molecular weight isoforms derived from the alternative initiation of translation in the 5 'NTR of FGF-2 at the level of the CUG initiation codon.

Figure 5: LucR-I-LucF bicistronic mRNA and Comparison of the IRES activity of the 5 'NTRs of human FGF-1 in different cell types

Figure 5A: Schematic representation of the bicistronic LucR-I-LucF mRNA. A looped rod structure was placed 3 'from LucR to limit the phenomenon of non-specific reinitiation of translation at the level of the second LucF cistron.

Figure 5B: Diagram of the IRES activities of the 5 'NTRs of human FGF-1 in different cell types. The ordinate values are obtained by calculating the ratio of LucF / LucR activities and are expressed in arbitrary units.

Figure 6: Representation of the murine FGF-1 gene and transcripts.

Figure 6 A: Gene structure of murine FGF-1. Exons 1, 2, and 3 constitute the 3 coding exons; the splicing of the non-coding exons -1A and -IB at the level of exon 1 generates the mRNAs A and B of FGF-1.

Figure 6B: Scheme of transcripts A and B of murine FGF-1. The transcripts have in common the reading frame and the 3 'NTR zone, and differ by their 5' NTR end.

Figure 7: Comparison between the human and murine 5 'NTR of FGF-1

Figure 7A: Alignment of the 5 'NTR A sequences of murine (mFGF-1) and human (h-FGF-1) FGF-1. Figure 7B: Diagram of IRES activities of 5 'NTR A and B of human (hFGFl) and murine (mFGFl) FGF-1 in different cell types. The ordinate values are obtained by calculating the ratio of LucF / LucR activities and are expressed in arbitrary units.

Figure 8 Diagram of IRES activities of different 5 'NTR subfragments of human FGF-1 in human SK-Hep-1 adenocarcinoma cells. The ordinate values are obtained by calculating the ratio of LucF / LucR activities and are expressed in arbitrary units.

Figure 8 IRES activity of 20 nucleotide sequences of fragment A (A, A2, A3, A4, A5, A6, A7, A8, A9, A10, Ai l, A12, A13, A14, A15, A16, A17, A18, A19, A20)

Figure 8B IRES activity of 4 nucleotide sequences of fragments B and C, respectively (B, B2, B3, B4) and (C, C2, C3, C4) and the nucleotide sequence of fragment D.

Figure 8C: IRES activity of 6 mutated A fragments (A26, A21, A22, A23, A24, A25) in the 911 lines (embryonic retinoblasts)

Figure 8D: IRES activity of 2 mutated B fragments (B10, B1) of 2 mutated C fragments (C10, Cl 1) in the 911 lines (embryonic retinoblasts)

Figure 8E: IRES activity of 3 mutated fragments A27, A28, A29 in the 911 lines (embryonic retinoblasts)

Figure 9 Diagram of IRES activities of different 5 'NTR subfragments of

Human FGF-1 in 911 lines (embryonic retinoblasts (Figure 9 A)), myoblasts (Figure 9B) and myoblasts differentiated into myotubes (Figure 9C). The ordinate values are obtained by calculating the ratio of the LucF / LucR activities and are expressed in arbitrary units. 5 nucleotide sequences of fragment A were thus tested (A, A2, A6, A7, A9), the nucleotide sequence of fragment B, including that of fragment B2 for FIG. 9A, the nucleotide sequence of fragment C, including that of the C2 fragment for FIG. 9A and the nucleotide sequence of the fragment D.

Figure 10: Diagram comparing the IRES activity of the transcript A of human FGF-1, in vivo (intramuscular injection in mice), with regard to IRES of EMCV, VEGF, FGF2, Bip and PDGF.

Figure 11: Diagram comparing the ERES activity of the transcript A of human FGF-1 and its subfragments A2, A6, A7 and A9, in vivo, with regard to the IRES of EMCV.

Figure 12: Diagram comparing the ERES activity of the transcript A of human FGF-1 and of its subfragment Al 1, of the transcript B of human FGF-1 and of its subfragments B2 and B3, of the transcript C of human FGF-1 and its fragments C2 and C3, with regard to TIRES of EMCV

EXAMPLE 1 ACTIVITY IRES OF THE 4 TRANSCRIPTS A, B, C AND D OF HUMAN AND MURIN FGF-1 ex vivo

A / Materials and methods

Extraction of mRNA: Extractions of RNA were carried out, for all cell types, on dishes 10 cm in diameter, at confluence. Each box is washed twice with sterile PBS, then treated with 1 ml of the RNAble lysis solution (Eurobio). Subsequently, the total RNAs undergo chloroform treatment, isopropanol precipitation and 80% ethanol washing. After drying, they are taken up in 50 μl of water and dosed at 260 nm. Reverse Transcription: The reverse transcription reactions were carried out according to the superscript II protocol (GIBCO BRL) using, for all cell types, 2 μg of total RNA extracts per reaction. For the synthesis of human FGF-1 cDNAs, degenerate hexamer primers were used. For mouse cDNAs, the 3 'primer specific for mouse FGF-1 was used.

