Purification and Characterization of Bile Acid-CoA: Amino Acid N ...

Vol. 266. No. 16, Issue of June 5, pp. 10227-10233,1991 Printed in U.S.A .

THEJOURNAL OF BIOLOGICAL CHEMISTRY 0 1991 by The American Society for Biochemistry and Molecular Biology, Inc

Purification and Characterization ofBile Acid-CoA:Amino Acid N-Acyltransferase from Human Liver* (Received for publication, November 2, 1990)

Martin R.Johnson$, Stephen BarnesSBII, Joseph B. Kwakye$,and Robert B. DiasioSIIll From the Departmentsof $Pharmacology and §Biochemistry and the llcomprehensiue Cancer Center, Universityof Alabama at Birmingham, Birmingham; Alabama 35294

The bile acid-conjugating enzyme, bile acid-CoA: amino acid N-acyltransferase, was purified 480-fold from the soluble fraction of homogenizedfrozen human liver. Purification was accomplishedby a combination of anion exchange chromatography, chromatofocusing, glycocholate-AH-Sepharose affinity chromatography, and high performance liquid chromatography (HPLC) gel filtration. Following purification, the reduced, denatured enzyme migrated as a single 50-kDa protein bandbysodium dodecyl sulfate-polyacrylamide gel electrophoresis. A similar molecular mass was obtained for the native enzyme by HPLC gel filtration. Elution from the chromatofocusing column suggested an apparent isoelectric point of 6.0 (20.2). Using a rabbit polyclonal antibody raised against the purified enzyme, Western blot analysis using 100,000 X g human liver supernatant confirmed that the affinity-purified polyclonal antibody was specific for human liver bileacid-CoA:aminoacid N-acyltransferase. The purified enzyme utilized glycine, taurine, and 2fluoro-&alanine (a 5-fluorouracil catabolite), but not ,&alanine, as substrates. Kinetic studies revealed apparent K,,, values for taurine, 2-fluoro-&alanine, and glycine of 1.1, 2.2, and 5.8 mM, respectively,with corresponding V,,,,, values of 0.33, 0.19, and 0.77 pmol/min/mg protein. These data demonstrate that a single monomeric enzyme is responsible for the conjugation of bile acids with glycine or taurine in human liver.

thioester by the microsomal enzyme, cholyl-CoA synthetase (EC 6.2.1.7) ( 5 , 6). R-COOH

+ ATP + COASH$ RCO-SCOA + AMP + PPi

(1)

In the second reaction (Equation 2), the bile acid moiety is transferredfromthe acyl-CoA thioester (RCO-SCoA) to either glycine or taurine (NH2R’) to form the respective N acyl bile acid conjugate (RCONHR’). This reaction is catalyzed bybile acid-CoA:amino acid N-acyltransferase’(EC 2.3.1.65, BAT)’ (7, 8). RCO-SCOA

+ NHzR’ $ RCONHR’ + COASH

(2)

There have been several studies characterizing BAT from mammals, such as cattle (S), rats (9), and dogs (lo), and one characterizing BAT from domestic fowl (11).However, very little is known about BAT in humanliver. Kimura et al. (12) reported a partial purification (93-113-fold) of glycine- and taurine-dependentBATactivities from human liver. The enzyme had an apparent molecular mass of 100 kDa by gel filtration, twice as high as that reported for BAT purified fromothermammalian species (8, 9). Theirfinal enzyme preparation utilized both glycine and taurine as substrates, as well as a-D-alanine and @-alanine, as determined by an indirect spectrophotometric assay procedure. Since formation of a-D-alanineand@-alanine bile acidconjugates was never verified, a question remains as to whether these amino acids are substrates of BAT. Recently, we reported a radioassay procedure (13) in which the conjugation of 3H-labeled amino acids with bile acids is detected by the partition of the hydrophobic bile acid conjuThe excretionof bile by the liver is of fundamental biolog- gate into a n-butyl alcohol phase, while the unreacted 3Hical importance in mostspecies, particularly mammals (1).In labeled amino acid remains in theacidic aqueous phase. This procedure is a direct measure of the formation of bile acidman, cholanoates (Cz4 bile acids), formedfrom the metabolism amino acid conjugates and overcomes the difficulties of preof cholesterol, are the major solutes in bile (2). Although divious assay procedures (13). and trihydroxy cholanoates are as freely soluble in water as Using the specific radioassay, we now report the purificatheir sodium salts, greater than 95% of the C24biliary bile tion (480-fold) to homogeneity of BAT from human liver. In acids are found as N-acyl conjugates with either glycine or contrast to theprevious study by Kimura et al. (12), we have taurine (3). These bile acid-amino acid conjugates serve as found that BAT is a monomeric protein with a molecular detergents in the gastrointestinal tract, solubilizing long chain mass of 50 kDa and that P-alanine is not a substrate of this fatty acids, mono- and &glycerides, fat-soluble vitamins, and enzyme. In addition, apolyclonal antibody, raised against cholesterol (4). Bile acid-amino acid conjugates are formed in purified BAT, was affinity-purified and shown to be specific the liver by two successive enzymic reactions. Initially (Equa- for BAT in100,000 x g human liver supernatant. tion l), the bile acid (R-COOH) is converted to anacyl-CoA * This work was supported by United States Public HealthService Grant ‘2.4-40530. The costs of publication of this articlewere defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” inaccordancewith 18 U.S.C. Section 1734 solely to indicate this fact. 11 To whom correspondence should be addressed Box 600, Volker Hall, University of Alabama at Birmingham, Birmingham,AL 35294. Tel.: 205-934-4578; Fax: 205-934-8240.

