Multistep autoactivation of asparaginyl endopeptidase in vitro and in vivo

Mammalian asparaginyl endopeptidase (AEP) or legumain is a recently discovered lysosomal cysteine protease that specifically cleaves after asparagine residues. How this unusually specific lysosomal protease is itself activated is not fully understood. Using purified recombinant pro-enzyme, we show that activation is autocatalytic, requires sequential removal of C- and N-terminal pro-peptides at different pH thresholds, and is bimolecular. Removal of the N-terminal propeptide requires cleavage after aspartic acid rather than asparagine. Cellular processing, either of exogenously added AEP precursor or of pulse-labeled endogenous precursor, introduces at least one further cleavage to yield the final mature lysosomal enzyme. We also provide evidence that in living cells, there is clear compartmental heterogeneity in terms of AEP activation status. Moreover, we show that human monocyte-derived dendritic cells harbor inactive proforms of AEP that become activated upon maturation of dendritic cells with lipopolysaccharide.     

Isolation and characterization of a stable activation intermediate of the lysosomal aspartyl protease cathepsin D.

Procathepsin D is the intracellular aspartyl protease precursor of cathepsin D, a major lysosomal enzyme. Procathepsin D is rapidly processed inside the cell, and, thus, examination of its proteolyic activation and structure has been difficult. To study this proenzyme, a nonglycosylated form of the human fibroblast procathepsin D was expressed in Escherichia coli, refold in vitro, and purified by affinity chromatography on pepstatinyl agarose. Sequence analysis of the refolded, autoactivated enzyme allowed determination of the autoproteolytic cleavage site. The sequence surrounding this cleavage site between residues LeuP26 and IleP27 (in the "pro" region) resembled the first cleavage site found during activation of other aspartyl proteases. Thus, the autoactivated procathepsin D is analogous to the pepsin activation intermediate, which has been termed pseudopepsin. The enzymatic activity, thermal and pH stability, and fluorescence spectra of pseudocathepsin D were compared to mature, predominantly two-chain, cathepsin D isolated from human placenta. The results indicated that pseudocathepsin D and mature enzyme have a similar Km toward a peptide substrate and cleave a protein substrate at identical sites. Temperature stability of the recombinant enzyme was similar to that of the tissue-derived enzyme. However, the recombinant enzyme had increased stability at low pH when compared to the glycosylated tissue-derived two-chain cathepsin D. Fluorescence spectra of the recombinant and tissue-derived enzymes were identical. Thus, the absence of asparagine-linked oligosaccharides and the presence of the remaining segment of propeptide did not significantly alter the structural and enzymatic properties of the enzyme.     

Characterization of legumain

The mammalian legumain, also called asparaginyl endopeptidase (AEP), is critically involved in the processing of bacterial antigens for MHC class II presentation. In order to investigate the substrate specificity of AEP in the P1' position, we created a peptide library and digested it with purified pig kidney AEP. Digestion was less efficient only when proline was in the P1' position. Maximum AEP activity was found in lysosomal fractions of different types of antigen presenting cells (APC). When the multiple sclerosis-associated autoantigen myelin basic protein (MBP) was digested with AEP, the immunodominant epitope 83-99 was destroyed. Myoglobin as an alternative substrate was AEP resistant. These results suggest an important, but not necessarily critical role for AEP in lysosomal antigen degradation.     

Neuroprotective actions of PIKE-L by inhibition of SET proteolytic degradation by asparagine endopeptidase

Ischemia and seizure cause excessive neuronal excitation that is associated with brain acidosis and neuronal cell death. However, the molecular mechanism of acidification-triggered neuronal injury is incompletely understood. Here, we show that asparagine endopeptidase (AEP) is activated under acidic condition, cuts SET, an inhibitor of DNase, and triggers DNA damage in brain, which is inhibited by PIKE-L. SET, a substrate of caspases, was cleaved by acidic cytosolic extract independent of caspase activation. Fractionation of the acidic cellular extract yielded AEP that is required for SET cleavage. We found that kainate provoked AEP activation and SET cleavage at N175, triggering DNA nicking in wild-type, but not AEP null, mice. PIKE-L strongly bound SET and prevented its degradation by AEP, leading to resistance of neuronal cell death. Moreover, AEP also mediated stroke-provoked SET cleavage and cell death in brain. Thus, AEP might be one of the proteinases activated by acidosis triggering neuronal injury during neuroexcitotoxicity or ischemia.     

