WO2000059493A2 - Improved dialysis solutions and methods - Google Patents

Abstract

The present invention provides improved dialysis compositions and methods for dialysis comprising utilizing the disclosed AGE inhibitors, together with methods to reduce dialysis-related complications and disorders.

Description

Improved Dialysis Solutions and Methods

Cross Reference

This application claims priority to U.S. Provisional Patent Application Serial No. 60/127,906 filed April 6, 1999, is a continuation in part of pending U.S. application Serial No. 08/971,285 (filed November 17, 1997), which is a continuation in part of U.S. Application Serial No. 08/711,555 (filed September 10, 1996), now U.S. Patent No. 5,985,857, and is related to pending U.S. Patent Application Nos. 09/322,569 (Filed May 28, 1999); 09/520,933 (filed March 8, 2000); 09/414,877 (filed October 8, 1999); 09/416,915 (filed October 13, 1999); and 09/422,615 (filed October 21, 1999), all incorporated herein by reference in their entirety.

Background of the Invention

The instant invention is in the field of dialysis methods and solutions, and Advanced Glycation End-products (AGEs) inhibition and inhibitors thereof.

Protein Aging and Advanced Glycosylation End-products

Nonenzymatic glycation by glucose and other reducing sugars is an important post-translational modification of proteins that has been increasingly implicated in diverse pathologies. Irreversible nonenzymatic glycation and crosslinking through a slow, glucose-induced process may mediate many of the complications associated with diabetes. Chronic hyperglycemia associated with diabetes can cause chronic tissue damage which can lead to complications such as retinopathy, nephropathy, and atherosclerotic disease. (Cohen and Ziyadeh, 1996, J. Amer. Soc. Nephrol 7:183-190). It has been shown that the resulting chronic tissue damage associated with long-term diabetes mellitus arise in part from in situ immune complex formation by accumulated immunoglobulins and/or antigens bound to long-lived structural proteins that have undergone Advanced Glycosylation End-product (AGE) formation, via non-enzymatic glycosylation (Brownlee et al, 1983, J Exp. Med. 158:1739-1744). The primary protein target is thought to be extra-cellular matrix associated collagen. Nonenzymatic glycation of proteins, lipids, and nucleic acids may play an important role in the natural processes of aging. Recently protein glycation has been associated with β-amyloid deposits and formation of neurofibπllary tangles m Alzheimer disease, and possibly other neurodegenerative diseases involving amyloidosis (Colaco and Harrington, 1994, NeuroReport 5: 859-861). Glycated proteins have also been shown to be toxic, antigemc, and capable of triggering cellular injury responses after uptake by specific cellular receptors (see for example, Vlassara, Bucala & Stπker, 1994, Lab. Invest 70:138-151; Vlassara et al., 1994, PNAS(USA) 91: 11704-11708; Daniels & Hauser, 1992, Diabetes 41 : 1415-1421, Brownlee, 1994, Diabetes 43:836-841 ; Cohen et al, 1994, Kidney Int. 45:1673-1679, Brett et al, 1993, Am. J Path 143:1699-1712; and Yan et al., 1994, PNAS(USA) 91-7787-7791). The appearance of brown pigments duπng the cooking of food is a universally recognized phenomenon, the chemistry of which was first descπbed by Maillard in 1912, and which has subsequently led to research into the concept of protem aging It is known that stored and heat-treated foods undergo nonenzymatic browning that is characteπzed by crosslinked proteins which decreases their bioavailibility. It was found that this Maillard reaction occurred in vivo as well, when it was found that a glucose was attached via an Amadoπ rearrangement to the amino-termmal of the -chain of hemoglobin.

The instant disclosure teaches previously unknown, and unpredicted mechanism of formation of post-Amadoπ advanced glycation end products (Maillard products; AGEs) and methods for identifying and characteπzing effective inhibitors of post- Amadori AGE formation. The instant disclosure demonstrates the unique isolation and kinetic characterization of a reactive protein intermediate competent in forming post- Amadoπ AGEs, and for the first time teaching methods which allow for the specific elucidation and rapid quantitative kinetic study of "late" stages of the protein glycation reaction. In contrast to such "late" AGE formation, the "early" steps of the glycation reaction have been relatively well characteπzed and identified for several proteins (Harding, 1985, Adv Protein Chem. 37:248-334; Monmer & Baynes eds., 1989, The Maillard Reaction in Aging, Diabetes, and Nutrition (Alan R. Liss, New York); Finot et al., 1990, eds. The Maillard Reaction in Food Processing, Human Nutrition and Physiology (Birkhauser Verlag, Basel)). Glycation reactions are known to be initiated by reversible Schiff-base (aldimine or ketimme) addition reactions with lysme side-cham ε- ammo and terminal α-amino groups, followed by essentially irreversible Amadoπ rearrangements to yield ketoamine products e.g. 1 -amino- 1-deoxy-ketoses from the reaction of aldoses (Baynes et al, 1989, in The Maillard Reaction in Aging. Diabetes. and Nutrition, ed. Monnier and Baynes, (Alan R. Liss, New York, pp 43-67). Typically, sugars initially react in their open-chain (not the predominant pyranose and furanose structures) aldehydo or keto forms with lysine side chain ε-amino and terminal α-amino groups through reversible Schiff base condensation (Scheme I). The resulting aldimine or ketimine products then undergo -Amadori rearrangements to give ketoamine Amadori products, i.e. 1 -amino- 1 -deoxy-ketoses from the reaction of aldoses (Means & Chang, 1982, Diabetes 31, Suppl. 3:1-4; Harding, 1985, Adv. Protein Chem. 37:248- 334). These Amadori products then undergo, over a period of weeks and months, slow and irreversible Maillard "browning" reactions, forming fluorescent and other products via rearrangement, dehydration, oxidative fragmentation, and cross-linking reactions. These post-Amadori reactions, (slow Maillard "browning" reactions), lead to poorly characterized Advanced Glycation End-products (AGEs). As with Amadori and other glycation intermediaries, free glucose itself can undergo oxidative reactions that lead to the production of peroxide and highly reactive fragments like the dicarbonyls glyoxal and glycoaldehyde. Thus the elucidation of the mechanism of formation of a variety of AGEs has been extremely complex since most in vitro studies have been carried out at extremely high sugar concentrations. In contrast to the relatively well characterized formation of these "early" products, there has been a clear lack of understanding of the mechanisms of forming the "late" Maillard products produced in post-Amadori reactions, because of their heterogeneity, long reaction times, and complexity. The lack of detailed information about the chemistry of the "late" Maillard reaction stimulated research to identify fluorescent AGE chromophores derived from the reaction of glucose with amino groups of polypeptides. One such chromophore, 2-(2-furoyl)-4(5)-(2-furanyl)-lH-imidazole (FFI) was identified after nonenzymatic browning of bovine serum albumin and polylysine with glucose, and postulated to be representative of the chromophore present in the intact polypeptides. (Pongor et al, 1984, PNAS(USA) 81:2684-2688). Later studies established FFI to be an artifact formed during acid hydrolysis for analysis.

A series of U.S. Patents have issued in the area of inhibition of protein glycosylation and cross-linking of protein sugar amines based upon the premise that the mechanism of such glycosylation and cross-linking occurs via saturated glycosylation and subsequent cross-linking of protein sugar amines via a single basic, and repeating reaction. These patents include U.S. Patents 4,665,192; 5,017,696; 4,758,853 4,908,446; 4,983,604; 5,140,048; 5,130,337; 5,262,152; 5,130,324; 5,272,165: 5,221,683; 5,258,381 ; 5,106,877; 5,128,360; 5,100,919; 5,254,593; 5,137,916 5,272,176; 5,175,192; 5,218,001 ; 5,238,963; 5,358,960; 5,318,982; and 5,334,617. (All U.S. Patents cited are hereby incorporated by reference in their entirety).

The focus of these U.S. Patents, are a method for inhibition of AGE formation focused on the carbonyl moiety of the early glycosylation Amadori product, and in particular the most effective inhibition demonstrated teaches the use of exogenously administered aminoguanidine. The effectiveness of aminoguanidine as an inhibitor of AGE formation is currently being tested in clinical trials.

Inhibition of AGE formation has utility in the areas of, for example, food spoilage, animal protein aging, and personal hygiene such as combating the browning of teeth. Some notable, though quantitatively minor, advanced glycation end-products are pentosidine and N^ε -carboxymethyllysine (Sell and Monnier, 1989, J. Biol. Chem. 264:21597-21602; Ahmed et al., 1986, J. Biol. Chem. 261:4889-4894).

The Amadori intermediary product and subsequent post-Amadori AGE formation, as taught by the instant invention, is not fully inhibited by reaction with aminoguanidine. Thus, the formation of post-Amadori AGEs as taught by the instant disclosure occurs via an important and unique reaction pathway that has not been previously shown, or even previously been possible to demonstrate in isolation. It is a highly desirable goal to have an efficient and effective method for identifying and characterizing effective post-Amadori AGE inhibitors of this "late" reaction. By providing efficient screening methods and model systems, combinatorial chemistry can be employed to screen candidate compounds effectively, and thereby greatly reducing time, cost, and effort in the eventual validation of inhibitor compounds. It would be very useful to have in vivo methods for modeling and studying the effects of post-Amadori AGE formation which would then allow for the efficient characterization of effective inhibitors.

Inhibitory compounds that are biodegradeble and/or naturally metabolized are more desirable for use as therapeutics than highly reactive compounds which may have toxic side effects, such as aminoguanidine

SUMMARY OF THE INVENTION

The present invention provides improved dialysis methods and compositions for dialysis that compπse utilizing an amount effective to inhibit AGE formation of a compound of the general formula

wherein R, is CH₂NH₂, CH₂SH, COOH, CH₂CH₂NH₂, CH₂CH₂SH, or CH₂COOH; R₂ and Ro is H, OH, SH, NH₂, C 1-6 alkyl, alkoxy or alkene, -R and Rs are H, C 1-6 alkyl, alkoxy or alkene;

Y is N or C, such that when Y is N R₃ is nothing, and when Y is C, R₃ is NO? or another electron withdrawing group, and salts thereof

In further aspects, the present invention provides methods for inhibiting dialysis- related cardiac morbidity and mortality, dialysis-related amyloidosis. limiting dialysis- 1 elated increases in permeability of the pentoneal membrane in a dialysis patient, inhibiting renal failure progression in a patient, and inhibiting ultrafiltration failure and peritoneal membrane destruction in a patient, compnsmg introducing into the patient a dialysis solution that compπses an amount effective to inhibit or limit the specified endpoint of a compound of the general formula

wherein R, is CH₂NH₂, CH₂SH, COOH, CH₂CH₂NH₂, CH₂CH₂SH, or CH₂COOH;

R₂ and R-s is H, OH, SH, NH₂, C 1-6 alkyl, alkoxy or alkene; t and R₅ are H, C 1-6 alkyl, alkoxy or alkene;

Y is N or C, such that when Y is N R₃ is nothing, and when Y is C, R₃ is NO? or another electron withdrawing group, and salts thereof.

In another aspect, the present invention comprises a method for inhibiting AGE formation in a dialysis patient comprising administering to the patient a dialysis solution comprising an amount effective amount to inhibit AGE formation of a compound of the general formula:

wherein R, is CH?NH₂, CH₂SH, COOH, CH₂CH?NH₂, CH?CH₂SH. or CH₂COOH:

R? and R,, is H, OH, SH, NH?, C 1-6 alkyl, alkoxy or alkene; 4 and R₅ are H, C 1-6 alkyl, alkoxy or alkene;

Y is N or C, such that when Y is N R₃ is nothing, and when Y is C, R3 is NO₂ or another electron withdrawing group, and salts thereof.

The instant invention encompasses pharmaceutical compositions which comprise one or more of the compounds of the present invention, or salts thereof, in a suitable carrier. The instant invention encompasses methods for administering pharmaceuticals of the present invention for therapeutic intervention of pathologies which are related to AGE foπnation in vivo. In one preferred embodiment of the present invention the AGE related pathology to be treated is related to diabetic nephropathy. BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 is a series of graphs depicting the effect of vitamin Bβ derivatives on AGE formation in bovine serum albumin (BSA). Figure 1A Pyridoxamine (PM); Figure IB pyridoxal phosphate (PLP); Figure 1C pyridoxal (PL); Figure ID pyridoxine (PN).

Figure 2 is a series of graphs depicting the effect of vitamin B \ derivatives and aminoguanidine (AG) on AGE formation in bovine serum albumin. Figure 2A Thiamine pyrophosphate (TPP); Figure 2B thiamine monophosphate (TP); Figure 2C thiamine (T); Figure 2D aminoguanidine (AG).

Figure 3 is a series of graphs depicting the effect of vitamin B6 derivatives on AGE formation in human methemoglobin (Hb). Figure 3A Pyridoxamine (PM); Figure 3B pyridoxal phosphate (PLP); Figure 3C pyridoxal (PL); Figure 3D pyridoxine (PN).

Figure 4 is a series of graphs depicting the effect of vitamin B] derivatives and aminoguanidine (AG) on AGE formation in human methemoglobin. Figure 2A Thiamine pyrophosphate (TPP); Figure 2B thiamine monophosphate (TP); Figure 2C thiamine (T); Figure 2D aminoguanidine (AG).

Figure 5 is a bar graph comparison of the inhibition of the glycation of ribonuclease A by thiamine pyrophosphate (TPP), pyridoxamine (PM) and aminoguanidine (AG).

Figure 6A is a graph of the kinetics of glycation of RNase A ( 10 mg/mL) by ribose as monitored by ELISA. Figure 6B is a graph showing the dependence of reciprocal half-times on ribose concentration at pH 7.5.

Figure 7 are two graphs showing a comparison of uninterrupted and interrupted glycation of RNase by glucose (7B) and ribose (7A), as detected by ELISA.

Figure 8 are two graphs showing kinetics of pentosidine fluorescence (arbitrary units) increase during uninterrupted and interrupted ribose glycation of RNase. Figure 8A Uninterrupted glycation in the presence of 0.05 M ribose. Figure 8B Interrupted glycation after 8 and 24 hours of incubation. Figure 9 is a graph which shows the kinetics of reactive intermediate buildup.

Figure 10 are graphs of Post-Amadori inhibition of AGE formation by ribose. Figure 10A graphs data where aliquots were diluted into inhibitor containing buffers at time 0. Figure 10B graphs data where samples were interrupted at 24h, and then diluted into inhibitor containing buffers.

Figure 11 is a graph showing dependence of the initial rate of formation of antigenic AGE on pH following interruption of glycation. Figure 12 are two graphs showing the effect of pH jump on ELISA detected

AGE formation after interrupted glycation. Interrupted samples left 12 days at 37°C in pH 5.0 buffer produced substantial AGEs (33%; Figure 12 B) when pH was changed to 7.5, as compared to the normal control sample not exposed to low pH (Figure 12 A).

Figure 13 is a series of graphs depicting the effect of vitamin B6 derivatives on AGE formation during uninterrupted glycation of ribonuclease A (RNase A) by ribose. Figure 13A Pyridoxamine (PM); Figure 13B pyridoxal-5 '-phosphate (PLP); Figure 13C pyridoxal (PL); Figure 13D pyridoxine (PN).

Figure 14 is a series of graphs depicting the effect of vitamin Bi derivatives and aminoguanidine ( AG) on AGE formation during uninterrupted glycation of ribonuclease A (RNase A) by ribose. Figure 14A Thiamine pyrophosphate (TPP); Figure 14B thiamine monophosphate (TP); Figure 14C thiamine (T); Figure 14D aminoguanidine

(AG).

Figure 15 is a series of graphs depicting the effect of vitamin B derivatives on AGE formation during uninterrupted glycation of bovine serum albumin (BSA) by ribose. Figure 15A Pyridoxamine (PM); Figure 15B pyridoxal-5 '-phosphate (PLP); Figure 15C pyridoxal (PL); Figure 15D pyridoxine (PN).

Figure 16 is a series of graphs depicting the effect of vitamin B] derivatives and aminoguanidine (AG) on AGE formation during uninterrupted glycation of bovine serum albumin (BSA) by ribose. Figure 16A Thiamine pyrophosphate (TPP); Figure 16B thiamine monophosphate (TP); Figure 16C thiamine (T); Figure 16D aminoguanidine

(AG).

