IL124016A - Trisulfated disaccharide and pharmaceutical compositions comprising it - Google Patents

Abstract

A compound of the formula in which X is H+ or a pharmaceutically acceptable cation. The compounds are for use for the inhibition of the production of an active cytokine.

Description

124016/2 9238 DIV.

A TRISULFATED DISACCHARIDE AND PHARMACEUTICAL COMPOSITIONS COMPRISING IT o >Dttii tmpy > vss}) VHsbw Jiwap υ^υι η» »ίΐ tm ^ Yeda Research and Development Co. Ltd.

Inventors: Irun R. Cohen, Ofer Lider, Liora Cahalon Oded Shoseyov and Raanan Margalit ,Ϊ^ΓΟ rniN> ,-i*p Ί£5ΐ ,ΙΓΡ "ρ : o wsoo 1. FIELD OF THE INVENTION The present invention relates to a disaccharide, pharmaceutical compositions comprising said disaccharide and to the use thereof for inhibiting the production of an active cytokine. 2. BACKGROUND OF THE INVENTION 2.1. Tumor Necrosis Factor Alpha TNF-a, a cytokine produced by monocytes (macrophages) and T lymphocytes, is a key element in the cascade of factors that produce the inflammatory response and has many pleiotropic effects as a major orchestrator of disease states (Beutler, B. and Cerami, A., Ann. Rev. Immunol. (1989) 7:625-655).

The biologic effects of TNF-a depend on its concentration and site of production: at low concentrations, TNF-a may produce desirable homeostatic and defense functions, but at high concentrations, systemically or in certain tissues, TNF-a can synergize with other cytokines, notably interleukin-1 (IL-1) to aggravate many inflammatory responses .

The following activities have been shown to be induced by TNF-a (together with IL-1) ; fever, slow-wave sleep, hemodynamic shock, increased production of acute phase proteins, decreased production of albumin, activation of vascular endothelial cells, increased expression of major histocompatibility complex (MHC) molecules, decreased lipoprotein lipase, decreased cytochrome P450, decreased plasma zinc and iron, fibroblast proliferation, increased synovial cell collagenase, increased cyclo-oxygenase activity, activation of T cells and B cells, and induction of secretion of the cytokines, TNF-a itself, IL-1, IL-6, and IL-8.

Indeed, studies have shown that the physiological effects of these cytokines are interrelated (Philip, R. and Epstein, L. B. , Nature (1986) 323 ( 6083 ): 86-89 ; Wallach, D. et al. , J. Immunol. (1988) 140(9) :2994-2999) .

How TNF-a exerts its effects is not known in detail, but many of the effects are thought to be related to the ability of TNF-a to stimulate cells to produce prostaglandins and leukotrienes from arachidonic acid of the cell membrane.

TNF-a, as a result of its pleiotropic effects, has been implicated in a variety of pathologic states in many different organs of the body. In blood vessels, TNF-a promotes hemorrhagic shock, down regulates endothelial cell thrombomodulin and enhances a procoagulant activity. It causes the adhesion of white blood cells and probably of platelets to the walls of blood vessels, and so, may promote processes leading to atherosclerosis, as well as to vasculitis.

TNF-α activates blood cells and causes the adhesion of neutrophils, eosinophils, monocytes/macrophages and T and B lymphocytes. By inducing IL-6 and IL-8, TNF- augments the chemotaxis of inflammatory cells and their penetration into tissues. Thus, TNF-a has a role in the tissue damage of autoimmune diseases, allergies and graft rejection.

TNF-a has also been called cachectin because it modulates the metabolic activities of adipocytes and contributes to the wasting and cachexia accompanying cancer, chronic infections, chronic heart failure, and chronic inflammation. TNF-a may also have a role in anorexia nervosa by inhibiting appetite while enhancing wasting of fatty tissue.

TNF-a has metabolic effects on skeletal and cardiac muscle. It has also marked effects on the liver: it depresses albumin and cytochrome P450 metabolism and increases production of fibrinogen, 1-acid glycoprotein and other acute phase proteins. It can also cause necrosis of the bowel.

In the central nervous system, TNF-a crosses the blood-brain barrier and induces fever, increased sleep and anorexia. Increased TNF-a concentration is associated with multiple sclerosis. It further causes adrenal hemorrhage and. affects production of steroid hormones, enhances collagenase and PGE-2 in the skin, and causes the breakdown of bone and cartilage by activating osteoclasts .

In short, TNF-a is involved in the pathogenesis of many undesirable inflammatory conditions in autoimmune diseases, graft rejection, vasculitis and atherosclerosis. It may have roles in heart failure and in the response to cancer. For these reasons, ways have been sought to regulate the production, secretion, or availability of active forms of TNF-α as a means to control a variety of diseases.

The prime function of the immune system is to protect the individual against infection by foreign invaders such as microorganisms. It may, however, also attack the individual's own tissues leading to pathologic states known as autoimmune diseases. The aggressive reactions of an individual's immune system against tissues from other individuals are the reasons behind the unwanted rejections of transplanted organs. Hyper-reactivity of the system against foreign substances causes allergy giving symptoms like asthma, rhinitis and eczema.

The cells mastering these reactions are the lymphocytes, primarily the activated T lymphocytes, and the pathologic inflammatory response they direct depends on their ability to traffic through blood vessel walls to and from their target tissue. Thus, reducing the ability of lymphocytes to adhere to and penetrate through the walls of blood vessels may prevent autoimmune attack, graft rejection and allergy. This would represent a new therapeutic principle likely to result in better efficacy and reduced adverse reactions compared to the therapies used today.

Atherosclerosis and vasculitis are chronic and acute examples of pathological vessel inflammation. Atherosclerosis involves thickening and rigidity of the intima of the arteries leading to coronary diseases, myocardial infarction, cerebral infarction and peripheral vascular diseases, and represents a major cause of morbidity and mortality in the Western world. Pathologically, atherosclerosis develops slowly and chronically as a lesion caused by fatty and calcareous deposits. The proliferation of fibrous tissues leads ultimately to an acute condition producing sudden occlusion of the lumen of the blood vessel .

TNF- has been shown to facilitate and augment human immunodeficiency virus (HIV) replication in vitro (Matsuyama, T. et al., J. Virol. (1989) 63 (6) : 2504-2509 ; Michihiko, S. et al., Lancet (1989) 1 (8648) : 1206-1207) and to stimulate HIV-1 gene expression, thus, probably triggering the development of clinical AIDS in individuals latently infected with HIV-1 (Okamoto, T. et al. , AIDS Res. Hum. Retroviruses (1989) 5(2) :131-138) .

Hence, TNF-a, like the inflammatory response of which it is a part, is a mixed blessing. Perhaps in understanding its physiologic function, one may better understand the purpose of inflammation as a whole and gain insight into the circumstances under which "TNF-a deficiency" and "TNF-a excess" obtain. How best to design a rational and specific therapeutic approach to diseases that involve the production of this hormone may thus be closer at hand. 2.2. Heparin Heparin is a glycosaminoglycan, a polyanionic sulfated polysaccharide, which is used clinically to prevent blood clotting as an antithrombotic agent. In animal models, heparin has been shown to reduce the ability of autoimmune T lymphocytes to reach their target organ (Lider, 0. et al., Eur. J. Immunol. (1990) 20:493-499). Heparin was also shown to suppress experimental autoimmune diseases in rats and to prolong the allograft survival in a model of skin transplantation in mice, when used in low doses (5 /ig for mice and 20 μg for rats) injected once a day (Lider, O. et al., J. Clin.

Invest. (1989) 83:752-756).

The mechanisms behind the observed effects are thought to involve inhibition of release by T lymphocytes of enzyme (s) necessary for penetration of the vessel wall, primarily the enzyme heparanase that specifically attacks the glycosaminoglycan moiety of the sub-endothelial extracellular matrix (ECM) that lines blood vessels (Naparstek, Y. et al. , Nature (1984) 310:241-243). Expression of the heparanase enzyme is associated with the ability of autoimmune T lymphocytes to penetrate blood vessel walls and to attack the brain in the model disease experimental autoimmune encephalomyelitis (EAE) .

European Patent Application EP 0114589 (Folkman et al.) describes a composition for inhibition of angiogenesis in mammals in which the active agents consist essentially of (1) heparin or a heparin fragment which is a hexasaccharide or larger and (2) cortisone or hydrocortisone or the 11-a isomer of hydrocortisone. According to the disclosure, heparin by itself or cortisone by itself are ineffective; only the combination of both gives the desired effects. Although there is no proof in the literature that there is a connection between angiogenesis and autoimmune diseases, the description on page 5 of the patent application connects angiogenesis with psoriasis and with arthritis, indicating the use of high doses of 25,000 units to 47,000 units of heparin per day (i.e., about 160 to about 310 mg per day) .

Horvath, J. E. et al.,in Aust. N.Z.J. Med. (1975) 5(6) :537-539, describe the effect of subanticoagulant doses of subcutaneous heparin on early renal allograft function. The daily dosage is high (5000 U or about 33 mg) and the conclusion of the study is that heparin in subanticoagulant doses has no effect on early graft function or graft survival and S that it may be associated with increased hemorrhagic complications.

Toivanen, M. L. et al. , Meth . and Find. Exp. Clin. Pharmacol. (1982) 4 (6) : 359-363 , examined the effect of heparin in high dosage (1000 U/rat or about 0 7 mg/rat) in the inhibition of adjuvant arthritis in rats and found that heparin enhanced the severity of the rat adjuvant arthritis.

PCT Patent: Application PCT/AU88/00017 published under No. WO88/05301 (Parish et al.) S describes sulphated polysaccharides that block or inhibit endoglycosylase activity, such as heparanase activity, for use as antimetastatic and anti-inflammatory agents. Heparin and heparin derivatives, such as periodate oxidized, reduced 0 heparins, that had negligible anticoagulant activity, were shown to have antimetastatic and anti-inflammatory activity when used in dosages within the range of 1.6-6.6 mg per rat daily, administered by constant infusion (corresponding .to 5 75-308 mg daily for an adult human patient) .

Heparin and heparan sulfate are closely related glycosaminoglycan macromolecules . The degradation products of these polymeric macromolecules, which are termed low molecular weight Q heparins (LMWH) , may have the same or greater pharmacologic effects on the blood clotting system as the parent macromolecules. Furthermore, because there is extensive but incomplete post-synthetic processing of the polymer's basic disaccharide subunit, 5 glucuronic acid and N-acetyl glucosamine, the LMWH will be a heterogeneous mixture not only of sizes but also of chemical compositions (See Goodman and Gilman's The Pharmacological Basis of Therapeutics, 8th Ed., (Pergamon Press, New York, 1990) pp. 1313-1315. Methods to obtain low molecular weight products from heparin, which are useful as anticoagulants, are described in the art. These methods seek to optimize the persistence in vivo or the extent of hemorrhagic side effects of their products (See, for example, Alpino R. R. , et al., U.S. Patent No. 5,010,063; Choay, J., et al. , U.S. patent No. 4,990,502; Lopez, L. L., et al., U.S. Patent No. 4,981,955). Others teach the use of affinity chromatographic methods to obtain low molecular weight products (See, for example, Rosenberg, R. D. , et al., U.S. Patent No. 4,539,398 and Jordan, R. E. , et al., U.S. Patent No. 4,446,314) .

Psuja, P., as reported in Folio Haematol. fLejpz) . (1987) 114:429-436, studied the effect of the heterogeneity of heparins on their interactions with cell surfaces. Psuja reported that there are moderate affinity receptors for LMWH (Dd= 5.6 μΜ) found on cultured endothelial cells, but he determined that the upper limit of the fraction of LMWH bound to these receptors was less than 1% of total LMWH.

Other workers have demonstrated effects of LMWH on the metabolism of a variety of cultured cell types. Asselot-Chapel, C, et al., in Biochem.

