Int. J . Peptide Protein Res. 23,1984,559-564

Importance of methionine residues in the enzymatic carboxylation of biotincontaining peptides representing the local biotinyl site

of E. coli acetyl-CoA carboxylase

HlROKl KONDO, SHINGO UNO, YOSHlYUKl KOMlZO and JUNZO SUNAMOTO

Depar tment of Industrial Chemistry, Faculty of Engineering, Nagasaki University, Nagasaki, Japan

Received 31 May, accepted for publication 30 September 1983

phase synthesis

Acetyl-CoA carboxylase plays a key role in the synthesis of long-chain fatty acids in biological systems ( 1 . 2). This carboxylation reaction takes place in two discrete steps, as shown in eqns. l a and l b . Thus,

E-Biotin + ATP + HCO; += E-biotin-CO;

+ A D P + P i (1 a)

+ Malonyl-CoA (1b)

E-Biotin-CO; + Acetyl-CoA + E-Biotin

A biotin-containing hexapeptide Ac-Glu-Ala-Met-Bct-Met-Met ( 1) that represents the local biotin-containing site of Escherichia coli acetyl-CoA carboxylase has been prepared by the solid phase method. Peptide 1 is carboxylated by the biotin carboxylase subunit dimer of E. coli acetyl-CoA carboxylase with the following kinetic parameters; K, 12 mM , V,, 2.8 PM . min-' , These compare with the parameters for biotin of K, 214 mM and V,,, 2 8 ~ ~ .min-'. Hence, the overall reactivity (Vmax/Km) of 1 is 1.8 times greater than that of free biotin. When all rnethionines in 1 are re ( 2 ) retains a similar binding ability but with a found that peptide 3, which carries an NE-be 5iocytin in 1, decreases the K, of biotin three

Ke.v words: acetyl-CoA carboxylase; carboxylation; me

be separated and ;hat the separated subunits retain physiological activity (3,4). Furthermore, biotin carboxylase can carboxylate not only BCCP but also free biotin, though the activity of biotin is much smaller than that of the former. According t o the sequence data of this

All amino acids used are of Ltonfiguration. The abbreviations are: Bct, M-biotinyllysine; DCC, di- cyclohexylcarbodiimide; DCUrea, dicyclohexylurea; EEDQ, Nethoxycarbonyl-2ethoxy-l,2dihydroquino- line; TFA, trifluoroacetic acid; CbzC1, benzyloxy- carbonyl chloride; Cbz, benzyloxycarbonyl; C1- Cbz, othlorobenzyloxycarbonyl; Bzl, benzyl(ester);

Boc, t-butoxycarbonyl; DMAP, 4-dimethylamino- pyridine; DIEA, diisopropylethylamine; DEAE, di- ethylaminoethyl; NADH, nicotinamide adenine di- nucleotide, reduced form; ATP, adenosine 5'-tri- phosphate; ADP, adenosine 5'diphosphate; SDS- PAGE, sodium dodecyl sulfate polyacrylamide gel electrophoresis.

559

H. Kondo et al.

enzyme and other biotin-dependent enzymes (5-7), the coenzyme binding sites constitute a so-called sulfide cluster. We are deeply .interested in it, because this sulfide cluster might have a certain chemical or physiological significance. For example, it might be required for a tight binding of the coenzyme to the enzyme active site. To prove this, we have undertaken a synthesis of peptides such as 5 and 6 , which represent the coenzyme binding site of E. coli acetyl-CoA carboxylase (8,9).

Ac-Met-Bct-Met-Met-OMe 5

Boc-Glu- Ala-Met -Bct-Met 6

The synthetic strategy adopted was to prepare a biotin-free peptide first, and coenzyme was incorporated at the final stage of the syn- thesis. The problem was the poor yield of the biotinylation step. To circumvent this and also to cut time needed for such a synthesis, we have attempted for the first time the solid phase synthesis of biotin-containing peptides. The peptides prepared in this article are the following:

Ac-Glu-Ala-Met-Bct-Met-Met 1

Ac-Glu- Ala- Ala-Bct- Ala- Ala 2

Ac-Glu-Ala-Met-Lys(Cbz)-Met-Met 3

EXPERIMENTAL PROCEDURES

Apparatus 1.r. and 'H n.m.r. spectra were determined on a Jasco A-100 and Jeol JNM-MH-100 spectro- meter, respectively. Electronic absorption spectra were taken on a Hitachi 200 or a Shimadzu UV-150-20 spectrophotometer, while fluorescence spectra were on a Hitachi 650-10 spectrophotometer. High-voltage paper electro- phoresis was run with a Toyo Kagaku Sangyo HPE-406. Amino acid analysis was carried out on a Jeol J-200 amino acid analyzer.