Figure 17 is a series of graphs depicting the effect of vitamin B6 derivatives on AGE formation during uninterrupted glycation of human methemoglobin (Fib) by ribose. Figure 17A Pyridoxamine (PM); Figure 17B pyridoxal-5 '-phosphate (PLP); Figure 17C pyridoxal (PL); Figure 17D pyridoxine (PN).

Figure 18 is a series of graphs depicting the effect of vitamin B6 derivatives on post-Amadori AGE formation after interrupted glycation by ribose. Figure 18A BSA and Pyridoxamine (PM); Figure 18B BSA and pyridoxal-5 '-phosphate (PLP); Figure 18C BSA and pyridoxal (PL); Figure 18D RNase and pyridoxamine (PM).

Figure 19 are graphs depicting the effect of thiamine pyrophosphate on post- Amadori AGE formation after interrupted glycation by ribose. Figure 19A RNase, Figure 19B BSA.

Figure 20 are graphs depicting the effect of aminoguanidine on post-Amadori AGE formation after interrupted glycation by ribose. Figure 20A RNase, Figure 20B BSA.

Figure 21 is a graph depicting the effect of N^α-acetyl-L-lysine on post-Amadori AGE formation after interrupted glycation by ribose.

Figure 22 are bar graphs showing a comparison of post-Amadori inhibition of AGE formation by thiamine pyrophosphate (TPP), pyridoxamine (PM) and aminoguanidine (AG) after interrupted glycation of RNase (Figure 22A) and BSA (Figure 22B) by ribose. Figure 23 is a bar graph showing the effects of Ribose treatment in vivo alone on rat tail-cuff blood pressure. Treatment was with 0.05 M, 0.30 M, and 1 M Ribose (R) injected for 1, 2 or 8 Days (D).

Figure 24 is a bar graph showing the effects of Ribose treatment in vivo alone on rat creatinine clearance (Clearance per 100 g Body Weight). Treatment was with 0.05 M, 0.30 M, and 1 M Ribose (R) injected for 1, 2 or 8 Days (D).

Figure 25 is a bar graph showing the effects of Ribose treatment in vivo alone on rat Albuminuria (Albumin effusion rate). Treatment was with 0.30 M. and 1 M Ribose (R) injected for 1, 2 or 8 Days (D).

Figure 26 is a bar graph showing the effects of inhibitor treatment in vivo, with or without ribose, on rat tail-cuff blood pressure. Treatment groups were: 25 mg/100 g body weight aminoguanidine (AG); 25 or 250 mg/100 g body weight Pyridoxamine (P); 250 mg/100 g body weight Thiamine pyrophosphate (T), or with 1 M Ribose (R).

Figure 27 is a bar graph showing the effects of inhibitor treatment in vivo, with or without ribose, on rat creatinine clearance (Clearance per 100 g body weight). Treatment groups were: 25 mg/100 g body weight aminoguanidine (AG); 25 or 250 mg/100 g body weight Pyridoxamine (P); 250 mg 100 g body weight Thiamine pyrophosphate (T), or with 1 M Ribose (R). Figure 28 is a bar graph showing the effects of inhibitor treatment in vivo without ribose, and ribose alone on rat Albuminuria (Albumin effusion rate). Treatment groups were: 25 mg/100 g body weight aminoguanidine (AG); 250 mg/100 g body weight Pyridoxamine (P); 250 mg/100 g body weight Thiamine pyrophosphate (T), or treatment with 1 M Ribose (R) for 8 days (D). Control group had no treatment.

Figure 29 is a bar graph showing the effects of inhibitor treatment in vivo, with 1

M ribose , on rat Albuminuria (Albumin effusion rate). Treatment groups were: 25 mg/100 g body weight aminoguanidine (AG); 25 and 250 mg/100 g body weight

Pyridoxamine (P); 250 mg/100 g body weight Thiamine pyrophosphate (T), or treatment with 1 M Ribose (R) for 8 days (D) alone. Control group had no treatment.

Figure 30A depicts Scheme 1 showing a diagram of AGE formation from protein. Figure 30B depicts Scheme 2, a chemical structure of aminoguanidine. Figure 30C depicts Scheme 3, chemical structures for thiamine, thiamine-5 '-phosphate, and thiamine pyrophosphate. Figure 30D depicts Scheme 4, chemical structures of pyridoxine, pyridoxamine, pyridoxal-5 '-phosphate, and pyridoxal. Figure 30E depicts Scheme 5, kinetics representation of AGE formation. Figure 30F depicts Scheme 6, kinetics representation of AGE formation and intermediate formation.

Figure 31 shows a 125 MHz C-13 NMR Resonance spectrum of Riobonuclease Amadori Intermediate prepared by 24 HR reaction with 99% [2-C13]Ribose. Figure 32 are graphs which show AGE intermediary formation using the pentoses Xylose, Lyxose, Arabinose and Ribose.

Figure 33 is a graph showing the results of glomeruli staining at pH 2.5 with Alcian blue.

Figure 34 is a graph showing the results of glomeruli staining at pH 1.0 with Alcian blue.

Figure 35 is a graph showing the results of immunofluroescent glomeruli staining for RSA.

Figure 36 is a graph showing the results of immunofluroescent glomeruli staining for Heparan Sulfate Proteoglycan Core protein. Figure 37 is a graph showing the results of immunofluroescent glomeruli staining for Heparan Sulfate Proteoglycan side-chain.

Figure 38 is a graph showing the results of analysis of glomeruli sections for average glomerular volume

Figure 39 is a graph demonstrating AGE formation (BSA model) and AGE inhibition by pyπdoxamme in 4 25% DIANEAL® + IM glucose at pH 7 5 conducted at 37°C for 52 days

Figure 40 is a graph demonstrating AGE formation (myoglobin model) and AGE inhibition by pyπdoxamine in DIANEAL® post-dialysis fluid for 12 hours at 60°C.

Figure 41 is a graph demonstrating AGE formation on metmyoglobm in 4.25% DIANEAL® post-dialysis fluid at pH 7 5 conducted at 37° for varying time penods.

DETAILED DESCRIPTION

Animal Models for Protein Aging

Alloxan induced diabetic Lewis rats have been used as a model for protem agmg to demonstrate the in vivo effectiveness of inhibitors of AGE formation The correlation being demonstrated is between inhibition of late diabetes related pathology and effective inhibition of AGE formation (Brownlee, Cerami, and Vlassara, 1988, New Eng J. Med. 318(20):1315-1321) Streptozotocin induction of diabetes in Lewis rats, New Zealand White rabbits with induced diabetes, and genetically diabetic BB/Worcester rats have also been utilized, as descnbed in, for example, U.S. Patent 5,334,617 (incorporated by reference) A major problem with these model systems is the long time peπod required to demonstrate AGE related injury, and thus to test compounds for AGE inhibition. For example, 16 weeks of treatment was required for the rat studies descnbed in U S. Patent 5,334,617, and 12 weeks for the rabbit studies Thus it would be highly desirable and useful to have a model system for AGE related diabetic pathology that will manifest in a shorter time peπod, allowing for more efficient and expeditious determination of AGE related injury and the effectiveness of inhibitors of post-Amadon AGE formation

Antibodies to AGEs

An important tool for studying AGE formation is the use of polyclonal and monoclonal antibodies that are specific for AGEs elicited by the reaction of several sugars with a vaπety of target proteins. The antibodies are screened for resultant specificity for AGEs that is independent of the nature of the protein component of the AGE (Nakayama et al., 1989, Biochem. Biophys. Res. Comm. 162: 740-745; Nakayama et al., 1991, J. Immunol. Methods 140: 1 19-125; Horiuchi et al., 1991, J. Biol. Chem. 266: 7329-7332; Araki et al., 1992, J. Biol. Chem. 267: 1021 1-10214; Makita et al., 1992, J. Biol. Chem. 267: 5133-5138). Such antibodies have been used to monitor AGE formation in vivo and in vitro.

Thiamine - Vitamin Bl

The first member of the Vitamin B complex to be identified, thiamine is practically devoid of pharmacodynamic actions when given in usual therapeutic doses; and even large doses were not known to have any effects. Thiamine pyrophosphate is the physiologically active form of thiamine, and it functions mainly in carbohydrate metabolism as a coenzyme in the decarboxylation of -keto acids. Tablets of thiamine hydrochloride are available in amounts ranging from 5 to 500 mg each. Thiamine hydrochloride injection solutions are available which contain 100 to 200 mg/ml. For treating thiamine deficiency, intravenous doses of as high as 100 mg / L of parenteral fluid are commonly used, with the typical dose of 50 to 100 mg being administered. GI absorption of thiamine is believed to be limited to 8 to 15 mg per day, but may be exceed by oral administration in divided doses with food.

Repeated administration of glucose may precipitate thiamine deficiency in under nourished patients, and this has been noted during the correction of hyperglycemia.

Surprisingly, the instant invention has found, as shown by in vitro testing, that administration of thiamine pyrophosphate at levels above what is normally found in the human body or administered for dietary therapy, is an effective inhibitor of post- Amadori antigenic AGE formation, and that this inhibition is more complete than that possible by the administration of aminoguanidine. Pyridoxine - Vitamin B

Vitamin B6 is typically available in the form of pyridoxine hydrochloride in over-the-counter preparations available from many sources. For example Beach pharmaceuticals Beelith Tablets contain 25 mg of pyridoxine hydrochloride that is equivalent to 20 mg of Bg, other preparations include Marlyn Heath Care Marlyn

Formula 50 which contain 1 mg of pyridoxine HCI and Marlyn Formula 50 Mega Forte which contains 6 mg of pyridoxine HCI, Wyeth-Ayerst Stuart Prenatal® tablets which contain 2.6 mg pyπdoxme HCI, J&J-Merck Co . Stuart Formula® tablets contain 2 mg of pyndoxine HCI, and the CIBA Consumer Sunkist Children^'s chewable multivitamms which contain 1.05 mg of pyndoxine HCI, 150%o of the U.S. RDA for children 2 to 4 years of age, and 53% of the U.S RDA for children over 4 years of age and adults. (Physician's Desk Reference for nonprescnption drugs, 14th edition (Medical Economics Data Production Co., Montvale, N.J., 1993)

There are three related forms of pyndoxine, which differ in the nature of the substitution on the carbon atom in position 4 of the pyndine nucleus: pyndoxine is a pπmary alcohol, pyndoxal is the corresponding aldehyde, and pyndoxamme contains an ammomethyl group at this position Each of these three forms can be utilized by mammals after conversion by the liver into pyndoxal-5'-phosphate, the active form of the vitamin. It has long been believed that these three forms are equivalent in biological properties, and have been treated as equivalent forms of vitamin Bg by the art. The

Council on Pharmacy and Chemistry has assigned the name pyndoxine to the vitamin. The most active antimetabo te to pyndoxine is 4-deoxypyndoxιne, for which the antimetabohte activity has been attnbuted to the formation in vivo of 4-deoxypyndoxme- 5-phosphate, a competitive inhibitor of several pyndoxal phosphate-dependent enzymes. The pharmacological actions of pyndoxine are limited, as it elicits no outstanding pharmacodynamic actions after either oral or intravenous administration, and it has low acute toxicity, being water soluble. It has been suggested that neurotoxicity may develop after prolonged ingestion of as little as 200 mg of pyndoxine per day Physiologically, as a coenzyme, pyndoxine phosphate is involved in several metabolic transformations of amino acids including decarboxylation, transammation, and racemization. as well as in enzymatic steps in the metabolism of sulfur-contammg and hydroxy-amino acids. In the case of transammation, pyndoxal phosphate is ammated to pyndoxamme phosphate by the donor ammo acid, and the bound pyndoxamme phosphate is then deammated to pyndoxal phosphate by the acceptor α-keto acid. Thus vitamin B complex is known to be a necessary dietary supplement involved m specific breakdown of ammo acids. For a general review of the vitamin B complex see The Pharmacological Basis of Therapeutics. 8th edition, ed. Gilman, Rail, Nies, and Taylor (Pergamon Press, New York, 1990, pp. 1293-4; pp. 1523-1540).

Surpnsingly, the instant invention has discovered that effective dosages of the metabo cally transitory pyndoxal amme form of vitamin Bβ (pyndoxamme), at levels above what is normally found m the human body, is an effective inhibitor of post- Amadon antigemc AGE formation, and that this inhibition may be more complete than that possible by the administration of aminoguanidine

Formation of Stable Amadori/Schiff base Intermediary

The typical study of the reaction of a protein with glucose to form AGEs has been by ELISA using antibodies directed towards antigemc AGEs, and the detection of the production of an acid-stable fluorescent AGE, pentosidme, by HPLC following acid hydrolysis. Glycation of target proteins (1 e. BSA or RNase A) with glucose and nbose were compared by momtonng ELISA reactivity of polyclonal rabbit anti-Glucose- AGE - RNase and anti-Glucose-AGE-BSA antibodies The antigen was generated by reacting 1 M glucose with RNase for 60 days and BSA for 90 days The antibodies (R618 and R479) were screened and showed reactivity with only AGEs and not the protein, except for the earner immunogen BSA

Example 1

Thiamine Pyrophosphate and Pyridoxamine Inhibit the Formation of Antigenic Advanced Glycation End-Products from Glucose: Comparison with Aminoguanidine

Some B₍5 vitamers, especially pyndoxal phosphate (PLP), have been previously proposed to act as "competitive inhibitors" of early glycation, since as aldehydes they themselves can form Schiff bases adducts with protem ammo groups (Khatami et al., 1988, Life Sciences 43 1725-1731) and thus limit the amount of amines available for glucose attachment. However, effectiveness in limiting initial sugar attachment is not a predictor of inhibition of the conversion of any Amadon products formed to AGEs. The instant invention descnbes inhibitors of "late" glycation reactions as indicated by their effects on the in vitro formation of antigemc AGEs (Booth et al , 1996, Biochem Biophvs Res Com 220 113-119)

Chemicals

Bovine pancreatic ribonuclease A (RNase) was chromatographically pure, aggregate-free grade from Worthmgton Biochemicals. Bovine Serum albumin (BSA, fraction V. fatty-acid free), human methemoglobin (Hb), D-glucose. pyndoxine, pyndoxal, pyndoxal 5 'phosphate, pyndoxamme, thiamine, thiamine monophosphate, thiamme pyrophosphate, and goat alkaline phosphatase-conjugated anti-rabbit IgG were all from Sigma Chemicals. Aminoguanidine hydrochlonde was purchased from Aldnch Chemicals

Uninterrupted Glvcation with Glucose

Bovine serum albumin, ribonuclease A, and human hemoglobin were incubated with glucose at 37°C in 0 4 M sodium phosphate buffer of pH 7.5 contaimng 0.02% sodium azide The protein, glucose (at 1 0 M), and prospective inhibitors (at 0.5, 3, 15 and 50 mM) were introduced into the incubation mixture simultaneously. Solutions were kept m the dark in capped tubes Aliquots were taken and immediately frozen until analyzed by ELISA at the conclusion of the reaction. The incubations were for 3 weeks (Hb) or 6 weeks (RNase. BSA)

Preparation of polvclonal antibodies to AGE proteins

Immunogen preparation followed earlier protocols (Nakayama et al, 1989, Biochem Biophys Res Comm 162.740-745; Honuchi et al., 1991, J. Biol Chem. 266-7329-7332, Makita et al, 1992, J Biol Chem. 267.5133-5138). Briefly, immunogen was prepared by glycation of BSA (R479 antibodies) or RNase (R618 antibodies) at 1.6 g protem m 15 ml for 60-90 days using 1 5 M glucose in 0 4 M sodium phosphate buffer of pH 7 5 containing 0 05% sodium azide at pH 7 4 and 37°C New Zealand white rabbit males of 8-12 weeks were immunized by subcutaneous administration of a 1 ml solution containing 1 mg/ml of glycated protem in Freund's adjuvant. The pnmary injection used the complete adjuvant and three boosters were made at three week intervals with Freund's incomplete adjuvant. Rabbits were bled three weeks after the last booster. The serum was collected by centnfugation of clotted whole blood. The antibodies are AGE- specific, being unreactive with either native proteins (except for the earner) or with Amadoπ intermediates. The polyclonal anti-AGE antibodies have proven to be a sensitive and valuable analytical tool for the study of "late" AGE formation in vitro and in vivo. The nature of the dominant antigemc AGE epitope or hapten remains in doubt, although recently it has been proposed that the protem glycoxidation product carboxymethyl lysine (CmL) may be a dominant antigen of some antibodies (Reddy et al., 1995. Biochem. 34:10872-10878). Earlier studies have failed to reveal ELISA reactivity with model CmL compounds (Makita et al., 1992, J. Biol Chem. 267:5133- 5138).