Pharmacol. (1989) 38:895-899 and Biochem. Biophys. Acta. (1989) 993:240-244, report that LMWH cause cultured smooth muscle cells to decrease the ratio of type III to type I collagen and fibronectin synthesis. Rappaport, R. in U.S. Patent No. 4,889,808, teaches that LMWH can cause human diploid pulmonary fibroblasts, cultured in the absence of serum, to respond to LMWH by increased secretion of tissue plasminogen activator and related proteins.

Effects of LMWH on complex multicellular systems have been reported. The work of Folkman et al. and Lider et al., in EPO Application 0114589 and J. Clin. Invest. (1989) 83:752:756, have been noted above. In addition, Diferrante, N. , in published International Application WO 90/03791, teaches the use of LMWH to inhibit the reproduction of HIV in cultures of C8166 transformed human lymphocytes (ALL) .

However, none of the prior art experiments that have studied the effects of LMWH on cellular metabolism has considered that the heterogeneity of LMWH may produce antagonistic effects. Furthermore, none has shown or suggested a regulatory effect on cytokine activity based on the use of substantially pure oligosaccharide substances. 3. SUMMARY OF THE INVENTION The present invention provides a disaccharide of the formula (II) in which X is H+ or a pharmaceutically acceptable cation.

The disaccharide and the pharmaceutically acceptable salts thereof are capable of inhibiting the production of an active cytokine, and thus can be used as active ingredients of pharmaceutical compositions, together with a pharmaceutically acceptable carrier, useful for the inhibition of the production of an active cytokine. The cytokine may be, but is not limited to, IL-1, IL-6, IL-8 and, in particular, TNF-a.

The pharmaceutical compositions of the present invention are useful in methods of treating a host, such as a mammalian subject, suffering from a medical condition the severity of which can be affected by the activity of a cytokine in the host comprising administering to such host said pharmaceutical compositions comprising the disaccharide in substantially purified form or the pharmaceutical compositions that can be prepared from same. Depending on the medical condition of the particular host, the disaccharide or the composition can be administered in Wi which reduce the availability or activity of TNF-a . Such compositions may be administered at low dosage levels daily or at intervals of up to about 5-8 days, preferably, once a week. Pharmaceutical compositions containing the disaccharide for parenteral, oral, or topical administration may be administered daily according to convenience and effectiveness and at dosages that would be readily determined by routine experimentation by one of ordinary skill in the art.

Furthermore, it is an object of the present invention to provide pharmaceutical compositions that may be administered in any manner as dictated by the particular application at hand including, but not limited to, enteral administration (including oral or rectal) or parenteral administration (including topical or inhalation with the aid of aerosols) . In preferred embodiments, the pharmaceutical compositions of the present invention are administered orally, subcutaneously, intraperitoneally or intravenously.

Thus, the pharmaceutical compositions of the present invention are useful, for example, in delaying or preventing allograft rejection and treating or preventing a variety of pathological processes such as those related to immune diseases, allergy, inflammatory diseases (in particular, inflammatory bowel diseases), or acquired immunodeficiency syndrome (AIDS) . The allograft may, of course, include an organ transplant, including, but not limited to, heart, liver, kidney or bone marrow transplants, and skin grafts. The present invention also finds utility in the treatment of diabetes type I, periodontal disease, skin diseases, liver diseases, uveitis, rheumatic diseases (in particular, rheumatoid arthritis), atherosclerosis, vasculitis, or multiple sclerosis .

The preparation of a pharmaceutical composition of the invention comprises combining the disaccharide or a salt thereof with a pharmaceutically acceptable carrier to provide a unit dose, preferably of low dosage, having an effective amount of the substance. The pharmaceutical composition may also comprise a stabilizing agent, for example, protamine, in an amount sufficient to preserve a significant, if not substantial, proportion of the initial activity of the substance over an extended period, e.g., about 100 percent over about 3 days. At storage temperatures below room temperature, e.g., about -10 to about 10 °C, preferably 4 °C, more of the initial activity is preserved, for up to about 4 months.

Because the pharmaceutical compositions of the present invention are contemplated for administration into humans, the pharmaceutical compositions are preferably sterile.

Sterilization is accomplished by any means well known to those having ordinary skill in the art, including use of sterile ingredients, heat sterilization or passage of the composition through a sterile filter.

The compositions of the present invention are capable of inhibiting experimental delayed type hypersensitivity (DTH) reactions to an applied antigen as evidence by a reduction in the induration observed after the application of the antigen to the skin up to about five to seven days after the administration of the substance or pharmaceutical composition of same relative to the induration observed after the application of the antigen to the skin in the absence of or after recovery from the administration of the substance or pharmaceutical composition of same. Examples of the applied antigen include, but are not limited to, tetanus, myelin basic protein, purified protein derivative, oxazolone, and the like.

The invention further provides the use of the disaccharide of formula (II) and pharmaceutically acceptable salts thereof for the preparation of a pharmaceutical composition for the inhibition of the production of an active cytokine, in particular TNF-a.

In a preferred embodiment, the secretion of TNF-a is inhibited by the disaccharide of the invention.

A bioassay for quantifying the effect of a test substance on the secretion of active TNF-a comprises the steps of preincubating human CD4+ T cells in a medium with varying concentrations of a test substance, adding a constant amount of an activator effective to elicit the secretion of TNF-a by the T cells in the absence of said test substance, collecting the medium after a sufficient period of time, and subsequently testing the activity of the TNF-a in the medium. Preferably, the human CD4+ T cell: are obtained from peripheral blood mononuclear leukocytes (PBL) . Suitable immune effector cell activators include, but are not limited to, T cell-specific antigens, mitogens macrophage activators, residual extracellular matrix (RECM defined in Section 4, below), laminin, fibronectin, and th like.

Specific regulatory activity of particular substances can be determined by in vitro and in vivo bioassays described in greater detail, below. Briefly, the substances described in the parent Patent Application No. 107563 display an inhibitory activity relating to the induction of the secretion of active TNF-a which is dose dependent.

That is, a plot of the percent inhibition versus the dose (e.g., pg/ml of substance) gives rise to a bell-shaped curve from which a. maximum percent inhibition (Inhmax) is readily apparent.

Thus, for every point on such a plot, a "ratio" between the percent inhibition and the concentration or dose can be calculated. In the present case, a "specific regulatory activity" or "R" value can be obtained from the ratio of the maximum percent inhibition (i.e., Inhmax) and the concentration or dose of test substance which gave rise to such maximum percent regulatory value. Furthermore, an "R" value can be obtained for each bioassay. Hence, an "R" value can be associated from an in vitro mouse spleen cell assay, an ex vivo mouse spleen assay, an in vitro human PBL assay, and an in vivo assay based on experimental DTH reaction. If no effect is observed, an "R" value of zero is assigned .

Further objects of the present invention will become apparent to those skilled in the art upon further review of the following disclosure, including the detailed descriptions of specific embodiments of the invention.

All parts of the description that are not encompassed by the claims are not part of the present invention.

BRIEF DESCRIPTION OF THE FIGURES Fig. 1 illustrates the adjuvant arthritis (AA) scores obtained from groups of rats which were treated with weekly administrations of Fragmin at various doses relative to a control group that received only phosphate buffered saline (PBS) .

Fig. 2 illustrates the AA scores obtained from groups of rats which received a constant 20 microgram dose of Fragmin under various dosage regimens, including single treatment, daily treatment, five day intervals, and weekly.

Fig. 3 compares the effectiveness of weekly administration of Fragmin versus Heparin and control (PBS) .

Fig. 4 illustrates the results of daily administration of Fragmin, Heparin or PBS.

Fig. 5 illustrates the AA scores obtained from groups of mice that were treated either weekly or daily with various low molecular weight heparins including^ Fraxiparin, Fraxiparine, and Lovenox.

Fig. 6 plots the percentage of survival rate of rats that had undergone allogeneic heart transplants and had also received either weekly administration of Fragmin or PBS.

Fig. 7 presents bar graphs illustrating the bTood glucose levels of two groups of NOD mice, one group receiving Fragmin and the other receiving only PBS.

Fig. 8 illustrates the results of a DTH experiment involving a human volunteer.

Fig. 9 illustrates the "bell-shaped" dose to response curve exhibited by active Fragmin.

Fig. 10 illustrates the loss of inhibitory activity displayed by inactivated Fragmin.

Fig. 11 shows the absorption at 206 nanometers of various fractions obtained from the gel filtration of inactivated Fragmin, including fractions F2, F8, F10 and F15.

Figs. 12, 12A and 12B illustrate the effects of active Fragmin, fraction F15, and Fraction F10, - -5-8- - respectively, at various doses on the sensitivity of mice to the DTH reaction.

Fig. 14 illustrates the absorption at 206 nanometers versus fraction number for a number of fractions obtained from the Sepharose 4B column separation of Fragmin and heparanase-degraded ECM .

Figs. 13 and 15 compare the elution profiles of fractions obtained from the Sepharose 4B column separation of Fragmin and heparanase-degraded ECM.

Fig. 16 shows that an oligosaccharide product (fraction 5 from Fig. 13) demonstrated a similar bell-shaped dose/response curve in its ability to inhibit the secretion of active TNF-a.

Fig. 17 shows that the areas of greatest anti-TNF-a effect lie in the subfraction between about 5.65 and about 5.8.

Figs. 18A and 18B illustrate the chromatogram obtained from the HPLC separation of Fragmin and heparanase-degraded ECM, respectively.

Fig. 19 illustrates the absorption at 206 nanometers of two fractions, F5 and F8 , obtained from the Sepharose 4B column separation of heparanase- degraded ECM.

Figs. 2OA and 2OB, on the other hand, illustrate the absorption .at 206 and 232 nanometers, respectively, of peak obtained from the HPLC separation of. fraction F5.

Figs. 21A and 2 IB illustrate the uv absorption of additional HPLC fractions obtained from fraction F5.

Fig. 22 illustrates the uv absorption of fractions F7 and F8 obtained from the Sepharose 4B column separation of heparanase-degraded ECM.

Fig. 23 illustrates the substantially pure peak obtained from the SAX-HPLC chromatography of combined fractions F7 and F8.

Fig. 24 illustrates another peak labeled "A23/4" obtained from desalted preparations of the peak labeled "1" from Fig. 23.

Figs. 25A, 25B and 25C illustrate the chromatograms that are obtained from the SAX-HPLC column separation of dxsaccharide standards obtained from Sigma labeled H-0895, H-1020 and H-9267, respectively .

Fig. 26 illustrates the Sepharose 4B column separation of a mixture obtained from the heparanase (MM 5) treatment of Heparin, yielding fractions F7 and F8.

Fig. 27 illustrates the absorption at 206 nanometers of various fractions obtained from the Sepharose 4B chromatography of PC3 heparanase alone and Heparin + PC3.

Figs. 28A and 28B illustrate additional fractions obtained from the HPLC separation of fraction F7 from Figure 26.

Fig. 29 illustrates fraction F90 obtained from the HPLC separation of Fragmin.

Fig. 30 , on the other hand, illustrates the chromatogram that is obtained from- a SAX-HPLC separation of an aged sample of A23/4.

• Fig. 31 illustrates the proton NMR spectrum of a 20 microgram sample of an ECM-derived disaccharide obtained from HPLC chromatography, as shown in Fig. 23.

Fig. 32 illustrates a two-dimensional COSY spectrum of the sample o Fig. 31.

Fig. 33 illustrates an expanded portion of the NMR spectrum of Fig. 31, showing the signal for the anomeric proton.

Figs. 34 and 35 illustrate the FTIR spectra obtained from two separate samples, one indicating the presence of a sulfated compound (Fig. 34) and the other indicating the presence of a partially desulfated analog (Fig. 35) .

Fig. 36A illustrates the mass spectrum of a methylated derivative of the sample obtained from Fig. 23 in a solvent matrix comprised of DTT : thioglycerol (1:1) .

Fig. 36B illustrates the mass spectrum of the solvent matrix only.

