Background

Lowering elevated total and low density lipoprotein (LDL) cholesterol levels is an established therapeutic method of controlling plasma lipids to reduce the risk of coronary heart disease and progression of atherosclerosis. Some drugs used to reduce cholesterol levels are referred to as bile acid sequestrants. Bile acid sequestrants can be used to treat hypercholesterolemia by binding bile acids in the intestine, carrying them through the small intestine and causing them to be excreted in the feces. Since the bile acid salts are not reabsorbed, the conver¬ sion of cholesterol to bile acids is accelerated to main- tain a constant pool of bile acids causing a concomitant decrease in blood cholesterol level. Thus, bile acid sequestrants are widely recognized as therapeutics for the treatment of elevated plasma lipids.

Cholestyramine, which is a synthetic anion exchange resin based on a styrene-divinylbenzene copolymer, has been shown to inhibit the reabsorption of bile acids, to alter the composition of the LDL particle thereby enhanc¬ ing its affinity for the LDL receptor, and to increase uptake and degradation of LDL. Although cholestyramine can lower cholesterol by up to 40 percent, it is difficult to maintain a patient on a strict cholesterol lowering regimen with cholestyramine due to its offensive taste and frequent side effects of constipation and nausea. Studies indicate that only approximately two percent of the pa- tients who are initially prescribed cholestyramine contin¬ ue to take the drug after two months.

In addition to cholestyramine, numerous other materi¬ als have been reported as displaying affinity for bile acids. Among the insoluble synthetic materials are poly¬ mers derived from poly(diallylmethylamine) derivatives (Buntin et a_l. , U.S. Pat. 4,759,923); metal coordinated synthetic polyamines derived from poly(methyl acrylates) (St. Pierre et a_l. , U.S. Pat. 5,114,709); and vinylimida- zole ethylene glycol dimethacrylate copolymers (Pirclo et al. ■ Eur. Pat. Appl. Al 0,162,388). Soluble derivatives, such as those based on quaterna¬ ry poly[ (alkylimino)alkylene] polymers or poly(diallyl- methyl ammonium chloride) derivatives, have also been reported, but were found to be toxic in mammals (U.S. Pat. 4,027,009) . The disadvantage of the non-systemic synthetic cho¬ lesterol lowering agents is their unfavorable amine odor or palatability characteristics (grittiness, sand-like outhfeel) . In addition, the solids of such compositions (e.g., cholestyramine) do not readily disperse in water, but tend to clump, and also rapidly settle in aqueous medium. Thus, the consumer needs to stir the compositions for several minutes before the drink can be consumed. At¬ tempts have been made to overcome these shortcomings, e.g., by blending cholestyramine with methylcellulose (Schulz, Eur. Pat. Appl. 0,278,464 Al,), or adipic acid (Hussein et al. , Eur. Pat. Appl. 0,447,362 Al) , or by presenting cholestyramine in a gelatin matrix (Killeen, Eur. Pat. Appl. 0,366,255). However, these methods were not satisfactory, as the active resin is effectively diluted. Large doses of these compositions are required for ingestion.

Numerous natural materials have been reported as being capable of lowering cholesterol levels, fibers such

as psylliu , bran, guar gum and konjac glucomannan, and other carbohydrates, including pectin, methyl cellulose and cyclodextrin. The disadvantage of the natural materi¬ als is their low level of effectiveness for lowering plasma cholesterol levels. For example, clinical trials of guar gum demonstrated only a 10-15% total cholesterol reduction after several months, while other studies showed no significant response at all. In addition, no clinical¬ ly or statistically significant changes in serum triglyc- eride levels have been observed.

The natural materials also tend to absorb substantial amounts of water, resulting in gel-like materials with concomitant gastrointestinal pain, nausea, undesirable bloating, flatulence and laxation potential. Others do not have an approved regulatory status and may have unde¬ sirable side effects, such as complexation of essential proteins and hormones. Still others (e.g., guar gum) are not uniform products and display variable viscosities which may affect their pharmacological properties. The dosage required for efficacy has been another drawback of fiber- or resin-based cholesterol lowering drugs. Large quantities of bile acid sequestration drugs are typically required to effect a meaningful decrease in lipoprotein plasma levels. This high dosage discourages patient compliance.

Summary of the Invention

This invention pertains to novel derivatized polysac¬ charide bile acid sequestrant compounds, to compositions containing the novel bile acid sequestrants and to methods for using and making the novel bile acid sequestrants. The compounds comprise hydrophobic, cationic ligand(s) coupled to a polysaccharide substrate. Preferred polysac-

charide substrates include starch, cellulose and glucan, such as whole glucan particles. The ligand(s) can be an alkylamine or alkanediamine where the N-methylated form of each is preferred. N-Methylated 1, 12-diaminododecane is most preferred. The ligand(s) can be coupled to the oxidized or carboxyalkylated substrate. Periodate oxida¬ tion is the preferred oxidation methodology.

