CHIMERIC POLYPEPTIDE COMPRISING THE FRAGMENT B OF SHIGA TOXIN AND PEPTIDES OF THERAPEUTIC INTEREST

Abstract

The invention pertains to methods for using chimeric polypeptides of the formula:

B-X

wherein B represents the B fragment of Shiga toxin or a functional equivalent thereof, and X represents one or more polypeptides of therapeutic significance. Compositions for therapeutic use comprising the polypeptide B-X are also included.

Claims

1-18. (canceled)

19. A polynucleotide coding for the chimeric polypeptide:
B-X wherein B is the B fragment of Shiga toxin or a B toxin fragment that binds to a globotriaosylceramide (Gb3) receptor; and X represents one or more polypeptides of therapeutic significance, wherein said polypeptides are compatible with retrograde transport mediated by B to ensure processing or correct addressing of X.

20. The polynucleotide according to claim 19, wherein X is an epitope which can be presented by the class I major histocompatibility complex.

21. The polynucleotide according to claim 19, wherein X is an epitope of a polypeptide or protein wherein the expression of said epitope is desired at the surface of cells of the immune system.

22. The polynucleotide according to claim 21, wherein said protein or polypeptide is a cancer cell protein, a viral protein, a viral protein from a cancer or an oncogene.

23. The polynucleotide according to claim 22, wherein said viral protein from a cancer are selected from the group of peptides from E6 proteins of HPV16, peptides from E7 proteins of HPV, peptides from a Hbs protein of HBV, peptides from EBV and peptides from cytomegalovirus.

24. The polynucleotide according to claim 19, wherein X is a human epitope from an autoimmune disease.

25. The polynucleotide according to claim 19, wherein X is a human epitope from an infectious disease.

26. The polynucleotide according to claim 25, wherein the human epitope from an infectious disease is an epitope from HTLV-I-associated myelopathy/tropical spastic paraparesis.

27. The polynucleotide according to claim 19, wherein X is an epitope derived from melanoma cell proteins.

28. The polynucleotide according to claim 27, wherein the melanoma cell proteins are selected from the group of BAGE from tyrosinase, GAGE from gp75, tyrosinase, p15 from A/MART-1 melanoma, MAGE-1 and MAGE-3.

29. The polynucleotide according to claim 19, wherein X is a parasitic antigen, a bacterial antigen or a cancerogenic viral protein.

30. The polynucleotide according to claim 19, wherein X is a polypeptide which can restore or activate an intracellular transport function by interacting with proteins of the cellular machinery, wherein said polypeptide competes in the endoplasmic reticulum with a mutated form of a protein involved in intracellular transport or by supplementing a function, which is deficient in said transport.

31. The polynucleotide according to claim 30, wherein X is the domain of interaction of a cystic fibrosis transmembrane regulator (CFTR) protein with calnexin.

32. A composition comprising the polynucleotide coding for the polypeptide according to claim 19.

33. The composition according to claim 32, wherein said composition stimulates the immune defenses of the organism towards viral, parasitic, infectious or cancerous antigens.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0076] FIG. 1A: HeLa cells in which expression of the Gb.sub.3 receptor has been inhibited. The internalised B fragment is not transported to the Golgi apparatus but is accumulated in the vesicular structures.

[0077] FIG. 1B: Control HeLa cells in which the B fragment is transported to the Golgi apparatus.

[0078] FIG. 2: Biochemical test showing the defect in B fragment transport.

[0079] FIG. 3: MHC class I restricted presentation of Shiga B-Mage 1 fusion proteins at peripheral blood monocytic cells (PBMC): role of the KDEL sequence. The PBMC (510.sup.4) were primed overnight with either Mage 1 peptide (1 M) or Mart 1 peptide (1 M) or Shiga B-Mage 1 fusion proteins (1 M) with a sequence which is active (B-Mage 1-Glyc-KDEL) or inactive (B-Mage 1-Glyc-KDELGL) for recycling to the endoplasmic reticulum. After washing, 210.sup.4 cytotoxic T cells specific for the Mage 1 epitope (clone 82/30) were incubated with PBMC cells primed for 24 hours. The supernatants were then collected and tested for the production of interferon.