Materials A chloromethyl resin, made of polystyrene crosslinked with 1% of divinylbenzene, was obtained from Pierce. Boc amino acids were either prepared or purchased from Protein Research Foundation. DCC, EEDQ, TFA, CbzCl

and 25% HBr/AcOH were obtained from the same source. d-Biotin and DMAP were from Wako Pure Chemical Ind. and Aldrich, respect- ively. Phosphoenol pyruvate, glutathione, NADH, ATP and bovine serum albumin were purchased from Seikagaku Kogyo Co., while lactate dehydrogenase and pyruvate kinase were from Boehringer Mannheim. Sephadex G-10 and DEAE-Sephadex A-25 were obtained from Pharmacia.

Peptide synthesis The chloromethyl resin (substitution level 0.67 mmol/g), was converted to a hydroxymethyl resin by way of an acetoxymethyl resin (10). Boc-Met was incorporated to the resin by means of DCC coupling in the presence of DMAP. In a typical run, 3.24g of the hydroxy- methyl resin swelled in 33ml CH2C12 was allowed to react with 1.62 g (6.5 mmol) of Boc- Met, 0.80g (6.5mmol) of DMAP and 1.34g (6.5mmol) of DCC for 2 h at room tempera- ture. The resin was washed as usual and then dried in VQCUO. The degree of amino acid sub- stitution was determined as 0.46 mmol/g, following acid hydrolysis of the resin (5.5 M HC1-propionic acid (1 : 1, v/v), 1 30°, 3 h), either by amino acid analysis or fluorometry after treatment with fluorescamine.

The peptide synthesis was carried out by the manual solid phase method using the symmetric anhydride of amino acids (1 1) with the excep- tion of Bct. This particular amino acid was coupled by the EEDQ method (12). The coupling procedures are summarized in Table 1. Amino acids employed are the following: Boc-Met, Boc-Bct, Boc-Ala, Boc-Glu(Bz1) and Boc-Lys(C1-Cbz). Following deprotection with TFA and neutralization with DIEA, the N-terminal amino acid of the growing peptide chain was allowed to react with 3 equiv. of the symmetric anhydride of the next amino acid, which was prepared by the DCC coupling of Boc amino acid. Six equiv. of Boc amino acid is allowed to react with 3 equiv. DCC in 10ml of CH2C12 for 20min at 0". The precipitated DCUrea was filtered off and washed with a small volume of the solvent. The symmetric anhydride thus prepared was used in step 10 of Table 1. Three equiv. of acetic anhydride was employed for the incorporation of the N-

560

Peptide-bound biotin

TABLE 1 Protocol for the solid phase synthesis of peptides I . 2 and 4 by the symmetric anhydride method. The scale of

the synthesis was 0.60 mmol and 1.28g Boc amino acid resin was used

Step Reagent Volume (ml) Time (min) Repetition

1 2 3 4 5 6 7 8 9 10 1 1 12 13 14

CH,Cl, 50% TFA/CH, C1, CH,CI, i-PrOH CH,Cl, 7% DIEA/CH,Cl, CH,Cl, 7% DlEA/CH,CI, CH,Cl, Symmetric anhydride/CH, C1, a

DIEA (1.1 equiv.)/CH,Clla CH,CI, EtOH CH,Cl,

16 20 16 20 16 20 16 20 16 15 2 16 20 16

1 30 1 1 1 3 1 3 1 10 30 1 1 1

3 1 3 1 2 1 2 1 4 1 1 3 2 3

aCoupling of Boc-Bct was accomplished by the EEDQ method. Steps 10 and 1 1 were replaced by a coupling reaction of 3 equiv. Boc-Bct and 6 equiv. EEDQ in 15 ml DMF for 6 h.

terminal acetyl group. In the EEDQ coupling, the N-free peptide chain was allowed to react with 3 equiv. Boc-Bct and 6 equiv. EEDQ in DMF for 6 h. Completeness of each coupling was confirmed by the ninhydrin color test