Oligonucleotides: They are presented in the 5 'to 3' direction, the Spel and Ncol cloning sites are underlined in the 3 'and 5' primers respectively. The 3 ′ primer is common to all 5 ′ NTRs of the same organism, due to its pairing at the TATG level of Texonl.

Human FGF-1 primer

3 'primer: CTGCTGAGCCATGGCTGAAGGG SEQ ED 43 5' primer A: ACAACTAGTGGGCATACCAGTGTCAGCTGCA SEQ ID 44;

B: CTAACTAGTAGAGAGCCGGGCTACTCTGAGAA SEQ ID 45;

C: ACAACTAGTCCAGGCAGTGGGGCCGGCTGGCCAGACTCTTGGGGGATT

CCTTAG SEQ ID 46;

D: ACAACTAGTAGAGAGAGAGAAAAAATACT GTTGGCAGCAGCACAATG SEQ ID 47

FGF-1 mouse primer

Primer 3 's: ACCACCGCTGCTTGCTGCCGAGCCATGGCTGAA SEQ ED 48

Primer 5 'sA: CTCTACTAGTGTCCTTTAGTGCCAGCTAGCCTTACA SEQ ID

49 sB: TCTCCAACTAGTAGAGCTGCAGAAATCCTGAGG SEQ ED 50

PCR: PCRs were carried out with TAQ polymerase platinum (Gibco BRL) for its capacity to amplify DNA rich in GC and its very good specificity. For each PCR, 3 μl of the cDNAs made above were used. For the PCRs aimed at amplifying the 5 'NTR C and D, 50 ng of a plasmid fragment carrying a part of these different 5'NTRs supplied by I-M Chiu (11) was added. The amplifications of the murine A and B transcripts were carried out in the same way. The amplifications are carried out according to the “Touch Down PCR” method. The 5 'NTR A was amplified with the primer 5' A and the primer 3 ', under the following conditions: 94 ° C 5 min, then 3 cycles 94 ° C 30sec / 76 ° C 2 min, then 3 cycles 94 ° C 30sec / 74 ° C 2min then 25 cycles 94 ° C 30sec / 72 ° C 2 min, finally 72 ° C 5 min.

The 5 'NTR B was amplified under the following conditions: 94 ° C 5 min, then 3 cycles 94 ° C 30sec / 68 ° C 2min, then 3cycles 94 ° C 30sec / 66 ° C lmin / 68 ° C 1min then 25 cycles 94 ° C 30sec / 64 ° C lmin / 68 ° C 1min, finally 68 ° C 5 min, using the primers 3 'and 5' B. In the same way the 5 'NTR C, D, and the 5' Murine NTR A and B were amplified, with their corresponding primers, under conditions similar to 5 'NTR B with hybridization temperatures of 70 ° C, 70 ° C, 68 ° C and 66 ° C respectively.

Cellular culture. : The cell lines were obtained from T “American Type Culture Collection” (ATCC) or T “European Collection of Cell Cultures” (ECACC) and were cultivated according to their instructions.

Construction of the plasmids: The insertion of the various 5 'NTRs of FGF-1 into the bicistronic vectors was carried out using the Spel and Ncol sites located in the PCR primers 5' and 3 'respectively. The various plasmids were obtained by conventional molecular biology techniques, namely: ligation, transformation, selection of positive clones, sequencing and preparation of a concentrated solution at lμg / μl.

Transfections: All of the transfections were carried out with the FUGENE 6 reagent according to the protocol provided by Boeringher-Roche, at a rate of 2.5 μl per μg of plasmid DNA. The transfections for the luciferase assays are carried out in 12-well dishes with 80,000 cells and 1 μg of plasmid per well. The transfections carried out for the Western blots are carried out with 2 μg of plasmid DNA in dishes 5 cm in diameter seeded with 200,000 Cos-7 cells.

Luciferase assay: The activity of LucR and LucF was measured using the “dual luciferase reporter assay” system (Promega). The transfected cells, placed in 12-well dishes, are rinsed twice with PBS. Then they are detached and homogenized in 100 μl of the lysis buffer supplied with the kit, 48 hours after the transfection. The chemiluminescent signals of LucF and LucR are measured in a luminometer (Berthold) equipped with automatic injectors.

Western blot: 48 hours after transfection, the Cos-7 cells are washed twice with PBS, and are harvested in 1 ml of PBS. After a centrifugation of 2 min at 2000 g, the pellet is resuspended in a 1% SDS solution containing protease inhibitors, then lysed by ultrasound. The proteins are assayed, with the BCA reagent (Pierce), by spectrophotometry at 595 nm. 20 μg of the proteins from each cell extract are used for the Western Blot. In summary, the cell lysates are heated at 95 ° C. for 2 minutes in a 1% SDS solution, 0.1 M DTT, and deposition buffer, then the proteins are separated on a 12.5% polyacrylamide gel) and transferred to a membrane. nitrocellulose. The CAT and NuCAT proteins are immunodetected by an anti-CAT polyclonal rabbit antibody prepared in the laboratory (dilution 1/10000). After their fixation, anti-rabbit IgG antibodies coupled to peroxidase are detected using the chemiluminescence reagent (Amersham).

