Patent Application
20080063645
Endotoxin-adsorbent used for the treatment of patients with autoimmune diseases is required the following criteria: 1) a high endotoxin-binding capacity, 2) suitable physiological features for oral administration usage and 3) high margin of safety without any adverse effects. The number of microorganisms residing in the human gastrointestinal tract is believed to be approximately 100 trillion, and the number of endotoxin-containing microorganisms among these bacteria is also massive. Therefore, in order to remove large portions of endotoxin in the gastrointestinal tract, endotoxin-adsorbent must have high binding capacity of endotoxin.
In order to satisfy these requirements, new endotoxin-adsorbents, which have high endotoxin-binding capacity, high margin of safety, and suitable physical features for oral administration use, are provided in this invention.
The toxic core of an endotoxin molecule is located in the Lipid A region. This invention consists of a Lipid A-binding substance and insoluble carrier particles in order to adsorb and eliminate large amounts of endotoxin from the gastrointestinal tract into feces by oral administration, for preventing and treating autoimmune diseases such as rheumatoid arthritis.
The endotoxin adsorbent suitable for the usage of above purposes is as follows:
Several materials such as polymixin B or a peptide antibiotic, and endotoxin-binding peptides (JP-A 11-335396, JP-A 2002-263486, JP-A 2002-311029, JP-A 2004-292357 and JP-A 2002-512140) have been known as Lipid A binding substances. All these materials can be used for the purpose of this invention.
The carrier of endotoxin-binding substance is desired to be small particles or powder suitable and convenient for oral administration usage. A variety of polysaccharides and their derivatives, such as cellulose, agarose, mannan, glucan, and chitin, or variety of synthetic polymers, such as polyacrylate, polystyrene, polypropylene, polyamide, and polyvinyl, can be used as the carrier of an endotoxin-binding substance.
There is no definite method required to conjugate Lipid A binding substance to the carrier particles. The cross-linkers, which have been widely used for immobilizing enzymes onto the solid surfaces like water soluble carbodiimides such as ECDI, hexamethylene diisocyanate, propyleneglycol di-glycidylether which contain 2 epoxy residues, and epichlorohydrin, can be used.
The features of Lipid A binding substance-carrier complexes are desired to be non-toxic, non-absorbable, and non-digestible by digestion enzymes and by microorganisms and resistant to other intestinal components such as gastric acid. “Non-digestible” means resistant to both digestion enzymes of animals and enzymes produced by microorganisms.
The microorganisms residing in the gastrointestinal tract possess enzymes which are capable of digesting cellulose and other substances that are resistant to digestion enzymes of animals (Kopecny J et al. Detection of cellulolytic bacteria from the human colon. Folia Microbiol (Praha) 49:175-7, 2004, Nakajima N et al. Dietary-fiber-degrading enzymes from a human intestinal Clostridium and their application to oligosaccharide production from nonstarchy polysaccharide using immobilized cells. Appl Microbiol Biotechnol 59: 182-9, 2002). Therefore, in this invention, the materials used as a carrier of LPS-binding substance should be restricted to materials, which are resistant to bacterial digestion, and usage of polysaccharides such as cellulose, agarose, mannan, glucan, and chitin as a carrier of an endotoxin-binding substance should be excluded.
Compared to naturally occurring polymers, synthetic polymers are generally resistant to digestion enzymes secreted into gastrointestinal tract of animals and even to various bacterial enzymes. Therefore, it is desired to choose a synthetic polymer as a carrier of lipid A binding substance. In fact, synthetic polymers such as polystyrene sulfonate calcium and anion exchange resin are widely used as a potassium adsorbent and as a cholesterol adsorbent for treatment of patients with high potassium and high cholesterols, respectively.
