This invention relates to a nondenaturing zwitterionic detergent for proteins which, for example, consists of an effective amount of 3-[(3- cholamidopropyl)-dimethylammonio]-1-propanesulfonate (CHAPS) . This detergent is of extreme interest in the biological study of proteins due to its nondena¬ turing characteristic. Other examples of the group may be prepared from different alicyclic compounds, for example, utilizing cholic acid and in others deoxycholic acid and dehydroabietic acid. A process- for the preparation of these compounds starts with cholic or the equivalent and from this is prepared the triethylammonium salt in tetrahydrofuran (THF) . After the salt is completely dissolved in THF, ethyl chloroformate is added and the flask cooled to 0°C. Then the mixed anhydride which forms is reacted with dimethylaminopropylamine to form the dimethylamino- propyl derivative of a carboxylic acid amide.

Finally, the tertiary amine group is reacted with propanesulfone to give the sulfobetaine product.

This application is an improved procedure for preparation of these compounds and especially for the last step as for CHAPSO to react the N- (3-dimethyl- hydroxy-3-propanesulfonate.

This invention relates to a nondenaturing zwit- terionic detergent for marforane biochemistry, both in its design and synthesis. This detergent combines the useful properties of both the sulfdbetaine type detergents and the bile salt anions. The new detergent proves to be effective at solubilizing membrane proteins in a nondenatured state.

PRIOR ART STATEMENT

Gαnenne, et al., "Solubilizatiαn of t-feπbrane Proteins by Sulfdbetain-εs, Novel Zwitterionic Surfactants", .analytical Bio-chemistry, 87:28-38 (1978).

Hjelmeland, et al., "Electrofocusing of Integral Membrane Proteins in Mixtures of Zwitterionic and Ncnionic Detergents", itoalytical Biochemistry, 95:201-208 (1979).

Parris, et al., "Soap-Based Detergent Forrπulatioris: XII. Alternate Syntheses of Surface Active Sulfαbetaines", J. Amer. Oil Chem. Soc., 53:60-63 (1976)

Parris, et al., "Soap-Based Detergent Formulations: XVIII. Effect of Structure Variations on Surface-i^ctive Properties of Sulfur Containing Arπ hoteric Surfactants", J. Amer. Oil Chan, Soc., 53:97-100 (1976).

Parris, et al., "Soap Based Detergent Formulation: XXIV. Sulfobetaine Derivatives of Fatty Amides", J. Amer. Oil Chan. Soc., 54:294-296 (1977).

Parris, et al., "Soap Based Detergent Formulations: V. Amphoteric Lime Soap Dispersing Agents", J. Amer. Oil Chgn. Soc., 50:509-512 (1973).

Herrmann, "Micellar Properties of Some Zwitterionic Surfactants", ^' J. Colloid Interface Sci. , 22:352-359 (1966).

Parris-, et al., "Determination of Sulfo- betaine Amphoteric Surfactants by Reverse Phase High Performance Liquid Chromatography", Anal. Chem. , 49:2228-2231 (1977).

Konig, Z. Anal. Chem., 259:191-194 (1972).

British Patent 1,037,645

DEFINITION

In this specification the following defini¬ tion from Condensed Chemical Dictionary, 9_, page 107, is given. Bile acid and salt are defined as an acid and acid salt found in bile (the secretion of liver) ; bile acids are steroids having a hydroxy group and a 5-carbon atom side chain terminating in a carboxyl group; cholic acid is mentioned as the most abundant bile acid in. human bile.

INTRODUCTION

One of the more important aspects of the purification of membrane proteins is. the choice of a suitable detergent. This choice is usually based on the ability to preserve an enzymatic activity or some other native property. In this respect, non- ionic detergents such as Triton X-100 (octylphenoxy

polyethoxy ethanol; Rohm and Haas) and Lubrol PX (blend of nonyl phenol and ethylene oxide; ICI) , and the bile salts are the reagents of choice. Two additional considerations must also be made. The first- relates to the artifactual aggregation of proteins while in the presence of detergents to form-nonspecific protein complexes which have no biological relevance. A useful detergent should be capable of breaking such interactions to give maximally disaggregated species in solution.

Nonionics are generally less efficient in this respect than are ionic detergents or bile salt anions. The second consideration is the extent to which the detergent affects the charge properties of solubilized proteins. Anionic detergents, for example, add substantial amounts of negative charge which may completely overshadow the charge proper¬ ties of the native protein. This type of charge alteration profoundly affects the utility of con- ventional techniques such as ion exchange chroma- tography and isoelectric focusing which depend pri¬ marily on charge properties to effect protein separations..

