THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1988 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 263, No. 29, Issue of October 15, pp. 14948-14955, 1988 Printed in U.S.A.

Purification and Characterization of an Na-Acetyltransferase from Saccharomyces cerevisiae*

(Received for publication, April 11, 1988)

Fang-Jen S. Lee$#, Lee-Wen Lin$#, and John A. Smith$n(( From the Departments of $Molecular Biology and llPathology, Massachusetts General Hospital, and §Departments of Genetics and IIPathology, Haruard Medical School, Boston, Massachusetts 02114

Na-Acetyltransferase, which catalyzes the transfer of an acetyl group from acetyl coenzyme A to the a- NHz group of proteins and peptides, was isolated from Saccharomyces cerevisiae and demonstrated by pro- tein sequence analysis to be NHz-terminally blocked. The enzyme was purified 4,600-fold to apparent ho- mogeneity by successive purification steps using DEAE-Sepharose, hydroxylapatite, DE52 cellulose, and Affi-Gel blue. The M, of the native enzyme was estimated to be 180,000 f 10,000 by gel filtration chromatography, and the M, of each subunit was esti- mated to be 95,000 f 2,000 by sodium dodecyl sulfate- polyacrylamide gel electrophoresis. The enzyme has a pH optimum near 9.0, and its PI is 4.3 as determined by chromatofocusing on Mono-P. The enzyme cata- lyzed the transfer of an acetyl group to various syn- thetic peptides, including human adrenocorticotropic hormone (ACTH) (1-24) and its [Phe’] analogue, yeast alcohol dehydrogenase I (1-24), yeast alcohol dehydro- genase I1 (1-24), and human superoxide dismutase (1- 24). These peptides contain either Ser or Ala as NH2- terminal residues which together with Met are the most commonly acetylated NHz-terminal residues (Person, B., Flinta, C., von Heijne, G., and Jornvall, H. (1985) Eur. J. Biochem. 152, 523-527). Yeast enolase, con- taining a free NHz-terminal Ala residue, is known not to be Na-acetylated in vivo (Chin, C. C. Q., Brewer, J. M., and Wold, F. (1981) J. Biol. Chem. 256, 1377- 1384), and enolase (1-24), a synthetic peptide mimick- ing the protein’s NHz terminus, was not acetylated in vitro by yeast acetyltransferase. The enzyme did not catalyze the Nu-acetylation of other synthetic peptides including ACTH(11-24), ACTH(7-38), ACTH(18- 39), human &endorphin, yeast superoxide dismutase (1-24). Each of these peptides has an NHz-terminal residue which is rarely acetylated in proteins (Lys, Phe, Arg, Tyr, Val, respectively). Among a series of divalent cations, Cu2+ and Zn2+ were demonstrated to be the most potent inhibitors. The enzyme was inacti- vated by chemical modification with diethyl pyrocar- bonate and N-bromosuccinimide.

The covalent attachment of an acetyl group to proteins is the most common chemical modification of the a-NHz group of eukaryotic proteins (reviewed in Refs. 1 and 2). Acetylation is mediated by N“-acetyltransferases, which catalyze the transfer of an acetyl group from acetyl coenzyme A to the a-

* This work was supported by a grant from Hoechst Aktiengesell- schaft (Federal Republic of Germany). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

NH2 group of proteins and peptides. Since the first identifi- cation that an acetyl moiety was the NHz-terminal blocking group of tobacco mosaic virus coat protein (3) and a-melan- ocyte-stimulating peptide (4), a large number of proteins from various organisms have been shown to possess acetylated NHz-terminal residues. For example, mouse L-cells and Ehr- lich ascites cells have about 80% of their intracellular soluble proteins Ne-acetylated (5, 6). Na-Acetylation plays a role in normal eukaryotic translation and processing (7) and protects against proteolytic degradation (8, 9). For example, Na-acet- ylated NADP-specific glutamate dehydrogenase from a Neu- rospora crassa mutant and N*-acetylated ribosomal S5 protein from Escherichia coli are heat-stable, while non-acetylated forms are heat-unstable (10, 11). More recently, the rate of protein turnover mediated by the ubiquitin-dependent deg- radation system was shown to depend on the presence of a free a-NHz group at the NHz terminus of model proteins (12, 13), further suggesting that N“-acetylation may have a role in impeding protein turnover. During protein translation, Nu- acetylation occurs either cotranslationally or soon after com- pletion of polypeptide chain biosynthesis (14, 15). For pep- tides derived from proopiomelanocortin, N*-acetylation has a profound regulatory effect on the biological activity because the opioid activity of P-endorphin is completely suppressed, while the melanotropic effect of a-melanocyte-stimulating peptide is markedly increased (16-18).