The particle size of endotoxin-adsorbent is an important factor that should be considered, since it has been known that small size particles, such as yeast, with less than 5 μm in diameter, are phagocytized by M cells, which reside on the surface of Peyer's patches scattered along small and large bowel regions (Gerbert A. et al. M cells in Peyer's patches of the intestine. Int Rev Cytol. 167:91-159, 1996). Therefore, the particle size of endotoxin-adsorbent not more than 5 μm in a diameter is excluded according to the specification of polystyrene sulfonate calcium defined in Japanese Pharmacopoeia.
The molecular weight of endotoxin is more than 10,000 daltons, and assumed to bind mainly on the surfaces of entotoxin-adsorbent particles rather than the inside of particles. Therefore, if the particle size is smaller, the endotoxin binding capacity is larger due to larger surface area per unit weight of particles. This evidence is shown in EXAMPLE 12.
There are two classes of fine grinding techniques, dry and wet methods. Impact method, screen method, grind method and others are known as dry methods, whereas catalyst-stirring method is an example of a typical wet method. There are several other methods, but there is no limitation in the methods for grinding the particles of endotoxin-adsorbent, and any of these methods can be used for preparing fine powder or small particles of endotoxin-adsorbent.
The particle size of endotoxin-adsorbent was determined based on particle size distribution method. The particle distribution analysis was performed according to “The method for determining particle size distribution. Method 1: Microscopic method” in the second supplement of the general test procedures, section 65, Japanese Pharmacopoeia, 13th Issue. The 50% particle size (μm) was explained as the diameter of particles of the corresponding values of accumulative volume of particles is 50%.
LPS binding capacity of endotoxin-adsorbent in a test tube was determined according to the method described in “endotoxin test procedures” in Japanese Pharmacopoeia, in addition to a simple assay method of LPS by measuring OD values, which was developed during this invention. The experimental procedures and results are shown in EXAMPLE10 and 11 in detail.
The therapeutic effect of endotoxin-adsorbent on autoimmune diseases can be determined in mouse arthritis model as described in our previous invention, JP-A 2006-151914. Briefly, arthritis can be induced in 100% of mice by IP injection of enough amounts of anti-type II collagen monoclonal antibody cocktail (Chondrex Inc., Redmond, Wash., USA) within 3 days (Terato K et al. Induction of arthritis with monoclonal antibodies to collagen. J. Immunol. 148:2103-2108, 1992). By reducing the dose of monoclonal antibody cocktail to 2 mg, all mice remained normal without developing arthritis. However, oral administration of 3 mg of LPS on 3 consecutive days from day 0, day 1, day 2 and day 3 into these mice induced clinically apparent arthritis, which reached the peak on day 6-7. Furthermore, one group of mice was co-administered with indomethasin and ovoinhibitor, a protease-inhibitor purified from egg white. The combination of indomethasin and ovoinhibitor was used to increase the mucosal permeability of gastro-intestinal mucosa. In these mice, the effect of LPS was more significant, and more severe arthritis was induced by oral administration of a same dose of LPS. Using this arthritis model induced by a combination of monoclonal antibody and LPS, the therapeutic effect of endotoxin-adsorbent can be determined. The benefit of this model is multifold: time of experimentation is short compared with authentic collagen-induced arthritis model, the standard deviation of severity of arthritis among individual mice is much less, and the effect of LPS-adsorbent is clearly determined.
Since the endotoxin-adsorbent is used as a therapeutic for human patients by oral administration, various formulas, which are currently employed in medicines used by oral administration, can be applied: for example, powdered or suspended powdered form, capsule, tablet and solution. These formulas can be provided using authentic methods by mixing the endotoxin-adsorbent with various vehicles and additives within a range that is acceptable with respect to pharmaceutical guidelines.
The endotoxin-adsorbent can be administered orally at 10 mg −10 g per adult by a single or three administrations per day.
The individual examples of this invention are described in detail in the following sections, but the invention is not to be considered limited to these examples as described below.