[0080] FIG. 4: MHC class I restricted presentation of the Shiga B-Mage 1 fusion protein by different types of cells presenting antigens. B lymphoblastoid cells (), dendritic cells (+) or clonal T cells () were primed with the soluble Shiga B-Mage 1 fusion protein, as for FIG. 3. Presentation of Mage 1 peptides was tested using the 82/30 CTL line.

[0081] FIG. 5: Analysis of the specificity of the MHC class I restricted presentation of the Shiga B-Mage 1 fusion protein by lines of B lymphoblastoid cells. Cells from the BM21 (HLA-A1) or BV1 (HLA-A2) B lymphoblastoid line were primed overnight either with medium alone or with Mage 1 or Mart 1 synthetic peptides (1 M, or with ShigaB-Mage 1 fusion protein (1 M), or with Antp-Mage 1 fusion protein or with the B fragment of wild-type Shiga toxin. After washing, specific cells of Mage 1 TCL 82/30 (A) or Mart 1 CTL LB373 (B) were incubated for 24 hours with primed B-EBV cells. The supernatants were then collected and tested for the production of interferon.

ICONSTRUCTION OF A RECOMBINANT CHIMERIC POLYNUCLEOTIDE AND THE PRODUCTION OF THE CORRESPONDING POLYPEPTIDE

[0082] I-1) Construction of Plasmid

[0083] The X epitope selected was the MAGE epitope, present in cancer cells of patients with melanoma. The plasmid used was the pSU1O8 plasmid described by Su et al, 1992, Infect. Immun. 60: 33-45, 3359.

[0084] The PCR primers used were as follows:

TABLE-US-00001 SEQIDNo.1: 5-ACTAGCTCTGAAAAGGATGAACTTTGAGAATTCTGACTCAGAATAG CTC-3 SEQIDNo.2: 5-CTTTTCAGAGCTAGTAGAATTAGGATGATAGCGGCCGCTACGAAAA ATAACTTCGC-3

[0085] The primers were used with specific primers from the ShigaAtpE (5) vector

TABLE-US-00002 SEQIDNo.3: 5-CACTACTACGTTTTAAC-3 SEQIDNo.4: 5-CGGCGCAACTATCGG-3
to produce fragments which were cloned at the restriction sites SphI and SaII of the SU108 plasmid.

[0086] Adapter fragments containing the glycosylation site and the KDEL sequence composed of the oligonucleotides sulphate 1: ((5-phosphorylated; 5-GGCCGCCATCCTAATTCTACTTCT-3 (SEQ ID No. 5) and sulphate 2 (5-CTCAGAAGTAGAATTAGGATGGC-3 (SEQ ID No. 6)) or sulphate 3 (5-GAGTCTGAAAAAGATGAACTTTGATGAG-3 (SEQ ID No. 7)) were ligatured overnight at 16 C.

[0087] The resulting fragments were cloned at the NotI and EcoRI restriction sites of pSU108 and containing the cDNA coding for B-Glyc-KDEL.

[0088] I-2) Purification of Proteins

[0089] The recombinant fragments were also purified using the technique described by Su et al, 1991, cited above. In brief, E. coli cells containing recombinant expression plasmids obtained from pSU108 were cultured overnight at 30 C. The culture was then diluted 5 times in LB supplemented with 50 mg/ml of ampicillin, at 50 C. After incubation for 4 hours at 42 C., the cells were thoroughly washed with 10 mM Tris/HCL, pH 8, incubated for 10 minutes in 10 mM Tris/Hcl, pH 8; 25% sucrose, 1 mM EDTA, and finally rapidly re-suspended in a water-ice mixture containing 1 mM of PMSF and a protease inhibitor mixture (leupeptin, chymostatin, pepstatin, antipain and aprotinin). The final step led to rupture of the periplasm. After clarification, the supernatant was charged onto a QFF column (Pharmacia) and eluted with a linear gradient of NaCl in 20 mM Tris/HCl, pH 7.5. depending on the construction, the B fragment was eluted between 120 mM and 400 mM. The fractions containing the B fragment were then dialysed against 20 mM of Tris/HCl, pH 7.5, and re-charged onto a monoQ column (Pharmacia) and eluted in the same manner as before. The resulting proteins, estimated to have a degree of purity of 95% using polyacrylamide-SDS gel 15 electrophoresis, were then stored at 80 C. until use.