Serine and alanine are the NHz-terminal residues most frequently observed in acetylated proteins and these residues, together with methionine, glycine, threonine, valine, and as- partic acid account for most of the N”-acetylated residues (1, 2, 19-21). However, since not all proteins with these residues at their NH, termini are acetylated, the selective basis by which certain proteins become acetylated is presently not understood. In order to elucidate the functional role of the acetylation and the mechanism of the enzymatic reaction, we have isolated an N“-acetyltransferase capable of specifically acetylating the NHz-terminal residue of peptides and have partially investigated its substrate specificity. The presence of an Nu-acetyltransferase has been previously demonstrated in E. coli (22), rat liver (23-25), rat brain (26), rat pituitary (27-30), bovine pituitary (29), bovine lens (311, hen oviduct (32), and wheat germ (33). None of these enzymes was purified to homogeneity, although the enzyme from hen oviduct was partially purified (40-fold).

This paper describes the first complete purification of an acetyltransferase from Saccharomyces cereuisiae. The enzyme was used to characterize partially its substrate specificity using a series of synthetic peptides derived from human

14948

Purification and Characterization of an N*-Acetyltransferase 14949

TABLE 1 Purification of N"-acetyltransferase from S. cerevisiae

SteD Activity Protein Specific activity Purification Yield units w units/mg -fold %

1. Crude extract 30,200 17,700 1.7 1.0 100 2. DEAE-Sepharose (0.2 M KCI) 32,200 3,710 8.7 5.1 107" 3. DEAE-Sepharose (0.05-0.5 M KCI) 61,500 1,470 41.8 24.5 204" 4. Hydroxylapatite 19,300 53.6 360 210 64 5. DE52-cellulose 12,700 8.58 1,500 870 42 6. Affi-Gel blue 8,160 1.05 7,800 4,600 27

An apparent inhibitor was removed during these chromatographic steps.

ACTH', as well as synthetic peptides based on the sequences of @-endorphin, and superoxide dismutase from human and alcohol dehydrogenase I, alcohol dehydrogenase 11, enolase, and superoxide dismutase from yeast.

EXPERIMENTAL PROCEDURES~

RESULTS

Homogeneity and Molecular Properties-An acetyltransfer- ase was purified to apparent homogeneity from S. cerevisiae according to the procedure described under "Experimental Procedures." As shown in Table 1, a multiple step purification from 800 g of yeast cells resulted in a 4,600-fold purification. The enzyme consists of approximately 0.01% of total cellular protein. Sodium dodecyl sulfate-polyacrylamide gel electro- phoresis reveals a single Coomassie Blue-stained band with an MI = 95,000 -C 2,000 (Fig. 6, lane 6). Gel filtration chro- matography on Sepharose CL-4B shows that the M, of the native acetyltransferase is 180,000 k 10,000 (Fig. 7). These data indicate that the yeast acetyltransferase is composed of two subunits. Chromatofocusing on Mono P revealed a single peak at apparent PI = 4.3 (Fig. 8). Assays for determining the temperature optimum for yeast acetyltransferase were per- formed from 5 to 55 'C (Fig. 9A), and the yeast acetyltrans- ferase displays a maximum activity at temperatures from 30 to 42 "C. Irreversible denaturation occurred after 1 min at 65 "C. The enzyme was most stable when stored at 4 "C in HDG buffer containing 0.2 M KCl. Under these conditions the purified enzyme had a half-life of approximately 4-6 days. The enzyme was sensitive to freezing at all stages of purifi- cation and displayed a >90% loss of activity per freeze-thaw cycle. The pH dependence of yeast acetyltransferase was measured by assaying at pH values from 5 to 11 in the presence of 50 mM buffers of potassium phosphate, HEPES,

' The abbreviations used are: ACTH, adrenocorticotropic hormone; a-MSH, a-melanocyte-stimulating hormone; HDG buffer, 20 mM HEPES, pH 7.4,0.5 mM dithiothreitol, 10% (v/v) glycerol, and 0.02% NaN3; HEPES, N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid; CAPS, 3-(cyclohexylamino)-l-propanesulfonic acid; CHES, 2- (N-cyclohexylamino)ethanesulfonic acid SDS-PAGE, sodium dode- cy1 sulfate-polyacryamide gel electrophoresis; DEPC, diethyl pyro- carbonate; NBS, N-bromosuccinimide; HNBS(CH&-Br, dimethyl- (2-hydroxy-5-nitrobenzyl)sulfonium bromide; NEM, N-ethylmaleim- ide; IAA, iodoacetic acid; IAM, iodoacetamide; pCMB, p-chloromer- curibenzoate; TNBS, 2,4,6-trinitrobenzenesulfonic acid DTT, dithi- othreitol; FPLC, fast protein liquid chromatography; SOD, superox- ide dismutase; bis-Tris, bis(2-hydroxyethyl)iminotris(hydroxy- methy1)methane; ADH, alcohol dehydrogenase; PTH, phenylthiohy- dantoin.

Portions of this paper (including "Experimental Procedures," Figs. 1-5,7-9, and Tables 2-4) are presented in miniprint at the end of this paper. Miniprint is easily read with the aid of a standard magnifying glass. Full size photocopies are included in the microfilm edition of the Journal that is available from Waverly Press.

205

116

97

66

FIG. 6. Sodium dodecyl sulfate-polyacrylamide gel electro- phoresis of purified yeast acetyltransferase. The electrophoresis was performed according to the method of Laemmli (30) using an 8% gel. The gel was stained with Coomassie Blue. Lane I, crude extract; lane 2, DEAE-Sepharose (0.2 M KCl) pool; lane 3, DEAE-Sepharose (0.05 to 0.5 M KC1) pool; lane 4, hydroxylapatite pool; lane 5, DEAE- cellulose pool; lane 6, Affi-Gel blue pool; lane 7, molecular weight standards (from the top): myosin (205,000), E. coli @-galactosidase (116,000), rabbit muscle phosphorylase (97,000), bovine serum albu- min (66,000), and egg albumin (45,000).

CHES, and CAPS buffers (Fig. 9B). Maximum enzyme activ- ity was observed at pH 9. Enzyme activity was <25% below pH 6.5 and above pH 11. The presence of KC1 or NaCl up to 0.45 M did not apparently affect the enzyme activity, although a 50% reduction in enzyme activity was observed when the enzyme assay was performed at 0.6 M KC1 or NaCl (data not shown).

Amino Acid Analysis-Samples of electroeluted yeast ace- tyltransferase were subjected to amino acid analysis, and the amino acid composition of the enzyme is shown in Table 2. The amino acid composition is characteristic of a globular protein.

Protein Sequence Analysis-Two different preparations of the electroeluted protein (each -300 pmol) were submitted to protein sequence analysis. Except for background levels of amino acids at the first cycle, no protein sequence was de- tected. These results demonstrate that the NH2 terminus of the protein is blocked. Studies now underway are aimed at defining the chemical nature of the NH2-terminal modifica- tion.

Effect of Divalent Cations on Enzyme Activity-The effect

14950 Purification and Characterization of an Ne-Acetyltransferase

of various divalent cations on the enzyme activity was deter- mined (Table 3). At 1 mM concentration, Ca2+ and Mg2f had no effect, whereas pronounced, concentration dependent in- hibition occurred in the presence of other divalent ions (Zn2+ > Cu2+ >> Fez+ = Cd2+ > Co2+ > Mn2+). Cu2+ and Zn2+ completely abolished enzyme activity at >0.1 mM, but only partially inactivated enzyme at 0.01 mM. Further, it was demonstrated that the pronounced effects, observed with CuS04 and ZnS04, were not due to the SOT2 anion, since MgS04 did not affect enzyme activity as compared with the activity in the presence of MgC12.

Effect of Chemical Modifications on Enzyme Activity-In order to determine the possible catalytic role for different types of amino acid residues in the enzyme, various chemical modifications were carried out (Table 4). The reaction of acetyltransferase with diethyl pyrocarbonate, a histidine- modifying reagent (40), caused a nearly complete inactivation of the enzyme. After incubated for 8 h a t room temperature with 0.25 M hydroxylamine, a reagent capable of reversing the ethoxyformylation of a histidine residue, -60% of the original enzyme activity could be recovered, although prolonged ex- posure to hydroxyamine slowly inactivated the enzyme (data not shown). The presence of a catalytically important tryp- tophan residue was investigated by chemical modification with NBS (41) and HNBS(CH3)2-Br (42). NBS partially inactivated the enzyme at 0.5 mM and completely inactivated the enzyme a t 5 mM. Although HNBS(CH&Br partially inactivated the enzyme, the loss of enzyme activity was small in comparision to NBS. Since NBS can also modify histidine and tyrosine residues (43), it is possible that the inactivation might be due to a chemical modification of the same histi- dine(s), presumably modified by diethyl pyrocarbonate. Sulfhydryl-reducing agents (ix. 2-mercaptoethanol and dithi- othreitol) did not affect on the enzyme activity. Sulfhydryl- modifying reagents (i.e. NEM, iodoacetic acid, iodoacetamide, and p-chloromercuribenzoate were also without effect at 1 mM, but partial inactivation of the enzyme was observed with each reagent at 10 mM concentration. Succinic anhydride and 2,4,6-trinitrobenzenesulfonic acid, which modify lysine resi- dues and u-NH2 groups (44, 45), had no effect on enzyme activity. N-Acetylimidazole, a tyrosine-modifying reagent, was also without effect (44).

Substrate Specificity-Because ACTH( 1-24) was used pre- viously as an assay substrate for characterizing other acetyl- transferases from bovine lens (31), pituitary (27-29), and hen oviduct (32), this peptide was chosen as the substrate in the assay used for the yeast N*-acetyltransferase. All the other peptides in Table 5 were compared to ACTH(1-24).

The [Phe'] analogue of ACTH(1-24) was acetylated almost as well as ACTH(1-24). In contrast, ACTH(11-24), ACTH(7- 38) and ACTH( 18-39) were not acetylated. &Endorphin was ineffectively acetylated by yeast Ne-acetyltransferase, al- though this substrate was previously shown to be acetylated at its NH2-terminal Tyr residue by acetyltransferase from pituitary (29, 30) or rat brain (26).

Yeast alcohol dehydrogenase, either type I or 11, can be naturally acetylated at its NH2-terminal Ser residue (46, 47). As shown in Table 5, the tetradodecapeptides mimicking the NH2 termini of these isoenzymes are acetylated equally well and as effectively as for ACTH(1-24). Human superoxide dismutase is naturally acetylated at its NH2-terminal Ala residue (48) and is also acetylated when expressed as a recom- binant protein in yeast (49). However endogenous yeast su- peroxide dismutase, which has a 48% amino acid sequence identity with human superoxide dismutase, is not Ne-acety- lated (50). Whether or not NH2-terminal sequence differences

TABLE 5 Relative activity of yeast acetyltransferase for the NO-acetylation of

synthetic peptides and histones Substrate Activity'

ACTH(1-24) Ser-Tyr-Ser-Met-Glu-His-Phe-Arg-Trp- Gly-Lys-Pro-Val-Gly-Lys-Lys-Arg-Arg- Pro-Val-Lys-Val-Tyr-Pro

[Phe'] ACTH(1-24) Ser-Phe-Ser-Met-Glu-His-Phe-Arg-Trp-

Glxys-Pro-Val-Gly-Lys-Lys-Arg-Arg- Pro-Val-Lys-Val-Tyr-Pro

ACTH(11-24) Lys-Pro-Val-Gly-Lys-Lys-Arg-Arg-Pro-

Val-Lys-Val-Tyr-Pro ACTH(7-38) Phe-Arg-Trp-Gly-Lys-Pro-Val-Gly-Lys- Lys-Arg-Arg-Pro-Val-Lys-Val-Tyr-Pro- Asn-Gly-Ala-Glu-Asp-Glu-Ser-Ala-Glu- Ala-Phe-Pro-Leu-Glu

ACTH(18-39) Arg-Pro-Val-Lys-Val-Tyr-Pro-Asn-Gly- Ala-Glu-Asp-Glu-Ser-Ala-Glu-Ala-Phe- Pro-Leu-Glu

@-Endorphin (human) Tyr-Gly-Gly-Phe-Met-Thr-Ser-Glu-Lys- Ser-Gln-Thr-Pro-Leu-Val-Thr-Leu-Phe- Lys-Asn-Ala-Ilu-Ilu-Lys-Asn-Ala-Tyr- Lys-Lys-Gly-Glu

Alcohol dehydrogenase I(1-24) (yeast) Ser-Ile-Pro-Glu-Thr-Gln-Lys-Gly-Val-Ile- Phe-Tyr-Glu-Ser-His-Gly-Lys-Leu-Glu- Tyr-Lys-Asp-Ile-Pro

Alcohol dehydrogenase II(1-24) (yeast) Ser-Ile-Pro-Glu-Thr-Gln-Lys-Ala-Ile-Ile-

Phe-Tyr-Glu-Ser-&-Gly-Lys-Leu-Glu- - His-Lys-Asp-Ile-Pro

Superoxide dismutase( 1-24) (yeast) Val-Gln-Ala-Val-Ala-Val-Leu-Lys-Gly- Asp-Ala-Gly-Val-Ser-Gly-Val-Val-Lys- Phe-Glu-Gln-Ala-Ser-Glu

Superoxide dismutase( 1-24) (human) Ala-Thr-Lys-Ala-Val-Cys-Val-Leu-Lys- Gly-Asp-Gly-Pro-Val-Gln-Gly-Ser-Ile- Asn-Phe-Glu-Gln-Lys-Glu

Enolase(1-24) (yeast) Ala-Val-Ser-Lys-Val-Tyr-Ala-Arg-Ser-Val- Tyr-Asp-Ser-Arg-Gly-Asn-Pro-Thr-Val- Glu-Val-Glu-Leu-Thr

Histone (lysine-rich) (calf thymus) Histone (arginine-rich) (calf thymus)

% 100 + 5

90 + 9

0

0

0

2 + 2

101 + 5

102 + 4

0

86 * 6

4 + 2

0 0

"Data reported as mean activity + S.D. (n = 3-5).

between these proteins account for the differences in Na- acetylation remains unclear. Synthetic human superoxide dis- mutase (1-24) was acetylated by the yeast enzyme. However, synthetic yeast superoxide dismutase (1-24) was not acety- lated. In addition, native yeast enolase containing an Ala residue with a free ( Y - N H ~ group is known not to be N- acetylated in vivo (51) and indeed a synthetic peptide mim- icking the NH2-terminal 24 amino acid residues of the protein was also not acetylated by the yeast enzyme. Furthermore, neither calf thymus lysine- nor arginine-rich histones was acetylated by the yeast acetyltransferase.

DISCUSSION

N"-Acetyltransferases transfer an acetyl group from acetyl-

Purification and Characterization of an Ne-Acetyltransferase 14951

CoA to NHz-terminal residues of proteins and peptides. Al- though others have attempted to isolate this enzyme from a variety of sources including liver (23-25), pituitary (27-29), and wheat germ (33), only the enzyme from hen oviduct (32) and rat liver (25) have been partially purified. We purified yeast Ne-acetyltransferase (Fig. 6) 4,600-fold with an overall recovery of 27%, by successive chromatographic procedures utilizing DEAE-Sepharose, hydroxylapatite, DEAE-cellulose, and Affi-Gel blue (Table 1). The enzyme was NHp-terminally blocked, although the chemical nature of this blocking group has not yet been determined. The M , of the native enzyme is 180,000 f 10,000, and the enzyme consists of two subunits each with a M , of 95,000 k 2,000. These values agree with the values previously determined by Bradshaw and coworkers (25), who found that the rat liver enzyme has a M , of -200,000 and that the enzyme is a dimer, Driessen et al. (2), who found that the calf lens enzyme has a M, of 170,000, and Chapel1 et al. (30), who found an enzyme in the rat pituitary neurointer- mediate lobe has a M, > 200,000. The PI for the yeast enzyme was 4.3, and the pH optimum was 9.0. The isoelectric point for the enzyme from calf lens (PI = 5) is quite similar to that of the yeast enzyme. By contrast, the hen oviduct enzyme (32) differs from the yeast enzyme both in its M, (250,000) and pH optimum (pH 7.2). The yeast enzyme was inhibited markedly by cupric and zinc ions (Table 3) but was not activated by chloride ion as previously demonstrated for the wheat germ enzyme (33).

Studies involving other acetyltransferases have shown that chemical modification of cysteine in acetyl-CoA.arylamine N - acetyltransferase (52) and choline O-acetyltransferase (53) or of histidine in acetyl-CoA:a-glucosaminide N-acetyltransfer- ase (54) inactivated these enzymes. Our studies using DEPC (and possibly NBS) suggest that a histidine residue may be located within the active site of yeast Ne-acetyltransferase, but the failure of @-mercaptoethanol, dithiothreitol, N-ethyl- maleimide, p-chloromercuribenzoate, iodoacetic acid, and iodoacetamide to inactivate the enzyme indicates that a cys- teine residue is probably not involved in the catalytic mech- anism (Table 4).

It is found in nature that serine, alanine, methionine, glycine, threonine and aspartic acid are the dominant NHz- terminal residues of the Ne-acetylated proteins (1, 2, 19-21). Although this predominant use of certain amino acid residues suggests that an NHz-terminal residue is the primary recog- nition signal for N"-acetyltransferase, there are numerous examples of proteins having these amino acids as their NH, termini that are not acetylated. Thus, it is likely that the enzyme must also recognize some additional sequence or conformational features within the NHZ-terminal region of a protein or peptide, as well as a specific NHz-terminal residue. As shown in Table 5, the relative acetylation of the NH2- terminal Ser of ACTH(1-24) and [Phe'] ACTH were nearly identical and indicated that the hydroxyl group of the tyrosyl residue at amino acid position 2 does not have a major role in the recognition of this site by acetyltransferase. Other ACTH analogues including ACTH(l1-24), ACTH(7-38), ACTH(18- 39), which contain Lys, Phe, and Arg as their NHz-terminal residues, were not acetylated. Such residues are observed rarely to be acetylated in eukaryotic proteins, although it is not clear whether the failure of acetyltransferase to acetylate these particular peptides is due to the presence of these residues themselves or to the absence of certain conforma- tional features required for interaction with acetyltransferase. Although it was shown that P-endorphin can be acetylated by an endogenous acetyltransferase found in rat brain (26) or pituitaries (29, 30) which results in its inactivation of the

hormone (16, 17), the yeast acetyltransferase did not effec- tively acetylate this substrate. These findings suggest that these pituitary enzymes have a different substrate specificity than the yeast enzyme, but whether or not the pituitary- derived acetyltransferase will acetylate the various synthetic peptides listed in Table 5 is at present not known.

In our study, yeast alcohol dehydrogenase isoenzyme I, as a native protein, contains an N-Ac-Ser as its NHz-terminal residue (46), and a peptide containing the first 24 residues of the protein was efficiently acetylated by yeast acetyltransfer- ase. This observation indicates that the amino acid sequence and conformational feature(s) required for the enzyme to initiate acetylation of the native alcohol dehydrogenase I is contained within an initial NHz-terminal sequence. The acet- ylated form of yeast alcohol dehydrogenase isoenzyme I1 was observed to be acetylated less frequently than the alcohol dehydrogenase isoenzyme I (47), and this has been attributed to differences in the amount of acetyl CoA available to be transferred by the acetyltransferase to the alcohol dehydro- genase I1 under different physiological conditions. However, synthetic peptides mimicking residues 1-24 of both alcohol dehydrogenase isoenzymes were equally acetylated by the yeast acetyltransferase, suggesting that under conditions where acetyl-coA is not depleted that effective NHz-terminal acetylation of both isoenzymes will be carried out. Further- more, although these peptides differ from one another at residues 8, 9, 15, and 20, it is clear that variation of these amino acid residues did not affect interaction with acetyl- transferase and subsequent acetylation. It has been demon- strated that the yeast superoxide dismutase protein is not acetylated in vivo (50). In contrast, the human superoxide dismutase (48) and a recombinant form of the human enzyme expressed in yeast (49) are both acetylated. Studies with synthetic peptides mimicking the NHz-terminal of these su- peroxide dismutase proteins indicated that the yeast acetyl- transferase could not acetylate the yeast superoxide dismutase peptide substrate (yeast superoxide dismutase (1-24)). How- ever, the yeast acetyltransferase effectively acetylated the corresponding human peptide substrate (human superoxide dismutase (1-24)) (Table 5). The nature of the side chain of NHz-terminal residue may account for this differential acet- ylation, since Ala (seen in human superoxide dismutase) is among the most frequently acetylated amino acids and Val (seen in yeast superoxide dismutase) is among the most infre- quently acetylated residues. Since these two peptides (as well as the proteins themselves) bear an -50% sequence identity with one another, it is likely that amino acid sequence or conformational differences elsewhere in the peptides account for the differences in acetylation. However, not all yeast proteins containing an Ala as their NH, terminus will be acetylated (e.g. yeast enolase is known not to be acetylated (51)). Furthermore, the peptide substrate corresponding to the NHz-terminal 24 residues of enolase was poorly acetylated (Table 5).

Enzymatic t-NHz acetylation of lysine residues has been documented to occur in a variety of proteins, and this type of acetylation has been most extensively studied using histones (28,55-58). Recently, the histone acetyltransferase from yeast was partially purified by Travis et al. (59), and it had a pH optimum of 8.2, a PI > 9.0, and a M, of 110,000. The yeast acetyltransferase described in this paper did not acetylate either lysine- or arginine-rich histones or various peptide substrates containing lysine residues (ACTH( 11-24), ACTH(7-38), ACTH( 18-39), @-endorphin, and superoxide dismutase (yeast)), and we conclude that this enzyme is not

14952 Purification and Characterization of an N*-Acetyltransferase

capable of transferring an acetyl group to the eNH2 group of a lysyl residue.

An inhibitor of acetyltransferase has been previously dem- onstrated in a crude homogenate from pituitary (29) and a crude lysate from yeast (59). During the purification of yeast acetyltransferase we too detected the presence of an endoge- nous inhibitor of acetyltransferase as evidenced by the 204% recovery of the enzyme activity observed after the second DEAE-Sepharose chromatography (Table 1). We have sub- sequently purified this inhibitor and have begun to character- ize its biological activity.

Furthermore, the fact that the synthetic human superoxide dismutase (1-24) and human ACTH(1-24) peptides can be acetylated by yeast acetyltransferase indicates that the sub- strate specificity of Ne-acetyltransferase is likely to be highly conserved between yeast and human. Therefore, this purified acetyltransferase from yeast may provide a ready source of enzyme for studies of catalytic mechanism and substrate specificity of both the yeast and human enzyme.

NH2-terminal acetylation has been suggested to be involved in protein degradation. Hershko et al. (12) showed that Ne- acetylated protein are degraded by the ubiquitin/ATP-de- pendent system less rapidly than the corresponding proteins with free NH2 termini, suggesting that N*-acetylation of cellular proteins may prevent their degradation by this sys- tem. Recently, Bachmair et al. (13) showed that the occur- rence of a free, “destabilizing,” NHz-terminal residue (leucine, phenylalanine, aspartic acid, lysine, arginine, isoleucine, glu- tamine, glutamic acid, and tyrosine) correlated with a short half-life of the protein in the cell, whereas methionine, serine, alanine, glycine, threonine, and valine correlate with a long half-life. The latter data suggest the involvement of specific aminopeptidases in the initiation of intracellular protein deg- radation by exposing free destabilizing NHz-terminal resi- dues. It is also possible that the removal of N-acetylated amino acids by acyl-peptide hydrolases (60, 61) may be a trigger for the degradation of N-acetylated proteins, although this remains to be established.

In light of the work of Rubenstein and Martin (62), who have demonstrated that nascent polypeptide chain of Dro- sophila actins is initially acetylated, that the acetylated resi- due is subsequently removed (presumably by either a N - acetylmethionine aminopeptidase (63) or an acylpeptide hy- drolase (60, 61)), and that the ultimate residue is once again acetylated, it is likely that acetylation may be important in the regulation of NHz-terminal processing pathways of eu- karyotic proteins.

Partial sequence data for acetyltransferase has been deter- mined and will be utilized for the cloning of the encoding gene. Later, while using yeast as a genetic analysis system, the cloned gene will form the basis for a direct investigation of the functional role of protein NH2-terminal acetylation.

Acknowledgments-We wish to acknowledge the skilled technical assistance of Gene Hannigan and Paul Gallant.

REFERENCES 1. Tsunasawa, S., and Sakiyama, F. (1984) Methods Enzymol. 106,

165-170 2. Driessen, H. P. C., De Jong, W. W., Tesser, G. I., and Bloemendal,

H. (1985) CRC Crit. Reu. Biochem. 18 , 281-325 3. Narita, K. (1958) Biochim. Biophys. Acta 2 8 , 184-191 4. Harris, J. I. (1959) Biochem. J. 7 1 , 451-459 5. Brown, J. L., and Roberts, W. K. (1976) J. Biol. Chem. 2 5 1 ,

6. Brown, J. L. (1979) J. Biol. Chem. 254, 1447-1449 7. Wold, F. (1984) Trends Biochem. Sci. 9, 256-257

1009-1014

8. Jornvall, H. (1975) J. Theor. Biol. 55, 1-12 9. Rubenstein, P., and Deuchler, J. (1979) J. Biol. Chem. 2 5 4 ,

10. Siddig, M. A. M., Kinsey, J. A., Fincham, J. R. S., and Keighren,

11. Cumberlidge, A. G., and Isono, K. (1979) J. Mol Biol. 131 , 169-

12. Hershko, A., Heller, H., Eytan, E., Kaklij, G., and Rose, I. A.

13. Bachmair, A., Finley, D., and Varshavsky, A. (1986) Science 234 ,

14. Bloemendal, H. (1977) Science 197 , 127-138 15. Kasten-Jolly, J., and Taketa, F. (1982) Arch. Biochem. Biophys.

16. Smyth, D. G., Massey, D. E., Zakarian, S., and Finnie, M. D. A.

17. Smyth, D. G., and Zakarian, S. (1980) Nature 288,613-615 18. Ramachandran, J., and Li, C. H. (1967) Adu. Enzymol. 2 9 , 391-

19. Persson, B., Flinta C., Heijne G., and Jornvall H. (1985) Eur. J .

20. Augen, J., and Wold, F. (1986) Trends Biochem. Sci. 11 , 494-

21. Tsunasawa, S., Stewart, J. W., and Sherman, F. (1985) J. Biol.

22. Brot, N., Marcel, R., Cupp, L., and Weissbach, H. (1973) Arch.

23. Pestana, A., and Pitot, H. C. (1975) Biochemistry 14,1397-1403 24. Pestana, A., and Pitot, H. C. (1975) Biochemistry 14,1404-1412 25. Yamada, R., Kendall, R., and Bradshaw, R. A. (1987) 1st Sym-

posium of the Protein Society, p. 34 (Abstr. 625), San Diego, CA

11142-11147

M. (1980) J. Mol. Biol. 137 , 125-135

189

(1984) Proc. Nutl. Acad. Sci. U. S. A. 8 1 , 7021-7025

179-186

214,829-839

(1979) Nature 2 7 9 , 252-254

477

Biochem. 152,523-527

497

Chem. 260,5382-5391

Biochem. Biophys. 155,475-477

26. O’Donohue, T. L. (1983) J. Biol. Chem. 258,2163-2167 27. Woodford, T. A., and Dixon J. E. (1979) J. Biol. Chem. 254,

28. Pease, K. A., and Dixon J. E. (1981) Arch. Biochem. Biophys.

29. Glembotski, C. C. (1982) J. Biol. Chem. 2 5 7 , 10501-10509 30. Chappell, M. C., O’Donohue, T. L., Millington, W. R., and

31. Granger, M., Tesser, G. I., De Jong, W. W., and Blomendal, H.

32. Tsunasawa, S., Kamitani, K., and Narita, K. (1980) J. Biochem.

33. Kido, H., Vita, A., Hannappel, E., and Horecker, B. L. (1981) Arch. Biochem. Biophys. 208,95-100

34. Bradford, M. M. (1976) Anal. Biochem. 72,248-254 35. Laemmli, U. K. (1970) Nature 227,680-685 36. Hunkapiller, M., Lujan, E., Ostrander, F., and Hood, L. (1983)

Methods Enzymol. 9 1 , 227-236 37. Moore, S. (1972) in Chemistry and Biology of Peptides (Meien-

hofer, J., ed), pp. 629-652, Ann Arbor Science, Ann Arbor, MI 38. Hewick, R. M., Hunkapiller, M. W., Hood, L. E., and Dreyer, W.

39. Finnegan, A., Smith, M. A., Smith, J. A., Berzofsky, J., Sachs, D. H., and Hodes, R. J. (1986) J. Exp. Med. 164,897-910

40. Miles, E. W. (1977) Methods Enzymol. 47,431-442 41. Spande, T. F., and Witkop, B. (1967) Methods Enzymol. 11,506-

42. Horton, H. R., and Koshland, D. E. (1972) Methods Enzymol.

43. Witkop, B. (1963) Adu. Protein Chem. 16, 221-321 44. Riordan, J. F., and Vallee, B. L. (1972) Methods Enzymol 2 5 ,

45. Fields, R. (1972) Methods Enzymol. 25,464-468 46. Jornvall, H. (1977) Eur. J. Biochem. 72.425-442 47. Jornvall, H., Fairwell, T., Kratofil, P., and Wills C. (1980) FEBS

48. Jabusch, J. R., Farb, D. L., Kerschensteiner, D. A., and Deutsch, H. F. (1980) Biochemistry 19,2310-2316

49. Hellewell, R. A., Mills, R., Tekamp-Olson, P., Blacher, R., Ro- senberg, S., Otting, F., Masiarz, F. R., and Scandella, c. J. (1987) BiolTechnology 5 , 363-366

50. Johansen, J . T., Overballe-Petersen, C., Martin, B., Hasemann, V., and Svendsen, I. (1979) Carlsberg Res. Commun. 4 4 , 201- 217

4993-4999

212,177-185

Kempner, E. S. (1986) J. Biol. Chem. 2 6 1 , 1088-1091

(1976) Proc. Nutl. Acad. Sci. U. S. A. 73, 3010-314

(Tokyo) 87,645-650

J. (1981) J. Biol. Chem. 2 5 6 , 7990-7997

522

25,468-482

500-506

Lett. 111,214-218

Purification and Characterization of an No-Acetyltransferase 14953 51.

52.

53. 54.

55.

56.

57.

Chin, C. C. Q., Brewer, J. M., and Wold, F. (1981) J. Biol. Chem.

Jencks, W. P., Gresser, M., Valenzuela, M. S., and Huneeus, F.

Roskoski, R., Jr. (1974) J. Biol. Chem. 2 4 9 , 2156-2159 Bame, K. J., and Rome, L. H. (1986) J. Biol. Chem. 2 6 1 , 10127-

De Lange, R. J., Fambrough, D. M., Smith, E. I., and Bonner, J.

Pestana, A., Sudilovsky, O., and Pitot, H. C. (1971) FEBS Lett.

Marushige K. (1976) Proc. Natl. Acad. Sci. U. S. A. 73, 3937-

256 , 1377-1384

C. (1972) J. Biol. Chem. 247,3756-3760

10132

(1969) J. Biol. Chem. 244,319-334

19,83-86

3941

58. Prevelige P. E., Jr., and Fasman, G. D. (1987) Biochemistry 2 6 ,

59. Travis, G. H., Colavito-Shepanski, M., and Grunstein, M. (1984)

60. Tsunasawa, S., Narita, K., and Ogata, K. (1975) J. Biochem.

61. Kobayashi, K., and Smith, J. A. (1987) J. Biol. Chem. 2 6 2 ,

62. Rubenstein, P. A,, and Martin, D. J. (1983) J. Biol. Chem. 258,

63. Radhakrishna, G., and Wold, F. (1986) J. Biol. Chem. 261,9572-

2944-2955

J. Biol. Chem. 259, 14406-14412

(Tokyo) 77, 89-102

11435-11445

11354-11360

9575

cell Exuacnon

14954 Purification and Characterization of an N*-Acetyltransferase

I I E 4000 . - ._ c 2 3000 z > - 'E 2000 4

0)

E, 1000 N c W

0

50

40

30 H E

20 g 6 N

10

0 20 40 60 80 100 120

Fraction Number

I I - E 6000 c

- 2. 4000 - - 2 &, 2000

E, W

4

3

2

1

0 0 1 on 200

Fraction Number

r 0 4

T 4000 1 - 15

E E 0 N a

0.5

0.4

0.3

0.2

0.1

0.0 0 100 200

Fraction Number

T 4000 I

3000

2000

1000

0 0 100 200

Fraction Number

I 3000 0 22

- - 0.20 1 .o

z 2000 0.8 E . I

A - L

9 1000

ui

0.12 0.2

0 50 100 150

Fraction Number

0 10 0.0

6

Acetyltransferase

h z z 5 - 0 -I

4 I . 1.8 2.0 2.2 2.4 2.6 2

V e /Vo

B

800 7

600 6

400 5

200 4

I B

Purification and Characterization of an N"-Acetyltransferase Table 3

IO0

80

60

40

20

0 I I . I . I . I ' 1 '

0 10 20 30 40 50 60

Temperature ( "C )

200

150

100

50

0

Q HEPES + K- PO^ +- CHES * CAPS

4 6 8 10 12

PH

I n. I 0.01

." ."

I20 ~..

I 20

43 84

M) 86

42 90

53 78

0 58

0 40

2-mrraptoethanol

D l T

NEM

L4.4

IAM

pCMB

TNBS

Succmc anhydnds

N~acetylrm>dmlc

0 s 5 I1

I1 5 S O

10.0 I .o

10.0

10.0

1.0 10.0

10.0 I .o

10.0 I O

I .o l o o

10.0 I .o

10.0 10

I .o t o o

I IO

Ill1

92 13

11x1 73

98 63

11x1 5s

94 76

41 71

I(K1 63

14955