Polymixin B sulfate (3 million units, Maruko Pharmaceuticals) was dissolved in 200 mL of 0.1M NaCl, and then pH was adjusted to 8 by adding NaOH. Four grams of polyacryl resin with epoxy residue (Amberzyme, Rohm and Haas, USA) was added to the solution and stirred using a blade propeller for 72 hours. The resin was washed with 1000 mL of distilled water on a membrane filter with a 5 μm pore size, and suspended in 50 mL of 1M glycine solution, pH 8.0, adjusted by NaOH. After incubation overnight, the resin was washed with 2 liters of distilled water on a filter, and dried in a desiccator. Yield of polymixin B-conjugated resin (this is called as RPMB) was 3.9 g.
Polymixin B sulfate (3 million units, Maruko Pharmaceuticals) was dissolved in 200 mL of 0.1M NaCl, and adjusted pH to 8 by adding NaOH. Four grams of polyacryl resin with epoxy residue (Amberzyme, Rohm and Haas, USA) was transferred into a mortar, and grounded with a pestle by adding polymixin B sulfate solution drop-wise. Polymixin B and epoxyacryl resin was stirred for 72 hours using a magnetic stirring bar. The resin was washed with 1 liter of distilled water on a membrane filter with a 5 μm pore size, then suspended in 50 mL of 1M glycine solution, pH 8.0, adjusted with NaOH. After incubation overnight, the resin was washed with 2 liters of distilled water on a filter and dried in a desiccator. The yield of polymixin B-conjugated resin (this is called as RPMB-1) was 3.2 g.
One gram of polyacryl resin containing epoxy residue (Amberzyme, Rohm and Haas, USA) was grounded with a mortar and pestle. The powdered resin was suspended in 20 mL of distilled water, reacted with 0.6 g of epichlorohydrin and 0.3 mL of 50% NaOH for 2 hours. The resin was washed with 100 mL of distilled water on a membrane filter with a 5 μm pore size, and then mixed with in 5 mL of 1M phosphate buffer, pH 10.0, containing 3 million units of polymixin B, and stirred at 40° C. for 16 hours. After the reaction, the resin suspension was added by 50 mL of 1M glycine solution, pH 8.0, adjusted with NaOH, and kept overnight. The resin was washed with 2 liters of distilled water on a filter and dried in a desiccator. The yield of polymixin B-conjugated resin (this is called RPMB-2) was 0.5 g.
Four grams of weakly acidic cation-exchange resin with carboxyl residues (Dowex MAC-3), was suspended in 50 mL of 0.1M MOPS (3-morpholinopropanesulfonic acid) solution, pH 7.5, and then reacted with 1 g of 1-Ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (Sigma, USA), a coupling agent, at 4° C. for 2 hours with stirring. The activated resin was collected on a membrane filter with a 5 μm pore size and washed with 200 mL of distilled water. The washed activated resin was then suspended in 50 mL of 0.1M MOPS, pH 7.5, and then mixed with 3 million units of polymixin B dissolved in 10 mL of 0.1M MOPS solution, pH 7.5, and reacted at 4° C. for 16 hours with stirring. The resin was collected on a membrane filter with a 5 μm pore size and then suspended in 50 mL of 1M glycine solution, pH 8.0, adjusted with NaOH, and kept at 4° C. overnight, washed with 2 liters of distilled water and then dried in a desiccator. The yield of polymixin-conjugated weakly acidic cation exchange resin (This preparation is called 4/300) was 3.8 g.
Using the same resin (Dowex MAC-3) and same procedures shown in the example 4, except reducing the amount of resin from 4 to 1 g, 0.9 g of polymixin B-conjugated weakly acidic cation exchange resin was obtained. This preparation is called 1/300.