A survey of existing detergents demonstrates that no single compound which is presently available is adequately nondenaturing, disaggregating, and at the same time electrically neutral. The bile salts are both nondenaturing and effective in dis¬ aggregating protein but lack the charge neutrality necessary for compatibility with charge fractionation techniques. In contrast, Triton X-100 and other polyethoxy-type nonionics are electrically neutral

and nondenaturing but appear not to be efficient at breaking protein-protein interactions. N-alkyl sulfobetaines are neutral and efficient at disaggre¬ gating protein but are unfortunately strongly denaturing. One possibility of a detergent useful in the purification of membrane proteins is a combination of a bile salt hydrophobic group and a sulfobetaine type polar group. This invention describes the synthesis and properties of a sulfo- betaine derivative of cholic acid and evaluates its protential utility in membrane protein purification.

DESCRIPTION OF THE DRAWING

The figure presents data on the effective¬ ness of- CHAPS in solubilizing mouse liver microsomes and shows that it is nondenaturing with respect to cytochrome P-450 at concentrations up to 10 mM. The left-hand panel indicates that CHAPS is capable of solubilizing about 70% of the protein of mouse liver microsomes which can be repelleted in the absence of detergent. Data are given at total protein concen¬ trations of 1, 3 and 5 mg/ml. The sharp break in the three curves between 4 and 6 mM indicates that the critical micille concentration of CHAPS is possibly in this region. The right-hand panel of the figure gives the percentage of the total cytochrome P-450 found in the supernatant at each of .the same protein and CHAPS concentrations. The higher protein con¬ centrations yield recoveries of soluble and nondena- tured cytochrome P-450 in excess of 90%.

C57BL/6 mouse liver microsomes were pre¬ pared by differential centrifugation of tissue homogenized in a 140 mM KC1, 10 mM EDTA buffer at pH 7.25, 4°C. For solubilization experiments, the appropriate amount of protein was diluted to give solubilization media with the stated protein and detergent concentrations, and final concentrations of 20% (v/v) glycerol and 0.1 M K phosphate, pH 7.25. Aliquots of 5 ml were incubated for 30 min. at 25°C, and then centrifuged at 105,000 xg, 25°C for 2 hr.

The left panel shows the amount of protein solubilized at various concentrations of CHAPS, ex¬ pressed as a percentage of the amount of protein which could be recovered in the pellet in the absence of detergent. Protein was measured in both the pellet and the supernatant. The right panel shows the amount of cytochrome P-450 remaining in the supernatant at various detergent combinatio s. These amounts are expressed as percentages of the total P-450 content of the intact microsomes. P-450 was measured as the absorbance at 450 ran minus 490 nm in the reduced vs. reduced + carbon monoxide difference spectrum.

FIG.' 1 shows as a dosage an optimum range for solubilizing proteins by CHAPS here given as about 5 mM to 10 mN which gives 50-90% protein solubilization which is non-denaturing. This is a preferred range.

FIG. 2 shows an evaluation of 3-[ (3-cholamido- propyl)dimethylam onio]-2-hydroxy-l-propanesulfonate with a 2:1 advantage over CHAPS in solubilizing active opiate receptors. SOLUBILIZING PROTEIN WITH CHAPS

As a new detergent to be used in the solu¬ bilization of membranes, CHAPS combines the useful features of both the bile salts and the n-alkyl

sulfobetaines. Like the sulfobetaines, CHAPS proves to be better at solubilizing protein than structurally related carboxylic acid anions. Although it is difficult to compare solubilization results from different tissues under different experimental condi¬ tions, studies of the efficiency of solubilization by Na cholate suggest that CHAPS is a substantially better detergent. The data presented here were obtained with microsomes which were treated with 10 mM EDTA and 150 mM KCl to remove extrinsic proteins, which account for perhaps 30% of the most easily solubilized protein in these membranes. By this criterion, CHAPS behaves more like Na deoxycholate in its ability to solubilize total protein although it is structurally more related to Na cholate. CHAPS is, however, much more effective at breaking protein-protein interactions than either Na cholate or Triton X-100. Cytochome P-450 is normally highly aggregated in solutions containing either of these detergents, but CHAPS disaggregates P-450 to its monomeric form.

The increased capacity of CHAPS to solubilize protein and disaggregate complexes is not gained at the expense pf increased denaturing properties. CHAPS is nondenaturing with respect to P-450 under the con- ditions employed in this invention. This is in contrast to Na deoxycholate, which denatures P-450 under similar conditions. A recent study of the opiate receptor found CHAPS to be the only detergent capable of solu¬ bilizing the receptor in a state exhibiting reversible binding of opioids. These are indirect measures of the physical interactions of CHAPS with these proteins and the term"denaturing" is used in its loosest sense.