It has been discovered that these compounds have the acquired ability to bind bile acids, as determined by an in vitro bile acid binding assay. Further, these polysac¬ charide derivatives have a significantly enhanced ability to reduce total cholesterol, low density lipoprotein cholesterol and triglyceride levels in the blood of hyper- lipide ic hamsters fed a diet containing a bile acid sequestrant of this invention.

Based on these findings, the novel compounds of this invention can be used to lower plasma lipid levels in mammals. The derivatized polysaccharide bile acid se¬ questrants have similar or increased potency for lowering cholesterol compared to cholestyramine. They can be administered to mammals in reasonable quantities and are organoleptically acceptable. The properties should im¬ prove patient compliance, compared to currently available non-systemic cholesterol lowering regimens.

Detailed Description of the Invention

This invention pertains to the discovery of compounds that have the ability to bind bile acids and thus, in turn, reduce plasma lipids (e.g., plasma total cholester¬ ol, triglycerides and LDL cholesterol) . These compounds are polysaccharides, which when derivatized, become potent bile acid sequestrants. The derivatized polysaccharide bile acid sequestrants of this invention comprise li-

gand(s) having hydrophobic and cationic moieties chemical¬ ly coupled to a polysaccharide substrate. Suitable poly¬ saccharide substrates are polysaccharides that can be derivatized; i.e., coupled to the ligand(s) . Mixtures of polysaccharides can also be used. Examples of suitable polysaccharides include but are not limited to cellulose, starch, chitosan, guar gum, carrageenan, agar, alginate, a ylose, amylopectin, curdlan, dextran, dextrin, glycogen, laminarin, psyllium, pustulan, nigeran, konjac gluco- mannan, pectin, cyclodextrin and bran. Preferably, the polysaccharide substrate is a whole glucan particle, cellulose or starch.

Whole glucan particles have been described in detail in U.S. Patent Nos. 4,810,646, 4,992,540, 5,037,972, 5,082,936, 5,028,703, 4,962,094 and U.S. Patent Applica¬ tion Serial No. 07/675,913, filed April 26, 1991. The teachings of each reference are incorporated herein by reference. A whole glucan particle is a glucan which maintains the intact, three-dimensional .in vivo morphology of the cells from which it is derived, such as yeast cell walls.

The ligand should be a chemical moiety that has hydrophobic and cationic moieties. The hydrophobic region can contain from about 6 to about 18 carbon atoms. For example, the ligand can be an alkylamine (e.g., hexyl- amine) or an alkanediamine (e.g., 1,6-diaminohexane and 1, 12-diaminododecane) . The carbon atom chain length should be at least about 6 carbons with the upper limit being about 18, with about 6 to about 12 carbon atoms being preferred for alkylamine and alkanediamine moieties. The ligand(s) can be N-methylated or quaternized after being coupled to the polysaccharide substrate. This can be achieved using well known techniques, such as those

techniques described by Sommer et al. , J. Org. Che . , 3_6:824-828 (1971).

A multiplicity of ligands can be coupled to the substrate. It is preferred to maximize the number of ligands on the substrate. It has been demonstrated that multiplicity of ligands contributes to the compound's ability to efficiently bind bile acids. The degree of derivatization of the polysaccharide should be at least above about 20%. For alkanediamine derivatives the nitro- gen content of the resulting derivative should be above about 2 percent. Depending upon the polysaccharide sub¬ strate selected, it is possible to achieve 100 percent derivatization of the residues. For example, all of the free hydroxyl groups on the glucose residues that comprise cellulose or starch can be derivatized.

The compounds of this invention can also be made by chemically attaching ligand(s) of choice to a carbonyl containing polysaccharide. A carbonyl containing polysac¬ charide is intended to embrace polysaccharides which naturally contain carbonyl groups (e.g., alginate, pectin, gellan, xanthan, welan) and those generated by oxidation or carboxyalkylation. Examples of oxidation methodologies include but are not limited to periodate oxidation, bro¬ mine oxidation and Moffat oxidations and hypochlorite oxidation, using techniques well known in the art.

In the case of periodate oxidative polysaccharide activation which is the preferred method, the ligand will be either an alkylamine or an alkanediamine. Other known polysaccharide activation techniques can also be used in addition to those recited above.

In one embodiment, the compound comprises a whole glucan particle (WGP) which is derivatized with 1,12- diaminododecane (also referred to herein as dodecane-

diamine or C12) . The compounds are referred to herein as WGP-C12M and WGP-C12, where M represents the methylated version. The N-methylated version of this compound is preferred over the non-methylated version. It has been shown in Table 2 that the amount of bile acids bound by WGP-C12M is greater than that bound by WGP-C12. The preferred method of coupling the C12 ligand to WGP is by reductive amination to the periodate oxidized polysac¬ charide. This method facilitates the coupling of multiple ligands to the whole glucan particles.

