β-Aminopeptidases: Insight into Enzymes without a Known Natural Substrate

β-Aminopeptidases are unique enzymes found in a diverse range of microorganisms that can utilize synthetic β-peptides as a sole carbon source. Six β-aminopeptidases have been previously characterized with preferences for different β-amino acid substrates and have demonstrated the capability to catalyze not only the degradation of synthetic β-peptides but also the synthesis of short β-peptides. Identification of other β-aminopeptidases adds to this toolbox of enzymes with differing β-amino acid substrate preferences and kinetics. These enzymes have the potential to be utilized in the sustainable manufacture of β-amino acid derivatives and β-peptides for use in biomedical and biomaterial applications. This is important, because β-amino acids and β-peptides confer increased proteolytic resistance to bioactive compounds and form novel structures as well as structures similar to α-peptides. The discovery of new enzymes will also provide insight into the biological importance of these enzymes in nature.

ABSTRACT

β-Aminopeptidases have the unique capability to hydrolyze N-terminal β-amino acids, with varied preferences for the nature of β-amino acid side chains. This unique capability makes them useful as biocatalysts for synthesis of β-peptides and to kinetically resolve β-peptides and amides for the production of enantiopure β-amino acids. To date, six β-aminopeptidases have been discovered and functionally characterized, five from Gram-negative bacteria and one from a fungus, Aspergillus. Here we report on the purification and characterization of an additional four β-aminopeptidases, one from a Gram-positive bacterium, Mycolicibacterium smegmatis (BapA_Ms), one from a yeast, Yarrowia lipolytica (BapA_Ylip), and two from Gram-negative bacteria isolated from activated sludge identified as Burkholderia spp. (BapA_BcA5 and BapA_BcC1). The genes encoding β-aminopeptidases were cloned, expressed in Escherichia coli, and purified. The β-aminopeptidases were produced as inactive preproteins that underwent self-cleavage to form active enzymes comprised of two different subunits. The subunits, designated α and β, appeared to be tightly associated, as the active enzyme was recovered after immobilized-metal affinity chromatography (IMAC) purification, even though only the α-subunit was 6-histidine tagged. The enzymes were shown to hydrolyze chromogenic substrates with the N-terminal l-configurations β-homo-Gly (βhGly) and β³-homo-Leu (β³hLeu) with high activities. These enzymes displayed higher activity with H-βhGly-p-nitroanilide (H-βhGly-pNA) than previously characterized enzymes from other microorganisms. These data indicate that the new β-aminopeptidases are fully functional, adding to the toolbox of enzymes that could be used to produce β-peptides. Overexpression studies in Pseudomonas aeruginosa also showed that the β-aminopeptidases may play a role in some cellular functions.

IMPORTANCE β-Aminopeptidases are unique enzymes found in a diverse range of microorganisms that can utilize synthetic β-peptides as a sole carbon source. Six β-aminopeptidases have been previously characterized with preferences for different β-amino acid substrates and have demonstrated the capability to catalyze not only the degradation of synthetic β-peptides but also the synthesis of short β-peptides. Identification of other β-aminopeptidases adds to this toolbox of enzymes with differing β-amino acid substrate preferences and kinetics. These enzymes have the potential to be utilized in the sustainable manufacture of β-amino acid derivatives and β-peptides for use in biomedical and biomaterial applications. This is important, because β-amino acids and β-peptides confer increased proteolytic resistance to bioactive compounds and form novel structures as well as structures similar to α-peptides. The discovery of new enzymes will also provide insight into the biological importance of these enzymes in nature.