The protected peptide was cleaved from the resin by incubation of the peptide resin (0.6 mmol) with 20ml of 25% HBr/AcOH and 3.5 ml methyl sulfide for 1 h at room tempera- ture. The resin was filtered off under suction and washed with AcOH. The filtrate and washings were combined, diluted with water and then evaporated to dryness. The crude peptide was taken up in a small volume (- 10 ml) of 0.1 M NaOH (the insoluble material was removed by filtration) and then applied to a column of Sephadex G-10 (3.2 x 56cm). The peptide was eluted with 0.1 M ammonia sol- ution. The peak fractions were pooled and lyophilized. The solid was dissolved in 25mM AcONH4, pH 9.5, and chromatographed on DEAE-Sephadex A-25. The column (1.6 x 22 cm) was eluted with a linear gradient of 25 mM to 1 M AcONH4, pH 9.5. The desired fractions were lyophilized and then chromato- graphed repeatedly on Sephadex G-10 for desalting.

(13).

Ac-Glu-A Ia-Met-Lys(CBz)-Met-Met (3). Ac-Glu-Ala-Met-Lys-Met-Met (4) was synthesized and purified as above (Table 2). Peptide 4 (1 80 mg, 0.23 mmol) was dissolved in 10 ml of water containing 1 .O g of sodium bicarbonate. Benzyl- oxycarbonyl 'chloride (200 mg, 1.2 mmol) was added in small portions over 1Omin and the mixture was stirred for 3.5h at room tem- perature. After extracting the excess reagent with ether, the aqueous phase was brought to pH 4. The precipitate formed was collected by filtration and it was purified by chromato- graphy on Sephadex G-10 with 25 mM AcONH4, pH 9.5, as the eluate. Yield 35mg. This com- pound was ninhydrin-negative and its ab- sorption spectrum showed a shoulder at - 260 nm .

Enzyme preparation and assay The biotin carboxylase subunit of E. coliacetyl- CoA carboxylase was prepared according to Guchhait et al. (14). The preparation was homogeneous on SDS-PAGE. The enzyme con- centration was determined on the basis of AiZ 6.3 (4). Enzyme activity was determined by the coupled assay, in which ADP formation is coupled with the oxidation of NADH. The assay mixture (0.5 ml) contained 1 mM ATP, 8

56 1

H. Kondo et al.

mM MgCI2, 8mM NaHC03, 3mM glutathione. 0.5 mM phosphoenol pyruvate, 0 .2 mM NADH, 5 units of lactate dehydrogenase, 3 units of pyruvate kinase, 0.6 mg/ml bovine seruni albumin, 10% ethanol, biotin and enzyme in 100 mM triethanolamine buffer, pH 8.0. The reaction was started by the addition of 8 milli- units of enzyme and the decrease o f NADH absorption was followed continuously at 340 nm. One unit of enzyme activity is defined by the amount of enzyme required t o form 1 pmol of I-carboxybiotin per min. Biotin carboxylase does not possess any ATPase activity.

RESULTS

Peptide synthesis The solid phase synthesis of biotin containing peptides 1 and 2 and their analogue 4 was carried out according t o the method of Merrifield (1 5 ) . The problems inherent t o the synthesis are the presence of unusual amino acid Bct and of many sulfide groups that are prone t o oxidation. A difficulty had sometimes been encountered in the coupling of Boc-Bct by the DCC coupling or the mixed anhydride method (1 6), presumably because of inter- ference of the biotin’s urea moiety with the coupling reaction. But, the EEDQ coupling appears t o be free from such problems and hence this method was tested first for the coupling of this amino acid. Thus, coupling of the growing peptide with 3 equiv. of Boc-Bct and EEDQ proceeded smoothly, as deduced by

the ninhydrin color test. All other couplings were carried out by the symmetric anhydride method successfully. Methionine and biotin were used with their sulfide groups unprotected and no special care was taken to prevent oxidation during the synthesis and work-up with the exception of HBr cleavage of the peptide from the resin, in which a large excess of methyl sulfide was added as a scavenger. The sulfide groups remained intact throughout these manipulations as judged by ‘H n.ni.r., in which sulfide methyl groups exhibited no down-field shift that should have been brought about upon sulfoxide formation.

Crude peptides were purified by Sephadex G-10 and DEAE-Sephadex chromatography. Each peptide preparation contained one or more undesired peptides, as well as the desired product, in small quantities, but they can be removed readily by the latter chromatography. The purified peptides were analyzed by paper chromatography, high-voltage paper electro- phoresis and amino acid analysis (Table 2). All peptides were obtained in fair yield of 26-49%.