ELISA detection of AGE products

The general method of Engvall (1981, Methods Enzymol 70:419-439) was used to perform the ELISA. Typically, glycated protein samples were diluted to approximately 1.5 ug/ml in 0.1 M sodium carbonate buffer of pH 9.5 to 9.7. The protem was coated overnight at room temperature onto 96-well polystyrene plates by pippettmg 200 ul of the protein solution m each well (0.3 ug/well). After coating, the protein was washed from the wells with a saline solution containing 0.05% Tween-20. The wells were then blocked with 200 ul of 1% casein in carbonate buffer for 2 h at 37°C followed by washing. Rabbit anti-AGE antibodies were diluted at a titer of about 1 :350 in incubation buffer, and incubated for 1 h at 37°C, followed by washing. In order to minimize background readings, antibodies R479 used to detect glycated RNase were raised against glycated BSA, and antibodies R618 used to detect glycated BSA and glycated Hb were raised against glycated RNase. An alkaline phosphatase-conjugated antibody to rabbit IgG was then added as the secondary antibody at a titer of 1 :2000 or 1 :2500 (depending on lot) and incubated for 1 h at 37°C, followed by washing. The p- nitrophenylphosphate substrate solution was then added (200 ul/well) to the plates, with the absorbance of the released p-nitrophenolate being monitored at 410 nm with a Dynatech MR 4000 microplate reader.

Controls containing unmodified protem were routinely included, and their readings were subtracted, the conections usually being negligible. The validity of the use of the ELISA method in quantitatively studying the kinetics of AGE formation depends on the hneanty of the assay (Kemeny & Challacombe, 1988, ELISA and Other Solid Phase Immunoassavs, John Wiley & Sons, Chichester, U.K.). Control expeπments were earned out, for example, demonstrating that the linear range for RNase is below a coating concentration of about 0.2-0.3 mg/well.

Results Figure 1 A-D are graphs which show the effect of vitamin Bβ deπvatives on post- Amadoπ AGE formation in bovme serum albumin glycated with glucose. BSA (10 mg/ml) was incubated with 1 0 M glucose in the presence and absence of the vanous indicated deπvative in 0 4 M sodium phosphate buffer of pH 7 5 at 37°C for 6 weeks. Aliquots were assayed by ELISA using R618 anti-AGE antibodies. Concentrations of the inhibitors were 3, 15 and 50 mM Inhibitors used in Figures (1A) Pyndoxamme (PM), ( IB) pyndoxal phosphate (PLP), (1C) pyndoxal (PL), (ID) pyndoxine (PN)

Figure 1 (control curves) demonstrates that reaction of BSA with 1 0 M glucose is slow and incomplete after 40 days, even at the high sugar concentration used to accelerate the reaction The simultaneous inclusion of different concentrations of vanous Bβ vitamers markedly affects the formation of antigemc AGEs (Figure 1A-D) Pyridoxamine and pyndoxal phosphate strongly suppressed antigemc AGE formation at even the lowest concentrations tested, while pyndoxal was effective above 15 mM. Pyndoxine was slightly effective at the highest concentrations tested Figure 2 A-D are graphs which show the effect of vitamin Bi denvatives and aminoguanidine (AG) on AGE formation in bovme serum albumin. BSA (10 mg/ml) was incubated with 1 0 M glucose in the presence and absence of the vanous indicated denvative in 0 4 M sodium phosphate buffer of pH 7 5 at 37°C for 6 weeks. Aliquots were assayed by ELISA using R618 anti-AGE antibodies Concentrations of the inhibitors were 3, 15 and 50 mM. Inhibitors used m Figures (2A) Thiamine pyrophosphate (TPP), (2B) thiamine monophosphate (TP), (2C) thiamme (T), (2D) aminoguanidine (AG)

Of the vanous Bl vitamers similarly tested (Figure 2 A-D), thiamme pyrophosphate was effective at all concentrations tested (Figure 2C), whereas thiamme and thiamine monophosphate were not inhibitory Most significantly it is remarkable to note the decrease in the final levels of AGEs formed observed with thiamine pyrophosphate, pyndoxal phosphate and pyndoxamme. Aminoguanidine (Figure 2D) produced some inhibition of AGE formation in BSA, but less so than the above compounds Similar studies were earned out with human methemaglobm and bovme ribonuclease A

Figure 3 A-D are graphs which show the effect of vitamin B6 denvatives on AGE formation in human methemoglobin. Hb (1 mg/ml) was incubated with 1.0 M glucose m the presence and absence of the vanous indicated denvative in 0.4 M sodium phosphate buffer of pH 7.5 at 37°C for 3 weeks. Aliquots were assayed by ELISA using R618 anti-AGE antibodies. Concentrations of the inhibitors were 0 5, 3, 15 and 50 mM. Inhibitors used in Figures (3A) Pyndoxamme (PM); (3B) pyndoxal phosphate (PLP); (3C) pyndoxal (PL), (3D) pyndoxine (PN)

It had been previously reported that Hb of a diabetic patient contains a component that binds to anti-AGE antibodies, and it was proposed that this glycated Hb (termed Hb-AGE, not to be confused with HbAlc) could be useful in measunng long- term exposure to glucose. The in vitro incubation of Hb with glucose produces antigemc AGEs at an apparently faster rate than observed with BSA. Nevertheless, the different B6 (Figure 3 A-D) and B] (Figure 4A-C) vitamers exhibited the same inhibition trends in Hb. with pyndoxamme and thiamme pyrophosphate being the most effective inhibitors in each of their respective families. Significantly, in Hb, aminoguanidine only inhibited the rate of AGE formation, and not the final levels of AGE formed (Figure 4D). With RNase the rate of antigemc AGE formation by glucose was intermediate between that of Hb and BSA, but the extent of inhibition within each vitamer senes was maintained. Again pyridoxamine and thiamine pyrophosphate were more effective that aminoguanidine (Figure 5).

Figure 4 A-D are graphs which show the effect of vitamin B i denvatives and aminoguanidine (AG) on AGE formation in human methemoglobin. Hb (1 mg/ml) was incubated with 1 0 M glucose in the presence and absence of the vanous indicated denvative in 0 4 M sodium phosphate buffer of pH 7.5 at 37°C for 3 weeks. Aliquots were assayed by ELISA using R618 anti-AGE antibodies. Concentrations of the inhibitors were 0.5, 3, 15 and 50 mM. Inhibitors used in Figures (4A) Thiamme pyrophosphate (TPP); (4B) thiamme monophosphate (TP); (4C) thiamme (T); (4D) aminoguanidine (AG).

Figure 5 is a bar graph which shows a compaπson of the inhibition of the glycation of nbonuclease A by thiamme pyrophosphate (TPP), pyndoxamme (PM) and aminoguanidine (AG). RNase (1 mg/ml) was incubated with 1.0 M glucose (glc) m the presence and absence of the vanous indicated denvative in 0 4 M sodium phosphate buffer of pH 7 5 at 37°C for 6 weeks. Aliquots were assayed by ELISA using R479 anti- AGE antibodies. The indicated percent inhibition was computed from ELISA readings in the absence and presence of the inhibitors at the 6 week time point. Concentrations of the inhibitors were 0.5, 3, 15 and 50 mM.

Discussion These results demonstrate that certain derivatives of B l and Bβ vitamins are capable of inhibiting "late" AGE formation. Some of these vitamers successfully inhibited the final levels of AGE produced, in contrast to aminoguanidine, suggesting that they have greater interactions with Amadori or post-Amadori precursors to antigenic AGEs. The Amadori and post-Amadori intermediates represent a crucial juncture where the "classical" pathway of nonenzymatic glycation begins to become essentially iπeversible (Scheme I). In earlier inhibition studies "glycation" was usually measured either as Schiff base formed (after reduction with labeled cyanoborohydride) or as Amadori product formed (after acid precipitation using labeled sugar). Such assays do not yield information on inhibition of post-Amadori conversion steps to "late" AGE products, since such steps lead to no change in the amount of labeled sugar that is attached to the proteins. Other "glycation" assays have relied on the sugar-induced increase of non-specific protein fluorescence, but this can also be induced by dicarbonyl oxidative fragments of free sugar, such as glycoaldehyde or glyoxal (Hunt et al., 1988, Biochem. 256:205-212), independently of Amadori product formation. In the case of pyridoxal (PL) and pyridoxal phosphate (PLP), the data support the simple mechanism of inhibition involving competitive Schiff-base condensation of these aldehydes with protein amino groups at glycation sites. Due to internal hemiacetal formation in pyridoxal but not pyridoxal phosphate, stronger inhibition of post-Amadori AGE formation by PLP is expected by this competitive mechanism. This indeed is observed in the data (Figure IB, IC, Figure 3B, 3C). The inhibition by pyridoxamine is necessarily different, as pyridoxamine lacks an aldehyde group. However, pyridoxamine is a candidate amine potentially capable of forming a Schiff-base linkage with the carbonyls of open-chain sugars, with dicarbonyl fragments, with Amadori products, or with post-Amadori intermediates. The mechanism of inhibition of B\ compounds is not obvious. All the forms contain an amino functionality, so that the marked efficiency of only the pyrophosphate form suggests an important requirement for a strong negative charge. A significant unexpected observation is that the extent of inhibition by aminoguanidine. and some of the other compounds, is considerably less at late stages of the reaction, than dunng the early initial phase This suggests a different mechanism of action than that of pyndoxamme and thiamme pyrophosphate, suggesting that the therapeutic potential of these compounds will be enhanced by co-administration with aminoguanidine

Example 2

Kinetics of Non-enzymatic glycation: Paradoxical Inhibition by Ribose and Facile Isolation of Protein Intermediate for Rapid Post-Amadori AGE Formation

While high concentrations of glucose are used to cause the non-enzymatic glycation of proteins, paradoxically, it was found that ribose at high concentrations is inhibitor.' to post-Amadori AGE formation in nbonuclease by acting on the post- Amadoπ "late" stages of the glycation reaction This unexpectedly inhibitory effect suggests that the "early" reactive intermediates, presumably Amadon products, can be accumulated with little formation of "late" post-Amadon AGE products (AGEs; Maillard products). Investigation into this phenomenon has demonstrated: (1) ability to define conditions for the kinetic isolation of Amadon (or post-Amadon) glycated ιntermedιate(s), (2) the ability study the fast kinetics of buildup of such an intermediate; (3) the ability to study the surpnsingly rapid kinetics of conversion of such intermediates to AGE products in the absence of free or reversώh bound sugar, (4) the ability to use these intermediates to study and charactenze inhibition of post-Amadon steps of AGE formation thus providing a novel system to investigate the mechanism of reaction and the efficacy of potential agents that could block AGE formation, and (5) with this knowledge it is also further possible to use ^C NMR to study the reactive intermediates and monitor their conversion to vanous candidate AGEs (Kha fah et al., 1996. Biochemistry 35(15) 4645-4654)

Chemicals and Materials As m Example 1 above

Preparation of polvclonal antibodies to AGEs As m Example 1 above ELISA detection of AGE products As in Example 1 above.

Amino Acid Analvsis Amino acid analyses were earned out at the Biotechnology Support Facility of the Kansas University Medical Center. Analyses were performed after hydrolysis of glycated protein (reduced with sodium cyanoborohydnde) with 6 N HCI at 1 10°C for 18- 24 h. Phenyl isothiocyanate was used for denvatization, and PTH denvatives were analyzed by reverse-phase HPLC on an Applied Biosystems amino acid analyzer (420A denvatizer, 130A separation system, 920A data analysis system).

Pentosidine Reverse-Phase HPLC Analvsis

Pentosidine production in RNase was quantitated by HPLC (Sell & Monnier, 1989, J Biol Chem 264:21597-21602; Odetti et al., 1992, Diabetes 41 : 153-159). Ribose-modified protem samples were hydrolyzed m 6 N HCI for 18 h at 100°C and then dπed in a Speed Vac. The samples were then redissolved, and aliquots were taken into 0.1% tnfluoroacetic acid and analyzed by HPLC on a Shimadzu system using a Vydac C-18 column equilibrated with 0 1 % TFA. A gradient of 0-6% acetomtπle (0.1% in TFA) was run in 30 min at a flow rate of about 1 ml/mm. Pentosidine was detected by 335 nm excιtatιon- 385 nm emission fluorescence, and its elution time was determined by running a synthesized standard. Due to the extremely small levels of pentosidine expected (Grandhee & Monnier, 1991, J Biol Chem 266: 11649-1 1653, Dyer et al, 1991 , J Biol Chem. 266: 1 1654-1 1660), no attempt was made to quantitate the absolute concentrations. Only relative concentrations were determined from peak areas.

Glvcation Modifications

Modification with ribose or glucose was generally done at 37°C in 0.4 M phosphate buffer of pH 7 5 containing 0.02%> sodium azide. The high buffer concentration was always used with 0.5 M nbose modifications. The solutions were kept in capped tubes and opened only to remove timed aliquots that were immediately frozen for later carrying out the vanous analyses. "Interrupted glycation" expenments were earned out by first incubating protein with the nbose at 37°C for 8 or 24 h, followed by immediate and extensive dialysis against frequent cold buffer changes at 4°C The samples were then reincubated by quickly warming to 37°C in the absence of external ribose Aliquots were taken and frozen at vanous intervals for later analysis Due to the low molecular weight of RNase, protein concentrations were remeasured after dialysis even when low molecular weight cut-off dialysis tubmg was used An alternative procedure was also devised (see below) in which interruption was achieved by simple 100-fold dilution from reaction mixtures containing 0 5 M nbose Protein concentrations were estimated from UV spectra The difference in molar extinction between the peak (278 nm) and trough (250 nm) was used for RNase concentration determinations in order to compensate for the general increase in UV absorbance that accompanies glycation. Time-dependent UV-difference spectral studies were earned out to charactenze the glycation contnbutions of the UV spectrum

Data Analvsis and Numerical Simulations of Kinetics Kinetic data were routinely fit to monoexponential or biexponential functions using nonlinear least-squares methods. The kinetic mechanisms of Schemes 5-6 have been examined by numencal simulations of the differential equations of the reaction. Both simulations and fitting to observed kinetics data were earned out with the SCIENTIST 2 0 software package (Micromath, Inc ) Determination of apparent half- times (Figure 6B) from kinetic data fit to two-exponential functions (Figure 6 A) was carried out with the "solve" function of MathCAD 4 0 software (MathSoft, Inc )

RESULTS

Comparison of Glvcation by Glucose and Ribose The reaction of RNase A with nbose and glucose has been followed pnmaπly with ELISA assays, using R479 rabbit AGE-specific antibodies developed against glucose-modified BSA To a lesser extent, the production of pentosidine, the only known acid-stable fluorescent AGE, was quantiated by HPLC following acid hydrolysis. Preliminary studies using 0 05 M ribose at 37°C showed that the rate of antigemc AGE formation appears to be modestly increased (roughly 2-3 fold as measured by the apparent half-time) as the pH is increased from 5 0 to 7 5, with an apparent small induction penod at the beginning of the kinetics m all cases The glycation of RNase with 0.05 M ribose at pH 7.5 (half-time near 6.5 days) appears to be almost an order of magnitude faster than that of glycation with 1.0 M glucose (half-time in excess of 30 days; see Figure 7B, solid line). The latter kinetics also displayed a small induction period but incomplete leveling off even after 60 days, making it difficult to estimate a true half-time.