INTRODUCTION

Naturally occurring β-amino acids are constituents of important bioactive molecules such as microcystin, paclitaxel (originally named taxol), and also precursors of β-lactam antibiotics (Fig. 1) (1–6). As peptides made solely of β-amino acids are yet to be discovered in nature (1, 6, 7), the first β-peptide with proteinogenic side chains was chemically synthesized in 1996, and since then there has been extensive research into β-peptides and their potential applications (1, 8–11). The advantages of β-peptides include resistance to proteolytic degradation and the ability to form secondary structures such as helices, sheets, and turns similar to those of α-peptides (1, 9, 12–16). Furthermore, the additional carbon in the peptide backbone provides extra length and flexibility for structural isomeric conformations and various side chain modifications (Fig. 1). These unique chemical and structural features make β-peptides attractive as potential bioactive compounds, peptidomimetics, and biomaterials (8–10, 12, 17–30). The biological functions of synthetic β-peptides include crossing cell membranes, binding nucleic acids, formation of artificial ion channels, enzyme inhibition, and mimicking receptor ligands, antimicrobial activity, and anticancer activity (14, 15, 20, 27, 30–54).

FIG 1.

Structures of α- and β-amino acids. β-Amino acids are classified further based on the position of the R group. If the R group is positioned close to the carboxylic acid, it is β²-amino acid, and if it positioned close to the amino group, it is β³-amino acid. The structure of paclitaxel (originally named taxol) is also shown with a β-amino acid moiety in its structure.

Despite the absence of naturally occurring β-peptides, enzymes that degrade chemically synthesized β-peptides have been identified (5, 55–64). Initially, mixed microbial populations from different locations were shown to degrade β-peptides (65). Since then, a small number of the β-peptide-degrading enzymes known as β-aminopeptidases have been purified and characterized and shown to hydrolyze N-terminal β-amino acids from β- or β,α-peptides (66). Database searches using BLAST and MEROPS, based on homology comparison, have indicated that genes encoding β-aminopeptidases are widespread and present in plants, archaea, fungi, and bacteria (67). The enzymes are variants of the self-processing N-terminal nucleophile (Ntn) hydrolase family, which uses a terminal cysteine or serine residue in the catalytic mechanism (57, 66–68). Unlike other members of the Ntn family, β-aminopeptidases differ in the direction and connectivity of the α-helices and β-sheets (58, 69). These enzymes are translated as inactive preproteins with or without a signal sequence and undergo posttranslational autocatalytic cleavage into active heterodimers consisting of α- and β-subunits (56, 62, 64, 68, 70). Self-cleavage occurs adjacent to a conserved nucleophilic serine residue to form the active enzyme (56, 57, 64, 70).

Six β-aminopeptidases have been functionally characterized to date: BapA from Sphingosinicella xenopeptidilytica 3-2W4 (BapA_3-2W4), BapA from Sphingosinicella microcystinivorans Y2 (BapA_Y2), BapA from Pseudomonas sp. strain MCI3434 (BapA_Ps), BapF from Pseudomonas aeruginosa PAO1, DmpA from Ochrobactrum anthropi, and DamA from Aspergillus oryzae (7, 57, 63, 64, 71). Except for DamA, which was isolated from a fungus, all other characterized β-aminopeptidases were isolated from Gram-negative bacteria. DmpA was the first enzyme identified with N-terminal β-aminopeptidase activity and is an l-aminopeptidase, d-Ala-esterase/amidase as it hydrolyzes a d-residue (d-alanine) and l-α-amino acids from peptides, amides, and esters and β-amino acids from peptides (70, 71). The enzymes BapA_3-2W4, BapA_Y2, and BapA_Ps are active only on substrates that contain β-amino acids (56, 63, 66). At the amino acid level, these enzymes share homology with DmpA, with variations in the substrate binding pocket region (68). Like all hydrolases, DmpA, BapA_3-2W4, and BapA_Y2 can catalyze amide cleavage and amide-forming reactions but have various catalytic rates and catalytic preferences for different N-terminal β-amino acid residues (62, 66, 72). With continued incubation, the enzymatically synthesized β-peptides are degraded to the constituent β-amino acids (62, 72).