Structural basis of proteolytic activation of L-phenylalanine oxidase from Pseudomonas sp. P-501

The mature form of l-phenylalanine oxidase (PAOpt) from Pseudomonas sp. P-501 was generated and activated by the proteolytic cleavage of a noncatalytic proenzyme (proPAO). The crystal structures of proPAO, PAOpt, and the PAOpt-o-amino benzoate (AB) complex were determined at 1.7, 1.25, and 1.35A resolutions, respectively. The structure of proPAO suggests that the prosequence peptide of proPAO occupies the funnel (pathway) of the substrate amino acid from the outside of the protein to the interior flavin ring, whereas the funnel is closed with the hydrophobic residues at its vestibule in both PAOpt and the PAOpt-AB complex. All three structures have an oxygen channel that is open to the surface of the protein from the flavin ring. These results suggest that structural changes were induced by proteolysis; that is, the proteolysis of proPAO removes the prosequence and closes the funnel to keep the active site hydrophobic but keeps the oxygen channel open. The possibility that an interaction of the hydrophobic side chain of substrate with the residues of the vestibule region may open the funnel as a putative amino acid channel is discussed.     

The activation of human skin fibroblast procollagenase. Sequence identification of the major conversion products

Human skin collagenase is secreted by cultured fibroblasts in a proenzyme form and can be activated to a catalytically competent enzyme by a number of processes. All modes of activation studied lead to conversion of the proenzyme to a stable 42-kDa active enzyme, concomitant with removal of an 81-amino acid peptide from the amino-terminal end of the molecule. The sequence of events leading to the formation of this enzyme form has been determined by analyzing the primary structure of the conversion intermediates. Trypsin-induced activation of procollagenase occurs as a result of the initial cleavage of the peptide bond between Arg-55 and Asn-56, generating a major intermediate of 46 kDa. Treatment of the proenzyme with organomercurials, which have no intrinsic ability to cleave peptide bonds, initially results in activation of the enzyme without loss of molecular weight. This is followed by conversion to two lower molecular weight species of 44 and 42 kDa, the latter corresponding to the stable active enzyme form. The final cleavage producing this form of collagenase is not restricted to a single polypeptide bond but can occur on the amino-terminal side of any one of three contiguous hydrophobic residues, Phe-100, Val-101, Leu-102. The data suggest that both trypsin and organomercurials activate procollagenase by initiating an intramolecular autoproteolytic reaction resulting in the formation of a stable 42-kDa active enzyme species.     

Structural constraints on autoprocessing of the human nucleoporin Nup98

Nucleoporin Nup98, a 98-kDa protein component of the nuclear pore complex, plays an important role in both protein and RNA transport. During its maturation process, Nup98 undergoes post-translational autoproteolysis, which is critical for targeting to the NPC. Here we present high-resolution crystal structures of the C-terminal autoproteolytic domains of Nup98 (2.3 A for the wild type and 1.9 A for the S864A precursor), and propose a detailed autoproteolysis mechanism through an N-O acyl shift. Structural constraints are found at the autocleavage site, and could thus provide a driving force for autocleavage at the scissile peptide bond. Such structural constraints appear to be generated, at least in part, by anchoring a conserved phenylalanine side chain into a highly conserved hydrophobic pocket at the catalytic site. Our high-resolution crystal structures also reveal that three highly conserved residues, Tyr866, Gly867, and Leu868, provide most of the interactions between the autoproteolytic domain and the C-terminal tail. These results suggest that Nup98 may represent a new subtype of protein that utilizes autoprocessing to control biogenesis pathways and intracellular translocation.     

Maturation of human tripeptidyl-peptidase I in vitro

Tripeptidyl-peptidase I (TPP I, CLN2 protein) is a lysosomal aminopeptidase that cleaves off tripeptides from the free N termini of oligopeptides and also shows minor endopeptidase activity. TPP I is synthesized as a preproenzyme. Its proenzyme autoactivates under acidic conditions in vitro, resulting in a rapid conversion into the mature form. In this study, we examined the process of maturation in vitro of recombinant latent human TPP I purified to homogeneity from secretions of Chinese hamster ovary cells overexpressing TPP I cDNA. Autoprocessing of TPP I proenzyme was carried out at a wide pH range, from approximately 2.0 to 6.0, albeit with different efficiencies depending on the pH and the type of buffer. However, the acquisition of enzymatic activity in the same buffer took place in a narrower pH "window," usually in the range of 3.6-4.2. N-terminal sequencing revealed that mature, inactive enzyme generated during autoactivation at higher pH contained N-terminal extensions (starting at 6 and 14 amino acid residues upstream of the prosegment/mature enzyme junction), which could contribute to the lack of activity of TPP I generated in this manner. Autoprocessing was not associated with any major changes of the secondary structure of the proenzyme, as revealed by CD spectroscopy. Both the activation and proteolytic processing of the recombinant TPP I precursor were primarily concentration-independent. The addition of the mature enzyme did not accelerate the processing of the proenzyme. In addition, the maturation of the proenzyme was not affected by the presence of glycerol. Finally, the proenzyme with the active site mutated (S475L) was not processed in the presence of the wild-type enzyme. All of these findings indicate a primarily intramolecular (unimolecular) mechanism of TPP I activation and autoprocessing and suggest that in vivo mature enzyme does not significantly participate in its own generation from the precursor.     