B / Results

Bl / Presence of IRES sites in the various 5 'NTRs of human FGF-1

Cloning of the 5 'NTR of human FGF-1 cDNA: The first step consisted in cloning the various 5' NTRs of human FGF-1. To do this, the total RNA was extracted from human cell lines present in the laboratory. After reverse transcription, PCRs aimed at amplifying the sequences of the various 5 'NTRs of the FGF-1 have been performed. After amplification, the presence of the expected fragment informs about the existence of each transcript in the cell type tested. The amplification of the various 5 'NTRs of human FGF-1 was observed using cDNAs from SK-Hep-1, SK-N-BE and SK-N-AS cells. The more abundant SK-Hep-1 cell cDNA allowed cloning into bicistronic vectors.

Bicistronic vectors: The strategy of bicistronic vectors used to demonstrate the existence of IRES (14) was applied to test the IRES activity of the different 5 'NTRs of FGF-1 in different cell types (fig n ° 3). These constructions make it possible to express, from the same mRNA, two different reporter genes. The different 5 'NTRs of FGF-1 have been introduced between these 2 reporter genes. The first cistron provides the level of dependent translation. The second cistron is expressed in proportion to the internal entry activity of the ribosomes at the level of the various 5 'NTRs of FGF-1

Bl.l Test of the IRES activity of the 5 'NTR of FGF-1 in Cos-7 cells

CAT-I-NuCAT bicistronic vectors: CAT-I-NuCAT bicistronic vectors were first used to determine qualitatively, by Western Blot, the internal input activity of the various 5'NTRs of the FGF -1. Indeed, these vectors are provided with cDNA of chloramphenicol acetyl transferase (or CAT) in the first cistron and that of a fusion protein Nucleoline-CAT (NuCAT) in the second cistron (fig.n ° 4). Due to the difference in molecular weight between these two proteins, they will be easily separated on polyacrylamide gel in denaturing condition. The CAT and NuCAT proteins will be easily visualized, by the use of antibodies directed against the C-terminal region of CAT, after transfer to the membrane. The evaluation of the relationship between the second and the first cistron informs about the internal input capacity of the inserted sequence and defines its ERES activity.

Results: The different CAT-I-NuCAT vectors, containing the 5 'NTR of FGF-1, were transfected into Cos-7 cells. The Western Blot consecutive to these transfections is presented in fig n ° 4. At the level of the CAT band (first cistron), variations are observed due to differences in the efficiency of transfections. However, the expression of the protein NuCAT is observed in the presence of transcripts A, B and C. These results suggest a strong IRES activity of variant B of FGF-1, then a good activity of variant A (if it is related to l transfection efficiency which is lower in this case). The activity of the 5 'NTR C and D is much lower. The overexposure of the film also informs about the absence of alternative initiation of FGF-1 at the level of the four 5'NTRs. This phenomenon, described and observed for FGF-2 (1), results in the formation of isoforms of higher molecular weight (fig. N ° 4; blue arrow). This result is in agreement with the sequence published by Chiu et al., Which described the presence of a stop codon in exon 1, a few bases before TAUG of FGF-1 (11). The presence of IRES activity for variants A, B, C motivated the quantitative comparison of the IRES activity of these different transcripts in different cell types. For this study, the western blot not being an appropriate technique, other bicistronic vectors coding for luciferase reporter genes were used.

B.1.2 Comparison of ERES activities of the 5 'NTR of FGF-1 in different cell types

LucR-I-LucF bicistronic vectors: The LucR-I-LucF bicistronic vectors (4) express two luciferases, which use different substrates. The Rénillia luciferase (LucR) is the first cistron and the Firefly luciferease (LucF) is the second cistron (fig. 5A). LucF reports on internal entry activity and is very sensitive to dosing, like LucR. At 3 ′ of the LucR gene, a very stable loop rod structure (ΔG = -40 kCal / mol) was placed, making it possible to limit the non-specific reinitiation phenomena observed during the introduction of short sequences into the intercistronic space ( Figure 5A). The use of such a system, based on the high sensitivity of reporter genes, allows rapid, precise and concomitant quantification of the activity of each reporter. The calculation of the LucF / LucR ratio provides information on the internal input capacity of the 5 'NTR sequence and defines the IRES activity of this sequence. The IRES activity between the different cell types is calibrated, in the laboratory, using the vector RHL (4), which contains a loop rod structure but no IRES between the two cistrons. This vector makes it possible to determine the value of non-specific initiation of the second cistron and to overcome variations in the efficiency of transfections between the cell types.