The substrate specificity of β-aminopeptidases is different for each enzyme, with BapA_3-2W4 and BapA_Y2 having a preference for sterically demanding N-terminal β-amino acids such as β-homo-Ala (βhAla), βhVal, βhLeu, βhPhe, βhTyr, and βhTrp but not for βhPro and βhGlu at the N terminus (4, 7, 62). BapA_Ps and BapF prefer dipeptides with N-terminal β-homo-Gly (βhGly) (57, 63). On the other hand, DmpA from O. anthropi has a catalytic preference for N-terminal amino acid residues that are aliphatic (7).

While the physiological role of β-aminopeptidases is of interest, these enzymes may also prove valuable in biotechnology. Until now, β-peptides have been synthesized by chemical methods. However, aspects of chemical syntheses, such as the use of toxic reagents and solvents, excess waste, number of reaction steps, and lack of stereoselectivity, present significant challenges (73–76). An alternative to chemical production processes is the use of biocatalysis, whereby β-peptides can be produced with high stereoselectivity and control under benign conditions using water as the main solvent (72, 77–80). This has been demonstrated with the heterologous expression of the β-peptidases DmpA and BapA_3-2W4 in Escherichia coli and Pichia pastoris enabling production of commercially important β-peptide carnosine (β-alanine-l-histidine) at concentrations of up to 3.7 g liter⁻¹ (77). Recently, β-aminopeptidases have also been utilized for the posttranslational modification of natural proteins, i.e., insulin, with β-amino acids to impart enhanced proteolytic stability and enable labeling (81).

The characterized β-aminopeptidases from Gram-negative bacteria show preferences for different β-amino acid substrates, demonstrating natural enzyme diversity. In this study, we wanted to explore an Australian microbial community for microorganisms with β-peptidase activity and to isolate and characterize β-aminopeptidases from other types of microorganisms to provide greater insight into the occurrence and function of this curious group of enzymes. Furthermore, additional enzymes with varied substrate preferences can be utilized in future applications, such as the biocatalytic production of β-peptides or the kinetic resolution of β-peptides, amides, and esters to obtain enantiopure β-amino acids (62, 72, 78–80, 82).

Here we report the cloning and characterization of β-aminopeptidases from a Gram-positive bacterium (Mycolicibacterium smegmatis MC2 155 [basonym Mycobacterium smegmatis MC2 155]; BapA_Ms) and from a yeast (Yarrowia lipolytica; BapA_Ylip). We also report the genetic analysis, cloning, expression, and characterization of the β-aminopeptidases BapA_BcA5 and BapA_BcC1 from two Gram-negative Burkholderia species isolated in Australia from activated sludge (aerated sewage). These enzymes were named in reference to the originating microorganism and β-aminopeptidase activity (BapA) (56).

RESULTS

Identification of β-aminopeptidase genes.

β-Aminopeptidases from different types of microorganisms display preferences for various β-amino acids and occur in a wide range of organisms across different kingdoms, but only six enzymes have been characterized to date. Here we utilize genomic databases and microorganisms from the Australian biota to identify and characterize β-aminopeptidases from different types of microorganisms to add to this toolbox of unique enzymes which catalyze peptide bonds of β-amino acids.

BLAST homology searches with amino acid sequences of four of the functionally characterized β-aminopeptidases, specifically BapA_3-2W4, BapA_Y2, DmpA, and BapF, revealed β-aminopeptidase-like sequences in a diverse range of organisms (7, 66, 83). BLAST searches using amino acid sequences of known β-aminopeptidases enabled identification of putative enzymes in a Gram-positive bacterium, Mycolicibacterium smegmatis MC2 155 (gene accession no. YP_886452.1; 367 amino acids), a yeast, Yarrowia lipolytica (gene accession no. XP_504986; 370 amino acids), and an archaeon, Pyrococcus horikoshii OT3 (gene accession no, WP_010884198.1; 361 amino acids) (see Table S1 in the supplemental material). Comparative analysis of these genes and amino acid sequence alignments with previously characterized enzymes, along with confirmation by SIGNALP 4.0 (84) and TargetP v1.1 server analyses (85), indicated the absence of a signal sequence in these enzymes. These β-aminopeptidase genes were codon optimized and synthesized by GenScript for cloning, expression, and enzyme analysis.