Enzymatic carboxylation The biotinyl moieties of peptides 1 and 2 are carboxylated by the biotin carboxylase subunit dimer of E. coli acetyl-CoA carboxylase. Initial rate data obtained by the coupled assay over the substrate concentration range of 2-20 mM were analyzed in terms of Lineweaver-Burk formulation (Table 3). Peptide 1, representing

TABLE 2 Analytic data for synthetic peptides I , 2 and 4

Peptide Yield (%) M.p. (“C) Rf Amino acid analysis‘ PCa PEb Glu Ala Met Lys

1 26 192-194 0.96 0.62 0.95 1.00 2.60 1.04 (1) (1) (3) (1)

2 32 170-173 0.90 0.68 1.00 4.00 1.04 (1) (4) (1)

4 49 236-238 0.87 0.06 1.06 1.00 2.94 1.01 (1) (1) (3) (1)

aPaper chromatography was run on a Toyo Roshi filter paper No. 51 in BuOH:AcOH:H, 0 = 1:1:3. bMobility relative to Glu in pyridine:AcOH:H,O = 10: 1:89, pH 6.5. The sample was electrophoresed on a Toyo Roshi filter paper No. 5 1 at 2,500 V for 2 h. ‘Numbers in parentheses denote the theoretical amount of amino acid. Methionine was analyzed in the sulfone form (19).

562

Peptide-bound biotin

TABLE 3 Carboxylation of biotin derivatives by the biotin carboxylase subunit dimer of E. coli

acetyl-Con carboxylase at p H 8.0 and 30.0"

Substrate K,(mM) Vmax(pM-min- ' ) Relative rate

d-Biotin 214 Biocytin -

6 18 1 12 2 16

68 &Biotin + 5 mM 3

28 - 2.8 2.8 0.72 9.8

100

119 178

34 110

19'

aTaken from Polakis et al. (20)

the six consecutive amino acid residues of the BCCP subunit of E. coli acetyl-CoA carboxylase has a K, 18 times smaller than that of free biotin. Meanwhile, the K, for ATP with biotin as the substrate (130pM) is little changed upon substitution of biotin by peptide 1 (120pM), indicating that integration of the peptide to the coenzyme does not affect the binding property o f the nucleotide. When all methionines of peptide 1 are replaced by alanine. the binding ability o f the resulting peptide (2) is not altered significantly. However, the V,, does differ between the two substrates; peptide 1 is 4 times more effective a substrate than peptide 2. This indicates that the methionine residues play an important role for biotin to be carboxylated by the enzyme. Peptide 3 carrying a benzyloxy- carbonyl in place o f a biotinyl group was unable to serve as a n inhibitor against 10-50 niM biotin at the highest peptide concentration tested (5 mM). Of interest, however, is the observation that when peptide 3 and free biotin were used together, the K, for biotin is lowered considerably at the expense of the Vmm. As a typical example given in Table 3 indicates, the kinetic parameters for biotin in the presence o f 3 approached those for 1 or 2, as if coenzyme and peptide 3 were linked covalently.

DISCUSSION

It has been reported in this article that the syn- thesis o f biotin-containing peptides is sucessfully accomplished by the solid phase method. In the past, such peptides were prepared by the liquid phase method, and biotin was incorpor- ated t o a coenzyme-free peptide by the active

ester method in most cases (8, 9). The active esters employed most often were the N-hydroxy- succinimide and p-nitrophenyl esters of biotin, and their quantitative coupling with the peptide can be achieved only at the expense o f a large amount of expensive biotin (17, 18). It is, hence, clear that this technique is feasible for a small scale synthesis but not for a large scale preparation. An alternative synthetic route to the biotin-containing peptide is the incorpor- ation of coenzyme at the early stage of peptide synthesis (1 8). This method is time-consuming and the overall yield o f the target peptide is not high. In contrast, the solid phase synthesis reported herein provides biotinylated peptides in a short time and in fair yield. The synthetic peptides obtained were satisfactory by all criteria such as amino acid analysis and enzymatic carboxylation. In the latter, peptides 1 and 2 were carboxylated by the enzyme just like peptide 6, which was prepared by the liquid phase method. Although only the synthesis of hexapeptides was attempted, we are certain that this technique is readily extended to a synthesis of larger biotin-containing peptides.