Results: To study the variation of IRES activity as a function of cell type, we chose different lines according to the expression of the four types of FGF-1 transcripts, and known results for TIRES of FGF-2. We chose Cos-7 cells (already used with the vector CAT-I-NuCAT), a line where Ton was able to amplify the four 5'NTRs of FGF-1 (SK-Hep-1 cells) and a line or the amplification did not work (Hela cells). Saos-2 cells, derived from an osteosarcoma P53 ^" '^" , are the last model chosen for this study because TIRES of FGF-2 is very active there (4). The results of transfections with the vectors LucR-I-LucF, in the Cos-7, Hela, SK-Hep-1 and Saos-2 cells are presented in FIG. 5 B. It appears clearly, as observed in FIG. 4, that the 5 'NTR A, B and C have IRES activity. On the other hand, the activity of 5 'NTR D is very low, at least in the cell types tested. The activity of the 5 'NTR A, B and C is however very variable.

Generally, the 5'NTR A is the most active, with values varying from 5.5 to 12.7. Its activity, in Hela and Sk-Hep-1 cells, is greater than that of TIRES of FGF-2. In contrast, TERES of variant A works less well than that of FGF-2 in Cos-7 and Saos-2 cells (although the values recorded in these cell types are the highest).

Variant B of FGF-1 is more active than A in Cos-7 cells (10 units), and its activity is modest in Saos-2 cells. In the other cell types tested, 5 'NTR B has no activity. As for variant C, it has a medium activity in SK-Hep-1 and Saos-2 (5.8 and 7.5 respectively). It is in spite of everything in this latter cell type where the 5 'NTR C is the most active. Its activity in Cos-7 and Hela cells is extremely low.

We can also add that the information obtained in Cos-7 cells with CAT-I-NuCAT vectors is correlable with the values obtained with LucR-I-LucF vectors.

B2 / Internal entry of ribosomes in the various 5 'NTRs of murine FGF-1

We find a gene structure identical to that of humans in mice (fig. 6A), beef, pork and chicken. The alignment of the sequences of the 5'NTRs of the FGF-1 of these various organisms (9, 11, 16) has shown good conservation of these non-coding sequences. This conservation is particularly true between mice and humans, with almost 70% similarity for the 5 'NTR A (figure n ° 7 A).

The hypothesis of a conservation of the mechanism of IRES from mice to humans was considered, so we carried out a comparative study of variants A and B of human and murine FGF-1.

To do this, we amplified by RT-PCR and cloned the 5 'NTR A and B of murine FGF-1 (fig.n ° 6B) from cDNA of murine cell lines. Then we used the same study strategy: construction of LucR-I-LucF bicistronic vectors, and transfection into the cell types used for the study of human counterparts.

Cloning of murine FGF-1 5'NTR A and B cDNA: We used the same cloning strategy, by RT-PCR, as for human 5 'NTR. The same amplification problems were encountered and resulted in the use of enzymes and identical stringency conditions. Murine 5 'NTRs were amplified and recloned from cDNA from L929 cells.

Comparison of the internal entry activity of 5'NTR A and B of murine and human FGF-1 in human cells: The transfections were carried out in Cos-7, SK-Hep-1 and Saos-2 cells, to make a direct comparison with human sequences. The results following the luciferase assays, are presented in figure n ° 7B.

The IRES activity profiles of the human and murine variant A are similar. They have values of the same order of magnitude and seem to be regulated in the same way depending on the cell type, with the exception of Saos-2 cells. Indeed, the IRES activity of the murine variant A is greater in these cells and is close to the value observed for TERES of FGF-2. We also find internal input values of the same order of magnitude for the 5 'NTR B human and murine. In addition, the best IRES activity is observed in Cos-7 cells, for the construction containing the 5 'NTR B of human FGF-1: this is also the case with the murine counterpart.

EXAMPLE 2: CHARACTERIZATION OF THE ERES OF THE 4 TRANSCRIPTS A, B, C and D OF HUMAN FGF-1 ex vivo

Example 2A: Characterization in Human Adenocarcinoma Cells (SK-Hep-1) The characterization of the IRES of the 4 transcripts consists in determining which is the smallest nucleotide sequence of each of the fragments A to D exhibiting satisfactory IRES activity, which allows to locate in a relatively precise way the IRES site of each of the transcripts. This experiment is carried out on human adenocarcinoma cells (SK-Hep-1). The following were thus tested: - the entire 5 'NTR of the transcript A (A),

20 nucleotide sequences corresponding to continuous fragments of the 5 'NTR of transcript A (A2, A3, A4, A5, A6, A7, A8, A9, A10, Ai 1, A12,

A13, A14, A15, A16, A17, A18, A19, A20), - the entire 5 'NTR of transcript B (B),

- 4 nucleotide sequences corresponding to a continuous fragment of the 5 'NTR of transcript B (B2, B3, B4)

- the whole of the 5 'NTR of the transcript C (C), - 4 nucleotide sequence corresponding to a continuous fragment of the 5' NTR of the transcript C (C2, C3, C4),

- the entire 5 'NTR of the transcript D (D).