In addition, β-aminopeptidases from two Australian Gram-negative bacterial isolates were included in this study. Enrichment studies using mixed microbial cultures grown on β-peptides as the sole carbon and energy source provided evidence that β-peptides were degraded by microorganisms (65). Based on this evidence, poly-β-peptide-degrading microorganisms were isolated from activated sludge by sequentially passaging a microbial mixture in minimal medium with poly-β-alanine as the sole carbon, nitrogen, and energy source. After enrichment and purification, two bacterial strains were isolated that formed colonies on minimal medium supplemented with poly-β-alanine after 2 days of incubation at 30°C. On nutrient agar, the isolates formed circular smooth colonies with diameters of 1 to 2 mm that were either pale yellow (designated strain A5) or white (designated strain C1). Each isolate was identified using 16S rRNA gene sequences in a BLASTn search against the NCBI Reference RNA Sequence (RefSeq) database, which indicated that strain A5 shared 99% identity with the Burkholderia cepacia strain GG4 (accession no. NC_018514.1) and Burkholderia cenocepacia strain AU1054 (accession no. NC_008061.1) nucleotide sequences, with 100% query coverage and an E value of 0.0. Strain C1 shared 99% identity with the B. cenocepacia strain AU1054 (accession no. NC_008061.1) and Burkholderi ambifaria strain AMMD (accession no. NR_074687.1) sequences, with 100% query coverage and an E value of 0.0.

Full genomic DNA sequences of B. cepacia GG4 (NCBI reference sequence, chromosome 2; NC_018514.1) and B. cenocepacia AU1054 (NCBI reference sequence, chromosome 2; NC_008061.1) were available in GenBank. Searching the genomes revealed that both strains encoded a serine peptidase belonging to the MEROPS S58 DmpA aminopeptidase family, with 93.1% identity at the nucleotide level (86). Flanking primers to amplify the complete putative β-aminopeptidase genes from the isolated strains were designed using the S58 genomic sequence region of B. cepacia GG4 and B. cenocepacia AU1054. Due to 100% sequence similarity in the forward flanking region and 55% similarity in the reverse flanking gene region for the two strains, only one set of primers was designed for the amplification of β-aminopeptidase genes from both organisms. The primers amplified PCR products of 1,170 bp in size from the two Burkholderia strains. Sequencing the two genes amplified from strains A5 and C1 showed 93.9% nucleotide identity to each other. Amino acid alignments of S58 enzyme from B. cepacia GG4 and B. cenocepacia AU1054 with BapA_BcA5 and BapA_BcC1 showed 22 amino acid differences (see Fig. S1 in the supplemental material).

A Clustal Omega amino acid sequence alignment was performed to compare the different putative β-aminopeptidases in this study with β-aminopeptidases DmpA, BapA_3-2W4, BapA_Y2, and BapF (87). The nucleophilic serine and the active site residues important for self-processing and substrate catalysis are highly conserved, as described previously by Merz et al. (68), and have been highlighted (Fig. S2). Percentage amino acid identity values between the different β-aminopeptidase sequences were determined using ExPASy LALIGN (Table S2) (88, 89). BapA_BcA5 and BapA_BcC1 shared an amino acid identity of 97.3%, and other sequences shared amino acid identities in the range of 36 to 51%, with the exception of BapA_BcA5 and BapA_BcC1, which shared 71.1% and 71.4% amino acid identity, respectively, with BapF (Table S2) (88, 89).