X-ray structures of Aerococcus viridans lactate oxidase and its complex with D-lactate at pH 4.5 show an alpha-hydroxyacid oxidation mechanism

l-Lactate oxidase (LOX) belongs to a family of flavin mononucleotide (FMN)-dependent α-hydroxy acid-oxidizing enzymes. Previously, the crystal structure of LOX (pH 8.0) from Aerococcus viridans was solved, revealing that the active site residues are located around the FMN. Here, we solved the crystal structures of the same enzyme at pH 4.5 and its complex with d-lactate at pH 4.5, in an attempt to analyze the intermediate steps. In the complex structure, the d-lactate resides in the substrate-binding site, but interestingly, an active site base, His265, flips far away from the d-lactate, as compared with its conformation in the unbound state at pH 8.0. This movement probably results from the protonation of His265 during the crystallization at pH 4.5, because the same flip is observed in the structure of the unbound state at pH 4.5. Thus, the present structure appears to mimic an intermediate after His265 abstracts a proton from the substrate. The flip of His265 triggers a large structural rearrangement, creating a new hydrogen bonding network between His265-Asp174-Lys221 and, furthermore, brings molecular oxygen in between d-lactate and His265. This mimic of the ternary complex intermediate enzyme-substrate-O₂ could explain the reductive half-reaction mechanism to release pyruvate through hydride transfer. In the mechanism of the subsequent oxidative half-reaction, His265 flips back, pushing molecular oxygen into the substrate-binding site as the second substrate, and the reverse reaction takes place to produce hydrogen peroxide. During the reaction, the flip-flop action of His265 has a dual role as an active base/acid to define the major chemical steps. Our proposed reaction mechanism appears to be a common mechanistic strategy for this family of enzymes.     

The macrocyclizing protease butelase 1 remains autocatalytic and reveals the structural basis for ligase activity

Plant asparaginyl endopeptidases (AEPs) are expressed as inactive zymogens that perform maturation of seed storage protein upon cleavage‐dependent autoactivation in the low‐pH environment of storage vacuoles. The AEPs have attracted attention for their macrocyclization reactions, and have been classified as cleavage or ligation specialists. However, we have recently shown that the ability of AEPs to produce either cyclic or acyclic products can be altered by mutations to the active site region, and that several AEPs are capable of macrocyclization given favorable pH conditions. One AEP extracted from Clitoria ternatea seeds (butelase 1) is classified as a ligase rather than a protease, presenting an opportunity to test for loss of cleavage activity. Here, making recombinant butelase 1 and rescuing an Arabidopsis thaliana mutant lacking AEP, we show that butelase 1 retains cleavage functions in vitro and in vivo. The in vivo rescue was incomplete, consistent with some trade‐off for butelase 1 specialization toward macrocyclization. Its crystal structure showed an active site with only subtle differences from cleaving AEPs, suggesting the many differences in its peptide‐binding region are the source of its efficient macrocyclization. All considered, it seems that either butelase 1 has not fully specialized or a requirement for autocatalytic cleavage is an evolutionary constraint upon macrocyclizing AEPs.     

Intracellular transit of a yeast protease is rescued by trans-complementation with its prodomain.