Four grams of weakly acidic cation-exchange resin with carboxyl residues (Dowex MAC-3), was grounded with a mortar and pestle, and suspended in 50 mL of 0.1M MOPS, pH 7.5. The resin was reacted with 1 g of 1-Ethyl-3-(3-dimethyl-aminopropyl) carbodiimide hydrochloride (Sigma, USA), a coupling agent, at 4° C. for 2 hours with stirring. The activated resin was collected on a membrane filter with a 5 μm pore size and washed with 200 mL of distilled water. The resin was suspended in 50 mL of 0.1M MOPS, pH 7.5, and then added to 10 mL 0.1M MOPS, pH 7.5, containing 3 million units of polymixin B. Polymixin B and the resin were reacted at 4° C. for 16 hours with stirring. The resin was collected on a membrane filter with a 5 μm pore size and then suspended in 50 mL of 1M glycine solution, pH 8.0, adjusted with NaOH, and kept overnight, washed with 2 liters of distilled water, and then dried in a desiccator. The yield of polymixin B-conjugated weakly acidic cation exchange resin was 3.1 g. This preparatopm is called 4M/300.
Using the same resin and same procedures as described in example 6, except reducing the amount of resin from 4 to 1 g, 0.52 g of polymixin B-conjugated weakly acidic cation exchange resin was obtained. This preparation is called 1M/300.
Dowex Mac-3 resin (diameter: 300-1200 mm), 3300 liters, was grounded using a Dalton NeaMill, NEA-48 type. The yield of powdered Mac-3 was 1400 kg and the average particle size was 30 μm, ranging from 10 to 50 μm.
The particle size distribution was measured according to “Measurement of Particle Size Distribution. Method 1: Microscopic method” in the second supplement of the general test procedures, Section 65, Japanese Pharmacopoeia, 13th Issue. The microscope and camera used for this experiment was Nikon ECLIPSE E600 and Victor KY-F55B, respectively. The collected data was analyzed using Nano Hunter NS2K-Pro. The result of analysis of 1006 particles of RPMB prepared in Example 1 by this method is shown in Table 1. The 50% particles size of RPMB was 213 μm, and the content of small particles not more than 5 μm in a diameter was 0%.
Endotoxin was assayed by an end point calorimetric assay method using Endospecy-ES24S kit (Seikagaku Kogyo, Japan). Lipopolysaccharide (LPS) from E. coli O-111 (Sigma L4130) was dissolved in pyrogen free water at 5 μg/mL. One mL of this LPS solution was mixed with 50 μg and 100 μg of RPMB-1, and 100 μg of polymixin B-unconjugated resin (control) and incubated at 37° C. with stirring. Endotoxin levels were determined in the supernatant before, 10 and 20 minutes after adding the resins. As shown in Table 2, LPS was specifically adsorbed by RPMB-1.
The endotoxin binding capacity of polymixin B conjugates was also studied. LPS (Sigma L4130) was dissolved in pyrogen free water at 0.2 mg/mL, and 4 mL of this solution was added to a test tube containing 20.6 mg of RPMB-1, and incubated at 37° C. with stirring. The supernatant was collected every 30 minutes by centrifugation and the OD values at 210 nm was determined. The OD210 value was dropped from 1 to 0.6 within the first 30 minutes of incubation and remained unchanged afterwards. By adding 20.5 mg of fresh RPM-1 into the supernatant, the OD value was slightly reduced from 0.6 to 0.4. Therefore, it was assumed that 20.6 mg of LPS adsorbent added in the first test tube was saturated with LPS. Accordingly, LPS adsorption capacity of RPMB-1 was calculated based on the OD value changes of LPS solution. Since the LPS preparation used in this experiment is not pure and contaminated by DNA and proteins, it was assumed that the final OD value of 0.4 reflects the OD value of such contaminants. Based on this assumption, it was calculated that 1 g of RPMB is capable of binding approximately 25.9 mg of LPS (Sigma L4130) using the following formula:
(0.8 mg×0.4/0.6)/0.0206=25.9 mg
Polymixin B-conjugated resins prepared in Examples 1-7 were analyzed for their particle size distribution by the method described in Example 9 and assayed for the endotoxin binding capacity by the method described in Example 11 to study the relationship between particle size and LPS binding capacity. AffiPrep poplymixin B (BioRad, USA) was used as a reference.
