J. Biochem. 116, 818-825 (1994) Isolation and Characterization of the Prolyl Aminopeptidase Gene (pap) from Aeromonas sobria: Comparison with the Bacillus coagulans Enzyme1 Ana Kitazono,* Atsuko Kitano,* Daisuke Tsuru,•õ and Tadashi Yoshimoto*,2 *School of Pharmaceutical Sciences , Nagasaki University, 1-14 Bunkyo-machi, Nagasaki, Nagasaki 852; and •õ Department of Applied Microbiology, Kumamoto Institute of Technology, 4-22-1 Ikeda, Kumamoto, Kumamoto 860 Received for publication, May 16, 1994 The Aeromonas sobria pap gene encoding prolyl aminopeptidase (PAP) was cloned. It consists of 425 codons and encodes a homotetrameric enzyme of 205kDa. The purified enzyme showed an almost absolute specificity for amino-terminal proline. Proline and hydroxyproline residues from many peptide and amide substrates could be easily removed, while no activity was detected for substrates having other amino terminals. The enzyme was very similar to that from Bacillus coagulans in many aspects, such as the strong inhibition caused by PCMB and the weak or no inhibition caused by DFP and chelators, respectively. However, these enzymes show only 15% identity in their amino acid sequences. Differences were also observed in their molecular weight, stability and activity toward some peptide substrates. When aligning the deduced amino acid sequence with known sequences from other microorganisms, conserved sequences were found at the amino-terminal region; the significance of these conserved regions is discussed. Based on the results of this work, and on the studies available to date, the occurrence of at least two types of PAPs is postulated. One group would be formed by the Bacillus, Neisseria, and Lactobacillus enzymes, and the other by enzymes such as the Aeromonas PAP. Key words: Aeromonas sobria, aminopeptidase, iminopeptidase, proline. Prolyl aminopeptidase (PAP, proline aminopeptidase, Pro inhibitor, 3,4-dichloroisocoumarin, although no inhibition X aminopeptidase, proline iminopeptidase) [EC 3.4.11.5] by DFP or by any other related chemical was reported (6). catalyzes the removal of amino-terminal proline from On the other hand, Atlan et al. found high similarities peptides, and has been the object of many studies since the among the amino acid sequences of the Bacillus and first report by Sarid et al. of the Escherichia coli enzyme Lactobacillus PAP enzymes and a group of hydrolases from (1, 2). Different sources of the enzyme have been described Pseudomonas putida (5). The significance of this finding has implying its wide distribution in nature. However, reports not yet been discussed, and it is possible that these enzymes concerning microorganisms and plants are the most abun bear both hydrolase and aminopeptidase activities, as has dant, and attempts to isolate the enzyme from mammalian been demonstrated for leukotriene A4 hydrolase (LTA, tissues have so far failed, suggesting the absence of a true hydrolase) [EC 3.3.2.5]. The predicted zinc binding site of PAP activity in the latter sources. this enzyme had been found to have a striking similarity to To date, there are three pap genes with known primary the corresponding primary structures of certain aminopep structures: the Bacillus coagulans (3), Neisseria gonor tidases and neutral proteases. Later work revealed that the rhoeae (4), and Lactobacillus delbrueckii subsp. bulgaricus LTA, hydrolase (which is also an epoxide hydrolase whose (5) genes. Homologous regions at the amino terminal were substrate is a fatty acid), also possesses a significant found by aligning the deduced amino acid sequences. The peptidase activity functionally related to aminopeptidase identities range from 23% for the Bacillus and Neisseria M. The studies led to the conclusion that LTA, is a Zn2+ enzymes to 36% for the Lactobacillus and Bacillus en - metallohydrolase and a bifunctional enzyme with intrinsic zymes. These enzymes share similar characteristics with aminopeptidase activity (7, 8). The enzyme was also found regard to substrate specificity, molecular weight, and to be inhibited by bestatin (9), and by captopril (10). inhibition by sulfhydryl reagents. However, the enzyme If the PAP activity happens to be also a bifunctional from L. bulgaricus was also inhibited by a serine protease enzyme, the difficulties in isolating and identifying a true PAP in mammalian tissues would become understandable. 