The alkaline extracellular protease (AEP) of the yeast Yarrowia lipolytica is synthesized as a preproprotein. The precursor undergoes a complex maturation during its intracellular transit, successively involving signal peptide cleavage, dipeptidyl aminopeptidase processing, and cleavage at a dibasic site which results in the extracellular release of the active enzyme. It was previously shown that various deletions within the proregion affect the intracellular transit of the protease. Prodeleted precursors are translocated and have their signal sequences removed, but they accumulate in the secretion apparatus. We show here that the secretion of partially active proteins is restored when the prodomain is supplied in trans as an independent peptide. The secretion rescue and maturation processing that are reconstituted by the free propeptide do not reach wild type efficiency. The results of pulse-chase experiments indicate that a rate-limiting step occurs during the intracellular transit of the rescued precursors, before Kex2p proteolytic cleavage. This delayed maturation seems to be responsible for an overall slower release of the rescued polypeptides. Propeptide and AEP were secreted in equimolar amounts by both wild type and trans-complemented strains, but none could be detected in the supernatant when expressed alone. These experiments suggest that the prodomain of AEP initially acts as a crucial folding aid for the early secretory transit of the translocated precursor. They further suggest that the prodomain is also required for a second structural change of the AEP precursor during its activation.     

Identification of internal autoproteolytic cleavage sites within the prosegments of recombinant procathepsin B and procathepsin S. Contribution of a plausible unimolecular autoproteolytic event for the processing of zymogens belonging to the papain family

The steps involved in the maturation of proenzymes belonging to the papain family of cysteine proteases have been difficult to characterize. Intermolecular processing at or near the pro/mature junction, due either to the catalytic activity of active enzyme or to exogeneous proteases, has been well documented for this family of proenzymes. In addition, kinetic studies are suggestive of a slow unimolecular mechanism of autoactivation which is independent of proenzyme concentration. However, inspection of the recently determined x-ray crystal structures does not support this evidence. This is due primarily to the extensive distances between the catalytic thiolate-imidazolium ion pair and the putative site of proteolysis near the pro/mature junction required to form mature protein. Furthermore, the prosegments for this family of precursors have been shown to bind through the substrate binding clefts in a direction opposite to that expected for natural substrates. We report, using cystatin C- and N-terminal sequencing, the identification of autoproteolytic intermediates of processing in vitro for purified recombinant procathepsin B and procathepsin S. Inspection of the x-ray crystal structures reported to date indicates that these reactions occur within a segment of the proregion which binds through the substrate binding clefts of the enzymes, thus suggesting that these reactions are occurring as unimolecular processes.     

Solution structure and backbone dynamics of an endopeptidase HycI from Escherichia coli: implications for mechanism of the [NiFe] hydrogenase maturation

[NiFe] hydrogenases are metalloenzymes involved in many biological processes concerning the metabolism of hydrogen. The maturation of the large subunit of these hydrogenases requires the cleavage of a peptide at the C terminus by an endopeptidase before the final formation of the [NiFe] metallocenter. HycI is an endopeptidase of the M52 family and responsible for the C-terminal cleavage of the large subunit of hydrogenase 3 in Escherichia coli. Although extensive studies were performed, the molecular mechanism of recognition and cleavage of hydrogenase 3 remains elusive. Herein, we report the solution structure of E. coli HycI determined by high resolution nuclear magnetic resonance spectroscopy. This is the first solution structure of the apo form of endopeptidase of the M52 family reported thus far. The overall structure is similar to the crystal structure of holo-HybD in the same family. However, significant diversity was observed between the two structures. Especially, HycI shows an open conformation at the putative nickel-binding site, whereas HybD adopts a closed conformation. In addition, we performed backbone dynamic studies to probe the motional properties of the apo form of HycI. Furthermore, the metal ion titration experiments provide insightful information on the substrate recognition and cleavage processes. Taken together, our current structural, biochemical, and dynamic studies extend the knowledge of the M52 family proteins and provide novel insights into the biological function of HycI.     

Multiple proteolytic action of rat liver cathepsin B: specificities and pH-dependences of the endo- and exopeptidase activities.

Dipeptidylcarboxypeptidase, endopeptidase, and carboxypeptidase activities of rat liver cathepsin B were investigated using soluble denatured protein substrates, reduced and S-(3-trimethylammonio)propylated proteins and their derivatives. It was found that the soluble denatured proteins were degraded mainly by the dipeptidylcarboxypeptidase activity and in a few cases by the endopeptidase and carboxypeptidase activities. The eipeptidylcarboxypeptidase activity showed broad substrate specificity with broad pH optimum at 4-6. A peptide having the alpha-carboxyl group amidated with methylamine could also be a good substrate for this activity. These results suggest that this activity is dependent not upon the dissociated alpha-carboxyl group at the P2' site but upon the hydrogen-bonding abilities of the alpha-imino moiety and the protonated or amidated alpha-carboxyl moiety at P2'. On the other hand, the endopeptidase and carboxypeptidase activities were observed in a few cases, suggesting that special amino acid sequences in the substrates are responsible for these activities. These activities showed sharp pH optima at 6 and seemed to prefer basic amino acid residues at P1 site. Therefore, we suppose that cathepsin B has a carboxyl group with a pKa of about 5.5 at the S1 subsite which more effectively interacts with a positive charge at the P1 site of the substrate at pH 6 than at pH 5. Based on these results, a model of the binding subsites of this enzyme is proposed.     