The more important but less obvious advan¬ tage of CHAPS.as a detergent for solubilizing mem- branes is its compatibility with charge fractionation

OMPI

techniques. Sulfobetaines, while being zwitterionic, behave essentially as nonionic compounds. Specifically, they possess no net charge at any pH between 2 and 12, they exhibit no conductivity or electrophoretic mobility, and do not bind to ion exchange resins. This gives sulfobetaine-type detergents a tremendous advantage in both ion exchange chromatography and isoelectric focusing. The utility of n-alkyl sulfobetaines in isoelectric focusing has been demonstrated and pre- liminary experiments with CHAPS indicate no interfer¬ ence with the formation or stability of pH gradients in this technique.

GENERALIZED PROCEDURE

Below is a step-by-step outline of the synthesis described in this invention.

Step 1. The triethylammonium salt of cholic acid is formed in THF.

. Step 2. After the salt is completely dissolved in THF, ethyl chloroformate is added and the flask is cooled to 0°C. At this point, a precipitate is formed which is triethylamine hydro- chloride. This is filtered away from the mixed anhydride.

Step 3. The mixed anhydride then reacts with the dimethylaminopropylamine to form a dimethyl- aminopropyl derivative of a carboxylic acid amide, ethanol, and carbon dioxide as a gas.

Step 4. In the final step, the tertiary amine group is reacted with propane sultone to give the sulfobetaine.

Step 4a. More preferably, in the final step the tertiary amine group is reacted with sodium 1-chloro- propanesulfonate to give the final product.

Step 1

S ep 2

Step 4

The three important synthetic elements in the final molecule are:

1. The starting carboxylic acid

2. The polyamine used to generate the functional!zed amide

3. The alkylating agent used to quaternize the tertiary amine and _^ give .the final product.

OMPI

The carboxylic acids may be:

Cholic Acid

Deoxycholic Acid

or

droabietic

OMPI

The dehydroabietic acid is important since it represents the most abundant naturally occurring group organic acids, the rosin acids which are refined from wood rosin and tall oil and have wide industrial uses, among them, the synthesis of surfactants.

The polyamines .may be:

Polyethyleneamines

(1) NH N<

-N . (2) NH

or

Polypropyleneamines

Others - such as N-methyl piperazine

I O O ^R MP

The alkylating agents may be:

(1) Propanesultone,

(2) Butanesultone,

(3) Na ⁺ Sodium 3-C1, 2-OH- Propanesulfonate.

(4) Na Sodium 2-Br Ethanesulfonate, or

(5) Cl Sodium chloroacetate.

Specific operative compounds are;

(CHAPS)

(CHAPSO)

EXAMPLE 1

Synthesis of 3-[(3-cholamidopropyl)dimethyl- ammonio]-1-propanesulfonate.

A solution of 40.86 g (0.1 mole) of cholic acid in 500 ml of anhydrous THF was prepared in a one- liter round bottom flask equipped with a drying tube. To this solution was added 13.95 ml (0.1 mole) of anhydrous triethylamine. The flask was gently swirled, 9.56 ml (0.1 mole) of ethyl chloroformate was added, and the flask was immediately placed in an ice bath for 20 minutes. A voluminous white pre¬ cipitate was visible at this point.

To a one-liter side arm flask was added 12.54 ml (0.1 mole) of 3-dimethylaminopropylamine and 10 ml of anhydrous THF. The flask was equipped

_* with a 9 cm Buchner funnel and a number 1 Watman filter circle. The contents of the one-liter flask were then filtered into the side arm flask. Evolution of carbon dioxide was visible as the filtrate mixed in the side arm flask. The round bottom flask was rinsed with an additional 20 ml of THF which was subsequently used to wash the filter cake.

OMPI

The filtrate was then transferred to a one- liter round bottom flask and the THF removed by dis¬ tillation at reduced pressure on a rotary evaporator. The residue was taken up in 500 ml of dichloromethane and transferred to a two-liter separatory funnel. The organic phase was extracted thoroughly with 200 ml of 3 M sodium hydroxide and 15 minutes were allowed for complete phase separation. Small amounts of ethanol (10 ml or less) were used to break any remaining emulsions. The dichloromethane (bottom phase) was drawn off and dried for 30 min. over 50 g of magnesium sulfate. The dried dichloromethane solution was de¬ canted into a one-liter round bottom flask. The mag¬ nesium sulfate was rinsed with an additional 20 ml of dichloromethane, which was then added to the round bottom flask, and all solvent was subsequently removed at reduced pressure in a rotary evaporator. Excess water was then, removed by repeatedly adding 50 ml of a 2:1 mixture of toluene and absolute ethanol to the round bottom flask followed by distillation in the rotary evaporator, until no cloudiness was observed in the distilled solvent. Removal- of all solvents left N-(3-dimethylaminopropyl)cholamide as a gummy white solid at room temperature. The gummy white residue from the previous step was taken up in 500 ml of anhydrous DMF and transferred to .a one-liter Erlenmeyer flask equipped with a ground glass joint and stopper. To this solution was added 12.25 g (0.1 mole) of propane- sultone and the flask was stoppered and incubated in a water bath at 60°C for 2 hr. The solution was then cooled to room temperature in an ice bath and 500 ml