Figs. 37A and 37B illustrate the mass spectrum of the same sample in a different solvent matrix comprised of methylnitrobenzyl alcohol, Fig. 37A being the mass spectrum of the sample plus the solvent matrix and Fig. 37B being the mass spectrum of the solvent matrix only.

Fig. 38 illustrates the results of experiments comparing the effectiveness of disaccharide 9392 and 1020 to improve the AA scor.es of female Lewis rats suffering from experimentally induced adjuvant arthritis.

Fig. 38A illustrates the effect of disaccharide 0895 on the AA scores of rats suffering from experimentally induced AA relative to control (PBS) .

Fig. 38B illustrates the effects of glucosamine treatment in the improvement of the AA score of Lewis rats under various dosages of glucosamine.

Fig. 38C, similarly, shows the effect of galactosamine at different dosages on the AA score of Lewis rats.

Figs. 38D and 38E illustrate the results of further experiments carried out with disaccharide 9392 in which the disaccharide is administered either weekly or daily beginning at day zero (i.e., start of induction of AA) or at day 12 (i.e., when the rat is already suffering from AA) .

Figs. 38F and 38G illustrate the results of a separate comparative set of experiments that were carried out on groups of Lewis rats to determine the effectiveness of disaccharide 9392 administered weekly compared with the effectiveness of a known antiinflammatory agent, dexamethasone phosphate, on the suppression of experimentally induced adjuvant arthritis.

Fig. 39 illustrates the effectiveness of subcutaneously injected disaccharide 1020 against liposaccharide (LPS) induced inflammation of rat corneas.

Fig. 39A presents the results of experiments relating to the radioprotective effects of glucosamine at various dosages relative to control (PBS) .

Fig.. 39B presents the results of similar irradiation experiments involving the administration of disaccharide 9392 at various dosages relative to control (PBS) .

Figs. 40 and 4OA illustrate the results of experiments that illustrate the ability of selected substances of the present invention to suppress allograft rejection. The results presented in Fig. 40 show that a 3 nanogram dose of disaccharide 9392 by subcutaneous injection one day before grafting and weekly thereafter, delayed the level of skin graft - -3-2- - rejection at 50% by 5 days. However, a 300 nanogram dose of the same disaccharide failed to produce a significant difference at 50% rejection relative to control (PBS) .

Fig. 41 illustrates the incidence of IDDM in groups of female NOD mice which had been separately treated with either disaccharide 9392, glucosamine or saline.

Fig. 41A presents the mortality rate of female NOD mice that, again, had been treated separately with the disaccharide 9392, glucosamine, or saline. It should be noted that in both Figs. 41 and 41A, the female NOD mice were approximately 3-1/2 months old, meaning that the mice as a group already endured a 20% incidence of IDDM.

Fig. 42 presents the respiratory distress (RD) score of six immunized rats challenged with aerosolized antigen with (Bl, B2 and B3) and without (Al, A2 and A3) treatment by substance H-9392. See text for details.

Fig. 43 represents the structure of 4-0- (2- deoxy-6-O-sulfo-2-sulfoamino-a-D-glucopyranosyl) - (2-0- sulfo-/3-D-glucopyranoside) uronic acid, prepared .by the synthetic scheme of Example 6.23. 4. DETAILED DESCRIPTION OF THE INVENTION In one aspect of the present invention, it was found that treatment with low molecular weight heparins (LMWHs) inhibited the ability of T cells and macrophages to secrete active TNF-a. In another aspect of the present invention, other substances, comprising carboxylated and/or sulfated oligosaccharides in substantially purified form, are described which collectively represent a means for regulating the biological activity of cytokines, such as TNF-α, in a host. For simplicity, the term " substance ( s) " or "active substance (s) " will be used to denote LMWHs, as used in the method of treatment disclosed herein, as well as the substances comprised of carboxylated and/or sulfated oligosaccharides that have been isolated herein in substantially pure form unless otherwise noted.

One functional expression of this effect can be seen in the inhibition in mice and humans of the delayed type hypersensitivity (DTH) reaction, a T cell dependent inflammatory reaction that may also be triggered by cells involving macrophages and other inflammatory cells. Treatment with the active substances at doses affecting active TNF-a production also was able to inhibit a model of autoimmune arthritis called adjuvant arthritis (AA) . Active substance treatment also prolonged the survival of allogeneic heart transplants in rats and abrogated insulin dependent diabetes mellitus (IDDM) in NOD mice. Moreover, similar treatment prevented the induction of active TNF-a production by T cells and macrophages in response to the stimulus of damaged or residual subendothelial extracellular matrix. This residual extracellular matrix (RECM) that is responsible for signaling the onset of TNF-α induction (and resulting inflammation) is to be distinguished from the enzyme degraded extracellular matrix (DECM) , selected components of which have been isolated herein and have been shown to either shut down TNF-a activity or amplify it.

Since TNF-a at the site of vascular injury probably has a role in the process of atherosclerosis, inhibition of TNF-a activity at the site of damaged subendothelial ECM will ameliorate the pathogenic process of atherosclerosis. A most surprising aspect - -3-4- - of treatment with the LMWH active substances is that such treatment is most effective when administered at low doses at weekly intervals. High doses of the LMWH active substances or doses of the LMWH active substances given daily are not effective in inhibiting TNF-a secretion or immune reactions.

Low molecular weight heparins, produced by fractionation or controlled depolymerization of heparins, show improved antithrombotic performance but also different pharmacokinetic properties as compared to heparin: the half-life is doubled and the bioavailability is higher with respect to their anticoagulant effect after subcutaneous injection (Bratt, G. et al., Thrombosis and Haemostasis (1985) 53:208; Bone, B. et al. , Thrombosis Research (1987) 46:845) .

According to the present invention it has now been found that the LMWH active substances administered at subanticoagulant doses at several day intervals are effective in the prevention and/or treatment of pathological processes involving induction of active TNF-a. Moreover, it has now been found that discrete substances, comprising an oligosaccharide of from 1-10 sugar units, preferably 2-4 sugar units, can be identified which can either inhibit or augment the activity of TNF-a. These discrete substances can be obtained, for example, from the tissue of a living organism, for instance, from the soluble degradation products of substrate extracellular matrix. 4.1. Sources of Active Substances The LMWHs to be used according to the invention are derived from LMWHs with an average molecular weight of 3000-6000, such as, for example - &- - the LMWHs disclosed in European Patent EP 0014184.

Some LMWHs are commercially available under different trade names, e.g., Fragmin®, Fraxiparin®, Fraxiparine® , Lovenox®/Clexane® .

LMWHs can be produced in several different ways: enrichment by fractionation by ethanol and/or molecular sieving, e.g., gel filtration or membrane filtration of the LMWH present in standard heparin and controlled chemical (by nitrous acid, /3-elimination or periodate oxidation) or enzymatic (by heparinases) depolymerization. The conditions for depolymerization can be carefully controlled to yield products of desired molecular weights. Nitrous acid depolymerization is commonly used. Also employed is depolymerization of the benzylic ester of heparin by β-elimination, which yields the same type of fragments as enzymatic depolymerization using heparinases. LMWH with low anticoagulant activity and retaining basic chemical structure can be prepared by depolymerization using periodate oxidation or by removing the antithrombin-binding fraction of LMWH, prepared by other methods, using immobilized antithrombin for adsorption.

Fragmin® is a low molecular weight heparin with average molecular weight within the range of 4000-6000 dalton, produced by controlled nitrous acid depolymerization of sodium heparin from porcine intestinal mucosa. It is manufactured by abi Pharmacia, Sweden, under the name Fragmin®, for use as an antithrombotic agent as saline solutions for injection in single dose syringes of 2500 IU/0.2 ml and 5000 IU/0.2 ml, corresponding to about 16 mg and 32 mg, respectively.

Fraxiparin®, and Fraxiparine® are LMWHs with average molecular weight of approximately 4500 dalton, - -3-6- - produced by fractionation or controlled nitrous acid depolymerization, respectively, of calcium heparin from porcine intestinal mucosa. It is manufactured by Sanofi (Choay Laboratories) for use as an antithrombotic agent in single doses comprising ca. 36 mg, corresponding to 3075 IU/0.3 ml of water.

Lovenox® (Enoxaparin/e) , a LMWH fragment produced by depolymerization of sodium heparin from porcine intestinal mucosa, using 3-elimination, is manufactured by Pharmuka SF, France and distributed by Rhone-Poulenc under the names Clexane® and Lovenox® for use as antithrombotic agent in single dose syringes comprising 20 mg/0.2 ml and 40 mg/0.4 ml of water.

As shown in the present application, the novel properties of LMWHs that have been discovered and are described herein are common to all LMWHs regardless of the manufacturing process, the structural differences (created by depolymerization or those dependent on variation in the heparin used as raw material) or the anticoagulant activity, provided that the LMWH employed is capable of inhibiting active TNF-a secretion in vitro by resting T cells and/or macrophages in response to activation by contact with T cell-specific antigens, mitogens, macrophage activators, residual ECM or its protein components, such as fibronectin, laminin, or the like.

Another test useful for identifying the LMWHs that are effective for the purpose of the present invention is the inhibition of experimental delayed type hypersensitivity (DTH) skin reactions, a T lymphocyte dependent reaction, to a variety of antigens (for example, tetanus antigen, myelin basic protein (MBP) , purified protein derivative (PPD) , and oxazolone) . The LMWHs also inhibit T cell adhesion to ECM and its protein components.

The LMWHs effective according to the invention are incorporated into pharmaceutical compositions, for example, as water solutions, possibly comprising sodium chloride, stabilizers and other suitable non-active ingredients. The preferable way of administration is by injection, subcutaneous or intravenous, but any other suitable mode of administration is encompassed by the invention, including oral administration.

According to the invention, the LMWH is to be administered at intervals of up to about five to eight days, preferably once a week. The other substances of the present invention, particularly the lower molecular weight (below 2000) oligosaccharides, may be administered in any convenient, effective manner (e.g., by injection, orally, or topically) at dosage regimens that may include daily or weekly administration. 4.2. Loss of Activity of LMWH Preparations Over Time. Influence of Added Stabilizer Time course studies conducted by the inventors demonstrate that LMWH samples, such as Fragmin, lose their ability to inhibit the activity of TNF-a within 72 h at ambient temperature and within a few months at low temperature (e.g., 4 *C) .

Table XI, Section 6.1, below, indicates that about 53% of activity of Fragmin® is lost after a day at ambient temperature. After about two days, about 87% of activity is lost, and after about three days, no activity is shown. As shown in Table XII, Section 6.2, below, experiments have shown that Fragmin® loses its anti-DTH reactivity even at colder temperatures (4 - -3-8- - *C); the process only requires more time.

Conventional non-fractionated heparins, in contrast, do not lose their classic anti-coagulant activities at 4 *C.

In an effort to discover an agent capable of stabilizing or preserving the cytokine inhibitory activity of the disclosed LMWH preparations, the inventors turned to a well known heparin additive.

Protamine sulfate is known to neutralize the anti-coagulant effects of heparinoid molecules and is used clinically for that purpose (See, Goodman and Gilman's The Pharmacological Basis of Therapeutics, Eighth Edition, Pergamon Press, New York, 1990, p. 1317) . It has been discovered, however, according to the invention that added protamine sulfate does not neutralize the inhibition of TNF-a-dependent activity by LMWH; in fact, protamine sulfate actually stabilizes this activity (See, Entries in Table XII, below, containing added protamine sulfate) .

In summary, one can conclude that (i) diluted LMWH solutions lose activity quickly at 20 'C and more slowly at 4 "C (it should be noted that the activity loss at 4 'C is not a feature of the standard anti-coagulant and anti-thrombotic activities of heparin or LMWH) ; (ii) added protamine sulfate, the classic neutralizer ..of the standard activities of heparins, does not interfere with the novel activity of LMWHs against TNF-a described in the present disclosure. Indeed, the inventors have demonstrated that protamine sulfate actually preserves this novel activity. - - 4.3. Fractionation of LMWH and Preparation of Degraded ECH or Degraded Heparin.