Isolation and characterization of a 70-kDa metalloprotease (gelatinase) that is elevated in Rous sarcoma virus-transformed chicken embryo fibroblasts.

Chicken embryo fibroblasts (CEF) transformed by Rous sarcoma virus (RSVCEF) secrete a 70-kDa metallo-gelatinase at elevated levels over that of normal CEF. The 70-kDa enzyme has been purified from RSVCEF conditioned medium and represents 1-3% of the total protein in the RSVCEF conditioned medium. A 22-kDa protein, which appears to be the avian form of the tissue inhibitor of metalloproteases (TIMP), is co-isolated in association with the 70-kDa enzyme and can be separated from the enzyme by gel filtration carried out under denaturing conditions. The isolated 70-kDa species is in the zymogen form. It can be activated by treatment with the organomercurial, p-aminophenylmercuric acetate (APMA), yielding a 62-kDa active species derived by an apparent autoproteolytic cleavage from the 70-kDa proenzyme as determined by both substrate gel analysis and immunoblots using a monospecific antibody to the 70-kDa proenzyme. The proenzyme is poorly activated by trypsin and not activated by plasmin. The APMA-activated enzyme rapidly degrades denatured collagens but under identical conditions is unable to degrade native collagens, including basement membrane type IV collagen. Only at very high enzyme to substrate ratios (1:2) will native type IV collagen be hydrolyzed. Partial N-terminal amino acid sequencing of both the 70-kDa proenzyme and the 62-kDa active enzyme indicates that the avian enzyme is a member of the matrix metalloprotease family (MMP-2). When CEF cultures, infected with a temperature sensitive mutant of RSV, conditional for the expression of the transforming src oncogene, were incubated at the permissive and nonpermissive temperatures, differential levels of the 70-kDa enzyme were produced in direct proportion to the functioning of the src oncogene.     

Blood clotting factor IX BM Nagoya. Substitution of arginine 180 by tryptophan and its activation by alpha-chymotrypsin and rat mast cell chymase

Factor IX BM Nagoya (IX Nagoya) is a natural mutant of factor IX responsible for severe hemophilia B. A patient with this mutant is characterized by a markedly prolonged ox brain prothrombin time. IX Nagoya was purified from the patient's plasma by immunoaffinity chromatography with an anti-factor IX monoclonal antibody column. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis showed that treatment of IX Nagoya with factor XIa/Ca2+ resulted in cleavage only at the Arg145-Ala146 bond. Reversed-phase high performance liquid chromatography of a trypsin digest of IX Nagoya showed an aberrant peptide, which was further digested with proteinase Asp-N. Primary structure analysis of one of the Asp-N peptides revealed that Arg180 is replaced by Trp. An essentially complete (99%) amino acid sequence of IX Nagoya was obtained by sequencing fragments derived from a lysyl endopeptidase digest in which no other substitutions in the catalytic triad or substrate binding site were found. We also found that IX Nagoya is activated by alpha-chymotrypsin or rat mast cell chymase by monitoring the rate of factor X activation using a fluorogenic peptide substrate in the presence of factor VIII, phospholipids, and Ca2+. These results indicate that the substitution of Arg180 by Trp impairs the cleavage by factor XIa required for activation of this zymogen and that the substitution causes hemophilia BM.     

The influence of pH on the degradation of bovine myelin basic protein by bovine brain cathepsin