In another embodiment, cellulose is coupled to alkanediamine or alkylamine ligand(s) which can preferably be methylated. The preferred ligand is 1, 12-diaminodo- decane which is N-methylated after coupling to cellulose. This embodiment is referred to herein as cellulose-C12 or cellulose-C12M, where M represents the methylated version. Periodate oxidation is the preferred method for oxidizing cellulose thus enabling essentially all residues on the cellulose substrate to be coupled to dodecanediamine. In yet another embodiment, periodate-oxidized starch can be reductively coupled to alkanediamine or alkylamine ligands which can subsequently be methylated. The pre¬ ferred coupled ligand is methylated 1, 12-diaminododecane. This compound is referred to herein as starch-C12 or starch-C12M, where M represents the methylated version. Compounds were prepared and evaluated in vitro for chloride ion exchange capacity and bile acid binding capacity. The results of the chloride ion exchange assay indicate the extent of ligand derivatization. The results of the n vitro bile acid binding assay are predictive of the compound's ability to reduce plasma cholesterol lev¬ els. Table 1 shows chloride exchange and taurocholate binding capacities of polysaccharide substrates, with and

without the C12 ligand. The non-derivatized polysaccha¬ rides demonstrate essentially no chloride exchange or taurocholate binding capacity. Upon derivatization, whole glucan particles, starch and cellulose are transformed into bile acid sequestrants. The binding capacities were compared to cholestyramine.

Table 1

CHLORIDE TAUROCHOLATE TAUROCHOLATE:

EXCHANGE BINDING

CAPACITY CAPACITY CHLORIDE

SUBSTRATE (mmol/α) (mmol/q) RATIO

WGP 0.1 0.05 —

WGP-C12M 1.8 1.3 0.72

Cellulose 0 0 —

Cellulose-C12M 2.9 1.8 0.62

Starch 0 0 —

Starch-C12M a 1.6 —

Cholestyr¬ 3.4 1.6 0.47 amine

Not measured

Table 2 shows the chloride exchange and taurocholate binding capacities of various alkylamine and diaminoal- kanes coupled to periodate-oxidized whole glucan parti¬ cles. These derivatized compounds were optionally methyl¬ ated. The results indicate that methylation improves taurocholate binding capacity and that the length of the ligand affects both chloride exchange and taurocholate binding capacity.

Table 2

CHLORIDE TAUROCHOLATE TAUROCHOLATE EXCHANGE BINDING LIGAND CAPACITY CAPACITY :CHLORIDE

ALKYLAMINES METHYLATED (mmol/q) (mmol/cr) RATIO hexyl no 0.2* 0.1 0.5 hexyl yes 0.2 0.2 1.0 dodecyl no 0.2 ¹ 0.2 1.0 dodecyl yes 0.2 0.2 0.8 hexadecyl yes 0.3 0.2 0.7 octadecyl yes 0.2 0.1 0.5

DIAMINOALKANES hexane no 1.4 0.5 0.4 hexane yes 1.4 0.7 0.5 octane yes 1.2 0.7 0.6 decane yes 1.2 0.8 0.7 dodecane no 1.1" 0.5 0.5 dodecane yes 1.1 1.0 0.9

1. The chloride exchange capacity for non-methylated derivative is assumed to be the same as that of the methylated derivative.

Table 3 reports the chloride exchange and taurocho¬ late binding capacities of some non-methylated deriva¬ tives. The compounds were prepared by four different methods of coupling the ligand (e.g., C12) to the polysac¬ charide.

Table 3

Properties of Polysaccharide-Alkylamine Derivatives

TAURO-

CHLORIDE CHOLATE

EXCHANGE BINDING TAUROCHOLATE:

CAPACITY CAPACITY CHLORIDE

MATERIAL COUPLING (mmol/q) (mmol/q) RATIO

Starch-C12 Periodate 1.4 1.0 0.7

WGP-C12 Bromine 4.1 0.5 0.1

WGP-C12 DMSO-Acetii z 2.0 0.7 0.4

Anhydride

WGP-C12 Carboxy- 3.5 1.0 0.3 methylation

The ability of the compounds to lower plasma lipids in vivo was evaluated by determining percent reduction of total cholesterol, LDL cholesterol and triglycerides in the blood of hyperlipidemic hamsters. Plasma lipids were measured in hamsters fed a high fat diet (HFD) only and those fed a high fat diet with drug intervention. In each case, the plasma lipids were compared to hamsters fed cholestyramine with the high fat diet. The amount of bile acid sequestrant fed to the hamsters (approximately 1.5% of the diet) was equilibrated to cholestyramine which represented 0.5% of the diet in order for adequate compar- ison.

Table 4 (see Example 15) shows a direct comparison of several methylated and non-methylated polysaccharide derivatives or cholestyramine given to hamsters fed high fat diets. There was no observed reduction of total cholesterol, LDL cholesterol or triglycerides in hamsters

fed only the high fat diet. Significant reductions in plasma lipids were achieved in hamsters administered WGP- C12M, compared to cholestyramine. Methylated derivatized polysaccharide bile acid sequestrants showed better reduc- tion than did their non-methylated counterparts.