To gain insight into the potential biological role of these β-aminopeptidases, the neighboring gene regions were examined using sequences from the GenBank (NCBI) database (90). There was no common gene type flanking the β-aminopeptidase genes, and the function of the flanking regions did not appear to be related to peptide metabolism (Table S3).

β-Aminopeptidase expression.

The previously characterized β-aminopeptidase BapA_3-2W4 and BapA_Y2 genes were used as positive controls throughout this study. The putative β-aminopeptidase genes from M. smegmatis MC2 155, Y. lipolytica, P. horikoshii OT3, and Burkholderia sp. strains A5 and C1 were cloned into a T7-based expression vector such that the 6-histidine (His₆) tag was attached to the α-subunit of the β-aminopeptidase of the expressed protein.

The β-aminopeptidase genes in this study encode a single inactive preprotein containing around 360 to 370 amino acids. These proteins undergo self-processing to heterodimers (α- and β-subunits) to become active. These enzymes self-processed at different rates. Analysis of whole-cell lysates of recombinant E. coli strains expressing β-aminopeptidases using SDS-PAGE indicated that BapA_Ylip, BapA_BcA5, and BapA_BcC1 had undergone complete self-cleavage to the α- and β-subunits (Fig. 2). PH0078 (an uncharacterized protein from Pyrococcus horikoshii OT3 for which the crystal structure has been determined) had undergone partial self-cleavage where both the α- and β-subunits and the uncleaved preprotein were present. After separation by SDS-PAGE, only BapA_Ms was detected as the full-length preprotein; all other proteins showed clear evidence of self-processing (Fig. 2 and 3). N-terminal amino acid sequencing of the α-subunit protein bands of BapA_Ylip and BapA_Ms, and the β-subunit of BapA_Ylip, confirmed that both the α- and β-subunits were expressed. The theoretical molecular weights were calculated using ExPASy (88, 89). The average molecular weight of the uncleaved proteins was 38 kDa, and after self-processing, the α-subunit was ∼25 kDa and the β-subunit was ∼13 kDa. The molecular weight of each subunit differs slightly depending on the β-aminopeptidase. The His₆ tag and region up to the NdeI site in the pET28a(+) amino acid sequence (HHHHHHSSGLVPRGSH) attached to the α-subunit contribute 1.82 kDa to the molecular weight of the expressed β-aminopeptidases (Table S4).

FIG 2.

SDS-PAGE analysis of recombinant β-aminopeptidases. SDS-PAGE of crude cell lysates of recombinant E. coli BL21(DE3)/pLysS with expression vectors containing the different β-aminopeptidases. Lanes: 1, 10, and 15, prestained SeeBlue Plus2 standard molecular weight markers (Invitrogen, Australia); 2, 3, 16, and 17, crude cell lysate of E. coli BL21(DE3)/pLysS with the pET28a(+) vector; 4 and 5, crude cell lysate of E. coli containing the expression vector with BapA_Y2; 6 and 7, crude cell extract of E. coli containing the expression vector with Y. lipolytica β-aminopeptidase (BapA_Ylip); 8 and 9, crude cell extract of E. coli containing the expression vector with M. smegmatis strain MC2 155 β-aminopeptidase (BapA_Ms); 11 and 12, crude cell extract of E. coli containing the expression vector with Burkholderia sp. strain A5 β-aminopeptidase (BapA_BcA5); 13 and 14, crude cell extract of E. coli containing the expression vector with Burkholderia sp. strain C1 (BapA_BcC1); 18 and 19, crude cell extract of E. coli containing the expression vector with P. horikoshii OT3 (PH0078). U, uninduced; I, 4 h after induction with 2 mM IPTG. Based on molecular weights, the β-aminopeptidases have undergone self-cleavage to form the α-subunit (α) and the β-subunit (β). BapA_Ms had not completely self-processed (*). Both α- and β-subunits are coexpressed from the plasmids.

FIG 3.