1 This work was supported in part by Grants-in-Aid for Scientific Matsushima et al. reported the identity of the PAP they Research from the Ministry of Education, Science and Culture of had purified with leucyl aminopeptidase (11), and later the Japan. A. Kitazono is a recipient of a scholarship from the same same group characterized a proline ƒÀ-naphthylamidase ministry and expresses her gratitude. from porcine intestinal mucosa, describing it as a serine 2 To whom correspondence should be addressed . hydrolase, composed of three identical subunits each with a Abbreviations: Hyp-ƒÀNA, hydroxyproline ƒÀ-naphthylamide; LAP, leucyl aminopeptidase; PAP, prolyl aminopeptidase; Pro-ƒÀNA, molecular weight of 58,000 (12). However, this enzyme proline ƒÀ-naphthylamide; Xaa, any amino acid. was finally identified as a carbaxylesterase from studies of 818 J. Biochem. Prolyl Aminopeptidase Gene from Aeromonas sobria 819 amino acid sequence homology (13). This report has 0.1% potassium phosphate dibasic), with vigorous agita recently been revised by Heymann and Peter, who tried to tion. Chromosomal DNA was prepared following the specify which of the at least two carboxylesterase isoen method of Saito and Miura (19); plasmid pBR322 was used zymes present in pig liver was identical with proline as the vector and E. coli DH1 as the host. For selection, ƒÀ-naphthylamidase. They concluded that the hydrolysis of either ampicillin (50ƒÊg/ml) or tetracycline (25ƒÊg/ml) proline ƒÀ-naphthylamide (Pro ƒÀNA) was just a minor side was added to the medium. Individual colonies of DH1 cells activity of the amidase-type isoenzyme of pig liver carbox transformed with the pBR322 plasmids from the genomic ylesterase, and that the sequence reported by Matsushima library were picked from plates and transferred into the et al. could possibly be that of the main carboxylesterase wells of microtiter plates containing 180ƒÊl of broth. After isoenzyme of pig liver and not that of proline ƒÀ-naphthyl. overnight incubation, 50ƒÊl of the culture was transferred amidase or the amide-cleaving isoenzyme of carboxyl to another plate and mixed with 50ƒÊl of 2mM Pro-ƒÀNA. esterase (14). So, the presence of a PAP-like enzyme in The reaction was conducted overnight at 30•Ž and stopped mammalian tissues might require reconsideration in this by adding 50ƒÊl of the Fast Garnet GBC salt solution (1mg/ still speculative new context, as is the case for the PAPs as ml in 1M acetate buffer pH 4.0 containing 10% Triton a whole. X-100). Rapid red color development indicated the pres Looking for new sources of the enzyme, we screened soil ence of high enzymatic activity. microorganisms for PAP producers, isolating a strain with DNA Manipulation, and Subcloning-Plasmid DNA was high activity. On the basis of its biochemical, cultural, and isolated by the alkaline extraction procedure (20) or by morphological characteristics, this strain was identified as CsCl-ethidium bromide equilibrium density gradient cen Aeromonas sobria. Aeromonas is a genus from which many trifugation. Competent cells for transformation were other peptidases have been isolated: the well-known leucyl prepared by rubidium chloride treatment. All other proce aminopeptidase (15), and a neutral protease from A. dures were done following the standard procedures (21). The fragment containing the pap gene was subcloned based proteolytica (16, 17). Recently the cloning of the A. hydrophila prolyl endopeptidase gene has also beennn report on the restriction enzyme map constructed, and deletion ed (18). In this study, we describe the cloning, sequencing, analysis of these subclones indicated the location of the and expression of the A. sobria pap gene; and a comparison gene and its transcription direction. between this and the B. coagulans enzyme. Nucleotide Sequencing•\Sequence data were generated from single-stranded M13 or Bluescript subclones, contain ing inserts in both orientations, and deleted plasmids MATERIALS AND METHODS obtained by exonuclease III digestion (22), by the dideox Materials-Restriction enzymes, BAL 31 nuclease, T4 ynucleotide chain-termination method using Sequenase DNA ligase, other modifying enzymes, kilo-sequencing (United States Biochemical). To overcome compression, dGTP was replaced by dITP, 7-deaza-dGTP (United States deletion kits, and primers for sequencing were from Takara Biochemical), or 7-deaza-dGTP and dATP mixes (Phar Shuzo, or Toyobo, and [35S]dCTP and [32P]dCTP were macia P-L Biochemicals) in some reaction mixtures. from Amersham. Sequenase was obtained from U.S. Routinely, 5% polyacrylamide gels containing 20% v/v Biochemicals, and Agarose I was from Dojin Chemicals. formamide were used for running sequencing reaction Proline ƒÀ-naphthylamide (Pro-ƒÀNA), hydroxyproline ƒÀ-naphthylamide (Hyp-ƒÀNA), Fast Garnet GBC salt, mixtures. Sequence data were analyzed using Genetyx Mac software (Software Development) and data base lysozyme, and RNAse A were from Sigma. Alkaline search was done using the "findget" and "iden" systems phosphatase from calf intestine and Pseudomonas fragi from the National Institute of Genetics, Mishima. endoproteinase Asp-N were obtained from