Drug intervention studies over four, six and twelve week periods were conducted to determine whether the derivatized polysaccharide sequestrants could maintain or improve the reduction in plasma lipids over prolonged drug intervention. The results show that the novel seques¬ trants reduced total cholesterol and triglycerides at four weeks (Table 5) and at six weeks (Table 4) . This decrease was maintained at twelve weeks (Table 6) . The results suggest that use of cholesterol lowering regimens over extended time periods will provide an individual with sustained reduction in total cholesterol, LDL cholesterol and triglycerides. Reductions obtained with the novel sequestrants were similar or greater than those achieved with cholestyramine. Continued maximal reductions in plasma lipids as achieved with the novel bile acid sequestrants of this invention provide a significant advantage over cholestyr¬ amine. Cholestyramine has been shown to lower LDL levels to 20-40% where its dose dependent reduction then plateaus at 43 percent, at up to 2% of diet. Groot et al. Bioch.

Biophys. Acta. , 1123 :76-84 (1992). This plateau effect is not observed with the compounds described herein, particu¬ larly with WGP-C12M. Maximal reductions in plasma lipids beyond that achieved using cholestyramine are particularly significant for those individuals with dangerously high plasma lipid levels.

It is believed that the novel derivatized polysaccha¬ ride bile acid sequestrants of this invention reduce

plasma lipid and lipoprotein levels by increasing the fecal excretion of bile acids and their salts, such as deoxycholate. It is well known that increasing loss of bile salts leads to an increased oxidation of cholesterol to bile acids, decreased levels of lipoprotein (LDL and VLDL) and decreased cholesterol in serum. Based on the mechanism of bile acid binding it is reasonable to expect that the compounds of this invention can play an important role in cholesterol homeostasis. The novel drugs can be administered orally to living animals, e.g., humans, by any suitable means, in any suitable form. For example, the drugs can be incorporated into ordinary foodstuffs and beverages in an amount suffi¬ cient to produce the desired clinical effect. The drugs can also be incorporated into pharmaceutical compositions customarily employed for oral administration.

Pharmaceutical compositions containing the drug can be formulated into a liquid suspension or in solid form, for example, tablet, capsule, pill or packaged powder. These compositions can be prepared using pharmaceutically acceptable carriers or diluents, such as, for example, starch, glucose, lactose, gelatin, sucrose, water, aqueous dilute ethanol, propylene glycol, glycerol and sorbitol. Such formulations can also include flavoring and sweeten- ing agents such as fructose, inert sugar, aspartame, cocoa, citric acid, ascorbic acid and fruit juices. Sus¬ tained release forms of administration are also accept¬ able. The amount administered will vary depending among other things on the size of the animal, the type of animal (e.g., human) and the general health of the animal.

The drugs of this invention can be administered in combination with other known drugs which reduce serum cholesterol, such as niacin, competitive inhibitors of

HMG-CoA reductase, Probucol and a fibric acid derivative (Ge fibrozil) . One or a combination of these compounds can be co-administered with the sequestrants of this invention. The invention will be further illustrated by the following examples:

Examples

Example 1 - Synthesis of GP-C12

Whole glucan particles (1 equivalent) (described in U.S. Patent Nos. 4,810,646, 4,992,540, 5,037,972,

5,082,936, and 5,028,703) were suspended in water and aqueous sodium metaperiodate (1.7 equivalents) was added with stirring. The suspension was stirred for 4 hours and quenched with the addition of glycerol (0.7 equivalents). After 30 min, the suspension was centrifuged and the pellet was washed (i.e., suspended in water, centrifuged and the supernatant discarded) . The water wash was re¬ peated 3 additional times and the pellet was then washed with ethanol twice. The pellet was suspended in ethanol and an ethanol solution of 1,12-diaminododecane (2.6 equivalents) was added with stirring. The solution was stirred overnight, centrifuged and the pellet was suspend¬ ed in water. Aqueous sodium hydrogen carbonate (6 equiva¬ lents) was added, followed by solid sodium borohydride (3.6 equivalents). The suspension was stirred overnight and the reaction was quenched with the addition of ace¬ tone. The suspension was centrifuged and the supernatant decanted. The pellet was washed with water, followed by ethanol 3 times, giving wet pellet 1. The pellet was fur- ther washed in water 2 times, 2-propanol 2 times and

acetone 2 times and dried in vacuo to provide WGP-C12 having a nitrogen content of 4.0%.

Example 2 - Synthesis of GP-C12M

Wet pellet 1 (1 equivalent) was suspended in a sealed container with water. Methyl iodide (12 equivalents) was added and mixed for 3 days. The suspension was centri¬ fuged and the pellet was washed with ethanol 2 times. The pellet was then washed in aqueous saturated sodium chlo¬ ride 5 times, water 6 times, 2-propanol 2 times, and ace- tone 2 times, and dried in vacuo to give WGP-C12M having a nitrogen content of 4.2%.