Discovery of Distinct Augmentative and Inhibitory Activities For and Against TNF- Activity As already discussed above, Low Molecular Weight Heparin (Fragmin) inhibits secretion of active TNF-a. Maximal inhibition or Inh^ (90%) , was observed at a concentration of 1 pg/ml. (See, Fig. 9) . By contrast, inactivated Fragmin had no effect on TNF-a production (See, Fig. 10) . However, fractionation of the inactivated material by low-pressure size-exclusion gel chromatographic separation using a Sepharose 4B solid, support (See, Fig. 11 for a plot of the absorbance at 206 nm versus fraction number) revealed active fractions of both inhibitory (F-15) and augmentative (F8, F2) effects (See, Fig. 11 and Table XIII) . The inhibitory fraction of the inactivated Fragmin (F-15) also inhibited the DTH reaction (See, Figs. 12; relating to active Fragmin®, 12A; relating to fraction F15, and 12B; relating to fraction F10) . Fraction F10 had no effect on TNF-a production or DTH reactivity.

A Sepharose 4B size-exclusion gel chromatographic separation was also carried out on the degradation products obtained from heparanase-treated ECM labeled with 3SS-containing sulfate groups.

Several types of heparanase enzyme were used in the present investigation. These enzymes include MM5 (Mammalian heparanase from human placentas, obtained commercially from Rad-Chenticals, Weizmann Industrial Park, Ness Ziona, Israel) , PC3 (Bacterial endoglycosidase, as described in Shoseiov, O. et al., Biochem. Biophys. Res. Commun. (1990) 169:667-672), and an enzyme from a bacterial source obtained from ' IBEX Technologies, Quebec, Canada. A plot of the radioactivity (CPM) versus fraction number is - --θ presented in Fig. 13. Another plot superimposing the elution profiles of fractionated Fragmin and fractionated ECM-heparanase is shown in Fig. 14. The conditions for the Sepharose 4B low-pressure separation are listed in Table I, below.

Table I Sepharose 4B Chromatography Conditions Column: Sepharose 4B (35 cm x 0.7 cm ID) Load: 1 - 1.5 ml Flow: 5 ml/hr Solvent: PBS (pH = 7.4) Fraction: 0.2 - 0.5 ml/tube Detector Absorption Setting: 206 nm, 280 nm The various fractions were assayed for their effect on TNF-a production and these results are presented in Table XV, below. Interestingly, fractions of similar elution properties from the two sources (i.e., F-39 and F-42 from Fragmin and Heparanase-degraded ECM) were found to have similar qualitative biological effects on TNF-a production and/or activity.

Figs. 15 and 13 illustrate one way of presenting the elution profile, obtained on. Sepharose 4B columns, of LMWH (Fragmin) and MS-sulfate labeled oligosaccharides of ECM, produced by purified MM5 heparanase, respectively. It can be seen in Fig. 13 that the heparan sulfate of the ECM (the substrate of heparanase) is degraded by the enzyme to produce heparan sulfate fragments with elution properties comparable to fractionated LMWH.

Fig. 16 shows that an oligosaccharide product (Sepharose 4B fraction #5, Fig. 13), obtained - -4-i- - from the ECM + heparanase "soup" (i.e., the mixture obtained from the heparanase degradation of ECM) , has a substantially similar dose/response characteristic as LMHV7 in its effects on the secretion of active TNF-a: that is, both display a bell-shaped dose/response curve and both exhibit maximal inhibition of about 90% at a concentration of about 1 pg/ml with less activity at either lower or higher concentrations. It is advantageous, thus, that the administration of these active substances includes dosages falling within an easily determined "window" of physiological effect.

Fig. 17 shows that the anti-TNF-a effect of the ECM-degradation products is highest in the area of a subfraction (between about 5.65 and about 5.80) of the fragments under peak number 5 of Fig. 13.

Thus, heparan sulfate can be acted upon by heparanase to generate degradation products that, like LM H, feed back on the T cells and macrophages to shut off active TNF-a production and, consequently, TNF-mediated inflammation.

It has also been discovered that low- molecular weight oligosaccharide fragments, obtained from endoglycosylase treatment of intact heparin, exhibit the desired regulatory effect over TNF-a activity. 4.4. HPLC Separation of LMWH Fractions and Fragments obtained from DECK and DH High performance liquid chromatography ("HPLC") techniques were utilized to obtain better resolution of the fractions from the LMWH (e.g., Fragmin) , ECM-degradation, and heparin-degradation samples. Initially, two types of HPLC conditions were used. Under the first set of HPLC conditions, a CYTOKINE REGULATORY SCHEME number of individual fractions were separated and isolated; their ability to regulate the secretion of active TNF-a was then examined. To the great surprise of the present inventors, it was discovered that selected fractions can augment the activity of TNF-a in the host while others inhibited TNF-a activity. A second set of HPLC conditions was then utilized to better separate the various components according to their molecular weight.

In the first set of HPLC conditions, a TS -GEL® G-Oligo-PW column (30 cm x 7.8 mm I.D.) equipped with a Guardcolumn Oligo (4 cm x 6 mm I.D.) was used. The conditions ("HPLC I") are provided in Table II, below. A representative chromatogram for the HPLC I separation of Fragmin and ECM + MM5 Heparanase is illustrated in Figs. 18A and 18B, respectively. - -4-3- - Table II. HPLC I Chromatography Conditions Column: TSK-GEL G-Oligo-P 30 cm x 7.8 mm ID Guard Column: Guardcolumn Oligo 4 cm x 6.0 mm ID Loop: 200 μΐ Flow: 0.5 ml/min.

Solvent: 0.2 M phosphate buffer (pH = 7.0) Fraction: 0.5 ml/tube Detector Absorption Setting: 190 nm - 400 nm The second set of HPLC conditions ("HPLC II") are described in Table III, below, and utilized conditions similar to those described by Rice, . G. et al. in Analytical Biochem. (1985) 150:325-331.

Hence, two columns connected in series were used: a Toyo Soda TSK-Gel G3000SW (7.5 mm x 50 cm) column connected to a G2000SW (7.5 mm x 50 cm) column. These columns, together with a 7.5 mm x 10 cm guard column attached to the inlet end of the G2000 column, were obtained from Phenomenex. Further experimental details are described in Sections 6.11, 6.14 and 6.15, below.

- ^Hr - Table III. HPLC II Chromatography Conditions Column: Toyo Soda TSK-GEL G3000SW (50 cm x 7.5 mm ID) and a G2000S (50 cm x 7.5 mm ID) in series Guard Column: Guardcolumn (10 cm x 7.5 mm ID) Loop : 20 or 100 μΐ Flow: 1 ml/min.

Solvent degassed 0.5 M NaCl Fraction: 0.5 ml/tube Detector Absorption Setting: 205 nm, 232 nm Under these conditions, smaller substances are retained longer than larger molecules.

In yet another set of HPLC conditions ("HPLC III") , the purity of selected desalted HPLC fractions was examined with the aid of a strong anion exchange (SAX) HPLC column. Such SAX HPLC columns are known to separate similarly sized molecules according to the number of negatively charged groups which are present in the molecules. The greater the number of negatively charged groups in -a substance, the longer it is retained in "the column. The.BPLC III conditions are outlined in Table IV, below.

Table IV. HPLC III Chromatography Conditions Column: SAX-HPLC column (25 cm x 4.6 mm ID, packed with Spherisorb, 5 im particle size) Loo : 1 ml Flow: 1.5 ml/min.

Solvent: linear gradient, below Fraction: 1 ml/tube Detector Absorption Setting: 205 nm, 232 nm Linear Gradient (See, Section 6.15, below) It will also be apparent to one of ordinary skill in the art, after considering the disclosure presented herein, that other HPLC conditions can be contemplated and applied to the separation and purification of the active substances of the present invention. In particular, reverse-phase conditions can also be utilized to good advantage. See, for example, Rice, .G. et al., supra.

Again, without wishing to be limited by theory, it is suspected that the activity of TNF-a is augmented by either increasing the intracellular production of active TNF-a, increasing the amount of active TNF-a secreted by the host's immune effector cells, or enhancing the activity of the cytokine through the action of an agonist.

It also follows that the biological activity of TNF-a may be inhibited by converse processes, including not only competition offered by the active inhibitory substance for the receptors of TNF-a (e.g., the inhibitory substance acting as or inducing the - production of another substance that acts as an antagonist of TNF-a) but also the formation of a complex of TNF-a and the inhibitory substance which is less active than free TNF-a. Alternatively, it follows that a "souped-up" complex between TNF-a and the augmentative substance may be responsible for the observed increase in the activity of TNF-a. 4.5. Determination of Activity The active substances of the present invention, both those able to inhibit TNF-a activity and those able to augment TNF-a activity, have been isolated and purified from mixtures containing them. In some cases, these active substances have been purified to substantial homogeneity by the powerful HPLC techniques described herein.

As a further indication of the purity of these active substances, the specific regulatory activities of the various substances were determined.

Initially, however, a carbazole assay, performed in a manner similar to that disclosed by Carney, S. L. in Proteoglycan Analysis. A Practical Approach f Chaplin, M. F. and Kennedy, J. F. (Eds.) IRL Press, Oxford, Washington, D.C. (1986) p. 129, was utilized to determine the amount of oligosaccharide material present (e.g., amount of sugar present) in a given test sample. Picogram (pg) quantities of sugar can be quantified in this manner. The assay is performed as described in Section 5, below.

Next, the apparent activity associated with that quantity of substance is determined by one of the biological assays that are described in great detail in Section 5, below, to provide a dose/response profile. These bioassays may either be carried out in vitro or under in vivo conditions.

It has, thus, been found that the observed inhibition or augmentation of TNF-a activity, expressed as a percentage of the activity of TNF-a observed in the absence of the substances of the present invention, depends on the concentration or dose of such substance present in the test sample.

The apparent activity profile that results is approximately bell-shaped as illustrated in Figs. 9 and 16. The maximum value of percent inhibition or augmentation observed for each substance is designated Inh^ or Aug^, as the case may be.

As described further, below, the bioassay used to establish the "ideal" unit dose (i.e., the one that corresponds to Inh^ or Aug^J can be based on the in vitro or in vivo inhibition or augmentation of the activity of TNF-a or DTH assay in mice.

Alternatively, an in vitro assay based in human cells (described further, below) may also be used. The specific regulatory activity or "R" value is, as defined herein, the ratio of the Inh^ or Aug^ and the "ideal" dose that gave rise to that maximum percent inhibition or augmentation. For the in vitro assays, the "R" values are typically expressed in units of % x (pg/ml)-1.

As stated above, the specific regulatory activity can also be established under in vivo conditions by monitoring the inhibition of experimental DTH reaction in mice or humans. It was found that the ability of a particular dose of an inhibitory composition to inhibit secretion of active TNF-a is positively correlated with its ability to inhibit the delayed type hypersensitivity (DTH) reaction, although the same composition may be more potent under one assay versus another (i.e., between in vitro and in vivo bioassays) . Inhibitory or 4-8- - augmenting activity in this in vivo cell-mediated inflammatory reaction is of great importance because the DTH reaction is an expression of the processes involved in autoimmune diseases, graft rejection, some types of blood vessel inflammation and allergy. Thus, activity in this test is indicative of utility in these types of diseases and possibly others, as described further below.