It is noticeable that integration t o biotin of a short stretch of peptide that corresponds t o the local coenzyme-carrying sequence increased a substrate activity o f the coenzyme as much as - 180% of that of free biotin in the acetyl-CoA carboxylase reaction. This figure is significant, as biocytin is only one fifth as effective as biotin. Thus, if we compare biocytin and peptide 1, the latter is over nine times more efficient a substrate. That this improvement is brought about by methionines is verified by the replace- ment experiment, in which all methionines of

563

H. Kondo et al.

peptide 1 are substituted by alanine. The resulting peptide (2) is only twice as effective as biocytin. Several explanations could be invoked as to the role of the peptide group in 1. The sulfide of methionine and/or the amide groups may provide a coordination site to a metal ion, an indispensable constituent of this carboxylase reaction. In view of our previous observation that amides rather than sulfides appear to be the principal site of magnesium coordination (8), it is unlikely that the sulfide moieties serve as ligands to the metal. Since the methionine side chain is apolar, the coenzyme in 1 may be placed in a hydrophobic sulfide cluster. This microenvironment might be needed for the substrate or the active site residues of enzyme to align properly. An interesting finding is that peptide 3 alone, which lacks the coenzyme but retains hydrophobic residues, lowers the K, for biotin considerably. These observations remind us of a peculiar character of this enzyme in that the enzyme activity is maximal in the presence of - 1.5% of an organic solvent such as ethanol and dioxane (14). Taken together, the structure and function of E. coli carboxylase may rest on a subtle balance of hydrophilicity and hydrophobicity.

ACKNOWLEDGMENTS

The authors thank Professor D. TSUN of Nagasaki University for use of this laboratory facilities. This work was supported by a Grant-in-Aid for Special Project Research 57102007 from the Ministry of Education, Science and Culture.

REFERENCES

Moss, J. & Lane, M.D. (1971) Adv. Enzymol. 35,

Wood, H.G. & Barden, R.E. (1977) Ann. Rev. Biochem. 46,385-413 Alberts, A.W. & Vagelos, P.R. (1968) Proc. Nat. Acad. Sci. US 59,561-568 Dimroth, P., Guchhait, R.B., Stoll, E. & Lane,

321-442

M.D. (1970) Proc. Nat. Acad. Sci. US 67, 1353- 1360

5. Sutton, M.R., Fall, R.R., Nervi, A.M., Alberts, A.W., Vagelos, P.R. & Bradshaw, R.A. (1977) J. Biol. Chem. 252,3934-3940

6. Rylatt, D.B., Keech, D.B. & Wallace, J.C. (1977) Arch. Biochem. Biophys. 183, 113-122

7. Maloy, W.L., Bowien, B.U., Zwolinski, G.K., Kumar, K.G., Wood, H.G., Ericsson, L.H. & Walsh, K.A. (1979) J. Biol. Chem. 254, 11615- 11622

8. Kondo, H., Moriuchi, F. & Sunamoto, J. (1982) Bull. Chem. SOC. Japan 55,1579-1583

9. Kondo, H., Uno, S., Moriuchi, F., Sunamoto, J., Ogushi, S. & Tsuru, D. (1983) Bull. Chem. Soc. Japan 56,1176-1180

10. Bodanszky, M. & Sheehan, J.T. (1966) Chem. Ind., 1597-1598

11. Wieland, T., Birr, C. & Flor, F. (1971) Angew. Chem. Int. Ed. Engl. 10,336

12. Belleau, B. & Malek, G. (1968) J. Am. Chem. SOC.

13. Kaiser, E., Colescott, R.L., Bossinger, C.D. & Cook, P.1. (1970) Anal. Biochem. 34,595-598

14. Guchhait, R.B., Polakis, S.E., Dimroth, P., Stoll, E., Moss, J. & Lane, M.D. (1974) J. Biol. Chem.

15. Merrifield, R.B. (1965) Science 150, 178-185 16. Bodanszky, M. & Fagan, D.T. (1977) J. Am.

Chem. SOC. 99,235-239 17. Hofmann, K., Finn, F.M., Friesen, H.J., Diacon-

escu, C. & Zahn, H. (1977) Proc. Natl. Acad. Sci.

18. Flanders, K.C., Mar, D.H., Folz, R.J., England, R.D., Coolican, S.A., Harris, D.E., Floyd, A.D. & Curd, R.S. (1982) Biochemistry 21, 4244- 425 1

19. Moore, S. (1963) J. Biol. Chem. 238,235-237 20. Polakis, S.E., Guchhait, R.B., Zwergel, E.E. &

Lane, M.D. (1974) J. Biol. Chem. 249, 6657- 6667

90,1651-1652

249,6633-6645

US 74,2697-2700

Address:

Hiroki Kondo Department of Industrial Chemistry Faculty of Engineering Nagasaki University Nagasaki 852 Japan

564