Example 3 - Synthesis of Cellulose-C12M

Cellulose (Sigmacell, type 50) was oxidized, coupled to 1, 12-diaminododecane and reduced as described in Exam- pie 1 and methylated, washed and dried as described in

Example 2, to give cellulose-Cl2M, with a nitrogen content of 5.6%.

Example 4 - Synthesis of WGP Coupled to Diaminooctane and Diaminodecane Whole glucan particles were oxidized, coupled to the appropriate diaminoalkane and reduced as described in Example 1 and methylated, washed and dried as described in Example 2.

Example 5 - Synthesis of Cellulose-C6 Cellulose (Sigmacell, type 50) was oxidized, coupled to 1, 6-diaminohexane and reduced as described in Example 1, except that the coupling was performed in water, giving cellulose-C6 with a nitrogen content of 3.9%.

Example 6 - Synthesis of GP-C6M

WGP-C6 was methylated, washed and dried as described in Example 2, giving WGP-C6M with a nitrogen content of 3.4%.

Example 7 - Synthesis of Cellulose-C6M

Cellulose was oxidized, coupled to 1,6-diaminohexane and reduced as described in Example 5 and methylated, washed and dried as described in Example 2, giving cellu- loεe-C6M with a nitrogen content of 2.5%.

Example 8 - Synthesis of WGP Coupled to Alkylamines

Whole glucan particles were oxidized, coupled to the appropriate alkylamine and reduced as described in Example 1, except that the coupling was performed in 1,4-dioxane and methylated, washed and dried as described in Example 2.

Example 9 - Synthesis of Starch Starch-C12M via Periodate Oxidized Precursor Starch dialdehyde obtained as in Example 1 (10 equiv¬ alents) was dispersed in ethanol or aqueous ethanol and treated with 1,12-diaminododecane (3-24 equivalents) for 4 h and subsequently with sodium cyanoborohydride (20 equiv¬ alents) for a further period of 4-48 hours. Alternative¬ ly, the reduction was performed using sodium borohydride as reductant, with the reductant being added shortly after the alkanediamine ligand. Optionally, the material was also treated with a cross-linking agent, e.g., epichloro- hydrin (0.01 equivalents). The resulting mixture was then filtered, extensively washed with water, 2-propanol and acetone, and dried, yielding starch-alkanediamine deriva- tives with nitrogen contents of 2.1-9.1%. The starch-

alkanediamine derivatives were subsequently methylated as described in Example 2.

Example 10 - Synthesis of Whole Glucan Particle-Alkane- diamine Derivative via Bromine Oxidized Precur- sor

An aqueous suspension of whole glucan particles (1 equivalent) was treated with aqueous bromine (0.1M, 1 equivalent) at room temperature. After 15 minutes, NaOH (1.5N) was added to neutralize the suspension, and the neutralization was periodically repeated over the course of 6 hours, before the mixture was filtered, extensively washed with water and 2-propanol, and dried in vacuo at 50°C to give a white product. Alternatively, the oxida¬ tion was carried out over longer periods. The oxidized whole glucan particles were then coupled to 1, 12-diaminododecane as described in Example 1 to give the corresponding whole glucan particle-alkanediamine derivatives with nitrogen contents of 2.6%. The glucan- alkanediamine derivatives were subsequently methylated as descried in Example 2.

Example 11 - Synthesis of Whole Glucan Particle-Alkane- diamine Derivative via Dimethylsulfoxide-Acetic Anhydride Oxidized Precursor A suspension of whole glucan particles (1 equivalent) in dimethylsulfoxide was treated with acetic anhydride (1 equivalent) at room temperature. After 15 minutes addi¬ tional solvent was added to dilute the resulting gel-like material, which was then stirred for an additional period of 4-48 hours. The oxidized material was precipitated with acetone, filtered, extensively washed with acetone and dried in vacuo at 50°C. Alternatively, different

ratios of oxidant (e.g., 0.5 equivalents) were used to give products with different (e.g., lower) degrees of oxidation.

The oxidized whole glucan particles were then coupled to 1,12-diaminododecane as described in Example 1 to give the corresponding whole glucan particle-alkanediamine derivatives with nitrogen contents of 2.9%. The whole glucan particle-alkanediamine derivatives were subsequent¬ ly methylated as described in Example 2.

Example 12 - Synthesis of Whole Glucan Particle-Alkane- diamine Derivative via Carboxyalkyl Precursors A suspension of whole glucan paricles (1 equivalent) in 50% aqueous NaOH was treated with chloroacetic acid (1 equivalent) dissolved in 2-propanol at room temperature. After stirring for 24 h, another portion of chloroacetic acid (0.3 equivalents) was added, the mixture was stirred for 4 h, extensively dialyzed and lyophylized. Alterna¬ tively, the glucan particles (1 equivalent) dispersed in a minimum amount of water were treated with aqueous NaOH (3 equivalents) then treated with chloroacetic acid (1.5 equivalents) , dissolved in methanol at room temperature for 30 h, filtered, washed with alcohol, and dried.