This enzyme has been previously referred to as cholylCoA:glycine/taurine N-acyltransferase. However, since it catalyzes conjugation of C,, bile acids and a variety of amino acids (depending on the species), we have used the more generalized name, bile acidCoA:amino acid N-acyltransferase. The abbreviations used are: BAT, bile acid-CoAamino acid Nacyltransferase; SDS, sodium dodecyl sulfate; BSA, bovine serum albumin; CNBr, cyanogen bromide; HPLC, high performance liquid chromatography.

10227

10228

Purification of Bile Acid-CoA:Amino Acid N-Acyltrarzsferase TABLE I Purification table Specific activity Fraction Total protein Total

activity”

mg

2 31.87

nmoljmin

Supernatant, 0.067 100,0000.17 Xg 1,200.018,000.0 1,764.0 DEAE-cellulose 987.0 20.24 Chromatofocusing271.0 52.0 GC-Sepharose affinity‘ 246.7 7.70 Gel 63.50 filtration ’All values calculated for glycine. A similar result was obtained for taurine and glycine. Activity inhibited by glycocholate.

Taurine

Recoveryb

Glycine %

nmol/min/mg

1.91

-fold

100 82

0.56

1.0 8.4

”

A

MATERIALS AND METHODS~

B

C

94 k t

RESULTS

67 kt

Enzyme Purification-BAT activity was purified from the soluble fraction of homogenized frozen human liver. The purification scheme involved four chromatography steps. Initially, the 100,000 x g liver supernatant was chromatographed on a DEAE-anion exchange column. BAT activity eluted during a linear gradient of NaCl (40-200 mM) at an average NaCl concentration of 120 mM (Fig. 1,Miniprint). Theactive fractions were pooled and loaded onto a PBE-94 chromatofocusing column. BAT activity was subsequently eluted at pH 6.0 (f0.2) (Fig. 2, Miniprint).Fractionscontaining BAT activity werepooled and loaded onto a glycocholate-AHSepharoseaffinity column; unbound and nonspecifically bound proteins were eluted with a high salt buffer. BAT activity was recovered from the affinity column by elution with5 mM glycocholate (Fig. 3, Miniprint). Concentrated, affinity-purified BAT activity was finally chromatographed on an HPLC gel filtration column (Fig. 4, Miniprint) which separated BAT activity from glycocholate and other minor protein contaminants. Glycine- and taurine-dependent BAT activities did not separate during the course of the purification. The finalpreparation had a 480-fold enrichment of enzyme activity, with an overall recovery of 20% (Table I). Molecular Mass Determinatwn-The molecular mass of native BAT was determined by gel filtration chromatography (calibrated with known standards) tobe 50 f 2 kDa (Fig. 5A, Miniprint) and was not influenced by the presence of2mercaptoethanol. The 5%(v/v) glycerol used in the purification buffer reduced the ability of the gel filtration column to separate the molecular mass standards, andtherefore glycerol was avoided during determinationof the molecular mass. The denatured, reduced enzyme preparation gave a single, sharp protein band (Fig. 6) with an apparent molecular mass of 50 kDa on a 12% SDS-polyacrylamide gel (Fig. 5B, Miniprint). The protein yielded a more intense band staining with Coomassie Blue R-250 than by silver staining. Asingle band was also obtained under nondenaturing conditions which, after electroelution from the gel (following dialysis to remove glycine), yielded BAT activity. These data suggest that BAT is a monomeric protein without subunits. Amino Acid Composition-The amino acid composition of carboxymethylated BAT is listed in Table I1 (Miniprint). These data represent the mean of two separate BAT preparations.

’ Portions of this paper (including “Materials

andMethods,” Figs.

1-5 and 7-9, and Table 11) are presented in miniprint at the end of

this paper. Miniprint is easily readwith theaid of a standard magnifying glass. Full size photocopies are included in the microfilm edition of the Journal thatis available from Waverly Press.

43 kt 30 kD

14 kD

FIG. 6. SDS-polyacrylamide gel electrophoresis (12%)of purified bile acid-CoA:aminoacid N-acyltransferase from human liver. Lane A contains molecular mass markers. Lane B contains 10 pg of the purified enzyme stained using Coomassie Blue R250. Lune C also contains 10 pgof purified enzyme; however, this lane was cut from the gel and visualized using a silver stain technique (23). kD, kilodaltons.