When the dependence of the kinetics on ribose concentration was examined at pH 7.5, a most unexpected result was obtained. The rate of reaction initially increased with increasing ribose concentration, but at concentrations above 0.15 M the rate of reaction leveled off and then significantly decreased (Figure 6A). A plot of the dependence of the reciprocal half-time on the concentration of ribose (Figure 6B) shows that high ribose concentrations are paradoxically inhibitory to post-Amadori antigenic AGE formation. This unusual but consistent effect was found to be independent of changes in the concentration of either buffer (2-fold) or RNase (10-fold), and it was not changed by affinity purification of the R479 antibody on a column of immobilized AGE-RNase. It is also not due to effects of ribose on the ELISA assay itself. The measured inhibitory effect by ribose on post-Amadori AGE formation is not likely due to ribose interference with antibody recognition of the AGE antigenic sites on protein in the ELISA assay. Prior to the first contact with the primary anti-AGE antibody on the ELISA plates, glycated protein has been diluted over 1000-fold, washed extensively with Tween-20 after adsorption, and blocked with a 1 % casein coating followed by further washing with Tween-20.

Kinetics of Formation of post-Amadori Antigenic AGEs by "Interrupted Glycation "

In view of the small induction period seen, an attempt was made to determine whether there was some accumulation during the reaction, of an early precursor such as an Amadori intermediate, capable of producing the ELISA-detectable post-Amadori antigenic AGEs. RNase was glycated at pH 7.5 and 37°C with a high ribose concentration of 0.5 M, and the reaction was interrupted after 24 h by immediate cooling to 4°C and dialysis against several changes of cold buffer over a period of 24 h to remove free and reversibly bound (Schiff base) ribose. Such a ribose-free sample was then rapidly warmed to 37°C without re-adding any ribose, and was sampled for post- Amadori AGE formation over several days. The AGE antigen production of this 24 h "interrupted glycation" sample is shown by the dashed line and open triangles in Figure 7A, the time spent in the cold dialysis is not included. An uninterrupted control (solid line and filled circles) is also shown for companson. Dramatically different kinetics of post-Amadon antigenic AGE formation are evident in the two samples The kinetics of Λ.GE antigen production of the nbose-free interrupted sample now show (1) monoexponential kinetics with no induction penod, (2) a greatly enhanced rate of antigemc AGE formation, with remarkable half-times of the order of 10 h, and (3) production of levels of antigen comparable to those seen in long incubations m the continued presence of πbose (see Figure 6A). Equally significant, the data also demonstrate that negligible AGE antigen was formed dunng the cold dialysis penod, as shown by the small difference between the open tnangle and filled circle points at time 1 day in Figure 7A Very little, if any, AGE was formed by the "interruption" procedure itself These observations show that a fully competent isolatable intermediate or precursor to antigemc AGE has been generated dunng the 24 h contact with nbose pπor to the removal of the free and reversibly bound sugar.

Samples interrupted after only 8 h produced a final amount of AGE antigen that was about 72% of the 24 h interrupted sample. Samples of RNase glycated with only 0 05 M πbose and interrupted at 8 h by cold dialysis and remcubation at 37°C revealed less than 5% production of ELISA-reactive antigen after 9 days Interruption at 24 h, however, produced a rapid rise of ELISA antigen (similar to Figure 7A) to a level roughly 50% of that produced in the uninterrupted presence of 0 05 M ribose

The same general interruption effects were also seen with other proteins (BSA and Hemoglobin) Except for a somewhat different absolute value of the rate constants, and the amount of antigenic AGEs formed dunng the 24 h 0 5 M ribose incubation, the same dramatic increase in the rate of AGE antigen formation was observed after removal of O 5 M πbose

Glycation is much slower with glucose than with πbose (note the difference in time scales between Figure 7A and Figure 7B). However, unlike the case with πbose, interruption after 3 days of glycation by 1 0 M glucose produced negligible buildup of precursor to ELISA-reactive AGE antigens (Figure 7B, dashed curve).

Kinetics of Pentosidine Formation The samples subjected to ELISA testing were also assayed for the production of pentosidine, an acid-stable AGE. The content of pentosidine was measured for the same RNase samples analyzed for antibody reactivity by ELISA Glycation by nbose m 0 4 M phosphate buffer at pH 7 5 produced pentosidine in RNase A that was quantitated by fluroescence after acid hydrolysis Figure 8A shows that under uninterrupted conditions, 0 05 M πbose produces a progressive increase in pentosidine. However, when glycation is earned out under "interrupted" conditions using 0 5 M πbose, a dramatic increase m the rate of pentosidine formation is seen immediately after removal of excess nbose (Figure 8B), which is similar to, but slightly more rapid than, the kinetics of the appearance of antigemc AGEs (Figure 7A). A greater amount of pentosidine was also produced with 24 h interruption as compared with 8 h Reproducible differences between the kinetics of formation of pentosidine and antigenic AGEs can also be noted A significant amount of pentosidine is formed during the 24 h incubation and also dunng the cold dialysis, resulting in a jump of the dashed vertical line in Figure 8B. Our observations thus demonstrate that a pentosidine precursor accumulates dunng nbose glycation that can rapidly produce pentosidine after nbose removal (cf. Odetti et al., 1992, Diabetes 41.153-159)

Rate of Buildup of the Reactive Intermedιate(s) The "interrupted glycation" expenments descnbed above demonstrate that a precuisor or precursors to both post-Amadon antigemc AGEs and pentosidine can be accumulated dunng glycation with πbose. The kinetics of formation of this intermediate can be independently followed and quantitated by a vaπation of the expenments described above The amount of intermediate generated in RNase at different contact times with ribose can be assayed by the maximal extent to which it can produce antigemc AGE after interruption. At vanable times after initiating glycation, the free and reversibly-bound πbose is removed by dialysis in the cold or by rapid dilution (see below) Sufficient time (5 days, which represents several half-lives according to Figure 7A) is then allowed after warming to 37°C for maximal development of post-Amadon antigenic AGEs. The ELISA readings 5 days after each interruption point, representing maximal AGE development, would then be proportional to the intermediate concentration present at the time of interruption. Figure 9 shows such an expenment where the kinetics of intermediate buildup are measured for -RNase A in the presence of 0.5 M πbose (solid symbols and curve). For comparison, the amount of AGE present before nbose removal at each interruption point is also shown (open symbols and dashed lines). As expected (cf. Figure 7A), little AGE is formed pnor to removal (or dilution) of nbose, so that ELISA readings after the 5 day secondary incubation penods are mostly a measure of AGE formed after πbose removal. The results in Figure 9 show that the rate of buildup of intermediate in 0 5 M πbose is exponential and very fast, with a half-time of about 3.3 h. This is about 3-fold more rapid than the observed rate of conversion of the intermediate to antigemc AGEs after interruption (open symbols and dashed curve Figure 7A).

In these expenments the removal of nbose at each interruption time was achieved by 100-fold dilution, and not by dialysis. Simple dilution reduced the concentration of ribose from 0.05 M to 0.005 M. It was independently determined (Figure 6A) that little AGE is produced in this time scale with the residual 5 mM πbose. This dilution approach was pnmanly dictated by the need for quantitative point-to-point accuracy Such accuracy would not have been achieved by the dialysis procedure that would be earned out independently for each sample at each interruption point. Our results show that dilution was equivalent to dialysis.

A separate control expenment (see Figure 10 below) demonstrated that the instantaneous 100-fold dilution gave nearly identical results to the dialysis procedure. These control expenments demonstrate that the reversible πbose-protein binding (Schiff base) equihbnum is quite rapid on this time scale. This is consistent with data of Bunn and Higgins (1981, Science 213: 222-224) that indicated that the half-time of Schiff base formation with 0.5 M nbose should be on the order of a few minutes. The 100-fold rapid dilution method to reduce nbose is a valid method where quantitative accuracy is essential and cannot be achieved by multiple dialysis of many samples.

Direct Inhibition of Post-Amadori AGE Formation from the Intermediate bv Ribose and Glucose

The increase m the rate of AGE formation after interruption and sugar dilution suggests, but does not prove, that high concentrations of ribose are inhibiting the reaction at or beyond the first "stable" intermediate, presumably the -Amadon denvative (boxed m Scheme I) A test of this was then earned out by studying the effect of directly adding ribose, on the post-Amadon reaction. RNase was first incubated for 24 h in 0 5 M nbose in order to prepare the intermediate. Two protocols were then earned out to measure possible inhibition of the post-Amadon formation of antigemc AGEs by different concentrations of πbose. In the first expenment, the 24 h nbated sample was simply diluted 100-fold into solutions containing varying final concentrations of nbose ranging from 0 005 M to 0 505 M (Figure 10A). The rate and extent of AGE formation are clearly seen to be diminished by increasing πbose concentrations. Significantly, up to the highest concentration of 0.5 M ribose, the kinetics appear exponential and do not show the induction penod that occurs with uninterrupted glycation (Figures 6 A and 7 A) in high ribose concentrations.

A second expenment (Figure 10B) was also conducted in which the 24 h interrupted sample was extensively dialyzed in the cold to release free and reversibly bound ribose as well as any inhibitory products that may have formed duπng the 24 h incubation with nbose Following this, aliquots were diluted 100-fold into varying concentrations of freshly made nbose. and the formation of antigemc AGE products was monitored as above. There results were nearly identical to the expenment of Figure 10A where the dialysis step was omitted. In both cases, the rate and extent of AGE formation were diminished by increasing concentrations of nbose, and the kinetics appeared exponential with no induction period

The question of whether glucose or other sugars can also inhibit the formation of AGEs from the reactive intermediate obtained by interrupted glycation in 0 5 M πbose was also investigated. The effects of glucose at concentrations of 1.0-2.0 M were tested (data not shown). Glucose was clearly not as inhibitory as πbose. When the 24 h ribose interrupted sample was diluted 100-fold into these glucose solutions, the amount of antigenic AGE formed was diminished by about 30%, but there was little decrease in the apparent rate constant. Again, the kinetics appeared exponential.

Effect ofpH on Post-Amadon Kinetics of AGE Formation The interrupted glycation method was used to investigate the pH dependence of the post-Amadon kinetics of AGE formation from the reactive intermediate. In these expenments, RNase A was first reacted for 24 h with 0 5 M nbose at pH 7 5 to generate the reactive intermediate. The kinetics of the decay of the intermediate to AGEs were then measured by ELISA. Figure 1 1 shows that an extremely wide pH range of 5.0-9 5 was achievable when the kinetics were measured by initial rates. A remarkable bell- shaped dependence was observed, showing that the kinetics of antigenic AGEs formation are decreased at both acidic and alkaline pH ranges, with an optimum near pH 8.

A single "pH jump" experiment was also earned out on the pH 5 0 sample studied above which had the slowest rate of antigenic AGE formation. After 12 days at 37°C in pH 5 0 buffer, the pH was adjusted quickly to 7.5, and antigemc AGE formation was monitored by ELISA Within expenmental eπor, the sample showed identical kinetics (same first order rate constant) of AGE formation to interrupted glycation samples that had been studied directly at pH 7 5 (Figure 12). In this expenment, the relative amounts of antigemc AGE could not be directly compared on the same ELISA plate, but the pH-jumped sample appeared to have formed substantial though somehow diminished levels of antigenic AGEs. These results demonstrate that intermediate can be prepared free of AGE and stored at pH 5 for later studies of conversion to AGEs.

Inhibition of Post-Amadori AGE formation by Aminoguanidine

The efficacy of aminoguanidine was tested by this interrupted glycation method, l e., by testing its effect on post-Amadon formation of antigenic AGEs after removal of excess and reversibly bound ribose. Figure 20A demonstrates that aminoguanidine has modest effects on blocking the formation of antigemc AGEs in RNase under these conditions, with little inhibition below 50 mM Approximately 50% inhibition is achieved only at or above 100-250 mM. Note again that in these expenments, the protein was exposed to aminoguanidine only after interruption and removal of free and reversibly bound ribose. Comparable results were also obtained with the interrupted glycation of BSA (Figure 20B)

Amino acid analvsis of Interrupted Glvcation Samples

Amino acid analysis was earned out on RNase after 24 h contact with 0.5 M ribose (undialyzed), after extensive dialysis of the 24 h glycated sample, and after 5 days of incubation of the latter sample at 37°C. These determinations were made after sodium cyanoborohydnde reduction, which reduces Schiff base present on lysines or the terminal amino group. All three samples, normalized to alanine (12 residues), showed the same residual lysine content (4.0 ± 0.5 out of the original 10 in RNase). This indicates that after 24 h contact with 0.5 M ribose, most of the formed Schiff base adducts had been converted to Amadori or subsequent products. No arginine or histidine residues were lost by modification.

Discussion

The use of rapidly reacting ribose and the discovery of its reversible inhibition of post-Amadori steps have permitted the dissection and determination of the kinetics of different steps of protein glycation in RNase. Most previous kinetic studies of protein "glycation" have actually been restricted to the "early" steps of Schiff base formation and subsequent Amadori rearrangement. Some kinetic studies have been carried out starting with synthesized fructosylamines, i.e. small model Amadori compounds of glucose (Smith and Thornalley, 1992, Eur. J. Biochem. 210:729-739, and references cited therein), but such studies, with few exceptions, have hitherto not been possible with proteins. One notable exception is the demonstration by Monnier (Odetti et al., 1992, supra) that BSA partially glycated with ribose can rapidly produce pentosidine after ribose removal. The greater reactivity of ribose has also proven a distinct advantage in quantitatively defining the time course of AGE formation. It is noted that glucose and ribose are both capable of producing similar AGE products, such as pentosidine

(Grandhee & Monnier, 1991, supra; Dyer et al. 1991, supra), and some ^C NMR model compound work has been done with ADP-ribose (Cervantes-Laurean et al., 1993, Biochemistry 32:1528-1534). The present work shows that antigenic AGE products of ribose fully cross-react with anti-AGE antibodies directed against glucose-modified proteins, suggesting that ribose and glucose produce similar antigenic AGEs. The primary kinetic differences observed between these two sugars are probably due to relative differences in the rate constants of steps leading to post-Amadori AGE formation, rather than in the mechanism.