The nucleotide sequences to be tested are continuous sequences originating from messenger TARN and have been subcloned into a bicistronic vector according to a procedure identical to that described in Example 1. The bicistronic vector codes for two luciferases, renilla luciferase (LucR , first cistron) and firefly luciferase or firefly (LucF, second cistron). The nucleotide sequences to be tested were therefore inserted between the two luciferase genes, so as to quantify, by measuring the LucF activity, their capacity to generate an initiation of translation by internal entry of the ribosomes. The value of LucF activity is related to the value of LucR activity which is proportional to the amount of mRNA. To express the activity values of IRES in arbitrary units (AU), the LucF / LucR ratio is itself related to the value of a negative control (bicistronic vector without IRES) which represents the level of "leak". Luciferase LucR and LucF activity was measured 48 hours after transfection on cellular extracts of human adenocarcinoma (SK-Hep-1).

The results are shown in Figures 8 A and 8B.

IRES FGF-IA: as shown in Figure 8A, the 5'NTR A of the FGF-1 mRNA (nucleotides 1 to 435) allows the internal entry of ribosomes with an efficiency of approximately 14 AU (FIG. 8A, construction A), which therefore confirms the results of Example 1. The analysis of different segments of this 5 'NTR makes it possible to conclude that the internal input activity does not require nt 1 to 270, nor the nt 387 to 435. On the basis of this mapping, it can be concluded that the minimum IRES of the FGF-IA comprises nucleotides 269 to 366 (construction Al 8). The IRES activity is optimized by the presence of nt located on both sides, ie from 214 to 269 and from 366 to 435 (construction A5). It should be noted that an essential but not necessarily sufficient element is found between nt 270 and 304.

IRES FGF-1 B: FIG. 8B indicates that the 5'NTR B of the mRNA of FGF-1 (nt 1 to 147) has a relatively low IRES activity of the order of 3 AU, which therefore comes confirm the results of Example 1. Deletion of nt 0 to 39 completely inactivates TIRES (Fig. 8B, construction B3 and B4).

Other examples will indicate that TIRES of FGF-1 B has superior activity in other cell types.

IRES FGF-1C: Figure 8B indicates that the 5'NTR C of the FGF-1 mRNA (149 nt) has a strong IRES activity with an efficiency of 6 AU. Interestingly, the deletion of nt 103 to 149 (corresponding to the area common to the 4 NTRs of the FGF-1 mRNA) generates an increase in the activity of TIRES (4 AU): it can therefore be concluded that minimum TIRES of FGF1-C includes nt 1 to 103. Furthermore, the deletion of the first 59 nt inactivates the IRES activity.

IRES FGF-ID: Figure 8B indicates that the 5'NTR D of the FGF-1 mRNA has no IRES activity in SK-Hep-1 cells. However, other examples will show that this fragment has IRES activity in other cell types, in particular the C2 / 7 myoblasts.

Example 2B ^• characterization in cells 91 1 (r.e11nl ^pg ήp rptinr>hlastp; embryonic hymans) and in cells C / 7 (cells of murine myoblasts) Example 2 A is repeated with the fragments A2, A6, A7 and A9

Embryonic retinoblast cells and myoblasts are seeded one to two days before transfection into a 24-well plate. All transfections of retinoblast and myoblast cells were carried out with the reagent Lipofectamine according to the protocol provided by Life Technologies, at a rate of 3 μl per μg of plasmid DNA per well. The cells are lysed, 48 h post transfection, with 100 μl of lysis buffer (Dual Luciferase, Promega). The luciferase activities (Rénilla and Firefly) are measured on 20 μl of cell lysate (Mediator PhL, Luminometer) for 10 s after 50 μl of addition of reagent (Dual Luciferase, Promega).

The results are shown in Figures 9A and 9B. The hairpin is a reference sequence of looped rod structure, not containing IRES.

As shown in FIG. 9 A, the 5 ′ NTR of the transcript A as well as the fragments A2, A6, A7 and A9 exhibit satisfactory IRES activity. The same is true for the 5 ′ NTRs of transcripts B, C and D in myoblasts when these same sequences exhibited less activity in human adenocarcinoma cells (SK-Hep-1) (see example 2A)

Example 2C: Characterization in C2 / 7 Myoblast Cells Then Differentiated into Myotubes)

Example 2 A is repeated with the transcripts A, A9, B, C, D and this activity is compared with that of EMCV.

Myoblast cells are seeded one to two days before transfection into a 12-well plate. The transfections, carried out on the fragments containing hairpin, EMCV, A, A9, B, C, or D; are made in duplicate as described above. The cells are lysed, 24 h post transfection (Figure 9C in shaded) or 9 days post transfection (Figure 9C in dotted lines). During the 9 days of incubation, the myoblast cells are differentiated into myotubes in a DMEM medium supplemented with 5% horse serum.

As shown in FIG. 9C, the TIRES activity of the FGF-1 transcript A, of the A9 fragment and of the FGF-1 transcript B has a higher activity in the myotubes than in myoblasts. The FGF-1 transcript C and D has comparable activity in myoblasts and myotubes.