Optimization of Temperature and pH Conditions-In a series of 100 mM potassium phosphate buffers covering a pH range between 4.0 and 10.0 with glycine or taurine as substrates, thehighest BAT activity was observed at pH8.4 (Fig. 7, Miniprint). Similarly, when incubated at temperatures over a range between 5 and 70 “C, the highest BAT activity was observed at 37 “C (Fig. 8, Miniprint). Kinetic Properties-Initial experiments using 3H-labeled amino acids showed that glycine, taurine, and 2-fluoro-Palanine, but not p-alanine, were substrates for purified BAT. In the presence of saturating amounts of cholyl-CoA, BAT activity varied with increasing amino acid concentrations to conform to Michaelis-Menten kinetics (Fig. 9, Miniprint). The apparent K,,, values of glycine, taurine, and 2-fluoro-Palanine derived from nonlinear regression analysis were 5.8, 1.1,and 2.2 mM, respectively. The corresponding V,,, values were 0.77,0.33, and 0.19 pmol/min/mg, respectively (Table 111). Polyclonal Antibody-Immunoblot analysis of proteins in 100,000 X g liver supernatant using affinity-purified antiBAT polyclonal antibody demonstrated asingle 50-kDa band (Fig. lOA, lane 2). This same band was detected using 0.5 pg of purified human liver BAT (Fig. lOA, lane 3 ) . Preimmune serum from the same rabbit did not detect any band under the same conditions (Fig. 10B, lane 5 ) . Monospecificity of the polyclonal antibody was verified by removal of 100% of the BAT activity (using either glycine or taurine as substrates)by passage of 100,000 x g liver supernatant over a rabbit anti-BAT antibody-Sepharose column. In addition, immunoblot analysis of the protein eluted from the column showed the disappearance of the single 50-kDa band (Fig. 11, lanes 2 and 3 ) .

Purification of Bile Acid-CoA:Amino Acid N-Acyltransferase

10229

TABLE 111 Comparison of BATproperties from differentspecies Kinetics Ref. Molecular

Species

mass

PI Taurine

kDa 8, 14 9

47-51 8.8 NA 63-65 45.7 100-118 50 5.8

Bovine Rat Fowl Dog Human Human

10 11 12 Present study

6.6 7.2 NA NA 5.85 6.0

Glycine

K,

Vm*,

K,

Vm*,

InM

pmol/rnin/rng

mM

prnol/rnin/rng

31.0 -b 3.2

NA“ 6.42 NA 0.77

6.7 0.8 4.0 NA 0.6 1.1

NA 2.42 NA NA NA 0.33

NA, data not available. not a substrate.

* -. A

B 1

49.5

2

3

I

5

4

-

.

u

e

e

w

FIG. 10. Immunoblot analysis of bile acid-CoA:amino acid N-acyltransferase. In A , affinity-purified polyclonal antibody was used as the primary antibody. Lane 1 contains prestained molecular weight standards, lane 2 contains 170 pg of 100,000 X g human liver supernatant protein, and lane 3 contains 0.5 pg of purified bile acidCoAamino acid N-acyltransferase. In B, preimmune serum wasused as the primary antibody. Lane 1 contains prestained molecular weight standards, and lane 2 contains 170 pg of 100,000 X g human liver supernatant protein. Bound antibody in both A and B was detected with alkaline phosphatase-labeled goat anti-rabbit IgG as described under “Materials and Methods” (Miniprint).

1

2

3

106

80.0

49.

FIG. 11. Immunoblot analysis of the eluent from the rabbit anti-BATantibody-Sepharose column. Lane 1 contains prestained molecular weight standards. Lane 2 contains 170 pg of the 100,000 X g humanliver supernatant protein that was originally loaded onto the column. Lane 3 contains 170 pg of the protein eluted from the column. DISCUSSION

In the present study we describe the purification to homogeneity and characterization of BAT from human liver. This study represents a 4-fold improvement on a previous partial purification from human liver (12). In additiona specific polyclonal antibody has been raised against humanliver BAT. Purification was accomplished by a combinationof DEAE-

cellulose anion exchange chromatography, chromatofocusing, glycocholate-AH-Sepharoseaffinity chromatography, and gel filtration on a TSK-250 high performance liquid chromatography column. The final product, when analyzed by SDSpolyacrylamide gel electrophoresis, consisted of a single 50kDa protein band. Electroelution from a polyacrylamide gel of the single protein band obtainedundernondenaturing conditions demonstrated that it was the source of BAT activity. The molecular mass of BAT of 50 kDa was also obtained for the native enzyme by gel filtration. The molecular mass of BAT varies considerably with respect to the species of origin (Table 111). Mammalian liver BAT activities (bovine and dog) have approximate molecular masses of 45-51 kDa, while that of domestic fowl is 63-65 kDa. The earlier report of a molecular mass of 100 kDa for human liver BAT (12) suggested that under the conditions used, BAT behaved as a dimer. This is a t variance with data from the present study and previous reports (8, 9, 10, 14). Although BAT has 4 cysteine residues (Table 11, Miniprint), these do not take part in intermolecular bondingsince the presentstudy has shown that BAT exists asa monomer. Further demonstration of the homogeneous nature of the BAT preparation was achieved using a polyclonal antibody raised in rabbitsagainst purified BAT. This is the first published report of an anti-BAT antibody. The polyclonal antibody was purified by passage over a BAT-Sepharose affinity column. The affinity-purified anti-BAT antibody reacted with a single protein band when Western blot analysis was carried out on 100,000 X g human liver supernatant; preimmune serum from the same rabbit had no reaction with these proteins. In a similar experiment, BATwas specifically and completely immunoabsorbed to anaffinity purified antiBAT antibody-Sepharose affinitycolumn. There was no evidence of there being separate glycine- or taurine-dependent BATactivities from data from any of the chromatography steps or from experiments designed to determine optimum pH and temperature values for this enzyme. This is in accordance with experiments performed on bovine liver BAT (8) and rat liver BAT (9). Comparison of the amino acid composition of BAT from two mammalian species (man andbovine) demonstrated considerable similarity (Table 11, Miniprint). Significant deviation in composition occurred for threonine (more abundant in human liver BAT) and arginine (less abundant in human liver BAT). A feature of this study was the use of a specific radioassay to demonstrate that both glycine and taurine are substrates for BAT. Analysis of the n-butyl alcohol phase from the radioassay by HPLC confirmed that no p-alanine-bile acid conjugates were formed by purified human liver BAT. This is in contrast with the earlier studyof human liver BAT utilizing