The degradation of bovine myelin basic protein by bovine brain cathepsin D (ED 3.4.23.5) was studied over a pH range of 2.75 - 6.0. Throughout this pH range pepstatin, an inhibitor of cathepsin D, prevented the degradation. The degradation at a pH away from the optimum of pH 3.5 was predictably slower, but also resulted in more restricted cleavage. Above pH 4.5 bovine basic protein peptide 1 - 42 was not degraded further to peptide 1 - 36 as occurs at pH 3.5. Additionally, at pH 5.5 another fragment of basic protein, peptide 1 - 91, persisted indicating that under certain basic protein as well as basic protein peptide 43 - 169 may be cleaved in the molecular region of basic protein around the phenylalanyl-phenylalanine residues at position 88 - 89. The small amount of peptides 1 - 91 and 92 - 169 detected at pH 5.5 suggests that the bond between residues 91 and 92 in intact basic protein is a minor cleavage site. The options and variation in cleavage around residues 88 - 92 of basic protein presumably result from pH-dependent changes in conformation in the is region but could also be due to changes in conformation of cathepsin D. These results indicate that local tissue changes such a pH amy affect not only the velocity of the reaction but also the nature of th product formed by the degradation of basic protein by brain cathepsin D     

Identification of the peptide bond cleaved during activation of human Clr

CNBr cleavage of unreduced proenzyme Clr yielded fragment CP2b, isolated by gel filtration and highpressure gel permeation chromatography. This fragment (˜ M_τ 55000) comprised at least 4 disulphidelinked peptides, which were separated by gel filtration after reduction and alkylation. Peptide CP2bRA4, overlapping the A- and B-chain regions in proenzyme Clr was digested by V8 staphylococcal protease, and the digest separated by reversed-phase HPLC. N-terminal sequence analysis of peptide CP2bRA4SP9 established that Clr activation involves the cleavage of a single Arg-Ile bond, located in the sequence: Gln-Arg-Gln-Arg-Ile-Ile-Gly-Gly     

Cloning and expression of a region of vesicle associated membrane protein2 (VAMP2) gene and its use as a recombinant peptide substrate for assaying clostridial neurotoxins in contaminated biologicals

An assay for the endopeptidase activities of clostridial neurotoxins in contaminated biotherapeutic products has been developed. Based on a synthetic peptide substrate representing amino acid residues 60–94 of the intracellular vesicle associated membrane protein2 (VAMP2), RT-PCR was used to amplify the VAMP2 sequence. The extended insert was digested with EcoRI and SalI and ligated into pGEX4T-1 vector for construction of the pGEX4T-1/VAMP plasmid for expressing in Escherichia coli a fusion protein linked to glutathione S-transferase (GST). The fusion protein was purified by affinity chromatography and used in an ELISA assay for comparison with the commercially available synthetic VAMP peptide and rabbit polyclonal antiserum. The identity of the immunoreactivity of recombinant VAMP2 protein with the chemically synthesized peptide was demonstrated by western blot. Our results indicated that recombinant VAMP2 peptide not only reacted with specific polyclonal antibody in a dose-dependent manner, without any remarkable difference observed between the reactivity of the fusion protein and commercial VAMP2 segment peptide, but also cleaved by botulinum neurotoxin type B (BONT/B) after endopeptidase assay. Thus, recombinant VAMP2 could serve as a replacement for VAMP2 synthetic peptide, potentially useful in endopeptidase assays for replacement of the currently used mouse bioassay for clostridial neurotoxins contaminating biotherapeutic products.     

Acyl-coenzyme A: isopenicillin N acyltransferase from Penicillium chrysogenum: effect of amino acid substitutions at Ser227, Ser230 and Ser309 on proenzyme cleavage and activity

Abstract Using a high level Escherichia coli expression system for the Penicillium chrysogenum penDE gene, we have produced acyl-coenzyme A: isopenicillin N acyltransferase (AT) enzymes containing amino acid substitutions at three conserved Ser residues. Chosen for study based on amino acid sequence homologies to other proteins, Ser²²⁷, Ser²³⁰ and Ser³⁰⁹ were changed to Cys or Ala to assess amino acid side chain involvement in proenzyme cleavage and AT enzyme mechanism. Substitutions at Ser²³⁰ had no effect on proenzyme cleavage, acyl-coenzyme A: IPN acyltransferase (IAT) or acyl-coenzyme A: 6-aminopenicillanic acid acyltransferase (AAT) activities. While Ser²²⁷→Cys had no effect, Ser²²⁷→Ala produced uncleaved proenzyme lacking both AAT and IAT activities, suggesting that the presence of a nucleophilic side chain at this residue is required for proenzyme cleavage and AT activity. Substitution of Ser³⁰⁹→Cys did not appreciably prevent proenzyme cleavage, IAT or AAT activity. Recombinant AT (recAT) proenzyme containing Ser³⁰⁹→Ala was cleaved; however, IAT and AAT activities were not observed. This separation of proenzyme cleavage from IAT and AAT activities has not been previously observed, and suggests that Ser³⁰⁹ is involved in substrate acylation.