The resulting carboxymethylated whole glucan parti¬ cles were subsequently coupled to 1,12-diaminododecane as described in Example 1 to give the corresponding whole glucan particle-alkanediamine derivatives with nitrogen contents of up to 9.0%. The glucan-alkanediamine deriva¬ tives were subsequently methylated as described in Example 2.

Example 13 - Chloride Titration Assay

Bile acid sequestrant synthesized according to Exam¬ ples 1-12 or cholestyramine (10-15 mg) was suspended in aqueous silver nitrate (0.6-0.7 mL, 0.1 M) and water (0.8- 1.0 mL) was added. The suspension was stirred overnight and centrifuged. The supernatant (0.5-0.6 mL) was trans¬ ferred to an Erlen eyer flask and diluted with water (15 mL) . Aqueous iron (III) ammonium sulfate (1.0 mL of 2.5 g NH ₄ Fe(S0 ) ₂ .12H ₂ 0 dissolved in 25 mL of 6 M nitric acid) was added as an indicator. The resulting solution was titrat¬ ed with aqueous potassium thiocyanate (0.005 M) until an orange color persisted for 1 minute.

Example 14 - Bile Acid Binding Assay by Enzymatic Analysis The in vitro bile acid binding capacity of the com- pounds of this invention was evaluated using an enzymatic method. A reaction solution containing 6.38 g NaHC0 ₃ , 2.34 g NaCl and 2.24 g KC1 was added to deionized water to a total of 1000 mL. The pH was adjusted to 6.2 with concentrated HC1 (30 mM) . A series of labeled 15 mL screw cap tubes were set up, each tube containing 10.0 mg of either test sample, cholestyramine or underivatized poly¬ saccharide; 3.0 mL of 10 mM taurocholate (268.9 mg/50 mL reaction solution) and 3.0 mL of reaction solution. Each tube was capped, mixed and shaken at 37°C for 2 hours. The tubes were then centrifuged at 3000 g for 15 minutes. An 0.1 L aliquot of supernatant was transferred from each tube to a fresh set of tubes (1.5 mL) , and 0.4 mL of water was added to each tube to dilute the supernatant.

Each tube was separately analyzed using an enzymatic bile acid assay. Twenty μl supernatant (sample, control or standard) , 20 μl Bovine Serum Albumin (HyClone) and 100 μl test or blank bile acid reagent (Bile Acid Analysis

Kit, Sigma #450-A) were combined, mixed well and reacted at 37°C for 10 minutes in a 96 well microtiter plate. Optical density (OD) was read at 550 nm in a THERMO max microplate reader (Molecular Devices) . The taurocholate binding capacity in mmol/g was calculated from the difference in taurocholate concentra¬ tion of the supernatant before and after exposure to sequestrant.

The results of these analyses are tabulated in Tables 1-3 above.

Example IS - Effects of Bile Acid Sequestrant on Choles¬ terol Lowering In Vivo in Hyperlipidemic Ham¬ sters

Experiment I - Six Week Drug Intervention Male, Golden Syrian hamsters (Charles River Laborato¬ ries, Wilmington, MA, 6-8 weeks old, weighing 80-100 g) were used. They were fed a high fat diet (HFD) ad libi¬ tum. The high fat diet consisted of coconut oil (10%) , cholesterol (0.2%), and hamster chow (#5001, Purina Mills, Inc., St. Louis, MO, 89.8%), plus water (30% relative to the dry material) to form a cake. After 6 weeks of HFD feeding, total cholesterol, LDL-cholesterol, and triglyc- eride plasma levels were measured and the hamsters were distributed into groups of ten in such a manner as to give equal mean total cholesterol and body weights in each group. The diets were then changed to the above mixture with incorporated drug, except for the HFD control group which did not receive drug intervention. After 6 weeks of drug intervention, plasma lipids were again measured and the results are reported in Table 4 below.

Table 4

TOTAL LDL+VLDL

CHOLESTEROL CHOLESTEROL TRIGLYCERIDES

TREATMENT ¹ mg/dL %red'n mg/dL %red'n mg/dL %red'n

HFD 824 — 450 — 1291 —

Cellulose- 545 34 268 40 916 29

C6M

WGP 548 33 332 26 673 48

WGP-C6 572 31 420 7 575 55

WGP-C6M 500 39 272 40 500 61

WGP-C12M 113 86 30 93 124 90

Cholestyr¬ 459 44 248 45 461 64 amine

Polysaccharide derivative was 1.5% of diet; cholestyramine was 0.5% of diet.

Experiment II - Four Week Drug Intervention

The experiment was performed as described for Experi¬ ment I except that drug intervention and HFD began simul¬ taneously, 5 animals were used per group, and plasma samples were assayed after 4 weeks. The results after four weeks of drug intervention are reported in Table 5 below.