10230

Purification Acid-CoA:Amino Bileof N-Acyltransferase Acid

an indirect spectrophotometric assay, in which P-alanine was excellent technical assistance and Dr. Denise Shaw for advice in the production of the polyclonal antibody. reported to be a substrate (12). Recently, we have shownthat BAT (partially purified from REFERENCES human liver) catalyzes the conjugation of 2-fluoro-P-alanine 1. Haslewood, G. A. D. (1967) Bile Salts, Methuen, London (a catabolite of 5-fluorouracil) to bile acids (15). The present 2. Bjorkhem, I. (1985) in Steroids and Bile Acids (Danielsson, H., study confirms that purified human liver BAT catalyzes this and Sjovall, J., ed) pp. 231-260, Elsevier Science Publishing reaction. It is therefore interesting to note that 2-fluoro-PCo., Inc., New York alanine is a substrate for this enzyme while &alanine is not. 3. Barnes, S. (1991) in OxfordTextbook of ClinicalHepatology (McIntyre, N., Benhamon, J. P., Bircher, J., Rizzeito, M., and Since the size of a hydrogen atom is close to thatof a fluorine Rhodes, J., eds) Oxford University Press, Oxford, in press atom, the difference in the substrateaffinity could be a result 4. Hofmann, A. F. (1989) in Handbook of Physiology, Section on of the charge differences due to the greater electronegativity Gastrointestinal System (Schultz, S. G., ed), pp. 549-566, Amerof the fluorine atom or to the chiral centerit forms. To deduce ican Physiological Society, Bethesda, MD the exact reason will require additional information on the 5. Schersten, T. (1971) in MetabolicConjugationandMetabolic active site of human liver BAT. Hydrolysis (Fishman, W . H., ed) pp. 75-121, Academic Press, New York Kinetic experiments have clearly shown that taurine is 6. Killenberg, P. G. (1978) J . Lipid Res. 1 9 , 24-31 preferred as a substratefor human liver BAT compared with 7. Schersten, T. (1967) Biochim. Biophys. Acta 1 4 1 , 144-154 glycine. In this respect, human liver BAT has similar kinetic 8. Vessey, D. A. (1979) J. Biol. Chem. 2 5 4 , 2059-2063 properties to rat liver BAT (9). In rats, biliary bile acids are 9. Killenberg, P. G., and Jordan, J. T. (1978) J. Biol. Chem. 2 5 3 , predominantly taurine conjugates (9). However, in man, gly1005-1010 cine conjugates predominate, with the ratio of glycine to 10. Czuba, B., and Vessey, D. A. (1981) Biochem. Biophys. Acta6 6 5 , 612-614 taurine bile acid conjugates (G:T ratio) in bile being typically 3.5:l. Feeding humans large doses of taurine results in 95% 11. Czuba, B., and Vessey, D. A. (1981) Biochem. J. 1 9 5 , 263-266 M., Okuno, E., Inada, J., Ohyama, H., and Kido, R. of the biliary bile acids being conjugated with taurine (16). 12. Kimura, (1983) Hoppe-Seyler’s 2. Physiol. Chem. 3 6 4 , 637-645 These data suggest that the supply of taurine to the liver is 13. Johnson, M. R., Barnes, S., and Diasio, R. B. (1989) Anal. limited to thatderived from the diet (4). Following absorption Biochem. 1 8 2 , 360-365 from the gut and entry into the systemic circulation, taurine 14. Czuba, B., and Vessey, D. A. (1980) J. Biol. Chem. 2 5 5 , 52965299 rapidly concentrates in muscle. This has the effect of maintaining low levels of taurine in the blood and hence in the 15. Johnson, M. R., Barnes, S., Sweeny, D. J., and Diasio, R. B. (1990) Biochem. Pharmacol. 4 0 , 1241-1246 liver (17). Glycine, the more abundant amino acid, is then 16. Sjovall, J. (1959) Proc. SOC. Exp.Biol. Med. 1 0 0 , 676-678 utilized as the substrate. 17. Hardison, W . G. M. (1978) Gastroenterology 7 5 , 71-75 There has been considerable controversy concerning the 18. Siperstein, M. D., and Murray, A. W. (1956) Science 1 2 3 , 377378 subcellular location of BAT. Previous investigators have reported BAT activity was localized in the soluble fraction of 19. Bremer, J. (1956) Acta Chem. Scand. 1 0 , 56-71 liver homogenates from rat (9) and guinea pig (18).However, 20. Elliot, W. H. (1955) Biochim. Biophys. Acta 17,440-441 B. F., and Bjorkhem, I. (1989) J. Biol. Chem. 2 6 4 , 9220in otherstudies, BAT activity was localizedin the microsomal 21. Kase, 9223 fraction of liver homogenates from rat and guinea pig (19, 22. Achersten, T., Bjorntorp, P., Edkahl, P. H., and Bjorkerud, S. 20). In a recent study, Kase and Bjorkhem (21) claimed that (1967) Biochim. Bi~phys. Acta1 4 1 , 155-163 BAT activity was localized in the peroxisomal fraction of rat 23. Shah, P. P., and Staple, E. (1968) Steroids 12,571-576 liver. The enzyme has also been reported to be in the lysoso- 24. Killenberg, P. G., and Dukes, D. F. (1976) J . Lipid Res. 1 7 , 451455 mal fraction of human liver (22). In the present study, BAT 25. Lowry, 0.H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. was exclusivelyfound in the cytosol of homogenates of frozen (1951) J. Biol. Chem. 193,265-275 human liver (data not shown). However, the freezing and 26. Merril, C.R., Goldman, S., Sedman, S. A., and Ebert, M. H. thawing of the liver may have caused lysis of intracellular (1981) Science 2 1 1 , 1437-1438 organelles leading to a redistribution of BAT activity to the 27. Allen, G. (1989) in Laboratory Techniques in Biochemistry and Molecular Biology, Volume on Sequencing of Proteins and Pepcytosol fraction. These discrepancies in subcellular location tides (Burdon, R. H., and van Knippenber, P. H., eds) pp. 58may be resolvedby future studiesutilizing immunohistochem59, Elsevier Science Publishing Co., Inc., New York ical techniques in conjunction with the affinity-purified anti- 28. Towbin, H., Staehelin, T., and Gordon, J. (1979) Proc. Natl. Acad. BAT antibody. Sci. U. S. A. 76,4350-4354 Acknowledgments-We