It is very interesting to note that the activity in the myotubes versus myoblasts of TIRES of the transcript A of FGF-1 (ratio = 8.47), of the fragment A9 (ratio = 11.63) and of the transcript B of FGF-1 (ratio - 18.27) is much higher than that of EMCV (ratio = 6.44).

Example 2D: Characterization of Mutant Transcripts in 911 Cells (Human Embryonic Retinoblast Cells)

Example 2 A is repeated from different transcripts to which a given sequence is added between TERES ° and the AUG codon of the firefly gene.

The fragments obtained are as follows:

A26: SEQ ID 26 (nt 269 to 435 of the sequence SEQ ID 8+ 20 nt) A21: SEQ ID 21 (nt 269 to 435 of the sequence SEQ ID 8 + 40 nt) A22: SEQ ID 22 (nt 269 to 435 of the sequence SEQ ED 8 + 70 nt) A25: SEQ ID 25 (nt 269 to 435 of the sequence SEQ ED 8 + 135 nt) A23: SEQ ED 23 (nt 269 to 390 of the sequence SEQ ED 9 + 70 nt) A24: SEQ ED 24 (nt 269 to 390 of the sequence SEQ ID 9 + 135 nt) B10: SEQ ID 34 (nt 1 to 147 of the sequence SEQ ID 30 + 40 nt) Bl 1: SEQ ED 35 (nt 1 to 147 of the sequence SEQ ED 30 + 70 nt) C10: SEQ ED 40 (nt 1 to 149 of the sequence SEQ ID 36 + 40 nt) Cl 1: SEQ ID 41 (nt 1 to 149 of the sequence SEQ ED 36 + 70 nt)

The results are shown in Figures 8C and 8D.

The efficiency of TIRES of FGF1A (FIG. 8C) is increased for the mutants A26 and A21 and slightly decreased for the A22 and A25. However, the segment between nucleotides 390 and 435 is necessary to have an efficiency of TIRES (mutant A23 and A24). It is observed that the efficiency of TIRES of FGF1B and 1C is increased by 4 to 6 AU compared to wild TIRES (FIG. 8D). We also observe good plasticity of fragments A, B and C.

Example 2E: Characterization of Mutant Transcripts in 911 Cells (Human Embryonic Retinoblast Cells)

Example 2 A is repeated from different transcripts of the A8 fragment to which a 40 nt sequence different from one fragment to another is added, between TERES ° and the AUG codon of the feverfly gene.

The fragments obtained are as follows:

A27: SEQ ID 27 (nt 269 to 435 of the sequence SEQ ID 8 + 40 nt)

A28: SEQ ED 28 (nt 269 to 435 of the sequence SEQ ID 8 + 40 nt) A29: SEQ ED 29 (nt 269 to 435 of the sequence SEQ ED 8 + 40 nt)

In the 3 cases, a satisfactory ERES activity is obtained, which further demonstrates the good plasticity of the fragment A. Out of a hundred strains tested, it is observed that at least 80% of them have an efficiency equivalent to or greater than the cut A .

EXAMPLE 3 ACTIVITY OF THE 5 'NTR OF TRANSCRIPTS A, B C and D OF FGF-1

HUMAN in vivo

The experiments are carried out according to the following protocol.

The animals used are female mice weaned for around 6 weeks. The strain used is Balb / C (EFFA CREDO). The mice are anesthetized with a mixture of ketamine (100 mg / kg) and xylazine (10 mg / kg) intraperitoneally. The duration of the anesthesia is 20 minutes. The DNA is amplified and purified by maxi-preparation in the absence of endotoxin (Endofree, Qiagen). After precipitation, the DNA is taken up in a 150 mM NaCl solution (Sigma) at a concentration of approximately 50 μg in 30 μl.

DNA is injected into the left anterior tibia using a Hamilton syringe. The leg, previously shaved, is coated with conductive gel. The two electrodes are applied on either side of the tab to make Telectro transfer. As a general rule, 8 electrical pulses of 20 milli seconds are delivered at a frequency of 2Hz., The voltage being 200V / cm. These experiments are carried out on groups of three animals per plasmid.

A week later, the animals are sacrificed by intraperitoneal injection of 0.1 ml of sodium pentobarbital 6 and then by cervical dislocation. The anterior tibia is removed and frozen in liquid nitrogen. The muscles are then ground in 500 μl of lysis buffer added with a protein inhibitor cocktail at 16.7 μl / ml. The muscle lysates are thawed for 5 minutes at 37 ° C, vortexed and centrifuged at 4 ° C for 5 minutes at 15,000 G. The luciferase activities (Renilla and Firefly) are measured on 20 μl of muscle lysate (Mediators PliL, Luminometer) for 10 s after 50 μl of reagent addition (Dual Luciferase, Promega).

The control muscle is either a muscle into which 30 μl of 150 mM NaCl was injected and then electroporated, or a muscle which has not undergone anything.