The endotoxin binding capacity of individual batches of endotoxin-adsorbents prepared by conjugating with 3 million units of polymixin B was compared and expressed as endotoxin units (EU) per gram weight of resin as well as LPS weight per gram resin. The weight of LPS was obtained by converting the EU values based on the EU value per mg of LPS preparation (Sigma L4130, LPS preparation from E. coli O-111,B4, by trichloroacetic acid extraction) used for this experiment.
In spite of using the same amount of 3 million units of polymixin B to make conjugates as described in Example 1-7, it was apparent that endotoxin-binding capacity of polymixin B-conjugated resin is higher if the 50% particle size is smaller as shown in Table 3.
DBA/1JNCrj mice (Japan Charles River) were divided into 5 groups (G1-G5, 5 mice per group). In order to increase the mucosal permeability, all mice received 40 μg of indomethasin (Sigma) and 2 mg of ovomucoid (Sigma) for 5 consecutive days from day −6 to −2 by oral route. On day 0, all mice received an IV injection of 0.2 mL of arthritogenic monoclonal antibody cocktail (10 mg/mL). Endotoxin derived from E. coli O-111 (Phenol extracted LPS, Sigma) was dissolved in PBS at 7.5 mg/mL, and 0.2 mL of this solution was administered into G1-G4 mice by oral route for 3 consecutive days from day 0 to 2. Endotoxin-adsorbent, RPMB-1, was suspended in distilled water at 100 mg/mL, and deaerated by vacuum pump to keep the particles in uniform suspension by preventing the aggregation of particles. The RPMB-1 suspension was administered to mice at doses of 0.125 mL (G2), 0.25 mL (G3) and 0.5 mL (G4) twice a day for 4 consecutive days from day 0 to 3 after LPS administration. G1 received 0.25 mL of water alone. Mice in G5, a positive control of arthritis, received IP injection of 0.1 mL of LPS solution (0.5 mg/mL in PBS) on day 3.
All mice were observed for the development of arthritis every day from day 0 to 14. Severity of arthritis was scored by 5 grades, 0: normal without any swelling, 1: clinically apparent swelling of one digit, 2: moderate redness and swelling of more than 2 digits or moderate redness and swelling of the entire paw, 3: severe redness and swelling of the entire paws, and 4: maximum inflamed limb with involvement of multiple joints. The sum of arthritis score (maximum 16 per mouse) of individual animals was calculated. The effect of endotoxin-adsorbent was calculated based on the average score of 5 mice using the following equation:
Suppression of arthritis (%)=(1−Average score of test group/Average score of control group)×100
Since the arthritis scores reached a maximum on day 7, the effect of endotoxin-adsorbent was calculated using the scores on day 7. The suppression of arthritis by RPMB prepared in EXAMPLE 2 at 25 mg, 50 mg and 100 mg per mouse by oral administration is shown in Table 3. None of the five mice which received 100 mg of RPMB-1 developed arthritis, whereas 2 out of 5 mice which received 50 mg of RPMB-1 developed mild arthritis (average score: 2), and 4 out of 5 mice which received 25 mg of RPMB-1 developed moderate arthritis (average score: 9) (Table 3, experiment 1), indicating a dose response effectiveness of RPMB-1.
Similarly, RPMB-2, 1M/300 and 4M/300, which have higher binding capacities of LPS then RPMB-1, were also tested for their effect on arthritis at a dose of 10 mg per mouse. All three preparations were equally effective and suppressed the development of arthritis almost completely (Table 3, experiment 2).
The tablets were prepared by mixing 15 g of 1M/300, which was prepared in EXAMPLE 7, 2.5 g of lactose, 2.4 g of corn starch, and 0.1 g of magnesium stearate. These four components were mixed well and compressed by single punch tableting machine to make tablets containing 200 mg of 1M/300 per a tablet.
RPMB-2 powder shown in EXAMPLE 3 was dispensed into hard capsules at 150 mg per a capsule.