Western blot analysis of recombinant E. coli BL21(DE3)/pLysS crude cell lysates to detect His₆-tagged expression of β-peptidases. For Western blot analysis, the primary antibody, penta-His monoclonal antibody (Qiagen, Germany), the secondary antibody, goat anti-mouse IgG (H+L)–HRP conjugate, and the Western Lightning Plus ECL detection reagent were used for detection according to the manufacturer’s instructions. U, uninduced sample; I, induced samples collected 4 h after IPTG induction. Lanes: 1, 10, and 15, prestained SeeBlue Plus2 standard molecular weight markers (Invitrogen, Australia); 2 and 3, crude cell lysate of E. coli BL21(DE3)/pLysS with the pET28a(+) vector; 4 and 5, crude cell lysate of E. coli containing the expression vector with pGDBapA_Y2; 6 and 7, crude cell extract of E. coli containing pGDBapA_Ylip; 8 and 9, crude cell extract of E. coli containing pGDBapA_Ms; 11 and 12, crude cell extract of E. coli containing pGDBapA_BcA5; 13 and 14, crude cell extract of E. coli containing pGDBapA_BcC1. Based on the molecular weights on the gel, the β-aminopeptidases had undergone self-cleavage to the α-subunit (α) and the β-subunit (β). The β-aminopeptidase expressing BapA_Ms had not completely self-processed (*). Both α- and β-subunits coexpress.

To confirm β-aminopeptidase expression, Western blot analysis was performed using anti-His tag antibodies. As only the α-subunit polypeptide was His tagged, only this protein subunit was detected (Fig. 3).

Functionality of recombinant β-peptidases.

To confirm that the expressed enzymes were functional and capable of hydrolyzing β-peptide bonds, a qualitative enzyme assay was developed by adapting the assay described by Geueke et al. (7). Using whole cells and the chromogenic substrate H-βhGly-p-nitroanilide (H-βhGly-pNA), the release of 4-nitroaniline (yellow color) indicated that hydrolysis of the chromogenic substrate had occurred, and this was quantified by measuring absorbance at 405 nm (7). Hydrolyzed samples exhibited various intensities of yellow in the reaction mixtures, indicating that different amounts of 4-nitroaniline were released by each expression system. The release of 4-nitroaniline was quantitatively measured, and the kinetic constants were determined (Table 1). E. coli whole cells expressing BapA_Ylip, BapA_Ms, BapA_BcA5, and BapA_BcC1 rapidly hydrolyzed the chromogenic substrate. E. coli whole cells expressing BapA_Y2 and BapA_3-2W4 hydrolyzed the substrate at a lower rate (Tables 1 and 2). When recombinant whole cells expressing PH0078 were incubated with the substrate H-βhGly-pNA or H-β³hLeu-pNA, 4-nitroaniline release was not detected (data not shown). Activity was also measured at 98°C; however, the substrate was hydrolyzed in the absence of enzyme at the higher temperature. Therefore, PH0078 was not used in subsequent studies.

TABLE 1.

Kinetic constants of β-aminopeptidase enzymes^a

Enzyme	H-βhGly-pNA			β3-hLeu-pNA
	Km (mM)	kcat (s−1)	kcat/Km (M−1 · s−1)	Km (mM)	kcat (s−1)	kcat/Km (M−1 · s−1)
BapAY2 (control)	2.4 ± 0.4	4.1 ± 0.6	1,700 ± 100	32 ± 12	367 ± 43	1.2 × 104 ± 3,000
BapA3-2W4 (control)	ND	ND	ND	1.4 ± 0.2	1.77 × 104 ± 1,200	1.3 × 107 ± 1 × 106
BapAYlip	0.0198 ± 0.005	3.8 × 104 ± 600	1.9 × 109 ± 0.3 × 109	11.6 ± 2.8	174 ± 23	1.5 × 104 ± 1,700
BapAMs	0.5 ± 0.04	6,430 ± 400	1.3 × 107 ± 3 × 105	33.0 ± 3.1	295 ± 28	8,930 ± 160
BapABcA5	0.0128 ± 0.002	659 ± 12	5.2 × 107 ± 8 × 106	1.7 ± 0.2	182 ± 8	1.1 × 105 ± 1 × 104
BapABcC1	0.024 ± 0.002	703 ± 20	3.0 × 107 ± 3 × 106	2.7 ± 0.3	229 ± 9	1.5 × 104 ± 1,700