wish to thank Benjamin M. Tangye for

29. Czuba, B., and Vessey,D.A. 6263

(1986) J. Biol. Chem. 2 6 1 , 6260-


10231

SupplementalMaterial To PurlficatlonandCharacterization of B k Acld CoA. Amino Acid: N-ACyltranstelaSe from HumanLlver by

Martin R . Johnson, Stephen Barnes, Joseph B. Kwakye and Robert B. Diaslo

Materials andMethods

Chemicals: DEAE-celluloseanionexchangergelwasobtained from Whatman (Clifton, NJ).Polybufferexchangergel (PBE 94).poiybuffer74,molecularweightmarkers, AH, Sepharose48, CNBr-activated SepharoseandSephadexLH-20 Were obtained from Pharmacla (Piscataway. NJ). l-Ethyl-3-(3-d~m~thylam~~oprapyl)carbodi~m~de, unlabeled amlno aclds.glycochol~cacldandcholicacidwereobtalned from Calbiochem (Sa" Diego, CAI CoenzymeAand2-mercaptoethanal were obtalned from SQma (St. Louis. MO). CoomassiebiilhantblueR-250andacrytamldewerepurchased from B I O ~ Rad (Richmond. CAI.Alkallnephosphatase-labeledgoatant#-rabbltantibody.nltroblue tetrazolium and 5-bromo-4~chloro~3-~ndalyl phosphaten-toluldfne salt were obtalned from SouthernBiotechnology(B,rmlngham.AL).[Z-oHI-Taur~ne(20.9Cllmmol).[3-3H]were obtainedfrom New p~alanlne(21 0 Cilmmol)and[2-3H]-glycine(20.0C,lmmol) EnglandNuclearCorp(Boston,MA). [3-3H]-FBAL was solated from the urine of pattents recelwng[6-3H]-FUra (15). Radiochemical purity was determlnedbyHPLC to be .99% (13) All othersolventsand reagents were purchased ~n the highest grade

available

Prmrailon of CholvlGOA-

Bde acld CoA derivatwesweresynthesuedbythemethod of ShawandSlaple (23) The cholyl CoA was partiallypurllled as previouslydescilbed (13) The cholyl CoA used ~n theklnetlcstudleswas further purlfled by passlng over a Sephadex LH-20column ( 2 4 ) .

Glycochollcacidwas P r m r a t l o n of Glvcocholate A H - W h a r o s eA l l ~ n l t vR e s m incubatedwlthAH-Sepharose 4 6 ~n the presence of 0.15 M l~elhyl-3(3-dlmelhylamino 25OC The gel was propy1)carbodllmlde In 50%aqueousdjoxane for 25hoursat collected on asjnteredglassfllterandwashedwith 100 ml of 1 M NaCI. followed by 200 ml of dlstllled deionned water The afflnity gel was equ8hbrated In 50 mM potassium phosphate, pH 8 3, 24 hours prior to u5e

m:BAT actlwtywasmeasured acids wereconlugated

in a radloassay I" which 3H-labeled a m n o l o unlabeledcholylCoA. Iormmg 3H-labeledcholatecon~u9aIws