The results presented reveal a marked and paradoxical inhibition of overall AGE formation by high concentrations of ribose (Figure 6) that has not been anticipated by earlier studies. This inhibition is rapidly reversible in the sense that it is removed by dialysis of initially modified protein (Figure 7A) or by simple 100-fold dilution (as used in Figure 11 ) The expenments of Figure 10 demonstrate that it is not due to the accumulation of dialyzable inhibitory intermediates duπng the initial glycation, since dialysis of 24 h modified protem followed by addition of different concentrations of fresh ribose induces the same inhibition The data of Figure 10A,B show that the inhibition occurs by reversible and rapid interaction of ribose with protem intermediate containing reactive Amadoπ products The inhibition is unlikely to apply to the early step of formation of Amadoπ product due to the rapid rate of formation of the presumed A-madoπ intermediate that was determined in the expenment of Figure 9 The identification of the reactive intermediate as an Amadon product is well supported by the amino acid analysis earned out (after sodium cyanoborohydrate reduction) before and after dialysis at the 24 h interruption point The unchanged residual lysme content indicates that any dischageable Schiff bases have already been lπeversibly converted (presumably bv Amadori rearrangement) by the 24 h time

The secondary ribose suppression of "late" but not "early" glycation steps significantly enhances the accumulation of a fully-competent reactive Amadoπ intermediate containing little AGE. Its isolation by the interruption procedure is of importance for kinetic and structural studies, since it allows one to make studies m the absence of free or Schiff base bound sugar and their attendant reactions and complications For example, the post-Amadon conversion rates to antigenic AGE and pentosidine AGE products have been measured (Figure 7A, open symbols, and Figure 8B), and demonstrated to be much faster (t 1/2 - 10 h) than reflected m the overall kinetics under uninterrupted conditions (Figure 6A and Figure 8A) The rapid formation of pentosidine that was measured appears consistent with an earlier interrupted πbose expenment on BSA by Odetti et al. (1992, supra). Since nbose and denvatives such as ADP-πbose are normal metabolites, the very high rates of AGE formation seen here suggest that they should be considered more senously as sources of potential glycation in vanous cellular compartments (Cervantes-Laurean et al, 1993, supra), even though their concentrations are well below those of the less reactive glucose

Another new application of the isolation of intermediate is in studying the pH dependence of this complex reaction. The unusual bell-shaped pH profile seen for the post-Amadon AGE formation (Figure 11) is in stπkmg contrast to the mild pH dependence of the overall reaction The latter kinetics reflect a composite effect of pH on all steps in the reaction, including Schiff base and Amadoπ product formation, each of which may have a unique pH dependence. This complexity is especially well illustrated by studies of hemoglobin glycation (Lowery et al., 1985, J. Biol. Chem. 260: 1 161 1- 1 1618). The bell-shaped pH profile suggests, but does not prove, the involvement of two ionizing groups. If true, the data may imply the participation of a second amino group, such as from a neighboring lysine, in the formation of dominant antigenic AGEs. The observed pH profile and the pH-jump observations described suggest that a useful route to isolating and maintaining the reactive intermediate would be by the rapid lowering of the pH to near 5.0 after 24 h interruption. The kinetic studies provide new insights into the mechanisms of action of aminoguanidine (guanylhydrazine), an AGE inhibitor proposed by Cerami and co- workers to combine with Amadori intermediates (Brownlee et al., 1986, supra). This proposed pharmacological agent is now in Phase III clinical trials for possible therapeutic effects in treating diabetes (Vlassara et al., 1994, supra). However, its mechanism of AGE inhibition is likely to be quite complex, since it is multifunctional. As a nucelophilic hydrazine, it can reversibly add to active carbonyls, including aldehydo carbonyls of open-chain glucose and ribose (Khatami et al., 1988, Life Sci. 43:1725-1731 ; Hirsch et al., 1995, Carbohyd. Res. 267:17-25), as well as keto carbonyls of Amadori compounds. It is also a guanidinium compound that can scavange highly reactive dicarbonyl glycation intermediates such as glyoxal and glucosones (Chen & Cerami, 1993, J. Carbohyd. Chem. 12:731-742; Hirsch et al., 1992, Carbohyd. Res. 232: 125-130; Ou & Wolff, 1993, Biochem. Pharmacol. 46: 1139-1 144). The interrupted glycation method allowed examination of aminoguanidine efficacy on only post- Amadori steps of AGE formation. Equally important, it allowed studies in the absence of free sugar or dicarbonyl-reactive fragments from free sugar (Wolff & Dean, 1987, Biochem. J. 245:243-250; Wells-Knecht et al., 1995, Biochemistry 34:3702-3709) that can combine with aminoguanidine. The results of Figure 20 demonstrate that aminoguanidine has. at best, only a modest effect on post-Amadori AGE formation reactions, achieving 50% inhibition at concentrations above 100-250 mM. The efficacy of aminoguanidine thus predominantly arises either from inhibiting early steps of glycation (Schiff base foiτnation) or from scavenging highly reactive dicarbonyls generated during glycation. Contrary to the original claims, it does not appear to inhibit AGE formation by complexing the Amadon intermediate.

The use of interrupted glycation is not limited for kinetic studies. Interrupted glycation can simplify structural studies of glycated proteins and identifying unknown

AGEs using techniques such as ^C NMR that has been used to detect Amadon adducts i of RNase (Negha et al., 1983, J Biol Chem 258 14279-14283, 1985, J Biol Chem 260 5406-5410) The combined use of structural and kinetic approaches should also be of special interest. For example, although the identity of the dominant antigenic AGEs reacting with the polyclonal antibodies remains uncertain, candidate AGEs, such as the recently proposed (carboxymethyl)lysιne (Reddy et al., 1995, Biochemistry 34:10872- 0 10878, cf. Makita et al., 1992. J Biol Chem. 267 5133-5138) should display the same kinetics of formation from the reactive intermediate that we have observed. The availability of the interrupted kinetics approach will also help to determine the importance of the Amadoπ pathway to the formation of this AGE. Similarly, momtoπng of the interrupted glycation reaction by techniques such as ^C NMR should identify 5 resonances of other candidate antigenic AGEs as being those displaying similar kinetics of appearance. Table I lists the ^C NMR peaks of the -Amadon intermediate of RNase prepared by reaction with C-2 enriched nbose. The downfield peak near 205 ppm is probably due to the carbonyl of the Amadoπ product In all cases, the ability to remove excess free and Schiff base sugars through interrupted glycation will considerably 0 simplify isolation, identification, and structural characterization.

Table I lists the peaks that were assigned to the Post-Amadon Intermediate due to their invariant or decreasing intensity with time Peak positions are listed in ppm downfield from TMS.

Table I 125MHz C-13 NMR Resonances of Ribonuclease Amadori 5 Intermediate Prepared by 24 HR Reaction with 99% [2-C13]Ribose

216.5 ppm 108.5 ppm

211.7 105.9

208 103.9

103 0 172 95.8

165

163.9 73.65

162.1 70.2

69 7 Ribonuclease A was reacted for 24 hr with 0.5 M πbose 99% ennched at C-2, following which excess and Schiff base bound nbose was removed by extensive dialysis m the cold The sample was then wanned back to 37°C immediately before taking a 2 hr NMR scan. The signals from RNase reacted with natural abundance nbose under identical conditions were then subtracted from the NMR spectrum. Thus all peaks shown are due to ennched C-13 that onginated at the C-2 position. Some of the peaks aπse from degradation products of the intermediate, and these can be identified by the increase in the peak intensity over time. Figure 31 shows the NMR spectrum obtained.

Example 3

In Vitro Inhibition of the Formation of Antigenic Advanced Glycation End- Products (AGEs) by Derivatives of Vitamins Bl and B6 and Aminoguanidine. Inhibition of Post-Amadori Kinetics Differs from that of Overall Glvcation

The interrupted glycation method for following post-Amadon kinetics of AGE formation allows for the rapid quantitative study of "late" stages of the glycation reaction. Importantly, this method allows for inhibition studies that are free of pathways of AGE formation which anse from glycoxidative products of free sugar or Schiff base (Namiki pathway) as illustrated in Scheme I Thus the interrupted glycation method allows for the rapid and unique identification and charactenzation of inhibitors of "late" stages of glycation which lead to antigemc AGE formation.

Among the vitamin B\ and Bβ derivatives examined, pyndoxamme and thiamme pyrophosphate are unique inhibitors of the post-Amadon pathway of AGE formation. Importantly, it was found that efficacy of inhibition of overall glycation of protein, in the presence of high concentrations of sugar, is not predictive of the ability to inhibit the post-Amadon steps of AGE formation where free sugar is removed. Thus while pyndoxamme, thiamine pyrophosphate and aminoguanidine are potent inhibitors of AGE formation in the overall glycation of protein by glucose, aminoguanidine differs from the other two in that it is not an effective inhibitor of post-Amadon AGE formation. Aminoguanidine markedly slows the initial rate of AGE formation by nbose under uninterrupted conditions, but has no effect on the final levels of antigemc AGEs produced. Examination of different proteins (RNase, BSA and hemoglobin), confirmed that the inhibition results are generally non-specific as to the protem used, even though there are individual vanations in the rates of AGE formation and inhibition

Chemicals and Materials As in Example 1 above

Prepai ation of polvclonal antibodies to AGEs Λ-s in Example 1 above

ELISA detection of AGE products As in Example 1 above

Uninterrupted ribose glvcatwn assavs

Bovine serum albumin, ribonuclease A, and human hemoglobin were incubated with ribose at 37°C m 0 4 M sodium phosphate buffer of pH 7 5 containing 0 02% sodium azide The protein (10 mg/ml or 1 mg/ml), 0 05 M ribose. and prospective inhibitors (at 0 5, 3, 15 and 50 mM) were introduced into the incubation mixture simultaneously Solutions were kept in the dark in capped tubes Aliquots were taken and immediately frozen until analyzed by ELISA at the conclusion of the reaction. The incubations were for 3 weeks (Hb) or 6 weeks (RNase, BSA) Glycation reactions were monitored for constant pH throughout the duration of the expenments

Intel / upted (post-Amadori) ribose glycation assays

Glycation was first earned out by incubating protem (10 mg/ml) with 0 5 M ribose at 37°C in 0 4 M phosphate buffer at pH 7 5 containing 0 2% sodium azide for 24 h in the absence of inhibitors Glycation was then interrupted to remove excess and reversibly bound (Schiff base) sugar by extensive dialysis against frequent cold buffer changes at 4°C The glycated intermediate samples containing maximal amount of Amadori product and little AGE (depending on protem) were then quickly warmed to 37°C without re-addition of πbose This initiated conversion of Amadon intermediates to AGE products in the absence or presence of vanous concentrations (typically 3, 15 and 50 mM) of prospective inhibitors Aliquots were taken and frozen at various intervals for later analysis The solutions were kept m capped tubes and opened only to remove timed aliquots that were immediately frozen for later carrying out the vanous analyses

Numerical Analvsis of kinetics data Kinetics data (time progress curves) was routinely fit to mono- or bi-exponential functions using non-linear least squares methods utilizing either SCIENTIST 2.0 (MicroMath, Inc.) or ORIGIN (Microcal, Inc ) software that permit user-defined functions and control of parameters to iterate on. Standard deviations of the parameters of the fitted functions (initial and final ordinate values and rate constants) were returned as measures of the precision of the fits Apparent half-times for bi-exponential kinetics fits were determined with the "solve" function of MathCad software (MathSoft, Inc.).

RESULTS

Inhibition b, vitamin Bfi derivatives of the overall kinetics of AGE formation from

Ribose The inhibitory effects of vitamin B] and Bβ denvatives on the kinetics of antigemc AGE formation were evaluated by polyclonal antibodies specific for AGEs. Initial inhibition studies were earned out on the glycation of bovine nbonuclease A (RNase) in the continuous presence of 0 05 M ribose, which is the concentration of ribose where the rate of AGE formation is near maximal Figure 13 (control curves, filled rectangles) demonstrates that the formation of antigenic AGEs on RNase when incubated with 0 05 M πbose is relatively rapid, with a half-time of approximately 6 days under these conditions. Pyndoxal-5 '-phosphate (Figure 13B) and pyndoxal (Figure 13C) significantly inhibited the rate of AGE formation on RNase at concentrations of 50 mM and 15 mM. Surpnsmgly, pyndoxine, the alcohol form of vitamin Bβ, also moderately inhibited AGE formation on RNase (Figure 13D). Of the B6 denvatives examined, pyndoxamme at 50 mM was the best inhibitor of the "final" levels of AGE formed on RNase over the 6-week time penod monitored (Figure 13 A).

Inhibition bv vitamin Bl derivatives of the overall kinetics of AGE formation from Ribose

All of the B] vitamers inhibited antigenic AGE formation on RNase at high concentrations, but the inhibition appeared more complex than for the Bβ denvatives (Figure 14A-C). In the case of thiamme pyrophosphate as the inhibitor (Figure 14A), both the rate of AGE formation and the final levels of AGE produced at the plateau appeared diminished. In the case of thiamme phosphate as the inhibitor (Figure 14B), and thiamine (Figure 14C), there appeared to be little effect on the rate of AGE formation, but a substantial decrease in the final level of AGE formed in the presence of the highest concentration of inhibitor. In general, thiamme pyrophosphate demonstrated greater inhibition than the other two compounds, at the lower concentrations examined.

Inhibition bv aminoguanidine of the overall kinetics of AGE formation from Ribose

Inhibition of AGE formation by aminoguanidine (Figure 14D) was distinctly different from that seen in the B \ and Bβ expenments Increasing concentration of aminoguanidine decreased the rate of AGE formation on RNase, but did not reduce the final level of AGE formed The final level of AGE formed after the 6-weeks was nearly identical to that of the control for all tested concentrations of aminoguanidine

Inhibition of the overall kinetics of AGE formation in serum albumin and hemoglobin from Ribose

Comparative studies were earned out with BSA and human methemoglobin (Hb) to determine whether the observed inhibition was protem-specific The different denvatives of vitamin B6 (Figure 15A-D) and vitamin Bi (Figure 16A-C) exhibited similar inhibition trends when incubated with BSA as with RNase, pyridoxamine and thiamine pyrophosphate being the most effective inhibitors or each family Pyndoxine failed to inhibit AGE formation on BSA (Figure 15D) Pyndoxal phosphate and pyndoxal (Figure 15B-C) mostly inhibited the rate of AGE formation, but not the final levels of AGE formed Pyndoxamme (Figure 15A) exhibited some inhibition at lower concentrations, and at the highest concentration tested appeared to inhibit the final levels of AGE formed more effectively than any of the other Bβ denvatives In the case of Bi derivatives, the overall extent of inhibition of AGE formation with BSA (Figure 16A-C), was less than that observed with RNase (Figure 14A-C). Higher concentrations of thiamine and thiamine pyrophosphate inhibited the final levels of AGEs formed, without greatly affecting the rate of AGE formation (Figure 16C) Aminoguanidine again displayed the same inhibition effects with BSA as seen with RNase (Figure 16D), appearing to slow the rate of AGE formation without significantly affecting the final levels of AGE formed The kinetics of AGE formation was also examined using Hb in the presence of the Bβ and B ] vitamin denvatives, and aminoguanidine. The apparent absolute rates of

AGE formation were significantly higher with Hb than with either RNase or BSA However, the tested inhibitors showed essentially similar behavior The effects of the vitamin Bβ denvatives are shown in Figure 17 Pyndoxamme showed the greatest inhibition at concentrations of 3 mM and above (Figure 17 A), and was most effective when compared to pyndoxal phosphate (Figure 17B), pyndoxal (Figure 17C), and pyndoxine (Figure 17D) In the case of the Bj vitamin denvatives (data not shown), the inhibitory effects were more similar to the BSA inhibition trends than to RNase The inhibition was only modest at the highest concentrations tested (50 mM), being nearly 30-50%) for all three B \ denvatives The pπmary manifestation of inhibition was in the reduction of the final levels of AGE formed Inhibition bv vitamin B derivatives of the kinetics of post-Amadori ribose AGE formation

Using the interrupted glycation model to assay for inhibition of the "late" post- Amadon AGE formation, kinetics were examined by incubating isolated Amadoπ intermediates of either RNase or BSA at 37°C in the absence of free or reversibly bound nbose Ribose sugar that was initially used to prepare the intermediates was removed by cold dialysis after an initial glycation reaction period of 24 h. After AGE formation is allowed to resume, formation of AGE is quite rapid (half-times of about 10 h) in the absence of any inhibitors Figure 18 shows the effects of pyndoxamme (Figure 18A), pyndoxal phosphate (Figure 18B), and pyndoxal (Figure 18C) on the post-Amadon kinetics of BSA Pyndoxine did not produce any inhibition (data not shown) Similar experiments were earned out on RNase. Pyndoxamme caused nearly complete inhibition of AGE formation with RNase at 15 mM and 50 mM (Figure 18D) Pyndoxal did not show any significant inhibition at 15 mM (the highest tested), but pyndoxal phosphate showed significant inhibition at 15 mM Pyndoxal phosphate is known to be able to affinity label the active site of RNase (Raetz and Auld, 1972, Bιochemιstn> 11.2229- 2236)

With BSA, unlike RNase, a significant amount of antigemc AGE formed duπng the 24 h initial incubation with RNase (25-30%), as evidenced by the higher ELISA readings after removal of ribose at time zero for Figures 18A-C. For both BSA and RNase, the inhibition, when seen, appears to manifest as a decrease in the final levels of AGE formed rather than as a decrease in the rate of formation of AGE

Inhibition by vitamin B] derivatives of the kinetics of post-Amadori nbose AGE jormatwn

Thiamme pyrophosphate inhibited AGE formation more effectively than the other B \ denvatives with both RNase and BSA (Figure 19) Thiamme showed no effect, while thiamme phosphate showed some intermediate effect. As with the B6 assays, the post-Amadon inhibition was most apparently manifested as a decrease in the final levels of AGE formed

Effects of aminoguanidine and N°--acetyl-L-lvsιne on the kinetics of post-Amadori ribose AGE formation

Figure 20 shows the results of testing aminoguanidine for inhibition of post- Amadon AGE formation kinetics with both BSA and RNase. At 50 mM, inhibition was about 20%) in the case of BSA (Figure 20B). and less than 15% with RNase (Figure 20A) The possibility of inhibition by simple ammo-contaimng functionalities was also tested using N^α-acetyl-L-lysme (Figure 21), which contains onlv a free N^α-amιno group N^α-acetyl-L-lysιne at up to 50 mM failed to exhibit any significant inhibition of AGE formation