[0090] I-3) Induction of a CTL Response In Vitro

[0091] Dendritic cells (DC) were cultured using previously established protocols (Romani et al, 1994). Briefly, PBMC were taken up into suspension in Iscove medium and incubated for 2 h at 37 C. in 6-well trays. Cells which had not adhered were removed and the remaining cells were incubated at 37 C. in the presence of GM-CSF (800 U/ml) and IL-4 (500 U/ml). After 5 days culture, the IL-1 and IFN- in respective concentrations of 50 U/ml and 150 U/ml were added and incubation was continued at 30 C. for 24 h. The dendritic cells were then taken up into suspension in Iscove medium in the presence of increasing concentrations of B fragment coupled with the MAGE epitope and in the presence of 3 g/ml of human 2-microglobulin to improve the capacity of the cells to presentation of membrane epitopes. This mixture was incubated at 30 C. for 4 hrs. DCs which had internalised the B fragment coupled to this epitope were irradiated at 5000 rad, assembled by centrifuging, taken up into suspension, and mixed with CD8.sup.+ lymphocytes (prepared from PBMC). These DC, pulsed with an antigen, and the CD8.sup.+ were then kept in co-culture in the presence of 5 ng/ml of IL-7.

[0092] After 10 days, responsive CD8.sup.+ lymphocytes were re-stimulated by freshly prepared irradiated DCs which were then also incubated in the presence of increasing concentrations of fragment B coupled to the same epitope. The co-culture of DC and responsive CD8.sup.+ lymphocytes was continued in the presence of IL-2 and IL-7 in concentrations of 10 U/ml and 5 ng/ml respectively. This re-stimulation protocol was repeated 3 times.

[0093] In order to measure the introduction of a CTL response, responsive CD8.sup.+ lymphocytes pre-stimulated as described above were incubated in the presence of cancer cells or cells infected with a virus. These cells, which expressed the selected epitope, were labelled with Na.sub.2.sup.51CrO.sub.4 then brought into contact with responsive CD8.sup.+ for 5 hours (Bakker et al, 1994). The radioactivity released in the medium was then determined, enabling the cytotoxic activity of the pre-stimulated responsive CD8.sup.+ lymphocytes to be quantified.

[0094] Results

IIPRESENTATION OF THE MAGE 1 ANTIGEN BY PENA-EBV CELLS AND DENDRITIC CELLS PULSED BY A B FRAGMENT OF THE SHIGA TOXIN CARRYING THIS ANTIGEN

[0095] II-1) Morphological Study of Intracellular Transport of a B Fragment Carrying the MAGE-1 Epitope

[0096] We have demonstrated that it is possible to fuse a peptide sequence to the carboxy-terminal end of the Shiga toxin B fragment while retaining intracellular routing of this protein towards the endoplasmic reticulum (ER). This demonstration was carried out by constructing chimeric polypeptides comprising the B fragment, the N-glycosylation site and the KDEL retention signal. As a control, the KDELGL retention signal was used, namely the inactive version of the KDEL peptide, Misendock and Rothman, 1995, J. Cell. Biol. 129: 309-319. Morphological and biochemical studies have shown that the modified B fragment is transported from the plasmid membrane via the endosomes and the Golgi apparatus to the endoplasmic reticulum. This transport is inhibited by BFA (brefeldin A fungal metabolite) and reduced by nocodazole (a microtubule depolymerisation agent).