Moreover, the new quantity, the specific regulatory activity, which is defined as the ratio between the In ^ or Aug^ value and the amount or concentration of substance (the "ideal" dose) which gave rise to that maximum percent value, can serve to distinguish the novel active substances of the present invention from those substances that may have been known, but unrecognized in the art as possessing the cytokine regulatory activity disclosed herein. This specific ratio is referred to herein as the "R" value, for short. Hence, the novel substances or compositions of the present invention can be described in terms of a minimum "R" value, which can be calculated from the apparent activity versus dose profile, and which "R" value will exceed the "R" -value that can be associated, by reference to the teachings of the present disclosure, with known compositions. 4.6. Types of Disorders That May Benefit From the Present Invention The disorders that can be prevented or treated according to the invention are all disorders linked to pathological processes involving induction of active TNF-oc secretion, including atherosclerosis and vasculitis and pathological processes related thereto; autoimmune diseases, e.g., rheumatoid arthritis, diabetes type I (insulin-dependent diabetes mellitus or IDDM) , multiple sclerosis, lupus - - θ- - erythematosus, Graves disease; allergy; graft rejection; acute and chronic inflammatory diseases, e.g. uveitis, bowel inflammation; anorexia nervosa; hemorrhagic shock caused by septicemia, and HIV infection in AIDS. In AIDS, the active substances will suppress replication of HIV thereby preventing the development of AIDS-related complex (ARC) . Other disorders that may benefit from a treatment designed to regulate cytokine activity include, but are not limited to, psoriasis, pemphigus, asthma, renal diseases, liver diseases, bone marrow failure, vitiligo, alopecia, and myositis.

Further, augmentation of active TNF-a is useful in the treatment of tumors, bacterial infections and viral infections. Parenteral, oral or topical administration of the substances of the present invention which augment the production of active TNF-a in a pharmaceutically acceptable carrier may also help combat skin cancer, such as basal cell cancer, squamous cell cancer, or melanoma.

In the clinical application of the active substances of the present invention, it should be kept in mind that the successful treatment of certain -types of disease consists, in large part, in the restoration of homeostasis. To the endocrinologist, this implies the judicious administration or antagonism of specific hormones. For example, an insulin-dependent diabetic may be effectively treated by insulin replacement therapy; a patient with Graves' disease may be helped by pharmacological measures that inhibit thyroxine release. Only rarely can disease be alleviated by administration of hormones that were never deficient to begin with.

The use of cytokines, such as TNF-a, as antineoplastic agents provides one such instance. The rationale for administration of immunomodulatory agents to cancer patients may be quite slender. Many cytokines, like TNF- , exhibit toxicities that prove dose-limiting long before a therapeutic goal is achieved. In such an event, the augmentation of the activity of endogenously produced TNF-a may provide an approach that is both novel and, eventually, prove more effective than any previously contemplated therapeutic regimen.

Clearly, our understanding of the role for TNF-a is still evolving and, doubtless, new and useful uses of the hormone and the substances able to regulate its activity will be uncovered. While it goes without saying that all uses of the claimed compositions and pharmaceutical preparations are within the scope of the present invention, those uses that either alleviate the symptoms of disease, prevent the onset of disease, or provide a cure for the disease are especially contemplated. 4.7. Topical Applications of the Oligosaccharide Substances of the Present Invention The substances of the present invention also find use in topically administered compositions, such as those preparations for he treatment of edema or inflammation. Indeed, above and beyond a purely therapeutic application, the substances of the present invention may also find utility in supplementing the protective action of cosmetic compositions, such as sunscreen or suntan lotions. Few, if any, sunscreen preparations are fully effective in blocking out all the harmful wavelengths (e.g., 290-320 nm) present in the ultraviolet region of the electromagnetic spectrum. Hence, overexposure to the sun often gives rise to an acute condition known as solar erythema and - -Si- - prolonged, repeated exposure can, of course, lead to leathery looking skin or, worse, skin cancer.

Thus, the incorporation of the active substances of the present invention in cosmetic preparations is specifically contemplated both for the purpose of preserving and protecting the skin, as well as alleviating a medical condition, such as solar erythema. In sunscreen or suntan preparations, it would be advantageous to include an effective amount of the oligosaccharides of the present invention along with conventional sunscreen agents. Generally, an amount of active substance would be present to provide a dose of about 1 ^g to about 100 mg per kilogram of subject, preferably from about 0.01 mg to about 10 mg per kilogram of subject, and most preferably about 0.1 mg to about 1 mg per kilogram of subject.

The cosmetic compositions, may contain conventional ingredients known to those of ordinary skill in the art, such as those described in irk- Othmer, Encyclopedia of Chemical Technology. Third Edition (1979), Vol. 7, pp. 143-176. In sunscreen preparations, the addition of the active substances of the present invention increases the minimum erythemal dose (MED) and, consequently, the sun protection factor (SPF) . Specific ingredients, including typical sunscreens, are ..listed in Kirk-Othmer , supra, at pp. 153-154. -In addition, topical preparations and cosmetic formulations may be prepared as described in U. S. Patent Nos. 4,199,576, 4,136,165, and 4,248,861, the complete disclosures of which are incorporated by reference herein. It would, of course, be apparent to those of ordinary skill in the art of cosmetology that the resulting compositions can be in many forms, including, but not limited to, solutions, lotions, creraes, pastes, emulsions, sprays, or aerosols. 4.8. Exemplary Dosage Regimens It was thus established according to the invention that the lowest dose of LMWH per kg causing inhibition of TNF-a, production or inhibition of DTH reactivity by at least 50% is considered to constitute 12 mouse inhibitory units per kg (12 u/kg) - Because of the differences in surface area and metabolism between mice and humans, humans should be treated with a lower dose of LMWH, and 12 u/kg in mice is established to correspond to 1 u/kg in humans. For example, the dose of Fragmin® batch 38609 effective in inhibiting both TNF-a secretion and DTH reactivity is 5 jug per mouse administered weekly. Since each mouse weighs about 25 g, the dose of Fragmin® 38609 equivalent to 12 u/kg is 200 Mg/kg of mouse. The dose of 1 u/kg suitable for humans is therefore 200 g/kg ÷ 12 = 16.67 ^g/kg. A human weighing about 70 kg would then be treated by a dose of about 1.2 mg given in a single dose subcutaneously once every 7 days. Since individual humans vary biologically, the optimal dose may be different from about 1.2 mg and will lie generally below 5 mg, particularly within the range of 0.3 to 3 mg .

Hence a rough guide for conversion of the mice dosage regimen to human dosage is the following: Dose Human/kg = Dose Mouse/kg ÷ 10 or 12 The dose of LMWH that should be effective in rats can be derived from the fact that the dose of LMWH per kg of rats is one-half the dose per kg of mice, i.e. 6 u/kg. For example, if 12 u of Fragmin® batch 38609 is 200 /jg/kg, then the 6 u dose suitable for rats should be 100 Mg/kg or 20 μg per 200 g rat, administered once a week. - -5-3- - For most of the oligosaccharide substances of the present invention, which have been isolated from LMWH, degraded heparin and degraded ECM, the following is a way to predict the effective dose of these oligosaccharide substances for treatment of humans from the in vivo DTH bioassay.

Figure 12A shows that an isolated fraction (F15) in vivo inhibits the DTH in mice at a range of 0.1-5.0 μg/mouse/week. Since our mice weigh 25 gm, the in vivo dose is approximately (0.1 ÷ 0.025 kg) 4-200 g/kg mouse/week (the equivalent of 0.01-10 pg/ml in vitro) .

To correct for the surface area difference between mice and humans, we have to divide the mouse dose/kg by 12 : 4-200 μg/kg mouse → 0.33-16.67 g/kg human.

Thus, a 70 kg human should receive up to about 1.2 mg (about 1,200 g) . To be certain that we could cover the difference between people, we might increase this dose to about 5 mg, an amount that is well below any doses of heparinoids used for their effects on coagulation or thrombosis. Hence, the dose for a 70 kg human, will be about 5 mg or less, preferably -about 3 mg or less, more preferably about 1.5 mg or less, and most preferably about 1 mg or less.

In fact for the highly purified materials of the present invention, including those that have been obtained from HPLC chromatography, the preferred dosages may be even less. For example, the disaccharides, described in greater detail below, have been found to exhibit inhibitory activity, when administered by injection, at about 0.1 g to about 0.5 μg per kilogram mouse. Hence, the dosage for humans are estimated to be about 0.01 μg to about 0.05 μ% per kilogram man or about 0.7 μg to about 3.5 g Cells from mice treated with LMWH are obtained as follows: female mice of the BALB/c strain (25 grams, 2 months old) , at least 5 mice per group, are injected subcutaneously with various doses of LMWH, usually in the range of 0.5 to 20 μg per mouse. Five days later the mice are killed by cervical dislocation, the spleens are removed and suspensions of spleen cells, depleted of red blood cells, are assayed for the production of TNF-a in response to induction by residual extracellular matrix (RECM) , Concanavalin A (Con A) or lipopolysaccharide (LPS) . 5.2. In Vivo Bioassay of Inhibition of Experimental DTH Reactivity Groups of inbred BALB/c (Jackson Laboratories, Bar Harbor, ME) or of outbred CD1 ( eizmann Institute Animal Breeding Center, Rehovot, Israel) mice are sensitized on the shaved abdominal skin with 100 μΐ of 2% oxazolone (OX) in acetone/olive oil (4/1, v/v) applied topically. DTH sensitivity is elicited 5 days later as follows: mice are challenged with 20 μΐ of 0.5% OX (10. μΐ administered topically to each side of the ear) in acetone/olive oil. A constant area of the ear is measured immediately before challenge and 24 and 48 h later with a Mitutoyo engineer's micrometer. The individual measuring ear swelling is blinded to the identity of the groups of mice. The increment (Δ) of ear swelling is expressed as the mean in units of 10"2 mm or 10"4 inch (+SE) depending on the micrometer that is used. Percent inhibition is calculated as follows: % Inhibition = X- f , Treated - negative control v control - negative control j Mice are treated with LMWH as in Example 5.1, injected the day before primary sensitization to OX. On the fifth day after sensitization to OX, the mice are challenged to induce a DTH reaction, as described above.

The positive control is the DTH reaction elicited in immunized mice in the absence of treatment with LMWH. The negative control is the background swelling produced by the antigen in naive (non-immunized) mice. 5.3. Induction of TNF-a Secretion by T Cells and Macrophages In Vitro Microtiter plates were prepared as follows: fibronectin (FN) or laminin (LN) (Sigma) were added to flat bottom 96-well plates (Costar) at a concentration of 1 Mg/50 μΐ PBS per well and removed after 16 h.

Remaining binding sites were blocked with BSA/PBS (10 mg/ml) which was added to the wells for 2 h and washed out.

ECM-coated wells were prepared as follows: bovine corneal endothelial cells were cultured in flat bottom 96-well plates. The confluent layers of endothelial cells were dissolved and the ECM was left intact free of cellular debris (Gospodarowicz , D. et al., J. Biol. Chem. (1978) 253:3736). Disrupted or residual ECM (RECM) was prepared by gently scratching the ECM three times with a 27G syringe needle and the exposed sites were subsequently coated with BSA/PBS. Resting cloned rat CD4+ T cells, designated l, which recognize myelin basic protein (MBP) , were propagated - -Sf - and maintained in culture and were added to the wells, 105 cells per well with or without 3 x 10s syngeneic splenic macrophages, in 100 μΐ per well RPMI 1640 (Gibco) supplemented with 1% BSA and antibiotics.

The splenic macrophages were purified by removing the T and B cells using specific monoclonal antibodies (mAb) . Anti-murine TNF-a mAb was obtained from Genzyme (Cambridge, MA) , and was diluted 300-fold. A 10 μΐ aliquot of this diluted solution was added to each well. MBP (100 μg/ml) , Con A (2.5 /xg/ml) , LPS (1 /xg/ml) , FN (5 xg/ml), and LN (5 /xg/ml) were added to the wells where indicated.

The plates were incubated at 37 * C in a humidified incubator for 3 h. Subsequently, the contents of the wells (4 wells per experimental group) were collected, centrifuged, and the media were assayed for active TNF-a secretion as in the example described in Section 5.1: That is, supernatants of cultured macrophages and lymphocytes were added to cultures of HeLa cells, which are sensitive to killing by TNF-a, and death of these cells in the presence of the test media was calibrated in comparison to titration curves of exogenous added TNF-a. Cell -death is examined by the release of neutral red dye from the preincubated HeLa cells. The results shown here represent data obtained from a total of six experiments that produced essentially similar results.