EXAMPLE 3A: Comparison of the IRES Activity of the Human FGF-1 Transcript A Against Known IRES

This example aims to compare, in vivo, in mouse muscle, the ERES activity of the transcript A of human FGF-1 compared to other IRES such as that of EMCV, VEGF, FGF2, Bip and PDGF. As shown in Figure 10, the TERES activity of FGF-IA is equivalent to that of EMCV and greater than the ERES of VEGF, FGF-2, Bip and PDGF.

EXAMPLE 3 B: Comparison of the IRES Activity of the Transcript A of Human FGF-1 and Its Subfragments A, A2, A6, A7 and A9, in vivo, with Regard to TIRES of EMCV

As shown in Figure 11 the IRES of FGF-IA and those modified, A2, A6 and A7, has an IRES activity similar to that of T RES of EMCV, while minimum IRES of FGF-IA ( construction A9) has slightly higher activity.

EXAMPLE 3C: Comparison of the IRES Activity of Transcript A of Human FGF-1 and Its Sub Fragment A1, of Transcript B of Human FGF-1 and Its Subfragments B2 and B3 of Transcript C of Human FGF-1 and its fragments C2 and C3, with regard to TIRES of EMCV

As shown in Figure 12, TIRES of FGF-1B and FGF-1C has activity similar to that of EMCV. The minimum IRES of FGF-1B (fragments B3) and FGF-1C (fragment C2) have a much higher activity than that of EMCV.

The Al 1 fragment has no activity, this confirms that the minimum TERES of the FGF-IA is the A9 fragment (example 3B).

BIBLIOGRAPHY

1. Arnaud, E., C. Touriol, C. Boutonnet, M. C. Gensac, S. Vagner, H. Prats, and A. C. Prats. 1999. A new 34-kilodalton isoform of human fibroblast growth factor 2 is cap dependently synthesized by using a non-AUG start codon and behaves as a survival factor. Mol Cell Biol 19: 505-

14.

2. Bernstein, J., O. Sella, S. Y. Le, and O. EIroy-Stein. 1997. PDGF2 / c-sis mRNA leader contains a differentiation-linked internai ribosomal entry site (D-IRES). J Biol Chem 272: 9356-62. 3. Chetouani, F., P. Monestie, P. Thebault, C. Gaspin, and B. Michot.

1997. ES SA: an integrated and interactive computer tool for analyzing RNA secondary structure. Nucleic Acids Res 25: 3514-22.

4. Creditor, L., D. Morello, P. Mercier, and A.C. Prats. 2000. Fibroblast growth factor 2 ERES activity ex vivo and in transgenic mice reveals a stringent tissue-specific regulation. J.Cell.Biol. (in press) .

5. Duconge, F., and J. J. Toulme. 1999. In vitro selection identifies key determinants for loop-loop interactions: RNA aptamers selective for the TAR RNA element of HIV-1. Rna 5: 1605-14.

6. Huez, L, L. Creditor, S. Audigier, M. C. Gensac, A. C. Prats, and H. Prats. 1998. Two independent internai ribosome entry sites are involved in translation initiation of vascular endothelial growth factor mRNA. Mol Cell Biol 18: 6178-90.

7. Kozak, M. 1989. Circumstances and mechanisms of inhibition of translation by secondary structure in eukaryotic mRNAs. Mol Cell Biol 9: 5134-42.

8. Macejak, D. G., and P. Sarnow. 1991. Internai initiation of translation mediated by the 5 'leader of a cellular mRNA [see comments]. Nature 353: 90-4.

9. Madiai, F., KV Hackshaw, and IM Chiu. 1999. Characterization of the entire transcription unit of the mouse fibroblast growth factor 1 (FGF- 1) gene. Tissue-specific expression of the FGF-1. At mRNA. J Biol Chem 274: 11937-44.

10. Miller, D. L., S. Ortega, O. Bashayan, R. Basch, and C. Basilico. 2000. Compensation by fibroblast growth factor 1 (FGF1) does not account for the mild phenotypic defects observed in FGF2 null mice [published erratum appears in Mol Cell Biol 2000 May; 20 (10): 3752]. Mol Cell Biol 20: 2260-8.

11. Myers, R. L., R. A. Payson, M. A. Chotani, L. L. Deaven, and I. M. Chiu. 1993. Gene structure and differential expression of acidic fibroblast growth factor mRNA: identification and distribution of four different transcripts. Oncogene 8: 341-9.

12. Nanbru, C, I. Lafon, S. Audigier, M. C. Gensac, S. Vagner, G. Huez, and A. C. Prats. 1997. Alternative translation of the proto-oncogene c- myc by an internai ribosome entry site. J Biol Chem 272: 32061-6. 13. Payson, R. A., M. A. Chotani, and I. M. Chiu. 1998. Regulation of a promoter of the fibroblast growth factor 1 gene in prostate and breast cancer cells. J Steroid Biochem Mol Biol 66: 93-103.