-* Discussion

Numerous studies have demonstrated that aminoguanidine is an apparently potent inhibitor of many manifestations of nonenzymatic glycation (Brownlee et al , 1986; Brownlee, 1992,1994, 1995) The inhibitory effects of aminoguanidine on vanous phenomena that are induced by reducing sugars are widely considered as proof of the 0 involvement of glycation in many such phenomena. Aminoguanidine has recently entered into a second round of Phase III clinical tnals (as pimagedme) for ameliorating the complications of diabetes thought to be caused by glycation of connective tissue proteins due to high levels of sugar

Data from the kinetic study of uninterrupted "slow" AGE formation with RNase ^ induced by glucose (Example 1 ) confirmed that aminoguanidine is an effective inhibitor, and further identified a number of denvatives of vitamins Bi and B6 as equally or slightly more effective inhibitors However, the inhibition by aminoguanidine unexpectedly appeared to dim ish in effect at the later stages of the AGE formation reaction Due to the slowness of the glycation of protem with glucose, this surpnsmg 0 obse ation could not be fully examined Furthermore, it has been suggested that there may be questions about the long-term stability of aminoguanidine (Ou and Wolff, 1993, supra)

-Analysis using the much more rapid glycation by nbose allowed for the entire time-course of AGE formation to be completely observed and quantitated dunng 5 uninterrupted glycation of protein The use of interrupted glycation uniquely allowed further isolation and examination of only post-Amadon antigenic AGE formation in the absence of free and reversibly bound (Schiff base) πbose Compaπson of the data from these two approaches with the earlier glucose glycation kinetics has provided novel insights into the mechanisms and effectiveness of vanous inhibitors Figure 22 are bar 0 graphs which depict summanzed comparative data of percent inhibition at defined time points using vanous concentrations of inhibitor Figure 22A graphs the data for inhibition after interrupted glycation of RNase AGE formation in nbose Figure 22B graphs the data for inhibition after interrupted glycation of BSA AGE formation by ribose 5 The overall results unambiguously demonstrate that aminoguanidine slows the rate ot antigemc AGE formation in the presence of sugar but has little effect on the final amount of post-Amadon AGE formed Thus observations limited to only the initial "early" stages of AGE formation which indicate efficacy as an inhibitor may in fact be misleading as to the true efficacy of inhibition of post-Amadori AGE formation. Thus the ability to observe a full-course of reaction using ribose and interrupted glycation gives a more complete picture of the efficacy of inhibition of post-Amadon AGE formation.

Example 4

Animal model & testing of in vivo effects of AGE formation/inhibitors

Hyperglycemia can be rapidly induced (within one or two days) in rats by administration of streptozocin (aka. streptozotocin, STZ) or alloxan. This has become a common model for diabetes melitus. However, these rats manifest nephropathy only after many months of hyperglycemia, and usually just prior to death from end-stage renal disease (ESRD). It is believed that this pathology is caused by the irceversible glucose chemical modification of long-lived proteins such as collagen of the basement membrane. STZ-diabetic rats show albuminuria very late after induction of hyperglycemia, at about 40 weeks usually only just prior to death.

Because of the dramatic rapid effects of ribose demonstrated in vitro in the examples above, it was undertaken to examine the effects of ribose administration to rats, and the possible induction of AGEs by the rapid ribose glycation. From this study, a rat model for accelerated ribose induced pathology has been developed.

Effects of very short-term ribose administration in vivo

Phase I Protocol

Two groups of six rats each were given in one day either: a. 300 mM ribose (two intraperitoneal infusions 6-8 hours apart, each 5% of body weight as ml); or b. 50 mM ribose (one infusion) Rats were then kept for 4 days with no further ribose administration, at which time they were sacrificed and the following physiological measurements were determined: (i) initial and final body weight; (ii) final stage kidney weight; (iii) initial and final tail-cuff blood pressure; (iv) creatinine clearance per 100 g body weight.

Albumin filtration rates were not measured, since no rapid changes were initially anticipated. Past experience with STZ-diabetic rats shows that albuminuria develops very late (perhaps 40 weeks) after the induction of hyperglycemia and just before animals expire.

Renal Physiology Results a Final body weight and final single kidney weight was same for low and high πbose treatment groups. b Tail-cuff blood pressure increased from 66 ± 4 to 83 ± 3 to rats treated with low ribose ( 1 x 50 mM). and from 66 ± 4 to 106 ± 5 for rats treated with high nbose (2 x 300 mM) These results are shown in the bar graph of Figure 23 c Creatinine clearance, as cc per 100 g body weight, was decreased (normal range expected about 1 0-1.2) in a dose-dependent fashion to 0.87 ± 0.15 for the low nbose group, and decreased still further 30%> to 0.62 ± 0 13 for the high nbose group. These results are shown in the bar graph of Figure 24

Phase I Conclusion

A single day's πbose treatment caused a dose-dependent hypertension and a dose-dependent decrease in glomerular clearance function manifest 4 days later. These are significant metabolic changes of diabetes seen only much later in STZ-diabetic rats These phenomenon can be hypothesized to be due to ribose irreversible chemical modification (glycation) of protem in vivo.

Effect of exposure to higher ribose concentrations for longer time

Phase II Protocol Groups of rats (3-6) were intrapentoneally given 0.3 M "low πbose dose" (LR) or 1 0 M "high πbose dose" (HR) by twice-daily injections for either (I) 1 day, (n) a "short-term" (S) of 4 days, or (in) a "long-term" (L) of 8 days Additionally, these concentrations of nbose were included in dnnkmg water.

Renal Physiology Results a Tail-cuff blood pressure increased in all groups of πbose-treated rats, confirming Phase I results. (Figure 23). b Creatinine clearance decreased in all groups in a πbose dose-dependent and time-dependent manner (Figure 24) c. Albumin Effusion Rate (AER) increased significantly in a πbose-dependent manner at 1-day and 4-day exposures. However, it showed some recovery at 8 day relative to 4 day in the high-dose group but not m the low-dose group. These results are shown in the bar graph of Figure 25 d. Creatinine clearance per 100 g body weight decreased for both low- and high- πbose groups to about the same extent in a time-dependent manner (Figure 24). Phase II Conclusion

Exposure to nbose for as little as 4 days leads to hypertension and renal dysfunction, as manifest by both decreased creatinine clearance and increased albumin filtration. These changes are typical of diabetes and are seen at much later times in STZ- diabetic rats.

Intervention by two new therapeutic compounds and aminoguanidine

Phase III Protocol

Sixty rats were randomized into 9 different groups, including those exposed to 1 M ribose for 8 days in the presence and absence of aminoguanidine, pyndoxamme, and thiamine pyrophosphate as follows-

Control Groups-

( I ) no treatment;

( n) high dose (250 mg/kg body weight) of pyndoxamme ("compound-P"), (in) high dose (250 mg/kg body weight of thiamine pyrophosphate ("compound-T" or

"TPP"); and

(iv) low dose (25 mg/kg body weight) of aminoguanidine ("AG").

Test Groups:

(1) only 1 M nbose-sahne (2 x 9 cc daily IP for 8 days); (n) ribose plus low dose ("LP") of pyndoxamme (25 mg/kg body weight injected as 0.5 ml with 9 cc πbose);

(in) ribose plus high dose ("HP") of pyndoxamme (250 mg/kg body weight injected as

0 5 ml with 9 cc nbose);

(iv) ribose plus high dose ("HTN of thiamme pyrophosphate (250 mg/kg body weight injected as 0.5 ml with 9 cc nbose); and

(v) πbose plus low dose of amino guamdine (25 mg/kg body weight injected as 0.5 ml with 9 cc ribose)

Unlike Phase II, no πbose was administered in drinking water. Intervention compounds were pre-admimstered for one day pnor to introducing them with πbose.

Renal physiology Results a. Blood pressure was very slightly increased by the three compounds alone

(control group); nbose-elevated BP was not ameliorated by the co-admmistration of compounds. These results are shown in the bar graph of Figure 26. b. Creatmme clearance in controls was unchanged, except for TPP which diminished it. c Creatinine clearance was normalized when πbose was co-admmistered with low dose (25 mg/kg) of either aminoguanidine or pyndoxamme. These results are shown in the bar graph of Figure 27 d High concentrations (250 mg/kg) or pyndoxamme and TPP showed only partial protection against the nbose-mduced decrease in creatmme clearance (Figure 27) e. Albumin effusion rate (AER) was elevated by nbose, as well as by high dose of pyndoxamme and TPP, and low dose of aminoguanidine in the absence of nbose. These results are shown in the bar graph of Figure 28. f. Albumin effusion rate was restored to normal by the co-admmistration of low dose of both aminoguanidine and pyndoxamme. These results are shown in the bar graph of Figure 29

Phase III Conclusions

As measured by two mdicies of renal function, pyndoxamme and aminoguanidine. both at 25 mg/kg, were apparently effective, and equally so, in preventing the πbose-induced decrease m creatinine clearance and πbose-induced mild increase in albummuna.

(I) Thiamme pyrophosphate was not tested at 25 mg/kg; (u) thiamine pyrophosphate and pyndoxamme at 250 mg/kg were partially effective in preventing creatinine clearance decreases but possibly not in preventing mild protemuna; (in) at these very high concentrations and in the absence of nbose, thiamme pyrophosphate alone produced a decrease in creatinine clearance, and both produced mild increases in albuminuria.

Summary

Renal Function and Diabetes

Persistent hyperglycemia in diabetes melhtus leads to diabetic nephropathy in perhaps one third of human patients. Clinically, diabetic nephropathy is defined by the presence of: 1. decrease in renal function (impaired glomerular clearance)

2. an increase in unnary protein (impaired filtration)

3 the simultaneous presence of hypertension Renal function depends on blood flow (not measured) and the glomerular clearance, which can be measured by either muhn clearance (not measured) or creatimne clearance Glomerular permeability is measured by albumin filtration rate, but this parameter is quite variable It is also a log-distπbution function, a hundred-fold increase in albumin excretion represents only a two-fold decrease in filtration capacity

Ribose Diabetic Rat Model

By the above cnteπa, nbose appears to very rapidly induce manifestations of diabetic nephropathy, as reflected in hypertension, creatinine clearance and albummuna, even though the latter is not large. In the established STZ diabetic rat, hyperglycemia is rapidly established m 1-2 days, but clinical manifestations of diabetic nephropathy anse very late, perhaps as much as 40 weeks for albummuna. In general, albummuna is highly vanable from day to day and from animal to animal, although unlike humans, most STZ rats do eventually develop nephropathy.

Intervention b\ Compounds

Using the nbose-treated animals, pyndoxamme at 25 mg/kg body weight appears to effectively prevent two of the three manifestations usually attnbuted to diabetes, namely the impairment of creatmme clearance and albumin filtration. It did so as effectively as aminoguanidine. The effectiveness of thiamme pyrophosphate was not manifest, but it should be emphasized that this may be due to its use at elevated concentrations of 250 mg/kg body weight. Pyndoxamme would have appeared much less effective if only the results at 250 mg/kg body weight are considered

Effect of Compounds Alone

Overall, the rats appeared to tolerate the compounds well. Kidney weights were not remarkable and little hypertension developed. The physiological effects of the compounds were only tested at 250 mg/kg. Thiamme pyrophosphate, but not pyndoxamme, appeared to decrease creatmme clearance at this concentration. Both appeared to slightly increase albummuna, but these measurements were perhaps the least reliable. Human Administration

A typical adult human being of average size weighs between 66 - 77 Kg. Typically, diabetic patients may tend to be overweight and can be over 1 12 Kg. The Recommended dietary allowances for an adult male of between 66 - 77 Kg, as revised in 1989, called for 1 5 mg per day of thiamme, and 2.0 mg per day of Vitamm Bβ (Merck Manual of Diagnosis and Therapy, 16th edition (Merck & Co., Rathaway, N J., 1992) pp 938-939)

Based upon the rat model approach, a range of doses for administration of pyndoxamme or thiamme pyrophosphate that is predicted to be effective for inhibiting post-Amadon AGE formation and thus inhibiting related pathologies would fall m the range of 1 mg/100 g body weight to 200 mg/100 g body weight. The appropπate range when co-admimstered with aminoguanidine will be similar. Calculated for an average adult of 75 Kg, the range (at 10 mg/1 Kg body weight) can be approximately 750 mg to upwards of 150 g (at 2 g/1 Kg body weight). This will naturally vary according to the particular patient.

Example 5

In Vivo Inhibition of the Formation of Advanced Glycation End-Products (AGEs) by Derivatives of Vitamins Bi and B6 and Aminoguanidine. Inhibition of diabetic nephropathy.

The interrupted glycation method, as descnbed in the examples above, allows for the rapid generation of stable well-defined protein Amadon intermediates from nbose and other pentose sugars for use in in vivo studies. The effects of 25 mg/kg/day pyndoxamme (PM) and aminoguanidine (AG) on renal pathology induced by injecting Sprague-Dawley rats daily with 50 mg/kg/day of πbose-glycated Amadoπ-rat serum albumin (RSA), AGE-RSA, and unmodified RSA for 6 weeks. Hyperfiltration (increased creatinine clearance) was transiently seen with rats receiving Amadon-RSA and AGE-RSA, regardless of the presence of PM and AG. Individuals from each group receiving Amadon-RSA and AGE-RSA exhibited microalbuminuna, but none was seen if PM was co-administered. Immunostaining with anti-RSA revealed glomerular staining m rats treated with AGE-RSA and with Amadon- RSA, and this staining was decreased by treatment with PM but not by AG treatment. A decrease in glomerular sulfated glycosammoglycans (Alcian blue pH 1 0 stam) was also found in rats treated with glycated (Amadon and AGE) RSA This appears to be due to reduced heparan sulfate proteoglycans (HSPG), as evidenced by diminished stammg with mAb JM-403 that is specific for HSPG side-cham. These HSPG changes were ameliorated by treatment with PM. but not by AG treatment

Thus we conclude that pyndoxamme can prevent both diabetic-like glomerular loss of heparan sulfate and glomerular deposition of glycated albumin in SD rats chronically treated with πbose-glycated albumin

Materials and methods

Chemicals

Rat serum albumin (RSA) (fraction V, essentially fatty acid-free 0 005%, A.2018). D-πbose, pyndoxamme, and goat alkaline phosphatase-conjugated anti-rabbit IgG were all from Sigma Chemicals Aminoguanidine hydrochlonde was purchased from Aldπch Chemicals

Preparation ofribated RSA

Rat serum albumin was passed down an Affi-Gel Blue column (Bio-Rad), a heparm-Sepharose CL-6B column (Pharmacia) and an endotoxin-bmding affimty column (Detoxigel, Pierce Scientific) to remove any possible contaminants The punfied rat serum albumin (RSA) was then dialyzed in 0 2 M phosphate buffer (pH 7 5) A portion of the RSA (20 mg/ml) was then incubated with 0 5 M nbose for 12 hours at 37°C in the dark After the 12 hour incubation the reaction mixture was dialyzed m cold 0 2 M sodium phosphate buffer over a 36 hour penod at 4°C (this dialysis removes not only the free ribose, but also the Schiff-base intermedianes) At this stage of the glycation process, the πbated protein is classified as Amadon-RSA and is negative for antigemc AGEs, as determined by antibodies reactive with AGE protein (as descnbed previously; R618, antige glucose modified AGE-Rnase). The nbated protem is then divided into portions that will be injected either as: a)Amadon-RSA, b)NaBH -reduced Amadon-RSA, c)AGE-RSA The πbated protein to be injected as Amadon-RSA is simply dialyzed against cold PBS at 4°C for 24 hours A portion of the Amadon-RSA in 0 2 M sodium phosphate is reduced with NaBH₄ to form NaBH₄-reduced Amadon-RSA Bnefly, aliquots were reduced by adding 5 uL of NaBH stock solution (100 mg/ml in O l M NaOH) per mg of protein, incubated for 1 hour at 37°C, treated with HCI to discharge excess NaBH₄, and then dialyzed extensively in cold PBS at 4°C for 36 hours The AGE- RSA w as formed by remcubatmg the Amadon-RSA m the absence of sugar for 3 days The mixture was then dialyzed against cold PBS at 4°C for 24 hours All solutions were filtered (22 um filter) steπhzed and monitored for endotoxms by a hmulus amoebocyte lysate assay (E-Toxate, Sigma Chemical) and contained <0 2 ng/ml before being frozen (-70°C) down into individual aliquots until it was time for injection

Animal Studies

Male Sprague-Dawley rats (Sasco, lOOg) were used After a 1 week adaptation penod. rats were placed m metabolic cages to obtain a 24 hour unne specimen for 2 days before administration of injections Rats were then divided into expenmental and control groups and given tail vein injections with either saline, unmodified RSA (50 mg/kg), Amadon-RSA (50 mg/kg), NaBH₄-reduced Amadon-RSA (50 mg/kg), or AGE-RSA (50 mg/kg)

Rats injected with -Amadon-RSA and AGE-RSA were then either left untreated, or futher treated by the administration of either aminoguanidine (AG, 25 mg/kg), pyndoxamme (PM, 25 mg/kg), or a combination of AG and PM (10 mg/kg each) through the dnnkmg water Body weight and water intake of the rats were monitored weekly in order to adjust dosages At the conclusion of the expenmental study the rats were placed in metabolic cages to obtain 24 hour unne specimen for 2 days pnor to sacπficmg the animals

Total protein m the unne samples was determined by Bio-Rad assay Albumin in unne was determined by competitive ELISA using rabbit anti-rat serum albumin (Cappell) as pnmary antibody (1/2000) and goat anti-rabbit IgG (Sigma Chemical) as a secondary antibody (1/2000) Unne was tested with Multistix 8 SG (Miles Laboratones) for glucose, ketone, specific gravity, blook, pH, protem, mtnte, and leukocytes Nothing remaikable was detected other than some protein. Creatmme measurements were performed with a Beckman creatmme analyzer II.