Table V shows that T cells and macrophages cultured together can be induced to secrete TNF-a by contact with specific antigen MBP (group 4) , the mitogen Con A (group 6) or LPS (group 8) . However, in the absence of antigenic or mitogenic stimulus, the secretion of TNF-a was also induced by residual extracellular matrix ( ECM; group 10) or by the ECM components, fibronectin (FN; group 12) or laminin (LN group 14) . Intact EC was a weak inducer of TNF-a (group 16) .

Table V. TNF-a secretion by T cells and macrophages is induced by specific antigen MBP , Con A, LPS, RECM, or ECM components .

Kl cells cultured together with Secreted TNF-a (yes) or without TNF-a Group inducer (no) macrophages (pg/ml) 1 none no 50 2 yes 65 3 MBP antigen no 30 4 yes 950 5 Con A no 120 6 yes 1300 7 LPS no 50 8 yes 1500 9 RECM no 30 10 yes 900 11 FN no .20 12 yes 650 13 LN no 50 14 yes 500 15 ECM no 30 16 yes 120 5.4. Regulation of TNP-a Secretion by LMWHs T cell and accessory cell cultures were prepared as described in Section 5.3. LM H was added to the wells at the beginning of the cell culture. 59 ' - The levels of TNF-a were examined after 3 h of incubation.

Table VI shows that the presence of LMWH (Fragitiin® batch 38609) in vitro inhibited active TNF-a secretion induced by specific antigen (MBP; group 4) , mitogens (Con A and LPS; groups 6 and 8) , RECM or ECM components (groups 10, 12 and 14) . Since TNF-a secretion induced by RECM is likely to be involved in atherosclerosis, inhibition of TNF-a by LMWH will be beneficial in atherosclerosis. - -66-- Table VI. Induction of TNF-α secretion induced in vitro is inhibited by LMWH (Fragmin® batch 38609) .

Secretion of TNF-a by cultures of T cells and TNF-a LMWH macrophages Group Inducer (1 Mg/ml) (pg/ml) 1 none none 65 2 yes 30 3 MPB antigen none 950 4. yes 60 5 Con A none 1300 6 yes 80 7 LPS none 1500 8 yes 80 9 RECM none 900 10 yes 90 11 FN none 650 12 yes 90 13 LN none 500 14 none 70 5.5. Ex Vivo Experiments with LMWH-Treated BALB/c Mice To examine the effect of LMWH administered to mice in vivo on the secretion of TNF-a by spleen cells in vitro, the following experiment was conducted. BALB/c mice, 5 per group, were treated with various doses of LMWH (Fragmin® batch 38609) diluted in saline, injected subcutaneously. After one week, the animals were killed and their spleen cells, devoid of red blood cells, were examined for their ability to secrete TNF-a in response to control wells without RECM (A) or to wells coated with RECM (B) .

Measuring the levels of TNF-a secretion was done as described in Section 5.1. Table VII shows the results which indicate that an injection of 5 μ-q of LMWH given once, 7 days earlier, inhibited TNF-a secretion induced by RECM. Higher or lower doses of LMWH were less effective. Thus, an optimal dose of LMWH administered in vivo a week earlier was effective.

Table VII. Ex vivo inhibition of T cell mediated TNF-a secretion in response to residual ECM.

In vitro TNF-a secretion (pg/ml) by spleen cells LMWH cultured on: treatment of BALB/c mice A. None B. Residual ECM (weekly) (% Inhibition) 1 None 30 400 - 2 0.5 g 50 380 (5) 3 1 /ig 25 90 (78) 4 5 Mg 25 60 (85) 5 10 g 30 140 (65) 6 20 40 320 (20) · Table VIII shows that a 5 jug dose in vivo of the LMWH Fragmin® batch 38609 was also effective in inhibiting TNF-a secretion induced by LPS. BALB/c (4 mice per experimental group) mice were treated with the indicated amounts of LMWH diluted. in saline and injected subcutaneously . After one week, the mice were injected intraperitoneally with 10 mg LPS, killed 4 hours later and their spleen cells, devoid of red blood cells, were subsequently cultured in RECM coated wells for 3 hours in a humidified incubator. The levels of TNF-a secreted in response to the RECM was 5 measured in the supernatants of the cultures. The results are given in Table VIII.

Table VIII. Treatment of mice with LMWH inhibits LPS mediated secretion of active TNF- by macrophages.

In vitro TNF-a LMWH secretion by treatment macrophages (pg/ml) of mice ^g) in response to LPS % Inhibition 0 690 — 0.1 500 28 1 350 50 5 120 82 20 550 20 5.6. Experiments Using a Variety of LMWH Sources To examine the effect of different LMWHs on the inhibition of secretion of active TNF-a and on DTH responses, mice were treated with the indicated LMWH administered subcutaneously in different concentrations. After one week, some of the mice were killed and the induction of secretion of active TNF-a in response to Con A activation in vitro was measured (Table IX) . The remaining mice were examined for their ability to respond to the antigen oxazolone (Table X) . The results are expressed in the Tables as percent inhibition compared to the responses of the LMWH untreated mice.

Two conclusions can be made by inspecting the results shown in Table IX and Table X: 1. Different batches of LMWH, each calibrated for by similar antithrombotic effect (Factor X assay) have different optimal doses for Βλ 6~3 inhibition of secretion of active TNF-a. Moreover, there are preparations of LMWH, such as Clexane® batch 4096, which have no inhibitory effect on secretion of active TNF- , at any of the doses tried. Therefore, it may be concluded that the antithrombotic effect of a LMWH preparation is not related to the potential of the LMWH preparation for inhibition of secretion of active TNF-a. The two different bioassays are mediated by different factors present in the preparations . 2. The ability of a particular dose of LMWH to inhibit secretion of active TNF-a is positively correlated with its ability to inhibit DTH reaction, and the dose of a LMWH preparation optimally effective in inhibiting secretion of active TNF-a is also optimally effective in inhibiting the DTH reaction.

Table IX. Weekly Treatment of Mice with Different LMWHs Inhibits DTH Sensitivity of Mice.

DTH Inhibition Batch of Dose Response of DTH "R" value LMWH (pg/gm mouse) (lO^mm) (%) % x (pg/gm)"' Fragmin Batch 38609 None 25 (+) Control 2 (-) Control 0.02 21 12 0.04 23 10 0.2 6 73 (max) 365 0.4 6 20 2 0 0 Batch 45389 None 28 (+) Control 2 (-) Control 0.004 26 6 0.04 4 89 (max) 2225 0.2 24 13 0.4 26 6 2 29 0 Clexane Batch 2088 None 22 {+) Control 2 (-) Control 0.004 17 23 0.04 3 87 (max) 2175 0.2 13 41 0.4 23 0 Batch 2066 None 23 (+) Control 2 ( - ) Control 0.004 20 13 0.04 8 65 0.2 7 70 (max) 350 0.4 7 70 Batch 4096 None 24 (+) Control 2 (-) Control 0.04 27 No effect 0 0.2 26 No effect 0 0.4 24 No effect 0 5 - e — Table X. Weekly Treatment of Mice with Different LMWHs Inhibits Ex Vivo Secretion of Active TNF Using Mouse Spleen Cell Bioassay.

Con A-Induced Batch of Dose TNF secretion Inhibition "R" va LMWH (pg/gm mouse) (pg/ml) (%) % x {μg/gm) Fragmin Batch 38609 None 4S0 Control 0.02 425 5 0.04 400 12 0.2 68 85 (max) 425 0.4 350 22 2 435 8 Batch 4S389 None 320 Control 0.004 280 13 0.04 70 78 (max) 1950 0.2 260 18 0.4 290 10 2 310 4 Ciexane Batch 2088 None 400 Control 0.004 360 10 0.04 64 84 (max) 2100 0.2 152 38 0.4 380 4 Batch 2066 None 350 Control 0.004 338 6 0.04 185 54 0.2 192 57 (max) 285 0.4 186 55 Batch 4096 None 320 Control 0.04 335 No effect 0 0.2 325 No effect 0 330 No effect 0 5.7. Treatment of Adjuvant Arthritis (AA) in Rats with Low Doses of LMWHs Adjuvant arthritis is an experimental disease inducible in some strains of rats by immunizing them to antigens of Mycobacterium tuberculosis (Pearson, c. M. , Proc. Soc. Exp. Biol. Med. (1956) 91:91). This experimental disease is considered to be a model of human rheumatoid arthritis (Pearson, C. M. , Arthritis Rheum. (1964) 7:80). The arthritis appears to be caused by T lymphocytes that recognize an antigen of M. tuberculosis that is cross-reactive with structures in the joint tissues (Cohen, I. R. , et al., Arthritis Rheum. (1985) 28:841).

Lewis rats were immunized with M. tuberculosis (1 mg) in oil to induce adjuvant arthritis (Pearson, C. M. , Proc. Soc. Exp. Biol. Med. (1956) 91:91). Five days later the rats were inoculated subcutaneously as indicated with the doses of LMWH and/or heparin and scored for the development of arthritis on a scale of 0-16 as described (Holoshitz, J., et al., Science (1983) 219:56). All the experiments were performed with Fragmin®, Batch 38609.

In order to study the dose response to Fragmin® (Fig. 1) rats immunized to induce AA were injected subcutaneously weekly, starting on the 5th day after injection with 0.5 μg (o) , 1 /zg (♦) , 2 g (·), 10 μg (O) , 15 μg (Δ) 20 . (!) ; 30 /ig (A), 40 jig (x) and PBS control (□) . The 20 μg dose was maximally effective in inhibiting arthritis.

The effect of the 20 μg dose of Fragmin® on the course of AA is shown in Fig. 2: PBS control (□) ; single treatment on 5th day (A); daily (·) ; every 5th day (o) ; weekly (H) . It is shown that Fragmin administration both at 5 day intervals and at 7 day intervals inhibits arthritis.

Fig. 3 shows the effect of weekly administration of Fragmin® (batch 38609) as compared to standard heparin on AA. Lewis rats were immunized to induce AA. Beginning on <5Q - -69-- day 5, the rats were inoculated subcutaneously at weekly intervals with a 20 g dose of Fragmin® (·), heparin (o) or phosphate buffered saline (PBS) control (□) . The results show a dramatic difference in potency between Fragmin® and heparin: Fragmin® completely inhibited arthritis, while heparin had no inhibitory effect.

No inhibitory effect on AA was found with daily administration of a 20 μq dose of LMWH, although surprisingly the inhibitory effect of heparin was stronger than that of Fragmin® in daily administration, as shown in (Fig. 4: Fragmin® (batch 38609) (·) , heparin (o) , PBS control (□) ) - A similar inhibitory effect was observed with several other LMWHs administered to Lewis rats immunized to induce AA. Fig. 5 shows the results of the injection of a 20 μ dose of Fraxiparin® (daily (□) ; weekly (H) ) ; Fraxiparine® (daily (Δ) ; weekly (A)), Lovenox®/Clexane® (daily (·) ; weekly (o) ) , and PBS control (x) . All the three LMWHs of different types and sources showed a marked inhibition of arthritis, when administered weekly, but not daily. 5.8. Treatment With LMWH Prevents Rejection of Allografts Wistar rats were subjected to allogeneic BN heart transplant (Ono, K. and Linsay, E. S., J. Thorac.

Cardiovasc. Surg. (1969) 45:225-229). From the day before transplantation, the rats were injected subcutaneously at 7 day intervals with 20 of Fragmin® or PBS control (Fig. 6, • and o, respectively) and scored for survival. The day of rejection was determined as the day the transplanted heart stopped beating, assayed by palpation of the abdomen. Fig. 6 shows that the rats treated with the weekly dose of LMWH had a markedly increased survival of the heart allografts. 5- - -6-6- - 5.9. Biological Effect of LMWH on Insulin Dependent Diabetes Mellitus (IDDM) of NOD Mice Mice of the NOD strain spontaneously develop a form of type I insulin dependent diabetes mellitus (IDDM) that is the accepted model for human IDDM (Castano, L. and Eisenbarth, G. S., Annu . Rev . Immuno1. (1990) 8:647-679). The disease begins at 4-5 weeks of age by the appearance of inflammation of the pancreatic islets, insulitis. The insulitis progressively damages the insulin-producing beta cells which are sensitive to damage by TNF-a. At about 4-5 months of age, a sufficient number of beta cells are destroyed so that diabetes becomes overt.