Table 5

TOTAL CHOLESTEROL TRIGLYCERIDES

TREATMENT mg/dL % red'n mg/dL % red'n HFD 780 1568

WGP-C12 314 60 621 60 WGP-C12M 62 92 177 89

1. Polysaccharide derivative was 1.5% of diet.

Experiment III - Twelve Week Drug Intervention

The experiment was performed as described for Experi¬ ment I except that the cholesterol content in the HFD was decreased to 0.1% and that the plasma samples were assayed after 12 weeks of drug intervention (0.5% of diet). The results after twelve weeks of drug intervention are re¬ ported in Table 6 below.

Table 6

TOTAL LDL+VLDL

CHOLESTEROL CHOLESTEROL TRIGLYCERIDES

TREATMENT mg/dL %red'n mg/dL %red'n mg/dL %red'n

HFD 444 — 365 784 —

Cellulose- 132 70 66 82 152 81

C12M

WGP-C12M 210 53 126 66 222 72

Cholestyr¬ 336 24 263 28 488 38 amine

Example 16 - Dose Response Study in Hγpercholesterole ic Hamsters

One hundred and thirty-two golden Syrian hamsters (BioBreeders, Inc., Fitchburg, MA; 80-100g; 6-8 weeks old) were used. They were fed a high fat diet (HFD) ad libitum consisting of coconut oil (5%), cholesterol (0.1%) and hamster chow (#5001, Purina Mills, Inc., St. Louis, MO) in the form of a cake.

After 6 weeks of HFD feeding, the hamsters were divided into eight groups (12 hamsters per group) as follows: Group 1: HFD (control)

Group 2: 0.5% cholestyramine + HFD Group 3: 1.5% cholestyramine + HFD Group 4: 0.2% sequestrant + HFD Group 5: 0.5% sequestrant + HFD Group 6: 1.0% sequestrant + HFD Group 7: 1.5% sequestrant + HFD Group 8: 0.1% sequestrant + HFD Group 9: 0.2% sequestrant + HFD Group 10: 0.5% sequestrant + HFD Group 11: 1.0% sequestrant + HFD

The sequestrant administered in this study is described in Example 3. Groups 1-7 were fed the above diets for 8 weeks. Groups 8-11 were fed a diet of 1.5% sequestrant and HFD for 4 weeks and then the above diets for the next four weeks.

After the eight week period, total cholesterol (TC) , triglyceride, HDL-cholesterol, LDL-cholesterol, liver cholesterol and fecal bile acids were measured using the protocols described in Example 17. Hepatic cholesterol concentrations are shown below in Table 7.

Table 7

Liver Weight TC* FC %EC

1 (g) (mg/g liver)

Control 5.8 ± 0.9 ^b 30.3 + 7.7 9.1 ± 2.4 70 ± 3

Cholestyramine

0.5% 5.6 + 0.8 28.5 ± 4.5 10.1 + 1.8 65 ± 3

1.5% 4.1 ± 0.8 3.1 ± 0.3 2.3 ± 0.2 26 ± 4

Sequestrant

0.2% 5.5 ± 0.7 26.7 ± 0.3 9.3 ± 1.9 65 ± 4

0.5% 5.4 ± 0.6 12.8 + 6.2 4.2 ± 1.2 63 ± 13

1.5% 4.1 ± 0.6 3.4 ± 0.2 2.9 ± 0.2 14 ± 5

Abbreviations: TC, total cholesterol; FC, free cho- lesterol; EC, esterified cholesterol. Values are mean ± SD (n = 12)

The effect of sequestrant on total cholesterol levels was to decrease the TC concentration in plassma as the amount of sequestrant in the diet increased (Table 8) . The number of hamsters per group was twelve.

Table 8

Total Percent

Cholesterol Reduction

Diet (mg/dL) in TC

HFD 462 ± 211

HFD + 0.2% sequestrant 342 ± 135 -26%

HFD + 0.5% sequestrant 225 ± 48 -51%

HFD + 1.0% sequestrant 132 ± 17 -71%

HFD + 1.5% sequestrant 97 ± 12 -79%

Fecal bile acids were measured in the control animals (0.64 ± 0.13 mg/day/hamster); animals fed HFD + 1.5% cholestyramine (1.04 ± 0.34 mg/day/hamster); and animals

fed HFD + 1.5% sequestrant (3.52 ± 0.71 mg/day/hamster). Hamsters fed cholestyramine had 162% greater fecal bile acid content than control. Animals fed novel sequestrant had 551% greater fecal bile acid content than control. The results show that the novel sequestrant had the abili¬ ty to bind and sequester a significantly greater amount of fecal bile acids compared to the same amount of cholestyr¬ amine.

The effect of varying the dosage of sequestrant on total cholesterol levels in plasma is shown in Table 9.