Blood samples were collected by heart puncture before termination and were used in the determination of creatmme clearance, blood glucose (glucose oxidase, Sigma chemical), fructosamme (nitroblue tetrazolium, Sigma chemical), and glycated Hb (columns, Pierce chemicals). Aorta, heart, both kidneys and the rat tail were visually inspected and then quickley removed after perfusing with saline through the right ventricle of the heart. One kidney, aorta, rat tail, and the lower 2/3 of the heart were snap-frozen and then permanently stored at -70°C. The other kidney was sectioned by removing both ends (cortex) to be snap-frozen, with the remaining portions of the kidney being sectioned into thirds with two portions being placed into neutral buffered formalin and the remaining third minced and placed in 2.5% glutaraldehyde/2%> paraformaldehyde.

L igh t Microscopy

After perfusion with saline, kidney sections were fixed in ice-cold 10% neutral buffered formalin. Paraffin-embedded tissue sections from all rat groups (n = 4 per group) were processed for staining with Harris' alum hematoxylin and eosin (H&E), perodic acid/Schiff reagent (PAS), and alcian blue (pH 1.0 and pH 2.5) stains for histological examination. The alcian blue sections were scored by two investigators in a blinded fashion.

Electron Microscopy

Tissues were fixed in 2.5% glutaraldehyde/2%> paraformaldehyde (0.1 M sodium cacodylate, pH 7.4), post-fixed for 1 hour in buffered osmium tetroxide (1.0%), prestained in 0.5% uranyl acetate for 1 hour and embedded in Effapoxy resin. Ultrathin sections were examined by electron microscopy.

Immunofluorescence Panafin-embedded sections were deparaffinized and then blocked with 10% goat serum in PBS for 30 min at room temperature. The sections were then incubated for 2 hour at 37°C with primary antibody, either affinity purified polyclonal rabbit anti-AGE antibody, or a polyclonal sheep anti-rat serum albumin antibody (Cappell). The sections were then rinsed for 30 min with PBS and incubated with secondary antibody, either affinity purified FITC-goat anti-rabbit IgG (H+L) double stain grade (Zymed) or a Rhodamine-rabbit anti-sheep IgG (whole) (Cappell) for 1 hour at 37°C. The sections were then rinsed for 30 min with PBS in the dark, mounted in aqueous mounting media for immunocytochemistry (Biomeda), and cover slipped. Sections were scored in a blinded fashion. Kidney sections were evaluated by the number and intensity of glomerular staining in 5 regions around the penphery of the kidney Scores were normalized for the lmmunofluorescent score per 100 glomeruli with a scoπng system of 0-3

Preparation of Polyclonal Antibodies to AGE-Proteins

Immunogen was prepared by glycation of BSA (R479 antibodies) or Rnase (R618 antibodies) at 1.6 g protein in 15 ml for 60 - 90 days using 1.5 M glucose in 0.4 M phosphate containing 0.05% sodium azide at pH 7.4 and 37°C. New Zealand white rabbit males of 8-12 weeks were immunized by subcutaneous administration of a 1 ml solution containing 1 mg/ml of glycated protein Freund's adjuvant. The pnmary injection used the complete adjuvant and three boosters were made at three week intervals with Freund's incomplete adjuvant. The rabbits were bled three weeks after the last booster. The serum was collected by centnfiigation of clotted whole blood. The antibodies are AGE-specific, being unreactive with either native proteins or with Amadori intermediates.

ELISA Detection of AGE Products The general method of Engvall (21) was used to perform the ELISA. Glycated protein samples were diluted to approximately 1 5 ug/ml in 0 1 M sodium carbonate buffer of pH 9 5 to 9 7 The protein was coated overnight at room temperature onto a 96- well polystyrene plate by pippettmg 200 ul of protem solution into each well (about .3 ug/well) After coating, the excess protein was washed from the wells with a saline solution containing 0.05%o Tween-20. The wells were then blocked with 200 ul of 1% casein in carbonate buffer for 2 hours at 37°C followed by washing. Rabbit anti-AGE antibodies were diluted at a titer of 1 -350 in incubation buffer and incubated for 1 hour at 37°C, followed by washing. In order to minimize background readings, antibody R618 used to detect glycated RSA was generated by immunization against glycated Rnase. An alkaline phosphatase-conjugated antibody to rabbit IgG was then added as the secondary antibody at a titer of 1.2000 and incubated for 1 hour at 37°C, followed by washing. The /-^j-nitrophenolate being monitored at 410 nm with a Dynatech MR4000 microplate reader.

Results

The rats in this study survived the treatments and showed no outward signs of any gross pathology. Some of the rats showed some small weight changes and tail scabbing.

Initial screening of kidney sections with PAS and H&E stains did not reveal any obvious changes, and some EM sections did not reveal any gross changes in the glomerular basement membrane (GBM). However, upon Alcian Blue staining, striking differences were discovered. Alcian blue staining is directed towards negatively charged groups in tissues and can be made selective via changes in the pH of staining. At pH 1.0 Alcian blue is selective for mucopolysaccharides, and at pH 2.5 detects glucoronic groups. Thus negative charges are detected depending upon the pH of the stain.

At pH 2.5 Alcian blue staining showed that Amadori-RSA (p<0.05) and AGE- RSA (p<0.01 ) induced increased staining for acidic glycosaminoglycans (GAG) over control levels (Figure 33). For both AGE-RSA and Amadori-RSA, treatment with pyridoxamine (PM) prevented the increase in staining (p<0.05 as compared with controls). In contrast, treatment with aminoguanidine (AG) or combined PM and AG at 10 mg/kg each, did not prevent the increase. At pH 1.0 Alcian blue staining was significantly decreased by AGE-RSA

(p<0.001 ) (Figure 34). However, no significant difference was seen with Amadori-RSA. Due to faint staining, treatment with PM, AG and combined could not be quantitated.

Immunofluorescent glomerular staining for RSA showed elevated staining with Amadori-RSA and AGE-RSA injected animals (Figure 35). Significant reduction of this effect was seen in the rats treated with PM, and not with AG or combined AG & PM.

Immunofluorescent glomerular staining for Heparan Sulfate Proteoglycan Core protein showed slightly reduced staining with Amadori-RSA and AGE-RSA injected animals but were not statistically significant(Figure 36). A reduction of this effect was seen in the rats treated with PM, and not with AG or combined AG & PM. However, immunofluorescent glomerular staining for Heparan Sulfate Proteoglycan side-chain showed highly reduced staining with Amadori-RSA and AGE-RSA injected animals (Figure 37) A significant reduction of this effect was seen in the rats treated with PM, and not with AG or combined AG & PM.

Analysis of average glomerular volume by blinded scoring showed that Amadori- RSA and AGE-RSA caused significant increase in average glomeruli volume (Figure 38). A significant reduction of this effect was seen with treatment of the rats with PM. No effect was seen with treatment with AG or combined AG and PM at 10 mg/kg each.

Example 6

AGE Inhibitor Compounds

The present invention encompasses compounds, and pharmaceutical compositions containing compounds having the general formula:

Formula I

wherein R, is CH₂NH₂, CH₂SH, COOH, CH?CH₂NH₂, CH₂CH,SH, or CH₂COOH;

R? is OH, SH or NH₂;

Y is N or C, such that when Y is N R₃ is nothing, and when Y is C, R₃ is NO₂ or another electron withdrawing group; and salts thereof.

The present invention also encompasses compounds of the general formula

Formula II

wherem R, is CH₂NH₂, CH₂SH, COOH, CH₂CH₂NH₂, CH₂CH₂SH, or CH₂COOH; R₂ is OH, SH or NH?_;

Y is N or C, such that when Y is N R3 is nothing, and when Y is C, R3 is NO₂ or another electron withdrawing group;

R4 is H, or C 1-6 alkyl;

R₅ and R-₆ are H. C 1-6 alkyl, alkoxy or alkane; and salts thereof.

In addition, the instant invention also envisions compounds of the formulas

and

The compounds of the present invention can embody one or more electron withdrawing groups, such as and not limited to -NH₂, -NHR, -NR2, -OH, -OCH₃, -OCR, and -NH-COCH₃ where R is C 1-6 alkyl.

By "alkyl" and "lower alkyl" in the present invention is meant straight or branched chain alkyl groups having from 1-12 carbon atoms, such as, for example, methyl, ethyl, propyl, isopropyl, n-butyl. sec-butyl, tert-butyl, pentyl. 2-pentyl. isopentyl. neopentyl, hexyl, 2-hexyl, 3-hexyl, and 3-methylpentyl. Unless indicated otherwisev the alkyl group substituents herein are optionally substituted with at least one group independently selected from hydroxy, mono- or dialkyl amino, phenyl or pyridyl.

By "alkoxy" and "lower alkoxy" in the present invention is meant straight or branched chain alkoxy groups having 1 -6 carbon atoms, such as, for example, methoxy, ethoxy, propoxy, isopropoxy, n-butoxy, sec-butoxy, tert-butoxy, pentoxy. 2-pentyl, isopentoxy. neopentoxy. hexoxy, 2-hexoxy, 3-hexoxy. and 3-methylpentoxy.

By "alkene" and "lower alkene" in the present invention is meant straight and branched chain alkene groups having 1-6 carbon atoms, such as, for example, ethlene, propylene. 1-butene, 1-pentene, 1-hexene, cis and trans 2-butene or 2-pentene, isobutylene, 3 -methyl- 1-butene. 2-methyl-2-butene, and 2,3-dimethyl-2-butene. By "salts thereof in the present invention is meant compounds of the present invention as salts and metal complexes with said compounds, such as with, and not limited to, Al,Zn, Mg, Cu, and Fe

One of ordinary skill in the art will be able to make compounds of the present invention using standard methods and techniques.

The instant invention encompasses pharmaceutical compositions which compπse one or more of the compounds of the present invention, or salts thereof, in a suitable earner The instant invention encompasses methods for admimstenng pharmaceuticals of the present invention for therapeutic intervention of pathologies which are related to ΛGE formation in vivo In one prefened embodiment of the present invention the AGE related pathology to be treated is related to diabetic nephropathy

Example 7. Improved Dialysis Solutions and Methods

It has also been demonstrated that formation of AGE products occurs in dialysis fluid in vitro (Lamb et al., Kidney Intl. 47 1768-1774 (1995)) Furthermore, the level of various AGE species is increased in blood of patients on either CAPD (continuous ambulatory pentoneal dialysis) (See, for example, Degenhardt et al , Kidney Intl. 52.1064-1067 (1997), Shaw et al , Cellular and Molecular Biology 44 1061-1068 ( 1998)) or maintenance hemodialysis (HD) (Motomiya et al., Kidney Intl. 54 1357-1366 ( 1998)), regardless of whether the patient is hyperglycemic (Miyata et al , Kidney Intl. 55 389-399 ( 1999))

CAPD involves the use of dialysis solutions containing high sugar concentrations, while HD does not Thus, the precipitating factor in AGE formation in dialysis patients has been hypothesized to involve "carbonyl stress", resulting either from increased oxidation of carbohydrates and hpids ("oxidative stress"), or inadequate detoxification or inactivation of reactive carbonyl compounds deπved from both carbohydrates and hpids by oxidative and non-oxidative chemistry (Miyata et al., Kidney Intl. 55 389-399 (1999)) Other studies indicate that nonenzymatic glycosylation of pentoneal components occurs during pentoneal dialysis (See. for example. Fnedlander et al , J Clin. Invest. 1996 97 728-735; Nakayama et al.. Kidney Intl. 51 182-186 (1997), and Korbet et al., Am J Kidney Disease 22:588-591 (1993) These vanous studies have implicated accumulation of AGEs in the following pathologies m patients receiving dialysis

1 Increased cardiac morbidity and mortality (Korbet et al , 1993)

2 Dialysis-related amyloidosis (Motomiya et al , Kidney Intl 54 1357-1366, (1998)

3 Increased permeability of the pentoneal membrane (Nakayama et al, 1997)

4 Renal failure progression (Dawnay and Millar, Cell Mol. Biol. 44-1081- 1094 ( 1998) (increased rate to end-stage renal disease) 5 Ultrafiltration failure and pentoneal membrane destruction (Linden et al.,

Pent Dial. Int 18.290-293 (1998)

Thus, in another aspect, the present invention provides improved dialysis methods and compositions for dialysis that compπse utilizing an effective amount of one or more of the compounds of the invention to inhibit AGE formation, particularly due to carbonyl stress, including the conversion of Amadon compounds to advanced glycation endproducts and inadequate detoxification or mactivation of reactive carbonyl compounds

In further aspects, the present invention provides methods for inhibiting dialysis- related cardiac morbidity and mortality, dialysis-related amyloidosis, limiting dialysis- related increases in permeability of the peritoneal membrane m a dialysis patient, inhibiting renal failure progression in a patient, and inhibiting ultrafiltration failure and peritoneal membrane destruction in a patient, compnsmg introducing into the patient a dialysis solution that compnses an amount of one or more of the compounds of the invention sufficient to inhibit or limit the specified endpomt

In another aspect, the present invention compnses a method for inhibiting AGE formation in a dialysis patient comprising admmisteπng to the patient a dialysis solution comprising an effective amount of a compound of the invention to inhibit AGE formation As used herein, dialysis solutions compπse solutions for both pentoneal dialysis

(PD) and hemodialysis (HD) As used herein, PD differs from HD in that the patient's pentoneum, not an artificial kidney, forms the dialyzmg membrane

As used herein, the term "osmotically active agent" refers to a substance present in the dialysis solution which is capable of maintaining the osmotic gradient required to cause transport of water and toxic substances across the pentoneum into the dialysis solution The normal function of the mammalian kidney includes such activity as maintaining a constant acid-base and electrolyte balance, removing excess fluids and removing undesirable products of the body's metabolism from the blood (U.S. Patent No. 5.869.444. incorporated by reference herein in its entirety). In an individual with end stage renal disease, this functioning of the kidney may be reduced to as low as 5% or less of the normal level. When renal function has decreased to this point, dialysis is used in an attempt to replace kidney activity. This is accomplished clinically by the use of dialysis. One of the most common dialysis methods is hemodialysis ("HD"), in which the patient's blood is passed through an artificial kidney dialysis machine, wherein a synthetic non-permeable membrane acts as an artificial kidney with which the patient's blood is contacted on one side. On the opposite side of the membrane is a dialyzing fluid or dialysate, the composition of which is such that the undesirable products in the patient's blood will naturally pass across the membrane by diffusion, into the fluid. The blood is thus cleansed, in essentially the same manner as the kidney would have done, and the blood is returned to the patient's body. Examples of HD solutions can be found in U.S. Patent Nos. 5,474,992; and 5,211,643; both incorporated by reference herein in their entirety. The dialysis solutions for HD are manufactured in the form of a suitable solution by standard procedures. The osmotic pressure and pH of the liquid preparation are preferably adjusted within the respective ranges for HD solutions in general. The HD may contain a variety of other ingredients which are generally included in dialysis solutions for extracorporeal hemodialysis. for example various salts such as sodium chloride, potassium chloride, calcium chloride, magnesium chloride, sodium acetate, and sodium hydrogen carbonate.