To test whether treatment with LM H could affect the IDDM process, a group of 10 female NOD mice was treated with weekly subcutaneous injections of 5 /zg per mouse of Fragmin® (batch 38609) , the dose determined to represent 12 mouse units per kg. A group of 10 control mice were treated with injections of saline. At the age of 5 months all the mice were bled to determine the development of IDDM using a standard procedure (Elias, D. et al., Proc. Natl. Acad. Sci. U. S.A. (1990) 87:1576-1580). Fig. 7 shows that the control mice ("none") had abnormal blood glucose (400 mg/dl) . In contrast the mice treated with LMWH had a normal blood glucose (100 mg/dl) . Thus treatment with LMWH can indeed cure the IDDM process. 5.10. LMWH Treatment of Allergy In many allergic patients, intradermal challenge with specific antigen or anti-IgE induces an immediate wheal and flare reaction which is followed, 4-8 h later, by a period of persistent swelling and leukocyte infiltration termed the late phase cutaneous reaction. Late phase reactions (LPR) 2 were initially described in the skin (Sollev. G. O. et al.. J. Clin. Invest. (1976) 58:408-420). - -69- - However, it is now clear that late consequences of IgE-dependent reactions, notably including infiltration of the reaction sites with blood-borne leukocytes, also occur in the respiratory tract and other anatomical locations (Lemanski, R. F. and aliner, M. , in Allergy: Principles and Practice, Vol. 1 (1988), Middeton, Jr., E. et al. (Eds.), pp. 224-246). Indeed, it has been argued cogently that many of the clinically significant consequences of IgE-dependent reactions, in both the skin and the respiratory system, reflect the actions of the leukocytes recruited to these sites during the LPR rather than the direct effects of the mediators released at early intervals after antigen provocation (Kay, A. B. J. Allergy Clin. Immunol. (1991) 87 : 893-910) .

It has recently been widely recognized that chronic allergic diseases such as asthma and atopic dermatitis are a result of an underlying inflammatory process which includes the infiltration and activation mainly of eosinophils and T cells (Kay, A. B. J. Allergy Clin. Immunol. (1991) 87:893-910).

Several lines of evidence support the hypothesis that the leukocyte infiltration associated with LPRs occurs as a result of mast cell degranulation. In both -man and experimental animals, agents that induce cutaneous mast cell degranulation by either IgE-dependent of certain other mechanisms can also promote infiltration of the reaction sites with leukocytes (Solley, G. 0. et al., J. Clin.

Invest . (1976) 58:408-420; Lemanski, R. F. and Kaliner, M. , in Allergy: Principles and Practice, Vol. 1 (1988) , Middeton, Jr., E. et al. (Eds.), pp. 224-246; Kay, A. B. J^ Allergy Clin. Immunol. (1991) 87:893-910). A review of the mediators that can be elaborated by activated mast cells reveals many that might contribute to leukocyte infiltration in LPRs, including lipid mediators such as LTB4, LTC4, LTD4, PGD2, and PAF (platelet activating factor) , as well as - - θ- - several peptide or proteinaceous chemotactic factors (Holgate, S. T. et al. , in Allergy: Principles and Practice, Vol. 1 (1988), Middleton, Jr. E. et al. (Eds.), pp. 135-S 178) . The latter agents range in size from tetrapeptide "eosinophil chemotactic factors of anaphylaxis" to very high molecular weight "neutrophil chemotactic factors".

Even more candidate mast cell associated mediators of leukocyte infiltration recently have been identified, 0 including cytokines similar or identical to JNF-α, IL-la, and four members of the MIP-1 gene family of small secreted peptides (Gordon, J. R. et al, Immunol. Today (1990) 11:458- 464) . Four of these cytokines (TNF-a, IL-la, MIP-la and MIP-l/S) have been demonstrated to have the ability to 5 promote leukocyte infiltration.

More recently, ( ershil, B. K. et al., in J. Clin. Invest . (1991) 87:446-453, by using mast cell deficient mice have demonstrated that the recruitment of leukocytes during IgE dependent LPR is mast cell dependent and that this 0 inhibition was partially blocked by local administration of anti TNF-a antiserum. It is widely accepted today that the inhibition of the cellular infiltration/activation associated with IgE dependent LPR is a crucial therapeutic approach in alleviating various allergic diseases (Barnes, 5 p- J- N. Eng. J. Med. (1989) 321 : 1517-1527) .

To the surprise of the present applicants, it was found that LMWH significantly inhibited the leukocyte infiltration during IgE dependent cutaneous LPR in mice undergoing passive cutaneous anaphylaxis (PCA) .

Q Mice received an i.d., injection (into the ears) of monoclonal IgE anti DNP Ab (-20 ng) . A day later, the mice were i.v. injected with DNP3(W0-HSA in saline. Ear swelling was determined by measurement of ear thickness with a micrometer before and at various intervals after the 5 challenge with the DNP-HSA. In all experiments, tissues from sites of PCA reactions were obtained after sacrifice by cervical dislocation and were processed for Giemsa-stained sections. LMWH was given once by s.c. injections (5 ^g/mouse) on Day - 2.

Results Swelling developed rapidly at sites of PCA reactions (Δ of 35 x 10"* inch at 15 min.) but not at control sites (ears injected with diluent alone) . Swelling of PCA sites diminished markedly between 2 and 4 hours after i.v. antigen challenge.

PCA and control sites were assessed histologically at 6-8 hours after the i.v. antigen challenge. The majority of mast cells at PCA sites exhibited extensive or moderate degranulation. By contrast < 5% of mast cells at control sites exhibited marked degranulation. There was a significant neutrophil infiltration only in PCA sites at 6 hour post antigen challenge. This infiltration was markedly reduced (by 60%) in mice which had been pretreated with LMWH two days earlier. There was no effect of this drug on the magnitude of mast cell degranulation. There was no effect of the drug on the total and differential count of leukocytes in the peripheral blood of these animals. It can be concluded that LMW heparin inhibited the cellular infiltration associated with the IgE dependent late cutaneous reaction. Additionally, the applicants also anticipate that the administration of LMWH will exhibit a beneficial effect on cutaneous LPR in animals with active cutaneous anaphylaxis (specific IgE production will be induced with DNP-HSA Alum) . ' Similar therapeutic effects on pulmonary allergic inflammation are also anticipated (Tarayre, J. P. et al. Int. J. Immunopharmacol . (1992) 14 (5) :847-855. - -Ψ2 5.11. LMWH Treatment of Human DTH Figure 8 shows an experiment in which a 40 year old male volunteer weighing 85 kg was tested for DTH reactivity to tetanus antigen (Merieux skin test applicator) . About 18 mm of induration was measured at 24 and 48 hours. The volunteer was then treated with a subcutaneous injection of Fragmin® (batch 38609) 3 mg. Five days later the volunteer was again tested for his DTH response to tetanus and the induration was inhibited to about 5 mm. The volunteer was tested again for DTH 3 weeks later ("Recovery") and the test showed positive reactivity (23 mm of induration at 24 and 48 hours) . The volunteer was then treated with Fragmin® as before and the DTH reactivity was measured again 7 days later ("7 days post") . Again the DTH was inhibited to about 5 mm of induration. Recovery of DTH again was found 3 weeks later. Thus, LMWH at a dose of less than 5 mg can inhibit DTH in humans at treatment intervals of 5 and 7 days. 6. Experiments Using LMWH (Fragmin) And Other Active Substances 6.1. Stability Studies of LMWH (Fragmin) TNF-a Inhibitory Activity Fragmin batch 38609 was diluted in normal saline to a concentration of 5 μg/0.1 ml. Some of the vials were mixed with 'an equal amount of protamine sulfate (5 ^g) and the vials were stored at room temperature (21 *C) for 0 to 72 hours (Table XI) or were stored at 4 *C for 1 to 4 months (Table XII) . The Fragmin with or .without protamine sulfate was then used in vivo to inhibit the DTH T cell reaction in BA-LB/c mice as described above. For the present experiments, the positive control DTH was 17.5 ± 1.2 x 10"2 mm (0 % inhibition) and the fully inhibited DTH was 2.6 ± 0.5 x 10'2 mm (100 % inhibition). . - -= r - The results of the incubation of LMWH at 20 'c are listed in the Table XI. It is evident from Table XI that LMWH loses its inhibitory activity against TNF- -dependent, T cell mediated DTH reaction upon incubation at ambient temperature over 72 h. In contrast, Heparin and LMWH lose their anti-coagulant activity at ambient temperature only slowly .

Table XI. Stability of Inhibitory Activity of Fragmin (Batch 38609, 5 g 0.1 ml), Without Protamine Sulfate, at 20 *C Against DTH-Reaction.

No. Hours DTH Reaction % anti-DTH Reactivity None 17.5 ± 1.2 Control* 0 2.6 ± 0.5 100** 24 9.6 ± 1 47 48 15.7 ± 1.6 13 72 17 ± 0.8 0 no inhibition full inhibition 6.2. Loss of Anti-DTH Reactivity at Low Temperature and Stabilizing Effect of Added Protamine Table XII shows that Fragmin loses its ability to inhibit the DTH reactivity of mouse T cells in dilute solution within 4 months at 4 *C. The addition of an equal concentration of protamine sulfate does not interfere with inhibition of the DTH reaction, but actually preserves this activity intact after 4 months at 4 'C. Again, this result is contrary to the normal role of protamine sulfate,- when added to heparin or Fragmin, in which the protamine sulfate - -÷4- - neutralizes the anti-coagulant effects of the heparinoid substances .

Loss of Anti-DTH Activity Over Time at Low Temperature. Stabilizing Effect of Added Protamine.

Fragmin Months at Protamine DTH % Anti-DTH (38609) 4 * c Sulfate (lO"2 mm) Activity none none none 15 ± 1 control* yes 1 none 2.8 ± 0.5 100** yes 1 yes 3 ± 0.4 100 yes 2 none 4 ± 0.8 82 yes 2 yes 2.4 ± 1 100 yes 3 none 9.6 ± 0.8 55 yes 3 yes 3 ± 0.5 100 yes 4 none 14.8 ± 1.4 0 yes 4 yes 3 ± 0.4 100 yes 0 yes 3 ± 0.5 100 no inhibition full inhibition 6.3. Preparation of ECM-Coated Plates ECM-coated wells were prepared as follows.

Freshly dissected bovine eyes were obtained from a slaughter house within a few hours after slaughter. The eyes were dissected in a hood to remove the cornea. The cornea were then scratched or scraped with a scalpel to obtain the corneal endothelial cells. These cells were cultured on tissue culture plates with approximately 5 ml of media comprising DMEM supplemented with 10% fetal calf serum, 5% calf serum and antibiotics, such as 1% streptomycin or 1% neostatin, together with 1% glutamine as a stabilizer. The cells settled to the bottom of the plates after approximately 2 days of seeding, were fed with fresh media every four days, and incubated at 37 °C in 5% C02 humidified incubators. If desired, some fibroblast growth factor may also be added to the media, although the addition of FGF is not crucial. When the cells were confluent (approximately 2 weeks later) , the supernatant was aspirated off, and the cells were then trypsinized with 1-2 mis of trypsin.