[0097] These experiments clearly demonstrate that intracellular routing of the fusion protein towards the endoplasmic reticulum was retained. To evaluate the potential of the B fragment as an epitope vector for anti-tumoral vaccination in vitro, the MAGE-1 epitope was added to the B-Glyc-KDEL fragment under the experimental conditions described above. The novel protein was designated B-MAGE-Glyc-KDEL. The B-MAGE-Glyc-KDEL protein was coupled with the fluorophore DTAF in order to follow its intracellular transport by confocal microscopy. After internalistion, this protein was detectable in the Golgi apparatus and in the ER of HeLa cells and Pena-EBV cells (a B lymphocyte line immortalised using the Epstein-Barr virus). These results confirm the original observations concerning the intracellular transport of the B fragment modified at its carboxy-terminal end (described above) and affirm that certain presenting cells of the hematopoietic line are capable of intemalising the B fragment and transporting the protein to the ER.

[0098] We shall now describe these studies measuring the N-glycosylation of the B-MAGE-Glyc-KDEL protein. N-glycosylation is a modification which is carried out specifically in the ER, and we have demonstrated above that a B fragment carrying a recognition site for N-glycosylation is in fact glycosylated if it is transported to the ER.

[0099] II-2) Study of the Presentation of the MAGE-1 Antigen by Pena-EBV Cells and Dendritic Cells Pulsed by the B-MAGE-Glvc-KDEL Protein.

[0100] In order to evaluate the capacity of the fragment to act as an epitope vector, we used a cytotoxic T lymphocyte clone (CTL 82/30) specifically recognising the MAGE-1 epitope associated with MHC class I of cells presenting the HLA-A1 haplotype. These CTL were kept in the presence of Pena-EBV cells or dendritic cells pulsed with the B-MAGE-Glyc-KDEL protein. If the MAGE-1 epitope is presented by presenting cells, the CTL will be activated and will secrete interferon which (IFN) which is then assayed.

[0101] The quantity of IFN secreted is proportional to the amplitude of the stimulation of the CTLs by the presenting cells.

[0102] Presenting cells (Pena-EBV and dendritic, 20000 cells per round bottom microwell) were either fixed for 1 h with 4% PBS-paraformaldehyde, or were not fixed. They were than washed twice with OptiMEM before being incubated for 15 h with dilutions of the B-MAGE-Glyc-KDEL protein. The protein was tested at 4 dilutions, starting with a concentration of 10 mM final, and diluting 5 in 5 in the OptiMEM medium (medium without serum). After 15 h, the plates were washed twice by low speed centrifuging. The CTL (CTL 82/30) were added in an amount of 5000 CTL per well in 100 l of culture medium (ID-HS-AAG+25 U/ml of IL2). As a positive control, some CTL 82/30 were kept in the presence of line G43 (a B lymphocyte line transfected with an expression plasmid of MAGE-1). After 24 h of incubation, the supernatants were harvested to determine the quantity of IFN produced.

[0103] The results are shown in Table I.

TABLE-US-00003 TABLE I Concentration of B-MAGE-Glyc-KDEL Celltype 10 M 2 M 0.4 M 0.08 M Den- Fixed 446 293 821 64 661 18 312 181 dritic Non 1557 404 1315 91 1231 150 1174 478. fixed Pena- Fixed 70 47 68 48 23 32 4 5 EBV Non 1966 415 1960 206 1544 42 853 116 fixed

[0104] The results are represented by the amount of EFN produced under each set of conditions (averagestandard deviation; n=3).

[0105] We can see that the dendritic cells and the Pena-EBV cells pulsed with the B-MAGE-Glyc-KDEL protein were properly recognised by the CTL, even at low concentrations of the protein. In contrast, the dendritic cells and the Pena-EBV cells which had been previously fixed were not recognised. It thus appears that endocytosis and processing of the B-MAGE-Glyc-KDEL protein had taken place. These encouraging results will now be backed up by in vitro vaccination experiments.

IIIIN VIVO ANTI-TUMORAL AND/OR ANTIVIRAL ACTIVITY TEST IN THE MOUSE

[0106] Mouse dendritic cells were prepared and marked with an antigen derived from P21RAS, P53 or EP2/NER proteins to test the anti-tumoral activity, or HBV, EBV or HPV to test antiviral activity. This preparation of dendritic cells was carried out using the protocol described in 1-4) above.

[0107] These dendritic cells were then introduced into the mouse.

[0108] The antiviral or anti-tumoral effect was observed by subsequent treatment of these mice grafted with tumoral cells or virus expressing this antigen.

Conclusion

[0109] The polypeptide sequences or polynucleotide sequences of the invention can thus advantageously constitute an active principle in a pharmaceutical composition intended for the treatment of certain cancers or certain viral or bacterial infections, from the moment when a particular epitope of said virus or said cancer cell will have been integrated into the recombinant nucleotide sequence, leading to synthesis of a chimeric polypeptide which can be restricted by the MHC class I and can be expressed on the membrane surface of immune system cells.

IVRESTORATION OF INTRACELLULAR TRANSPORT OF THE MUTATED PROTEIN CFTR (F508) USING THE SHIGA TOXIN B FRAGMENT

[0110] The CFTR (cystic fibrosis transmembrane regulator) protein is a chlorine channel of the plasmic membrane. In the large majority of patients with cystic fibrosis, the CFTR gene carries mutations. The mutation (F508) which is the most frequently observed affects intracellular routing of the CFTR protein. In fact, the mutated protein CFTR(F508), which is functional as regards its ionic channel activity, remains blocked at the endoplasmic reticulum, instead of being transported to the plasmic membrane. Using the Shiga toxin B fragment, we have introduced a domain of the CFTR protein which is known to be the domain of interaction with the calnexin protein (ER chaperone) into the endoplasmic reticulum. This domain is fused to the carboxy-terminal end of the B fragment. We have tested whether this chimeric protein can displace N-glycosylated chains of the CFTR (F508) glycoprotein from the interaction site with calnexin, with the result that the CFTR (F508) protein is no longer retained in the endoplasmic reticulum and can be transported to the plasmic membrane and thus function normally.

[0111] Firstly, we constructed a chimeric protein composed of a B fragment and the interaction domain derived from the CFTR protein. A recycle signal (the KDEL peptide) was added to the carboxy-terminal end of this protein to increase its retention in the endoplasmic reticulum. It was first verified that the novel protein was also transported in the endoplasmic reticulum of target cells. Mobilisation of CFTR(F508) was studied in cells of a stable cell line, LLCPKI, transfected with the cDNA of CFTR(F508). This line was established by Mlle. M. A. Costa de Beauregard and M. D. Louvard (Institut Curie, Paris, CNRS UMR 144). The CFTR (F508) protein which was expressed in these cells was also endowed with an epitope tag. It was thus possible to detect the arrival of the CFTR (F508) protein in the plasmic membrane by immunofluorescence. In the absence of treatment, the plasmic membrane of LLCPKI cells of the line was depleted with specific CFTR (F508) tags in the plasmic membrane. While the results of these pilot experiments are promising, we are obliged to develop this approach within the context of cystic fibrosis therapy.

Conclusion

[0112] The experiment described above shows that the synthetic polypeptide in which X is constituted by an interaction domain between the CFTR protein and calnexin can advantageously constitute the active principle of a therapeutic composition intended to treat cystic fibrosis. In fact, competition between the mutated interaction domain in the mutant and the fragment of synthetic polypeptide for the interaction with calnexin can restore secretion of the mutated protein in the bronchia.

Antigenic Presentation Test:

[0113] Cells presenting the antigen (CMSP, B-EBV cells, T cells, dendritic cells) were incubated in 96-well microplates in a density of 10.sup.5 cells per well and pulsed at 37 C. for 4 hours or 15 hours with the antigen dissolved in 1001 of Iscove medium. After incubation, the medium was removed and 20000 CTL cells were added to each well in 100 l of CTL culture medium containing 25 U/ml of IL2. After 24 hours, 50 l of supernatant was collected and the interferon was measured by an ELISA (Diaclone) test. In some experiments, the cells were fixed with 1% paraformaldehyde for 10 minutes at ambient temperature and washed thoroughly before transfer into the microplates.