Alternatively, the patient's own peritoneum can be used as the required semipermeable membrane. The peritoneum is the membranous lining of the body cavity that contains large numbers of blood vessels and capillaries, thus allowing its function as a natural semipermeable membrane. (U.S. Patent No. 5,869,444) Dialysis solution is introduced into the peritoneal cavity, via a catheter in the abdominal wall. A suitable period of residence time for the dialysate is allowed to permit the exchange of solutes between it and the blood. Fluid removal is achieved by providing a suitable osmotic gradient, via inclusion of an osmotically active agent in the dialysate. from the blood to the dialysate to permit water outflow from the blood. Thus, the proper acid-base, electrolyte and fluid balance is returned to the blood and the dialysis solution is simply drained from the body cavity through the catheter. Although more than one type of peritoneal dialysis exists, the technique known as continuous ambulatory pentoneal dialysis (CAPD) is particularly favored, since it does not require the patient to remain tied to machinery while the solute and fluid exchange is accomplished. The only sedentary penod required is dunng infusion and draining of the dialysis solution.

The osmotically active agent which has cunently achieved the most widespread acceptance is glucose. Glucose has the advantage of being non-toxic, and is so readily metabohzable if it enters the blood. However, glucose is readily taken up into the blood from the dialysate, which may lead to vanous complications. (U.S. Patent No. 5,869.444) Among these complications is the build-up of advanced glycation end products discussed above.

Therefore, in one aspect the present invention provides improved dialysis solutions compnsing an amount effective to inhibit AGE formation m a patient who is to receive the dialysis solution of one or more compounds or pharmaceutical compositions comprising a compound of the general formula:

wherem R, is CH₂NH₂, CH₂SH, COOH, CH₂CH₂NH₂, CH₂CH₂SH, or CH₂COOH; R₂ and R*, is H, OH, SH, NH₂, C 1-6 alkyl, alkoxy or alkene; R-₄ and R₅ are H, C 1-6 alkyl, alkoxy or alkene;

Y is N or C, such that when Y is N R₃ is nothing, and when Y is C, R₃ is NO? or another electron withdrawing group, and salts thereof.

According to this aspect of the invention, the compound(s) is used as an additive to any type of dialysis solution in which inhibiting AGE formation is desirable, including but not limited to hemodialysis solutions and pentoneal dialysis solutions. In one embodiment, the dialysis solutions compnse: a an osmotically active agent that is capable of maintaining the osmotic gradient required to cause transport of water and toxic substances across the pentoneum into the dialysis solution; and b an amount of the compounds of the invention effective to inhibit the conversion of Amadon compounds to post Amadon advanced glycation endproducts m a patient who is to receive the solution.

In a preferred embodiment, the osmotically active agent is selected from the group consisting of nbose, lyxose. xylose, arabinose, glucose, fructose, maltose, lactose, mannose, fructose, and galactose, or polymers thereof, and polyamons (For examples of polymers, see Barre et al., Adv Pent. Dial. 15:12-16 (1999), Wang et al.. Pent. Dial. Int. 18: 193-203 ( 1998); Plum et al.. Am. J. Kidney Dis. 30:413-422 (1997); Ho-dac- Pannekeet et al, Kidney Intl. 50:979-986 (1996), Chen et al., Adv Pent. Dial. 14.116- 1 19 ( 1998); Dawnay et al.. Pent. Dial. Int. 17:52-58 (1997), Twardowski et al., Artif. Organs 7.420-427 (1983))

15 In a further prefened embodiment, the dialysis solution further compnses sodium in a concentration that is less than a sodium plasma concentration in a renal patient who is to receive the solution. In another preferred embodiment, the osmotic agent is glucose.

In a most prefened embodiment, the compound compnses pyndoxamme.

In a further aspect, the present invention compnses an improved method of 0 performing dialysis on a patient wherem the improvement compπses introduction into the patient in need of dialysis a dialysis solution that compnses an amount effective to inhibit AGE formation in the patient of one or more compounds or pharmaceutical compositions compnsing a compound of the general formula:

"> S wherein R, is CH?NH?, CH₂SH, COOH, CH₂CH₂NH₂, CH₂CH₂SH, or CH₂COOH; R₂ and R_<, is H, OH, SH, NH₂, C 1-6 alkyl, alkoxy or alkene; R-4 and R₅ are H, C 1-6 alkyl, alkoxy or alkene;

Y is N or C, such that when Y is N R₃ is nothing, and when Y is C. R₃ is NO₂ or another electron withdrawing group, and salts thereof.

In a prefened embodiment, the compound comprises pyndoxamme. The dialysis solutions for use in this aspect of the invention are as described above.

In other aspects, the present invention provides methods for inhibiting dialysis- related cardiac morbidity and mortality, dialysis-related amyloidosis. limiting dialysis- related increases in permeability of the peritoneal membrane in a patient, inhibiting renal failure progression in a patient, and inliibiting ultrafiltration failure and peritoneal membrane destruction in a patient, comprising introducing into the patient a dialysis solution that comprises an amount of one or more of the compounds of the invention sufficient to inhibit or limit the specified endpoint. In another aspect, the invention comprises a method for inhibiting AGE formation in a dialysis patient comprising administering to a patient undergoing dialysis an effective amount of one or more of the compounds of the invention to inhibit AGE formation. In a prefened embodiment of each of these methods, the compound is pyridoxamine.

The concentration of the compounds of the invention in the dialysis solutions is based on a variety of factors, including the composition of the dialysis solution, treatment of the dialysis solution (i.e.: sterilization, etc.), type of dialysis (CAPD vs. HD), type of condition, compound used, age, weight, sex, medical condition of the individual, and the severity of the condition. Thus, the concentration may vary widely, but can be determined routinely by a physician using standard methods. Concentration levels of the order of between 1 μM to 100 mM are useful for all methods of use disclosed herein.

Examples

Example 1. Inhibition of AGE formation in peritoneal dialysis fluid

Albumin was added to DIANEAL® peritoneal dialysis (PD) (Baxter Corp. Deerfield, IL) fluid after adjustment of the PD fluid pH to 7.5. The DIANEAL® PD fluid used in this experiment was composed of: sodium = 132 mEq/1 , calcium = 2.5 mEq/1 , magnesium = 0.5 mEq/1 , chloride = 95 mEq/ 1, lactate = 40 mmol/1,

4 25%o dextrose with an osmolaπty = 483 mOsmol/1 PD fluid containing glucose as an osmotic agent is generally prepared at a pH between 5 0 and 5 5 to prevent carmehzation of glucose upon heat stenhzation of the PD fluid (U S. Patent No 5,869,444) At this non-physiological pH, AGEs do not form. A-s the PD fluid enters the body, its pH changes very quickly to physiological pH.

The PD fluid-albumm samples were then incubated at 37°C for 52 days m the presence and absence of pyndoxamme (1 mM, 3 mM, and 15 mM) In addition, IM glucose was added to each of the samples treated with pyndoxamme as well as to one control sample Glucose addition was utilized to accelerate the process of AGE formation Antibodies specific for AGEs (carboxymethyl-lysme) were used to conduct ELISA to determine the amount of albumin AGEs in each sample. Figure 39 demonstrates that pyndoxamme significantly inhibits formation of protem (albumin) AGEs in PD fluid under these conditions.

Example 2 Inhibition of AGE formation from patient post-dialysis peritoneal dialysis fluid Post-dialysis fluid was collected from a non-diabetic pentoneal dialysis patient at the University of Kansas Medical Center, and had a pH of 7 5. Pnor to dialysis, the PD components were as descnbed above, except that the PD solution contained 2.5% dextrose, and no glucose was added However, the composition of a PD solution is altered by exchange with the peritoneum. Thus, the exact composition of the PD solution is difficult to determine.

Myoglobm was incubated with post-dialysis PD fluid for 12 hours at 60°C in the presence and absence of pyndoxamme. (3 mM, 0.5 mM, 0 1 mM, and 0.02 mM) This expenment was conducted at 60°C to accelerate AGE formation, which is temperature dependent. Antibodies were then used to conduct ELISAs as descnbed above The results of these expenments (Figure 40) demonstrate that pyndoxamme inhibits the formation of myoglobm AGEs in post-dialysis PD fluid under these conditions

In a similar expenment, post-dialysis fluid was collected from a diabetic pentoneal dialysis patient at the University of Kansas Medical Center, and had a pH of 7.5. Prior to dialysis, the PD components were as described above, except that the PD solution contained 4.25% dextrose, and no glucose was added.

Metmyoglobin was incubated with the PD fluid at 37°C for various periods of time, up to 42 days. The experiments were done in the presence and absence of 3 mM pyridoxamine). The data (Figure 41) demonstrate that pyridoxamine inhibits the formation of protein (metmyoglobin) AGEs in post-dialysis PD fluid under these conditions.

The instant invention may be embodied in other forms or carried out in other ways without departing from the spirit or essential characteristics thereof. The present disclosure and enumerated examples are therefore to be considered as in all respects illustrative and not restrictive, the scope of the invention being indicated by the appended claims, and all equivalency are intended to be embraced therein. One of ordinary skill in the art would be able to recognize equivalent embodiments of the instant invention, and be able to practice such embodiments using the teaching of the instant disclosure and only routine experimentation.

Claims

We claim:

1. An improved dialysis solution, wherein the improvement comprises an amount effective to inhibit AGE formation of one or more compounds of the general formula

wherein R, is CH₂NH₂, CH₂SH, COOH, CH₂CH₂NH,, CH₂CH₂SH, or CH?COOH;

R? and -R-s is H, OH, SH, NH₂, C 1-6 alkyl, alkoxy or alkene;

Rt and R₅ are H, C 1-6 alkyl, alkoxy or alkene;

Y is N or C, such that when Y is N R₃ is nothing, and when Y is C, R₃ is NO₂ or another electron withdrawing group, and salts thereof.

2. The dialysis solution of claim 1 further comprising an osmotically active agent that is capable of maintaining the osmotic gradient required to cause transport of water and toxic substances across the peritoneum into the dialysis solution.

3. The dialysis solution of claim 2, wherein the osmotically active agent is selected from the group consisting of ribose, lyxose, xylose, arabinose, glucose, fructose, maltose, lactose, mannose, fructose, and galactose, or polymers thereof, and polyanions.

4. The dialysis solution of claim 1 further comprising sodium in a concentration that is less than a sodium plasma concentration in a renal patient who is to receive the solution.

5. The dialysis solution of any of claims 1-4 wherein the compound is pyridoxamine.

6. A-n improved method of performing dialysis on a patient wherein the improvement comprises introducing into the dialysis patient a dialysis solution that comprises an amount effective to inhibit AGE formation in the patient of one or more compounds of the general formula:

wherein R, is CH₂NH?, CH SH, COOH, CH₂CH₂NH₂, CH₂CH₂SH, or CH₂COOH;

R₂ and R* is H, OH, SH, NH₂, C 1-6 alkyl, alkoxy or alkene;

Rt and R₅ are H, C 1-6 alkyl, alkoxy or alkene;

Y is N or C, such that when Y is N R is nothing, and when Y is C, R3 is NO₂ or another electron withdrawing group, and salts thereof.

7. A method for decreasing dialysis-related cardiac morbidity and mortality in a dialysis patient, comprising introducing into the patient a dialysis solution that comprises an amount effective to decrease dialysis-related cardiac morbidity and mortality in the patient of one or more compounds of the general formula:

wherein R, is CH₂NH₂, CH₂SH, COOH, CH₂CH₂NH₂, CH₂CH₂SH, or CH₂COOH;

R₂ and R_<, is H, OH, SH, NH₂, C 1-6 alkyl, alkoxy or alkene; i and R₅ are H, C 1-6 alkyl, alkoxy or alkene;

Y is N or C, such that when Y is N R₃ is nothing, and when Y is C, R₃ is NO₂ or another electron withdrawing group, and salts thereof.

8. A method for decreasing dialysis-related amyloidosis in a dialysis patient, comprising introducing into the patient a dialysis solution that comprises an amount effective to decrease dialysis-related amyloidosis in the patient of one or more compounds of the general formula:

wherein R, is CH₂NH₂, CH₂SH, COOH, CH₂CH₂NH₂, CH₂CH₂SH, or CH₂COOH;

R₂ and R^ is H, OH, SH, NH₂, C 1-6 alkyl, alkoxy or alkene;

R-t and R₅ are H, C 1-6 alkyl, alkoxy or alkene;

Y is N or C, such that when Y is N R₃ is nothing, and when Y is C, R₃ is NO₂ or another electron withdrawing group, and salts thereof.

9. A method for limiting dialysis-related increases in permeability of the peritoneal membrane in a dialysis patient, comprising introducing into the patient a dialysis solution that comprises an amount effective to limit dialysis-related increases in permeability of the peritoneal membrane in the patient of one or more compounds of the general formula:

wherem R, is CH₂NH₂, CH₂SH, COOH, CH₂CH₂NH₂, CH₂CH₂SH, or CH₂COOH; R₂ and R<, is H, OH, SH, NH₂, C 1-6 alkyl, alkoxy or alkene; Rt and R₅ are H, C 1-6 alkyl, alkoxy or alkene; Y is N or C, such that when Y is N R₃ is nothing, and when Y is C, R₃ is NO₂ or another electron withdrawing group, and salts thereof.

10. A method for inhibiting renal failure progression in a dialysis patient, comprising introducing into the patient a dialysis solution that comprises an amount effective to inhibit renal failure progression in the patient of one or more compounds of the general formula:

wherein R, is CH₂NH₂, CH₂SH, COOH, CH₂CH₂NH,, CH₂CH₂SH, or CH₂COOH;

R₂ and R«, is H, OH, SH, NH₂, C 1-6 alkyl, alkoxy or alkene;

R and R₅ are H, C 1-6 alkyl, alkoxy or alkene;

Y is N or C, such that when Y is N R₃ is nothing, and when Y is C, R₃ is NO₂ or another electron withdrawing group, and salts thereof.

11. A method for inhibiting ultrafiltration failure and peritoneal membrane destruction in a dialysis patient, comprising introducing into the patient a dialysis solution that comprises an amount effective to inhibit ultrafiltration failure and peritoneal membrane destruction in the patient of one or more compounds of the general formula:

wherein R, is CH₂NH₂, CH₂SH, COOH, CH₂CH₂NH₂, CH₂CH₂SH, or CH₂COOH;

R? and R^ is H, OH, SH, NH₂, C 1-6 alkyl, alkoxy or alkene;

-R4 and R<- are H, C 1-6 alkyl, alkoxy or alkene;

Y is N or C, such that when Y is N R₃ is nothing, and when Y is C, Ri is NO₂ or another electron withdrawing group, and salts thereof.

12 A method for inhibiting AGE formation in a dialysis patient compnsing admimstenng to the patient a dialysis solution compnsing an amount effective to inhibit AGE formation of one or more compounds of the general formula:

wherein R, is CH₂NH?, CH?SH, COOH, CH₂CH₂NH₂, CH₂CH,SH, or CH₂COOH;

R₂ and -R₆ is H, OH, SH, NH₂, C 1-6 alkyl, alkoxy or alkene;

R---1 and R^ are H, C 1-6 alkyl, alkoxy or alkene;

Y is N or C, such that when Y is N R₃ is nothing, and when Y is C, R3 is NO₂ or another electron withdrawing group, and salts thereof.

13. The method of any of claims 6-12 wherem the compound is pyndoxamme.