The JH-labeledcholateconlugaleswereseparatedfrom the unreactedlabeled ammo acld by extractton Into n-butanol (131 The assaymixturecontained100 mM potasslum phosphate. 1.0mM cholylCaA,0.25mMunlabeled amlno scad (ellherglycine or taunne) wlth 0.025 pCi of thecorfesponding3H-labeled amlno acid, in a totalvolume of 100 ( h a 1 p H 8.4). Incubations were performed at 37oC far 30 mlnutes.Reactlons were mtlatedbytheaddltion of cholyl CoA and terminated by the addltlon o f 1 ml 100 m M patasslumphosphate containing 1% ( w l v ) SDS, pH 2.0 (saturated wlth n-butanol1 Protem concenI" the presence of heat mcttvated BAT Assay blankswereperformed tiallon was determined by the method of Lowry (25). using BSA as a protelnstandard p o l m r v l a m t d eG e l ElectrSDS polyacrylam~degel electrophoreslswas carrled out ~n a0.75 m m , 12% ( w l v ) polyacrylamide gelcontalnlng 0 375 M Trts~HCI (pH 8 8)and 0.1% SDS Samples Were prepared by mlxmg them wlth an equal volume of sample buffer (0.0625 M Trls-HCI, pH 6.8: 10% glycerol, 0 2% SDS ( w l v ) : 80 mM 2msrcaptoethanol) and bolling for 5 mlnutes The electrophorests was conducted at a constant current of 30 mA for 30 minutes at 250C 1.) m m a s s # e blue The gel was flxed 10 a 5% methanol 1 7 % acetlc accd Solutm foc 30 mlnutesand Stacned overnlght usmg 0.01% (wlu) Coomassle brilhant blue R-250 ~n a 5% IrlChlOrOaCel~Cacid I 2 5% methanol I 3 5% aCetlC acld i I The gel was fixed m 40%methanol I 10% acetlc acld for solution 2 ) %k%-&nm 30 minutes The gelwasstalned USlng the 810-Rad sliver stain as derlvedfrom the method of Merril et al.(26). Briefly. I O I I O W I ~ Q flxatlon.the gel wasIncubated m oxtdizersolution for 20 minutes The gel was washedwith dis18lled detonlzed water and incubated wlthsilver so1ut1on for 30 minutes. The gel was againwashedwlth dlstllled deionized waterandIncubatedwlth the developing s01~t10nsupplied by the manufacturer EledroluttonFromNativePolvacrvlamtdeGel.Gelelectrophoresswascawed out on 200pg purlftedBATunder non-denaturing cond~tionsI" a12%(wlv) polyacrylamcde gel A 0 5 cm Strip from thegelwas cut andstained according to themethods described above. This strip was hned up wlth the unslalnedgelandthesingle corresponding bandcut out of the unstained gel. The gel was m m e d and electroeluted a Blo-RadModel422electro-eluter In 25 mM Trls1192 mM glyclnebuffer.pH 8.3, for l o u r hoursat 10 mA (Constant current) according to themanufacturer's InsifucllonsThesample was dralyzedovermght (40C) ~n 1 llter of 50 mM potasslum phosphate. pH 8.4, before being assayed. MolecularWelahtDetermtnatlon. The molecularweight 01 natlve BAT wasdetermlned by gel f,llratlon A 2 15 x 60 cm TSK-250gelflltrat~onHPLCcolumn(Blo-Rad, Rohmond, CA) was equilibrated with 35 mM potasswmphosphate Containing 10 mM NaCI. pH 8 1 . at a flow rate of 2 5 mllmln The column wascallbraledusingknown mole~ularWelght standardsandthe retention time Of lndivldual PlOtelnS determlned by thelr peaks of absorbance at 280 nm The retention tlme of purifledenzyme was thencompared to that of the mOlecular welghtstandards.Themolecularwelght Of reduced,denaturedBATwasdetermlnedby SDS-polyacrylam8de gelelectrophoresls. ~ n standard g protelns of knownmolecularwelghts. Initla1 reactIan rates. were determinedat various Concentrat8onS of each arnlno actd (0.25, 0 75, 1.0. 2 0 . and 10.0 mM) ~n thepresence of a saturatlng concentratcon of cholyl CoA (1.15 mM) React8ons were run in a 100 mM potasslum phosphatebutter [pH 8.4) at 37OC Theincubation time and protein concentiatm were adlusted so that no more than 10% 01 the Ihmit8ng substrate was consumedEstimat ~ o nof the apparent K, and V,, for each amino acldwasperformedby flttlng these data for severalconcentrat~ons o f glyclne. t a u m e or 2-fluoro-p-alanlne to the Michaelis-Menten equalton by "on-Itnear ragress~an analysts (15) Enzyme activity 8s expressed a5 nmol of productformed per m8n permgofprotein

. . Protein sampleswerecarboxymethylatedasdescrlbedby Allen (27). The amlnoacldcomposltlonwasdeteimlnedby flrst hydrolyzlngthesampleat 1 looc for 20 hours In the presence of 6 N HCI, the amlno acids were analyzed by reversed-phaseHPLC using the PIC0 TAG system (Waters Assoctates. Mllford. MA)