14. Pelletier, J., and N. Sonenberg. 1988. Internai initiation of translation of eukaryotic mRNA directed by a sequence derived from poliovirus RNA. Nature 334: 320-5.

15. Pennisi, E. 2000. Human Genome Project. And the number of gene is ....? [News]. Science 288: 1146-7.

16. Philippe, J. M., F. Renaud, Y. Courtois, and M. Laurent. 1996. Cloning of multiple chicken FGF1 mRNAs and their differential expression during development of whole embryo and of the lens. DNA

Cell Biol 15: 703-15.

17. Pyronnet, S., L. Pradayrol, and N. Sonenberg. 2000. A Cell Cycle- Dependent Internai Ribosome Entry Site. Mol. Cell 5: 607-616.

18. Vagner, S., MC Gensac, A. Maret, F. Bayard, F. Amalric, H. Prats, and AC Prats. 1995. Alternative translation of human fibroblast growth factor 2 mRNA occurs by internai entry of ribosomes. Mol Cell Biol 15: 35-44.

19. Vagner, S., A. Waysbort, M. Marenda, M. C. Gensac, F. Amalric, and A. C. Prats. 1995. Alternative translation initiation of the Moloney murine leukemia virus mRNA controlled by internai ribosome entry involving the p57 / PTB splicing factor. J Biol Chem 270: 20376-83.

20. West, M. J., N. F. Sullivan, and A. E. Willis. 1995. Translational upregulation of the c-myc oncogene in Bloom's syndrome cell Unes. Oncogene 11: 2515-24.

Claims

1 / Use for its IRES activity, of an isolated nucleic acid containing a sequence identical or homologous to that of the transcripts A, B, C, D of FGF1 corresponding respectively to the sequences SEQ ID 1 (A), SEQ ID 30 (B) , SEQ ED 36 (C), SEQ ED 42 (D). 2 / Use according to claim 1, characterized in that the nucleic acid contains a sequence identical or homologous to a sequence chosen from the group comprising SEQ ID 2 (A2), SEQ ED 3 (A3), SEQ ID 4 (A4) , SEQ ID 5 (A5), SEQ ID 6 (A6), SEQ ID 7 (A7), SEQ ED 8 (A8), SEQ ED 9 (A9), SEQ ID 12 (A12), SEQ ED 16 (A16), SEQ ED 17 (A17), SEQ ED 18 (A18), SEQ ID 19 (A19), SEQ ED 20 (A20), SEQ ED 26 (A26), SEQ ID 21 (A21), SEQ ID 27 (A27), SEQ ED 28 (A28), SEQ ID 29 (A29). 3 / Use according to claim 1, characterized in that the nucleic acid contains a sequence identical or homologous to a sequence chosen from the group comprising SEQ ID 31 (B2), SEQ ED 34 (B10). 4 / Use of an isolated nucleic acid according to claim 1, characterized in that it contains a sequence identical or homologous to a sequence chosen from the group comprising SEQ ED 36 (C2). SEQ ID 40 (C10), SEQ ED 41 (C11). 5 / Cassette for expression of at least one gene of interest in a eukaryotic cell, said cassette being characterized in that it contains at least one nucleic acid as defined in claims 1 to 4. 6 / Expression cassette according to claim 5, characterized in that it contains at least two genes of interest. Il Recombinant expression vector characterized in that it comprises the expression cassette which is the subject of claim 5. 8 / Expression vector according to claim 7, characterized in that it is a plasmid or viral vector chosen from the group comprising retroviruses, lentiviruses and adenoviruses. 9 / A pharmaceutical composition comprising the expression cassette which is the subject of claim 5 or a recombinant vector which is the subject of claim 7. 10 / Recombinant host cell characterized in that it comprises the expression cassette which is the subject of claim 5 or a recombinant vector which is the subject of claim 7. 11 / Transgenic animal characterized in that it comprises the expression cassette which is the subject of claim 5 or a recombinant vector which is the subject of claim 7. 12 / Isolated nucleic acid with IRES activity, consisting of a sequence having at least 60% identity with a sequence chosen from the group SEQ ED 1 (A), SEQ ED 30 (B), SEQ ID 36 ( C), SEQ ED 42 (D), SEQ ED 2 (A2), SEQ ID 3 (A3), SEQ ED 4 (A4), SEQ ED 5 (A5), SEQ ID 6 (A6), SEQ ED 7 (Al ), SEQ ED 8 (A8), SEQ ED 9 (A9), SEQ ED 12 (Al 2), SEQ ID 16 (Al 6), SEQ ED 17 (Al 7), SEQ ED 18 (Al 8), SEQ ID 19 (Al 9), SEQ ED 20 (A20), SEQ ED 26 (A26), SEQ ED 21 (A21), SEQ ED 27 (A27), SEQ ID 28 (A28), SEQ ID 29 (A29), SEQ ED 31 (B2), SEQ ID 34 (B10), SEQ ED 36 (C2). SEQ ID 40 (C10), SEQ ED 41 (Ci l), SEQ ED 42 (D).