Eighty percent of these primary cells (the fate of the remaining 20% of the primary cells is described immediately below) were taken and divided onto 5 flat-bottomed 96-well plates. The cells were cultured in DMEM supplemented with 4% dextran T-40, 10% fetal calf serum and 5% calf serum. After about 7 days of incubation at 37 °C in a 10% C02 humidified incubator, the resulting confluent layers of endothelial cells were lysed. The lysing buffer, comprising 0.025 M NH40H containing 0.25% Triton X in PBS, was allowed to remain over the cells for 10 minutes and then decanted. The contents of the plates were then washed three times with PBS chilled to 4 °C. The preceding procedure left the ECM intact, firmly attached to the entire area of the well. The resulting ECM was also free of nuclei and cellular debris. The ECM-coated plates can be stored at 4 °C for at least three months.

The remaining 20% of the primary cells were left on a single plate and cultured in approximately 5 ml of media comprising DMEM supplemented with 10% fetal calf serum, 5% calf serum and antibiotics as described above.

This secondary crop of cells was allowed to become confluent and was treated with trypsin as described above. Again, the trypsinized cells were divided, 80% being cultured in 5 plates in the growth media containing 4% Dextran T-40, and 20% being cultured in a single plate as before. It is 6 possible to perform this 80/20 division yet one more time from this single plate. 6.4. Degradation of Sulfated Proteoglycans 35 (S) 04-labelled ECM was incubated with 5 μΐ of MM5 heparanase (4 u/ml) in 1 ml PBS and 100 μΐ 8.2 M phosphate-citrate buffer (pH 6.2) for 48 hrs. at 37 °C. The medium was then collected, centrifuged at 10,000 g for 5 min. (optional) and analyzed by gel filtration on Sepharose 4B columns. Two ml fractions were eluted with PBS at a flow rate of 5 ml/hr and were counted for radioactivity using Bio-Fluor Scintillation fluid.

This 35 (S) 04-labelling experiment showed that the ECM was actually being degraded, that the resulting degradation products were successfully being released, and, furthermore, were being properly filtered through the Sepharose 4B columns. Subsequent experiments related to the degradation of sulfated proteoglycans were carried out on non-labeled ECM, with the degradation products being monitored by their absorption at 206 or 232 nm, instead.

Enzyme degradation experiments were carried out as above and, in addition, the degradation products (DECM) were purified further by loading the degraded proteoglycans that were eluted from the Sepharose columns onto HPLC columns. HPLC analysis of the Sepharose column fractions was carried out in a manner such as that described in Section 6.11 et seq. Detection of the degradation products was achieved by monitoring their absorption at 206 nm.

Additional enzyme degradation experiments were carried out with similar results using PC3 enzyme and heparanase obtained from IBEX. 6.5. Purification of Human CD4+ T Cells CD4+ T cells were obtained from peripheral blood mononuclear leukocytes obtained from healthy human donors as - - follows. The mononuclear cells were isolated on a Ficoll gradient, washed in RPMI supplemented with 10% FCS and antibiotics in petri dishes and incubated at 37 °C in a 10% C02 humidified atmosphere. After 1 h, the non adherent cells were removed and incubated on nylon-wool columns (Fenwall, IL) for 45-60 min at 37 °C in a 10% CO, humidified atmosphere. Non adherent cells were eluted and washed.

CD4+ T cells were negatively selected by exposure of the eluted cells to a mixture of the following monoclonal antibodies (mAb) : anti-CD8, CD19, and CD14 conjugated to magnetic-beads (Advanced Magnetics, Cambridge, MA) . Unbound cells were recovered and their phenotypes were examined. The resultant purified cells were predominantly (>90%) CD3+CD4+ as determined by FACScan analysis. 6.6. Bioassay of TNF-a Activity Using Human CD4+ T Cells Derived from PBLs.

Two hundred fifty thousand human CD4+ T cells were preincubated with 150 μΐ of ECM degradation products at various concentrations for 1.5 h at 37 °C, under a 7% C02 atmosphere. Then 100 μΐ of PHA (Wellcome Co., England, 1 Mg/ml) were added for 3 h incubation, in flat-bottomed 96-well plates (Costar) . Subsequently, the contents of the wells (3-6 wells per experimental group) were collected, centrifuged, and the media were assayed for TNF-a secretion as previously described in Section 5.1. Briefly, supernatants of cultured lymphocytes were added to cultures of mouse fibrosarcoma cell clones (BALB/c . CL7) . BALB/c. CL7 cells are sensitive to killing by TNF-a in the presence of actinomycin D (0.75 μg/ml) . Nophar, Y. et al. J. Immunol. (1988) 140(10) :3456-3460. The death of these cells, in the presence of the test media, was calibrated in comparison to titration curves of added exogenous TNF. Cell viability is determined by incubation with MTT tetrazolium (Sigma, Cat.

No. M2128) for two hours, extracting the dye that was taken 6? up by the cells with isopropanol-HCl mixture and o quantitating it cjilorimetrically (at 570 nm) with a Microelisa Autoreader. TNF-a typing was done by examining the neutralizing effect of anti-murine TNF-a mAb (diluted 1/400; Genzyme, MA). 6.7. Degradation of Heparin One milligram of heparin (Sigma) in 1 ml of PBS and 100 μΐ 25 M phosphate-citrate buffer (pH 6.2) was incubated with 20 μΐ MM5 (5 u/ml) for 48 hrs . at 37 °C. The products of the reaction were then analyzed by gel filtration on Sepharose 4B columns. Two ml fractions were eluted with PBS at a flow rate of 5 ml/hr. To further characterize the degradation products, the peaks eluted from the Sepharose column were subjected to HPLC separation using Toyo Soda-Gel G3000SW and G2000S HPLC columns, as described in Section 6.11 et seq.

Additional experiments were carried out using 20 μΐ of PC3. The PC3 enzymatic reaction was carried out with 1 mg of heparin under the same conditions as described above for the MM5 except that the reaction was incubated for 24 hrs instead of 48 hrs. The products were then analyzed by gel filtration on Sepharose 4B columns (Fig. 29) . The in vitro bioassay results are shown in Table XIX below. 6.8; Elicitation of DTH Response in Mice and Examining Inhibitory Effects BALB/c mice (at least 5 mice per group) were sensitized on the shaved abdomen with 3% 4-ethoxymethylene-2-phenyl oxazolone (OX; BDH Chemicals, GB) in acetone/olive oil applied topically. DTH sensitivity was elicited 5 days later as follows. Mice were challenged with 0.5% OX in acetone/olive oil. The ear was measured immediately before challenge and 24 h later with Mitutoyo engineer's micrometer (Japan). The individual measuring the swelling of the ear was blinded to the identity of the - -^τ - up by the cells with isopropanol-HCl mixture and quantitating it calorimetrically (at 570 nm) with a Microelisa Autoreader. TNF- typing was done by examining the neutralizing effect of anti-murine TNF-a mAb (diluted 1/400; Genzyme, MA). 6.7. Degradation of Heparin One milligram of heparin (Sigma) in 1 ml of PBS and 100 μΐ 25 M phosphate-citrate buffer (pH 6.2) was incubated with 20 μΐ MM5 (5 u/ml) for 48 hrs. at 37 «C. The products of the reaction were then analyzed by gel filtration on Sepharose 4B columns. Two ml fractions were eluted with PBS at a flow rate of 5 ml/hr. To further characterize the degradation products, the peaks eluted from the Sepharose column were subjected to HPLC separation using Toyo Soda-Gel G3000SW and G2000SW HPLC columns, as described in Section 6.11 et seq.

Additional experiments were carried out using 20 μΐ of PC3. The PC3 enzymatic reaction was carried out with 1 mg of heparin under the same conditions as described above for the MM5 except that the reaction was incubated for 24 hrs instead of 48 hrs. The products were then analyzed by gel filtration on Sepharose 4B columns (Fig. 29) . The in vitro bioassay results are shown in Table XIX below. 6.8. Elicitation of DTH Response in Mice and Examining Inhibitory Effects BALB/c mice (at least 5 mice per group) were sensitized on the shaved abdomen with 3% 4-ethoxymethylene-2-phenyl oxazolone (OX; BDH Chemicals, GB) in acetone/olive oil applied topically. DTH sensitivity was elicited 5 days later as follows. Mice were challenged with 0.5% OX in acetone/olive oil. The ear was measured immediately before challenge and 24 h later with Mitutoyo engineer's micrometer (Japan). The individual measuring the swelling of the ear was blinded to the identity of the groups of mice. To interfere with DTH response, the low molecular weight immuno-regulatory fractions, diluted in PBS, were administrated subcutaneously into the back of the treated mice at the indicated time schedules and concentrations. Treated mice were inspected during and after (>2 months) the treatment and no major side-effects were observed clinically. 6.9. Separation of LMffH {Fragmin) on Size-Exclusion Gel Chromatography Column (Sepharose 4B) Fragmin (Batch 38609) and inactive Fragmin were fractionated by gel filtration on Sepharose 4B (Pharmacia) columns. Fractions of 0.5 ml were eluted with PBS at a flow rate of 5 ml/hr, and monitored for absorbance at 206 nm.

(No absorbance was detected at 280 nm) . A plot of the fraction number versus absorption at 206 nm appears on Fig. 11. The results of the bioassays for selected fractions are presented in Tables XIII and XIV, below. - -8-θ- Table XIII. Effect of Whole Fragmin, Sepharose 4B Fractions of Fragmin, and an HPLC Fraction a Sepharose 4B Fraction on the Secretion o Active TNF Using Human PBL Bioassay.

Bioassay of Test TNF Activity "R" value Material cone (pg/ml)

Claims (1)

1. 7. The pharmaceutical composition according to anyone of claims 2 to 6 which is in unit dosage form. 8. The pharmaceutical composition according to any one of claims 2 to 7 for daily administration. 9. The pharmaceutical composition according to any one of claims 2 to 7 for weekly administration. 10. The pharmaceutical composition according to any one of claims 2 to 9 for the treatment of a medical condition responsive to an inhibited production of TNF-a. 11. The pharmaceutical composition according to claim 10 in which said medical condition is selected from the group consisting of insulin-dependent diabetes mellitus, periodontal disease, inflammatory bowel disease, skin diseases, uveitis, rheumatic diseases, chronic inflammation, multiple sclerosis, lupus erythematosus, atherosclerosis, arthritis, vasculitis, allergy, autoimmune disease and allograft rejection. 12. The pharmaceutical composition according to claim 11 in which said allograft is an organ transplant or a skin graft . 13. The pharmaceutical composition according to claim 12 in which said organ is heart, liver, kidney or bone marrow. 14. Use of a compound of the formula (II) as defined in claim 1 for the preparation of a pharmaceutical composition substantially as described in the specification . 15. The use according to claim 14 in which said pharmaceutical composition is for the inhibition of the production of an active cytokine. 16. The use according to claim 15 in which said active cytokine is selected from the group consisting of active IL-1, IL-6, IL-8 and TNF-a. 17. The use according to claim 16 in which said active cytokine is active TNF-a. 18. The use according to any one of claims 14 to 17 in which said pharmaceutical composition is adapted for parenteral, oral or topical administration. 19. The use according to any one of claims 14 to 18 in which said pharmaceutical composition is in unit dosage form. 20. The use according to any one of claims 14 to 19 in which said pharmaceutical composition is for daily administration. 21. The use according to any one of claims 14 to 19 in which said pharmaceutical composition is for weekly administration . 22. The use according to any one of claims 14 to 21 in which said pharmaceutical composition is for the treatment of a medical condition responsive to an inhibited production of TNF-a. 23. The use according to claim 22 in which said medical condition is selected from the group consisting of insulin-dependent diabetes mellitus, periodontal disease, inflammatory bowel disease, skin diseases, uveitis, rheumatic diseases, chronic inflammation, multiple sclerosis, lupus erythematosus, atherosclerosis, arthritis, vasculitis, allergy, autoimmune disease and allograft rejection . 24. The use according to claim 23 in which said allograft is an organ transplant or a skin graft. 25. The use according to claim 24 in which said organ is heart, liver, kidney or bone marrow. For the Applicants, Paulina Ben-Ami Patent Attorney