Pr ration of PolvctonalAntibodv.Male.NewZealand rabbts were mmunlred with subcutaneousmjectfons of punfiedBAT.The first inlection conssled of 100 pg of purlfted antigen. mlxed wlth an equalvolume of Freun#scompletead)uvant:tendays later lhcs lnlection was repeated. Twenty days followlng the secondInfectionthe cycle was repeated with the antigen mlxed in an equal volume of Freund'sincomplete adjuvant.Allquots from ear nlcks were screened for antlbody formation. Ten days followlngthethirdcycleofinjections, the rabbttsweresacrificedbycardiac puncture andtheblood collected. The blood was allowed l o clot and spun at 2000 RPM for 15 mln The Serum was loaded on a 1 x 10 cm protein A-Sepharose 4 Fast Flow column (Sigma Chemlcal Co , St L o u s MO), previouslyequ#llbrated wtth phosphate-bufferedsaline. The c o l ~ r n nwas washedwlth4 Column VOlUmeS Of phosphate-bufferedsallneandthe IgG ant8bodLes wereelutedwllhan a a d wash C O W ~ofS0.2 I IM ~glycine-HCI Q contaln8ng 0 075 M NaCI. pH 2 5 lmmed8ately upon elution from the column the fractions were neutralized with 1.0 M Trls~HCI.pH 10.1 SDS-polyacrylamidegelelectrophoreslswas h m l ~ n o b l o t AnalvsIL performed on x g human l h e l supernatant (from tlaction 1) andpurlfled freshlyprepared100,000 BAT (from fraction 5). The protelnwastransferred from thegelto a nitrocellulose filter fallowingthemethod of Towbln ( 2 8 ) The nitrOcellUloSe filter was incubated overnlghtat 40C wlth the aftlnitypurlfledpolyclonalantibody(dlluted 1:400) In a 120 mMborate-salme solution contalnmg 1% (wlv) BSA. pH8.5 The nitrocellulose fllter was washedwlthborate-sallne contatnmg 0 1 % Tween20 ( v l v ) and mubated wltha secondary.alkatmephosphatase-labeledgoat antWabb8t antcbody. The nltlocetIuIose l~ltsrwasdeveloped in a0.1 M sodlumcarbonatebuffer (100 ml, pH 9 5) containing 30 mg n ~ t r oblue tetrazoI,um(addedas a 1 ml solution d,ssolved In 70%dlmelhylformamlde)and15 rng 5-bromo-4-chloro-3-~ndolylphosphate p-toluidine salt (added as a1 ml solution dlssalved m 100%dimethylformam~de). AfflmtvPUIiflcation of Polvclonnl Ant#kQPdli; PurlfledBAT was coupled to CNBractivated Sepharose 4 8 followlng InStrUCtlons suppliedbythemanufacturer The coupledBAT-Sepharose 48 alflnlty r e s f n was washedwlthphosphate-bufferedsaline and packed In a 2 cc syrlngeIgG antibodies (previously purlfled on the protern A Sepharosecolumn) were loaded onto the column and eluted following thesame procedureused I" the protein A-Sepharose pmfication proceduredescrlbedabove. Preoaratlon 01 Rabblt Ant,-BAT Seohdrose Allinitv Resln: Rabblt Ant(-BAT polyclonal antlbody was incubated with CNBr~actlvatedSepharose4B ~n the presence 01 coupling buffer (0 1 M NaHC0310 5 M NaCI. pH 8 3) tor 25 hours at 40C. The gel was blocked In 1 0 M ethanolamlne-HCI. pH 8.0, for 2 hours at room temperatureandwashedwith c o ~ p l ~ nbuffer g followed by 0.1 Msodium acetate buffer (containing 0.5 M NaCI), pH 4 0 The gelwas equllhbrated ~n phosphate-bufferedsallne24 hours prior to use.

AII procedures were perlormed at 4DC

Fractloo 2: DEAE-Cellulose An~on Eschanae ChromatoaraDhy. The dlalyzate from fractlan 1 was loaded onto aDEAE-cellulosecolumn(5 x 30 cm) which hadbeen prev~ouslyequlllbrated with 50 mMTrls-HCI. pH 8.25. Thecolumnwaswashedwlth 1000 ml 50 mM Trls-HCI. pH 8.25. to elute unbound proteln. Enzymeactiwtywas eluted wlthalineargradlent of NaCl (0 to 200 mM) Fractions contalnmgN-acyltransferase acttvrty werepooledanddiluted 1 1 wllhdlstllleddeionizedwater

Fraction 3 Chromatofocuslne The dllutedDEAE eluent from fractbon 2 was loaded onto achromatotocuslngcolumn(1 6 x 70cm)packedwlthPB-94previous!y equlllbratedwlth 25 mM mdazale-HCI. 10 m M Z~mercaptoethanoland 5% ( v l v ) g l y ~ e r ~ l , p H6.8. The column was re-equllfbratedwlth5columnvolumes 01 equilibration buffer.Enzymeactlwtywaseluted in a pH gradlentcreatedby Passage of polybuffer 74 diluted 1:8 wtth glass dE.1~11ed water ( h a 1 pH adlusted to 5 0 wlth HCI). BAT actlwtyeluted at a pH of approxlmately 6 0 ~ractlon4 Glvcocholate A H - a h a r o s e Alflnltv Chromatoarmhy The pooled AHChromatOfoCus~nQfractions from traction 3 were loaded onto aglycocholate Sepharose altmty column (1 x 10 m ) . previously equhbiated with 50 mM Potassium phosphate, pH 8 4 The column was washed wlth 10 column volumes Of 50 mM potass~umphosphate COntainlng 100 mM NaCl. pH 8.4. Enzyme actlvlty was eluted with 50 mM polasslum phosphate contalnlng 250 mM NaCl and 5 mM glycocholate, PH 8 4 . Fractions contalnlngBATactlwty were pooledandconcentrated10-foldin an Amlcon centrtpiep 10 Concentrator