Table 9

Total Cholesterol (mg/dL)

Diet -6 Weeks 0 Weeks 4 Weeks 8 Weeks

HFD 92 ± 11 432 ± 184 369 ± 159 462 ± 211

HFD + 0.5% Cholestryamine 92 ± 11 432 ± 181 306 ± 126 298 ± 132

HFD + 0.5% Sequestrant 92 ± 11 432 ± 144 266 ± 102 225 ± 48

HFD + 1.5% Sequestrant 92 ± 11 432 ± 139 135 + 22 97 ± 12

The maintenance of total cholesterol levels in plasma with varying concentrations of the novel sequestrant is shown in Table 10.

Table 10

Total Cholesterol (mg/dL)

Diet -6 Weeks 0 Weeks 4 Weeks 8 Weeks

HFD 92 ± 11 432 + 184 369 ± 159 462 ± 211

HFD + 1.5% Sequestrant 92 ± 11 432 ± 140 135 ± 22 97 ± 12

HFD + 0.1% Sequestrant 122 ± 15 185 ± 38

HFD + 1.0% Sequestrant 148 ± 30 115 ± 16

Example 17 - Analytical Methods A. Analysis of Fecal Bile Acids

Forty samples from 8 groups were qualitatively and quantitatively analyzed using a gas-liquid chromatographic method. In brief, 0.25 g of ground feces, with 5α-choles- tane added as an internal standard, were extracted with 3 mL of absolute ethanol at 100°C overnight, cooled and another 3 mL of ethanol added and shaken in a shaker for 3 hours. After centrifugation at 3,500 g for 20 minutes, 2 mL of supernatant was taken and dried and then subjected to derivatization of the bile acids and neutral sterols with Tri-sil at 80°C for 20 minutes and analyzed as their trimethysilyl ethers by gas-liquid chromatography (Hewlett-Packard HP 5700) using DB-5 fused silica column (15 m X 0.25 mm. J & #122-5012).

B. Analysis of Liver Cholesterol Sample preparation: Hepatic cholesterol was extract¬ ed by grinding a 100 mg portion of liver with anhydrous sodium sulfate and extracting 3 times with chloroform: methanol (2:1 v/v) . The supernatant was evaporated under

a dry air stream, redissolved in 1 mL of chloroform and an aliquot evaporated and redissolved in isopropanol.

Sample analysis: Total and free cholesterol were determined using a microplate method, using the ako "Free cholesterol C" kit for free cholesterol and the Sigma kit #352 for total cholesterol. Esterified cholesterol was calculated as total minus free cholesterol concentration.

C. Analysis of Triglycerides

Plasma samples were diluted on an ELISA plate with 0.9% NaCl. Triglyceride reagent (lipoprotein lipase, glycerol kinase, ATP, glycerol phosphate oxidase, 4-amino- antipyrine, peroxidase and sodium-N-ethyl-N-3-sulfopropyl m-anisidine; Product #339, Sigma Chemicals) was dissolved in water and standard diluted to 50 mg/dL. A 96 well ELISA plate was set up in duplicate for each sample and 200 μL of triglyceride reagent were placed in each well. Twenty μL of water (blank) , calibrator (50 mg/dL) and sample were placed in the wells and mixed. The plates were incubated for 10 minutes at room temperature. The amount of triglyceride was read in a Microplate Reader (Model THERMO max, Molecular Devices) at 550λ.

D. Analysis of Total Cholesterol

Plasma samples were diluted on an ELISA plate using 0.9% NaCl. A 96 well ELISA plate containing 200 μL of cholesterol reagent (cholesterol esterase, cholesterol oxidase, p-hydroxybenzene, sulfonate, 4-aminoantipyrine peroxidase; Product #352, Sigma Chemicals) was set up. Twenty μL of water (blank) , calibrator (cholesterol cali¬ brator, Product #C9908, Sigma Chemicals) and samples were placed in appropriate wells and mixed. The plate was incubated for 10 minutes and the amount of total choles-

terol read in Microplate Reader (Model THERMO max, Molecu¬ lar Devices) at 490λ.

E. Analysis of HDL-Cholesterol

A series of 0.5 mL microcentrifuge tubes each con- taining 100 μL plasma were set up. Twenty μL HDL choles¬ terol reagent (phosphotungetic acid and magnesium chlo¬ ride. Product #352-4, Sigma Chemicals) was added to each tube and mixed with a vortex mixer. The tubes were al¬ lowed to stand for 5 minutes and were then centrifuged for 5 minutes at 5000 rpm in an Eppendorf Centrifuge (5415) . A 96 well ELISA plate was set up and the procedure for determining total cholesterol described above was fol¬ lowed.

F. Analysis of LDL-Cholesterol Amount of LDL-cholesterol was determined as the difference between TC and HDL.

Equivalents

Those skilled in the art will recognize, or be able to ascertain using not more than routine experimentation, many equivalents to the specific embodiments of the inven¬ tion described herein. Such equivalents are intended to be encompassed by the following claims: