Purification and Characterization of a Methionine Aminopeptidase from Saccharomyces cerevisiae”

Methionine aminopeptidase (MAP), which catalyzes the removal of NH&erminal methionine from proteins, was isolated from Saccharomyces cerevisiae. The enzyme was purified 472-fold to apparent homogeneity. The M, of the native enzyme was estimated to be 36,000 + 5,000 by gel filtration chromatography, and the M, of the denatured protein was estimated to be 34,000 f 2,000 by sodium dodecyl sulfate-poly- acrylamide gel electrophoresis. The enzyme has a pH optimum near 7.0, and its p1 is 7.8 as determined by chromatofocusing on Mono P. The enzyme was inactivated by inhibitors (EDTA, o-phen-anthroline and fying reagents (HgC12 and p-hydroxymercuribenzoic and Zn’+. Yeast MAP


Purification
and Characterization of a Methionine Aminopeptidase from Saccharomyces cerevisiae" (Received for publication, June 8, 1990)  During translation, newly synthesized proteins contain either an NHz-terminal formylmethionine (prokaryotic proteins) or a methionine (eukaryotic proteins) residue (l-11). In the case of prokaryotic proteins, the formyl group is usually removed by a deformylase, leaving a methionine residue bearing a free a-NH, group (7)(8)(9)(10)(11). NH,-terminal methionine is subsequently cleaved from many, but not all, prokaryotic and eukaryotic proteins by a methionine aminopeptidase (MAP)' (l-5).
Analyses of protein sequence data from disparate proteins, as well as of data from model protein systems indicate that MAPS from both prokaryotes and eukaryotes share similar substrate specificities. MAPS are capable of removing the NHz-terminal methionine when the penultimate amino acid residue is relatively small and/or uncharged (e.g. Gly, Ala, Pro, Ser, Thr, Val, and Cys) but not when the residue is relatively bulky and/or charged (l-5, 12-14; for a review, see Ref. 15). Although a functional role for methionine remains unclear, Varshavsky and co-workers (16, 17) have suggested that NHz-terminal methionine may play a role in stabilizing certain cytosolic eukaryotic proteins whose penultimate residues are bulky and/or charged.
The genes encoding MAP from both Escherichia coli and Salmonella typhimurium were cloned and sequenced (l&19). However, the purification of MAP from a eukaryotic organism has not been previously achieved due to the instability of MAP, as well as contamination with nonspecific aminopeptidases (15,20).
In this paper we report the first complete purification, as well as the partial characterization of a MAP from a eukaryotic organism, Saccharomyces cerevisiae. This purified enzyme allowed us to clone, sequence, and study the essentiality of the MAP1 gene.* The availability of this purified enzyme will also allow us to generate polyclonal antibodies to be used in studies aimed at determining the interaction of MAP with ribosomes and other acylating enzymes. Purification and Characterization of a Yeast MAP our initial purification (data not shown). However, the enzyme could be readily purified from yeast cells stored at -20 "C for at least 4 months. It is unclear why prolonged storage at -20 "C enhanced the apparent stability of MAP, but inactivation of proteases capable of degrading MAP or processing of native MAP to a more stable form are possible explanations.
As shown in Table 1, a multiple-step purification from 200 g (wet weight) of yeast cells resulted in a 472fold purification.
The overall low fold purification can be accounted for by the presence of several contaminating leutine aminopeptidases (LAPS) in the crude extract (26, 27), which each readily degraded Met-Ala-Ser, the substrate used for assaying MAP activity. These LAPS were separated from MAP by the CM-Sepharose chromatography step. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis reveals a single Coomassie Blue-stained band with an M, = 34,000 + 2,000 (Fig, 5, lane 2). Gel filtration chromatography shows that the M, of the native MAP is 36,000 f 5,000 (Fig.  6). These data indicate that MAP is monomeric and that its M, is similar to prokaryotic MAPS from E. coli and S. typhimurium (32,000 and 35,000, respectively) (18, 28). However, the calculated molecular weights for both prokaryotic MAPS, derived from their deduced amino acid sequences, are considerably smaller (29,333 and 29,292, respectively) (18, 28). Whether a similar discrepancy for yeast MAP will be observed and what is the nature of the presumed post-translational modification remain to be established. Chromatofocusing on Mono P revealed a broad peak at apparent p1 = 7.8 (Fig. 7), which is different from the apparent p1 (5.4) determined for the S. typhimurium MAP (28). The pH dependence of yeast MAP was measured by assaying at various pH values from 5 to 8.5 (Fig. 8A), and maximal enzyme activity was observed at pH 7. Assays for determining the temperature optimum for yeast MAP were performed from 4 to 55 "C ( Fig. 8B), and the yeast MAP displays maximum activity at temperatures from 30 to 45 "C. The amino acid composition of the enzyme shown in Table  2 is characteristic of a globular protein, and differs from the amino acid compositions of the E. coli and S. typhimurium MAPS (18,28).
MAP was readily inactivated by sulfhydryl-modifying reagents, HgCb and pHMB (Table 3). Hence, a sulfhydryl group may be present in catalytic residues or coordination sites for active-site metal ion binding.
MAP can also be completely inactivated by 10 mM ZnCln (Table 3). Recently, it was shown that Zn*+ ions can inhibit another metalloprotease, carboxypeptidase A, after the formation of ZnOH+, which binds to the carboxylate involved in metal ion binding and forms a stable hydrogen bond between the inactivating Zn2+ ion and the catalytic metal ion (30). Whether Zn2+ inhibits MAP by a similar mechanism remainsto be established.
To further examine the metal ion requirement of MAP, the endogenous metal ion(s) was removed by dialysis against 1 mM EDTA, which decreased the enzyme activity of MAP (Table 4). After incubation of inactivated MAP with 3 mM CoC12, MAPS enzyme activity increased to 37%, although activation was not observed with MgClz, MnCb, CuCl,, FeC12, and ZnClz (Table 4). These data indicate that CO*+ likely plays an important role in the catalysis and stability of the eukaryotic MAP, as it does for the MAP from E. coli or S. typhimurium (18,28). Substrate Specificity of MAP-Nine amino acid p-nitroanilides and 28 peptide substrates (differing in sequence and length) were used to measure the activity of yeast MAP in vitro ( Table 5). As shown in Table 5, yeast MAP showed no detectable activity toward any of the representative amino acid p-nitroanilides (pNA) (Ala, Arg, Glu, Gly, Leu, Lys, Met, Pro, or Val), indicating that MAP is different from all other yeast aminopeptidases (for review, see Ref. 27). Among 11 representative tripeptide substrates, Xaa-Ala-Ser (Xaa = Ala, Asp, Gln, Glu, Ile, Leu, Lys, Met, Phe, Pro, and Ser), MAP cleaved only the Met-Ala peptide bond in Met-Ala-Ser, indicating that MAP was capable of cleaving only NH*-terminal methionine.
In addition, MAP cleaved NHz-terminal methionine from other tripeptides whose penultimate amino acid residue is relatively small and/or uncharged (e.g. Ala, Pro, Gly, Val, Thr, or Ser) but not when the residue is relatively bulky and/or charged (Arg, His, Leu, Met, or Tyr) ( Table 5). These results are consistent with the substrate specificity of the MAPS from E. coli and S. typhimurium (18,19) and are, in general, consistent with the results obtained by other investigators, who have analyzed the NH,terminal sequence data of various mutant forms of recombinant proteins (12)(13)(14)(15).
Yeast MAP cleaved Met-Ala and Met-Ser, albeit inefficiently, although E. coli MAP did not cleave these substrates (18). Furthermore, Huang et al. (12) observed that methionine followed by Thr, Ser, or Ala was efficiently removed in uiuo from recombinant plant thaumatin variants expressed in yeast, although in vitro yeast MAP displayed about 20-40 times higher activity toward Met-Ala-Ser than Met-Thr-Ser or Met-Ser-Gly.
Whether this disagreement is simply due to differences in the in uivo versus in vitro assays or whether there is another MAP that is responsible for the specific removal of methionine followed by Thr or Ser is presently unclear.