OMP

of absolute methanol were added. The bulky precipi¬ tate was broken up, collected on a Buchner funnel by vacuum filtration and the filter cake washed with an additional 200 ml of absolute methanol. The crude product was subsequently triturated in 500 ml of boiling acetone and again collected by vacuum filtra¬ tion.; Thorough drying at room temperature yielded 45 to 50 g of 3-[ (3-cholamidopropyl)dimethylammonio)]- 1-propanesulfonate (CHAPS) (75-80% theoretical). The crude material was judged to be better than 95% pure by thin layer chromatography on silica gel G in a 95% methanol 5% ammonium hydroxide solvent system. The product appeared as a spot with an R _f of 0,32 which was visualized with iodine, phosphomolyb- date, or ninhydrin. The tertiary amine precursor appeared as a spot with R _f = 0.4.

Analytically pure material was obtained by repeated crystallization at 0°C from absolute methanol, followed by drying under high vacuum at room tempera- ture to a constant weight. The calculated analysis for ₃₂ ^H 5g ^N 2 ^S l ⁰ ₇ ^a ^ ^ter correction for 4.26% water determined by Karl Fischer analysis was: C 59.85%, H 9.58%, N 4,36, S 4.99; found: C 59.85%, H 9.19%, N 4.24%, S 5.06%.

OMPI

EXAMPLE

Synthesis of 3-[ (3-cholamidopropyl)dimethylammonio]- 2-hydroxy-l-propanesulfonate (see Fig. 2)

N-(3-dimethylaminopropyl)cholamide was prepared as described in the parent patent application ϋ.S ^, Serial No. 181,465). Sodium l-chloro-2-hydroxy-3- propanesulfonate was prepared according to the method described by Parris et al, J. Amer. Oil Chem. Soc, 53, 60-63 (1976). To a solution of 0.1 mole (49.3 g) of N-(3-di- methylaminopropyDcholamide in 300 ml of 40% aqueous methanol was added 0.1 mole (15.15 g) of sodium l-chloro-2-hydroxy-3-propanesulfonate. The mixture was refluxed with stirring for 3 hours, after which the solution was allowed to cool to room temperature and 100 g of RG-501 X8 mixed bed ion-exchange resin (Bio Rad) were added. The suspension of ion-exchange resin was gently stirred until the pH of the supernatant was approximately 7. The mixed bed ion-exchange resin was then filtered off and the solvent removed by distil¬ lation under reduced pressure to leave approximately 30 grams of 3-[ (3-cholamidopropyl)dimethylammonio]-2- hydroxy-1-propanesulfonate (50% theory) . This material was judged to be homogeneous by thin layer chroma- tography on Silica Gel G in a 95% methanol/5% ammonium hydroxide solvent system followed by iodine visualiza¬ tion.

Calculated analysis for:

^C 32 ^H 58 ^N 2 ^S0 8 ^*1H 2° ^{WaS: C} ' ⁵⁹ - ^22% ' H, 9.32%; N, 4.31%; S, 4.94%. The experimental analysis found: C, 59.12%; H, 9.31%; N, 4.31%; S, 4.99%.

Suggested trivial name CHAPSO .

OMPI

In an evaluation of 3-[ (3-cholamidopropyl)- dimethylammonio]-2-hydroxy-l-propanesulfonate relative to CHAPS in the opiate receptor solubilization assay, solubilization and Sepharose 6B chromatography of active opiate receptors was performed as described by Simonds, et al, in Proc. Nat. Acad. Sci. ϋ. S. A., 77, 4623-4627 (1980). Fig. 2 demonstrates that 3[ (3- cholamidopropyl)-dimethylammonio]-2-hydroxy-l- propanesulfonate (CHAPSO) is about twice as effective as CHAPS at solubilizing active opiate receptors. CHAPS itself may be prepared using as analogous alkylating agent sodium 1-chloro-propanesulfon- ate.

OMPI