DESCRIPTION The present invention refers to recombinant defective adenoviruses containing a DNA sequence coding for hIL-6 (human Interleukin 6) antagonists and/or superantagonists. Such recombinant adenoviruses are used for the preparation of pharmaceutical compositions, in particular in gene therapy for the treatment and/or prevention of diseases caused by hIL-6 overproduction.

Altered hIL-6 serum levels have been described in several pathologies, such as multiple myeloma (1) , Castleman's disease (2), mesangial glo erulonephritis (3) , osteoporosis (4) , EBV positive lymphomas (5) , rheumatoid arthritis (6) and systemic lupus erythematosus (7) . In multiple myeloma, IL-6 clearly plays an autocrine and paracrine role in the growth of malignant cells.

It is known from WO 95/26409 and WO 95/25508, in the name of Rhone-Poulenc Rorer S.A., the use of recombinant defective adenoviruses to achieve stable expression of a protein of interest (in particular a cytokine as in the case of WO 95/26409) , whose coding sequence is inserted in the defective adenovirus genome under the control of a strong promoter. It is also known that, in the case of adenovirus mediated gene therapy, inflammation represents a major obstacle, being the cause for tissue injury as well as, at least in part, for termination of gene expression (9) (10) . There are indications that IL-6 is the major inflammatory cytokine produced after adenovirus injection (11) (12) .

Now, it has been unexpectedly found that mutants of hIL-6 with antagonist or superantagonist activity over hIL-6 (disclosed in WO/00852 in the name of the present Applicant) produced by cells infected both in vitro and

in vivo with a recombinant defective adenovirus, containing a DNA sequence coding for said antagonists or superantagonists under the control of a strong promoter, inhibit hI -6 activity on a variety of human cells. Therefore, said recombinant adenoviruses can be used to prolong adenovirus mediated gene of interest expression in vivo.

The subject matter of the present invention is a recombinant defective adenovirus comprising a DNA or a cDNA sequence coding for a mutant of human Interleukin 6, a fragment thereof or a derivative thereof, with antagonist or superantagonist activity over hIL-6.

In preferred embodiments of the invention, the hIL-6 mutants with superantagonist activity over hIL-6 are Tyr31Asp, Gly35Phe, SerllθArg, Vall21Asp, Glnl75Ile, Serl76Arg, Glnl83Ala and Tyr31Asp, Gly35Phe, Leu57Asp, Glu59Phe, Asn60Trp, Gln75Tyr, Ser76Lys, Serll8Arg, Vall21Asp, Glnl75Ile, Serl76Arg, Glnl83Ala (hereinafter referred to also as Santl and Sant7 respectively) . The recombinant defective adenovirus of the invention can be put under the control of signals allowing its expression in different kind of human specific cells. The expression signals can be selected among the viral promoters . The viral promoter can be Rous Sarcoma Virus (RΞV) .

The invention also refers to the use of the recombinant defective adenovirus, comprising at least a DNA or a cDNA sequence coding for a mutant of hIL-6, a fragment thereof or a derivative thereof, with antagonist or superantagonist activity over hIL-6, for treating and/or preventing diseases such as multiple myeloma, Castleman's disease, mesangial glomerulonephritis, osteoporosis, EBV positive lymphomas, rheumatoid arthritis and systemic lupus erythematosus . Pharmaceutical compositions comprising at least one of the above recombinant defective adenoviruses are also the subject matter of the present invention. The

pharmaceutical compositions can be in injectable form. They can contain the recombinant defective adenovirus in the range from 104 to 1014 pfu/ml (plaque forming unit/ml) , preferably from 106 to 1010 pfu/ml. Mammalian cells, in particular human cells, infected by at least one of the above recombinant adenoviruses are a further subject of the present invention. These cells can additionally express a gene of interest coding for a therapeutic protein. These cells can be further infected with at least one virus expressing a therapeutic gene.

The infected human cells can be selected from the group comprising fibroblasts, myoblasts, hepatocytes, endothelial cells, glial cells and keratynocytes.

Another subject matter of the present invention is an implant comprising the cells infected by at least one of the above recombinant defective adenoviruses and an extracellular matrix. The extracellular matrix can be a gelifying compound which can be selected from the group consisting of collagen, gelatin, glucosaminoglycans, fibronectins and lectins. The extracellular matrix can comprise a support for anchoring the infected cells. The support can be preferably made of polytetrafluoroethylene fibers .

The objects, characteristics and advantages of the invention will be further illustrated, in a non limitative extent, by the following examples directed to embodiments thereof. EXAMPLE 1 Construction of a recombinant adenovirus . called AdRSVhlL-S(■ containing the. sequence coding tor t e mutant IL-6 form called Santl and characterised by havincr the mutations Tyr31Asp. Gly35Phe. Serll8Arσ. Vall21Asp,

Glnl75Ile. Serl76Arσ and Qlnl83Ala

For the construction of AdRSVhIL-6 (, the cDNA coding for Santl (described in WO95/00852 by the same applicant and characterised by having the seven a ino acid substitutions Tyr31Asp, Gly35Phe, SerllθArg, Vall21Asp,

Glnl75Ile, Serl76Arg and Glnl83Ala as compared with wt hIL-6) was subcloned together with hIL-6 leader sequence in the adenoviral vector pAdRSVβgal (20) , cleaved with the restriction enzymes Sail and EcoRV. The recombinant plasmid was cotransfected together with Clal restricted AdRSVβgal genome (20) into human embryonic kidney 293 cells (ECACC) . Recombinant plaques were isolated and amplified as described in (14) (15) . DNA pattern of recombinant AdRSVhIL-6 ( was controled by multiple restriction analysis.

All viral stocks were prepared in 293 cells and purified twice on isopicnic CsCl gradient . Desalting was performed using Pharmacia G50 columns. Viruses were aliquoted and kept in PBS-15% glycerol at -80 °C. Titers, calculated by plaque analysis on 293 cells, varied from 1 to 2 xlOll pfu/ml. EXAMPLE 2

Calls infected in vitro with t±i≤ recombinant adenovirus AdRSVhII.-6.. produce Santl in the cell culture supernatant

To test antagonist expression by the AdRSVhIL-6(, constructed as described in the above example, 35 mm falcon plates of subconfluent human embryonic kidney 293 cells were infected in 3 ml of medium with 10 plaque forming unit (pfu) /cell of a purified AdRSVhIL-6( stock, prepared as described in the example 1. 24 hours after infection, supernatants were collected and concentrated 5-fold with centricon 3 tubes (Amicon) . 25 μl of concentrated supernatants were loaded on acrylamide gel . The gel was electroblotted and nitrocellulose membrane was blocked overnight in TBS-5% milk-0.05% Tween 20.

After washing in TBS-0.05% Tween 20, the membrane was incubated two hours with polyclonal anti-hIL6 antibodies

(R&D, hIL6 ELISA kit) , coupled to horse-radish peroxidase, diluted 1;2, v/v, in TBS-5% milk-0.0 ^~ 5% Tween 20. After washing with TBS-0.05% Tween 20, the membrane was revealed with ECL bioluminescence kit (Amersham)

following the instructions furnished by the manufacturer. A single band, of the expected molecular mass, was detected in the supernatants of cells infected with AdRSVhIL-6(, but not in the supernatants of cells infected with AdRSVβgal, or in the supernatants of mock- infected cells. Therefore cells infected with AdRSVhIL-6( express and secrete in the cell culture supernatant a protein of the same size as hIL-6 and as Santl (which is a mutant form of hIL-6 itself) and which is recognised by antibodies raised against hIL-6. EXAMPLE 3

Santl expressed in vitro by cells infected with

AdRSVhIL-6 ( inhibits hlL-6 biological activity on human hepato a cells It is well known in the state of the art that IL-6 stimulates the transcription of a set of genes (known as Acute Phase Genes) in liver cells or in liver-derived hepatoma cell lines in culture. In particular, IL-6 stimulates the transcription by the C-reactive gene promoter in human Hep3B hepatoma -cells ((13) , and references therein) .

To test whether the Santl produced by the recombinant virus was able to inhibit the IL-6 biological activity described above, 35 mm falcon plates of subconfluent human embryonic kidney 293 cells were infected with AdRSVhIL-6( or AdRSVβgal (in 3 ml of medium with 10 plaque forming unit (pfu) /cell of a purified stock) , 24 hours after infection supernatants were collected and tested for inhibition of hIL-6 biological activity on human Hep3B hepatoma cells. The effectiveness of transcriptional stimulation was measured according to the state of the art (13) . Human Hep3B hepatoma cells were stimulated with 4 ng/ml of hIL-6, and this extent of stimulation was taken as 100%, or with 4 ng/ml of hIL-6 in the presence of serial dilutions of cell culture supernatant obtained from AdRSVhIL-6 ( infected 293 cells; in the latter cases the extent of -transcriptional

stimulation was expressed as percent of the stimulation obtained in cells incubated with 4 ng/ml of hIL-6 only. The results of the experiment are given in Table 1 below.

TABLE 1 Inhibition of hIL-6 biological activity on human Hep3B hepatoma cells by serial dilutions of cell culture supernatant obtained from AdRSVhIL-6 ( infected 293 cells.

Biological activity

Serum dilut:ion {% of the control

1:2916 85.8 0, o

1:972 87.7 "o

1:324 72.3 %

1:108 57.0 %

1:36 28.1 o, 0

1:12 23.8 %

1:4 8.1 9o-

It can be seen that the cell culture supernatant obtained from AdRSVhIL-6( infected 293 cells diluted 1:4 inhibit completely the bioactivity-of 4 ng/ml of hIL-6 on human Hep3B hepatoma cells. Inhibition was not due to toxicity, because in the presence of molar excess of hIL-6 (1000 ng/ml) the C-reactive protein gene promoter activity was rescued up to 100% (data not shown) . The data reported in table 1 above can be used to plot a dose-inhibition curve, from which it can be interpolated that 50% of inhibition of IL-6 activity is obtained at a serum dilution of 1:60. In a parallel inhibition experiment performed not with AdRSVhIL-6 ( infected 293 cells supernatant, but with recombinant, E. coli produced Santl, 50% of inhibition of IL-6 activity is obtained at a protein concentration of 10 ng/ml. Therefore, the Santl concentration in undiluted AdRSVhIL-6 ( infected 293 cells supernatant can be calculated multiplying the concentration of Santl needed to reach 50% inhibition of IL-6 activity on Hep3B hepatoma cells (10 ng/ml) times the AdRSVhIL-6 ( infected 293 cells supernatant dilution

factor (60) which gives 50% inhibition of IL-6 activity. The caculation gives:

Santl concentration in undiluted AdRSVhIL-6 ( infected 293 cells supernatant= = 10 ng/ml x 60 = 600 ng/ml.

EXAMPLE 4

Santl expressed in vitro b . CSJJLS infected with

AdRSVhIL-6 ( inhibits hTL-6 biological activity on human myeloma cells In the previous example it was shown that the cell culture supernatant obtained from AdRSVhIL-6 ( infected 293 cells was able to inhibit interleukin 6 biological activity (stimulation of transcription by an IL-6 inducible promoter) on human hepatoma cells. In the introductory part of the above description it is stated that altered IL-6 serum levels have been described in several pathologies, like various forms of multiple myeloma/plasmacytoma. We provide here a further example by showing that the cell culture supernatant obtained from AdRSVhIL-6 ( infected 293 cells fully inhibit the interleukin 6-dependent growth of a human myeloma cell line, called XG-1, derived from freshly isolated myeloma cells obtained from a patient with terminal disease. The XG-1 myeloma cell line growth is strictly dependent on exogenously added IL-6, similarly to what has been shown for fresh myeloma cells, therefore this cell line can be considered an excellent in vitro model of the multiple myeloma (16) .

To test mutants antagonism on wild type interleukin 6, XG-1 myeloma cells were cultured in 96-well microtiter plates at 6000 cell/microwell with hIL-6 at 0.1 nanograms per milliliter (ng/ml) of culture medium, in the presence of serial dilutions of cell culture supernatant obtained from AdRSVhIL-6( infected 293 cells as described in the previous example. After 7 days of culture, cell numbers were evaluated by colorimetric determination of hexosaminidase levels (17) . The following Table 2 shows

the inhibition of hIL-6 activity as a function of the dilutions of cell culture supernatant obtained from AdRSVhIL-6( infected 293 cells.

TABLE 2 Inhibition of hIL-6 biological activity on human XG-1 myeloma cells by serial dilutions of cell culture supernatant obtained from AdRSVhIL-6 ( infected 293 cells.

Biological activity

Serum dilut;ion (% of the control

1:2916 84.7 %

1:972 83.1 %

1:324 80.5 %

1:108 61.0 %

1:36 46.1 %

1:12 28.2 %

1:4 7.0 %

It can be seen that the cell culture supernatant obtained from AdRSVhIL-6 ( infected 293 cells diluted 1:4 inhibit completely the growth stimulatory effect of 0.1 ng/ml of hIL-6 on human XG-1 myeloma cells. As it was in the previous example, inhibition was not due to toxicity, because in the presence of molar excess of hIL-6 (100 ng/ml) XG-1 cell growth was rescued up to 100% (not shown) . The data reported ^" in table 2 above can be used to plot a dose-inhibition curve, from which it can be interpolated that 50% of inhibition of IL-6 activity is obtained at a serum dilution of 1:40. In a parallel inhibition experiment performed not with AdRSVhIL-6 ( infected 293 cells supernatant, but with recombinant., E. coli produced Santl, 50% of inhibition of IL-6 activity is obtained at a protein concentration of 20 ng/ml. Therefore, the Santl concentration in undiluted AdRSVhIL-6( infected 293 cells supernatant can be calculated multiplying the concentration of Santl needed to reach 50% inhibition of IL-6 activity on XG-1 myeloma

cells (20 ng/ml) times the AdRSVhIL-6( infected 293 cells supernatant dilution factor (40) which gives 50% inhibition of IL-6 activity. The caculation gives:

Santl concentration in undiluted AdRSVhIL-6 ( infected 293 cells supernatant= = 20 ng/ml x 40 = 800 ng/ml

which is in very good agreement with the estimate of Santl concentration in undiluted AdRSVhIL-6 ( infected 293 cells supernatant obtained in the previous example (600 ng/ml) .

EXAMPLE 5

Animals inj cted with the recombinant adenovirus AdRSVhIL-6( produce Santl

To test in vivo antagonist expression by the AdRSVhIL-6(, constructed as described in the example 1, 6 to 8 weeks old mice (purchased from IFFACREDO, France) were injected in the orbital vein with the standard dose of 2x109 pfu/mouse, diluted in lOOμl of PBS. Blood was collected at different times after the injection by puncture in the orbital vein after treating the animal with ether. Sera were prepared according to the state of the art and the antagonist concentration in the serum was measured according to the state of the art by means of a "sandwich" ELISA test, using a commercially available kit produced by the company "R&D Systems", and scrupulously following the manufacturer's instructions. The experiment demonstrated that intra-venous (iv) injection of 2x109 pfu of AdRSVhIL-6( in 6 to 8 weeks old mice allowed expresion of Santl at a concentration of 5 ng/ml, for at least two months .

EXAMPLE 6 iSanti produced in vivo in mice infecte — ith AdRSVhIL-6( inhibits hTL-6 induced APRF activation in human myeloma cells

It is well known in the state of the art that one of the first steps in IL-6 induced cell activation is tyrosine phosphorilation of the acute phase transcription factor (APRF) . Upon phosphorilation, APRF aquires the ability to bind specific DNA sequences (acute phase response elements: APREs) , and migrates to the nucleus (18) .

To test whether the Santl produced in vivo by the recombinant adenovirus was able to inhibit the IL-6 biological activity described above, sera of mice injected with 2x109 pfu of AdRSVhIL-6 ( or of control adenovirus were collected at day 40 after the injection as described in the previous example, concentrated and tested on human XG-1 myeloma cells for inhibition of hlL- 6 biological activity described above. After 4 hours of IL-6 deprivation, XG-1 cells were incubated for 15 minutes at 37° C with culture medium without IL-6. with 0.2 ng/ml of IL-6 in the presence of 50% serum of uninjected mice, with 0.2 ng/ml of IL-6 in the presence of 50% serum of mice injected with AdRSVhIL-6 ( and with 0.2 ng/ml of IL-6 in the presence of 50% serum of mice injected with control virus. Whole cell extract were prepared and APRF activation was monitored by gel retardation as described in the state of the art (18) . APRF activation for the various samples was quantified at the phosphoimager and the extent of activation in the various samples was expressed as percent of the activation obtained in cells stimulated with 0.2 ng/ml of IL-6 in the presence of 50% serum of uninjected mice. The results of the experiment are reported in table 3 below.

TABLE 3 Inhibition of hIL-6 induced APRF activation in human XG-1 myeloma cells by serum of mice injected with AdRSVhIL-6 ( or with control adenovirus

APRF activation Sample (% of the control) no IL-6 0%

0.2 ng/ml IL-6, 50% serum of uninjected mice 100% 0.2 ng/ml IL-6, 50% serum of AdRSVhIL-6( injected mice 5% 0.2 ng/ml IL-6, 50% serum of control adenovirus injected mice 75%

As it can be noticed, serum from mice injected with AdRSVhIL-6( recombinant adenovirus and containing Santl expressed in vivo as determined in the previous example inhibit very effectively APRF activation induced by 0.2 ng/ml of IL-6 in XG-1 cells. Again, inhibition was not due to toxicity, because in the presence of molar excess of hIL-6 (100 ng/ml) the IL-6 induced APRF activation was fully rescued (not shown) . Moreover, table 3 also show that serum from mice injected with control recombinant adenovirus failed to inhibit significantly APRF activation induced by 0.2 ng/ml of IL-6 in XG-1 cells.

REFERENCES

(1) Klein B, Brailly H. (1995) Immunol. Today 16:216- 220.

(2) Brandt, S. J. , Bodine, D.M., Dunbar, C. E. and Nienhuis, A. W. (1990) J. Clin. Invest. 86:592- 599.

(3) Horii, Y. et al . (1989) J. Immunol. 143:3949-3955.

(4) Jilka RL, Giao H, Girasole G, Passeri G, Williams DC, Abrams JS, Boyce B, Broxmeyer H and Manolagas SC (1992) Science 257:88-91.

(5) Durandy A, E ilie D, Peuchmar M, et al . (1994) J. Immunol. 152:5361-5367.

(6) Hermann E, Fleisher B, Mayer WJ, Poralla T, and Meyer Zura Beschernfelde KH (1989) Clin. Exp. Rheumatol. 7:411-414.

(7) Hirohata S, Myiamoto T. (1990) Arthritis Rheum. 33:644-649.

(8) Swaak A, van Rooyen A, Aarden LA. (1989) Rheumatol. Int 8:263-268.

(9) Engelhardt JF, Ye X, Doranz B, Wilson J. (1994) Proc. Natl. Acad. Sci . 91:6196-6200. (10) McCoy RD, Davidson BL, Roessler BJ, et al . (1995) Hum Gene Ther 6:1553- 1560.

(11) Ginsberg HS, Moldawer LL, Sehgal PB, et al . (1991) Proc. Natl. Acad. Sci. 88:1651-1655.

(12) Crystal RG, McElvaney NG, Rosenfeld MA, et al . (1994) Nature Genetics 8:42-51.

(13) Gregory B, Savino R and Ciliberto G. (1994) J. Immunological Methods 170:47-56

(14) Graham FL, Van der Eb EJ. (1973) Virology 52:456- 467. (15) Graham FL, Smiley J, Russel WC, Nairn R. (1977) J.

Gen. Virol. 36:59-72. (16) Jourdan M, Zhang X-G, Portier M, Boiron J-M,

Bataille R and Klein B. (1991) J. Immunol.

147:4402-4407. (17) Landegren, U. (1984) J. Immunol. Methods 67:379-388

(18) Lϋtticken C, Wegenka UM, Buschmann J, Schindler C, Ziemiecki A, Harpur AG, Wilks AF, Yasukawa K, Taga T, Kishimoto T, Barbieri G, Pellegrini S, Sendtner M, Heinrich PC and Horn F (1994) Science 263:89-92.

(19) Stratford-Perricaudet LD, Makeh I, Perricaudet M, Briand P. (1992) J. Clin. Invest. 90:626-630.