10232

Fraction 5: ' Thepooledconcentratedaffinity fractions fromfraction 4 were injec!ed Onto aBiorad TSK-250 gelliltrationcolumn (2.15 x 50 cm), previously equilibratedwith 35 mM potassium phosphate containing 10 mM NaCI, 10 mM 2-mercaptoethanol and 5% (vlv) glycerol, pH 8.1. Enzyme activity eluted in 40 minutes. Activefractions Were pooledandconcentrated10-fold in an Amicon centriprep 10 concentrator.

40000

A

TAELE II of Blle Acid CoA: Amino Acid: N-Acvtransferase

Leu

55.8

36 Ala

46.9

4Gix. 3

45.7

3 5GIY

40.5

Thr

31.7

2Asxb 7

31.6

Pro 17

0

10

30

40

50

60

70

FRACTION

22.7

Ile

15

16

20

Elution of glycine-(A) and taurine-(@) dependent BAT activitiesfromglycocholateAH-Sepharoseaffinitychromatography. Enzyme activlty was elutedby washing the column with 50 mM potassium phosphate1 250 mM NaCll 5 mM glycocholate, pH 8.4. Fraction size was 2.5 ml with allowrate of 10 mllhr.

25.3

Phe

10

30

27.8

22 Ser 3

42

31.1

LY 5 2 4Val

L0

21.7

2 4Arg

20.5

10 His

15.6

1 2Tyr

13.4

Met

4.7

5

CMO

3.7

5

30000

7

aGlutamlc acidandglutamlne bAspartic acldandasparaglne Carboxymethylatedcystine

ii

dReference 29 Ail valuesarecorrected

for hydrolysis loss at 1lOoC for 20 hours

30000

NaCl (mM)

n

200

20000

Ip 0

,

10000

E

0

80

40

120

160

200

FRACTION FIGURE 4 : Gel filtration chromatography demonstratlng glycine-(A) and taurine-( ) dependent BAT activities and absorbance at 280 nm (-). Enzyme activlty eluted in 46 minutes using a mobile phaseof 35 mM potassium phosphate110 mM NaCIIIO mM 2-mercaptoethanol and 5% (vlv) glycerol. pH 8.1. Fraclton s u e was 1.25 ml with a constant llow rate 01 150mllhr

0 0

20

40

60

80

100

FRACTION FIGURE I; Glycine-(A) and taurme-( @ ) dependent BAT actlwtles from DEAE-cellulose. The bound enzyme was eluted with a lhnear gradient of 0200 mM NaCl (-----). Protein was measured by absorbance at 280 nm (-). Fractlon size was 10 mi wlth a flow rate ot 75 mlihr.

20000

........ RETENTION TIME (min)

a d

3

0 W

0

20

40

60

80

100

FRACTION FIGURE 2 : Elutlon pattern of glycine-(A) and taume-( @ ) dependent EA7 activities from a PB-94 chromatofocusing Column. Enzyme activity was elutedinapH gradient with polybulfer 74 diluted 1 8 pH 5.0 with HCI (...... ). Protein was measured by absorbance at 280 nrn (-), Fraction Slze was 7 ml with a flow rate of 75 mllhr.

0

155

10

20

M O B I L I T Y( c m )

!KUELZ (A)Molecularweight

determinatton ofnatwepurlfled BAT ( 0 ) by HPLC gel fbllrallon chromatography. (8)Molecular wetght determinatlon 01 purlfled BAT ( 0 ) by SDS-polyacrylamide gel electrophoresis. The molecular weight standards range from 97 kDa lo 31 kDa ( 0 )


10233

0.040

0.035

100 -

0.030

80 -

-z

60:

E

40:

u 4

20 I

0

1

2

3

4

PH FIGURE 7 : glycine^( ) and taurme-(A) dependent EAT acitiv~t~es of purified BAT were evaluated m a sertes of 100 mM polassIum phosphate buffers coveringa pH range between 4 and I O . Samples werelncubated for 30 minutes at 37oC. Act~v~ty !s expressed as a percentage 01 the value ootatned at pH 8 4

I/AMINO ACID (mM)

FIGURE 9: Double reciprocal plot of the $nltlal reaction rate as a function of the COnCentratlOn o f QlyClne. IaUrlne and 2-lluoro-palanine. The rates 01 blle acldconpgate formallon were determlned m the Presence of 1.15 mM choiyl CoA using the fOllOWlnQ fixedconcentrations (0 25, 0 75, 1.00, 2 00, 10.00 mM) of glycine taurlne ( 0 ] and 2-fluor0 P-alanlne).(

(A),

0

0

10

20

30

40

50

60

70

80

TEMPERATURE ("C) Eiaure 8 : Glyclne ( 0 )andlaurlne (& conjugating actlvmes of purlfled BAT wereevaluated lor 30 mlnutes in 100 mM potasslumphosphate bulfer,pH 8.4. overatemperaturerange of 4 to 70oC. Activity 1s expressed as a percentage 0 1 the value obtained at 37OC