^aThe kinetic constants (k_cat and K_m) of purified recombinantly expressed β-aminopeptidase enzymes BapA_Y2, BapA_3-2W4, BapA_Ylip, BapA_Ms, BapA_BcA5, and BapA_BcC1 were determined with different substrate concentrations of H-βhGly-pNA and H-β³hLeu-pNA. The kinetic constants were measured at 405 nm and 25°C. The kinetic parameters k_cat and K_m were determined using GraphPad Prism 6. Values represent an average of results of three replicates with the standard error of the mean shown (n = 3). ND, not determined.

TABLE 2.

Substrate specificity of β-aminopeptidase enzymes^a

Enzyme	Sp act (U · mg−1)
	l-Ala-pNA	βhGly-pNA	β3hLeu-pNA
BapAY2	<0.005	0.13 ± 0.01	0.63 ± 0.04
BapA3-2W4	<0.005	0.26 ± 0.02	10.70 ± 0.63
BapAMs	<0.005	3.26 ± 0.05	0.01 ± 0.0
BapAYlip	0.34 ± 0.02	36.71 ± 1.22	2.34 ± 0.03
BapABcA5	0.35 ± 0.05	14.42 ± 0.56	2.79 ± 0.18
BapABcC1	0.19 ± 0.01	10.69 ± 0.97	1.86 ± 0.06

^aThe substrate specificities of purified enzymes BapA_Y2, BapA_3-2W4, BapA_Ms, BapA_Ylip, BapA_BcA5, and BapA_BcC1 to H-βhGly-pNA, H-β³hLeu-pNA, and an α-amino acid (H-l-Ala-pNA) were determined at a substrate concentration of 2.5 mM. The enzyme activity was measured over the initial 5 min following enzyme addition. The spectrophotometric measurements were carried out at 405 nm. Enzyme activity was measured in triplicate with 2.5 mM substrate and 10 μg ml⁻¹ enzyme for each reaction.

Purification of recombinant β-aminopeptidase enzymes.

The recombinant β-aminopeptidases were purified using immobilized-metal affinity chromatography (IMAC) with Talon resin charged with cobalt. Despite undergoing self-cleavage in vivo, active protein comprised of α- and β-subunits was detected after purification, even though the His₆ tag was attached only to the α-subunit (Fig. 4). SDS-PAGE analysis confirmed that the purified enzymes were comprised of two subunits with molecular weights similar to those of the characterized enzymes, BapA_3-2W4 and BapA_Y2 (Fig. 4). A third unprocessed band of a higher molecular mass (∼41 to 43 kDa) was also present in all lanes, and this was equivalent to the sum of the two subunits of each enzyme, indicating unprocessed preprotein (Fig. 4).

FIG 4.

SDS-PAGE analysis of purified recombinant β-aminopeptidases. Purified recombinant β-aminopeptidases from different strains were subjected to SDS-PAGE and stained with SimplyBlue SafeStain. Lanes: 1, SeeBlue Plus2 prestained standard molecular weights; 2, BapA_3-2W4; 3, BapA_Y2; 4, BapA_Ms; 5, BapA_Ylip; 6, BapA_BcA5; 7, BapA_BcC1. Polypeptides with molecular masses of ∼27 to 28 kDa and ∼13 to 14 kDa were deemed likely to represent the α- and β-subunits.

Enzyme kinetics and substrate specificity.

Kinetic analysis using BapA_3-2W4, BapA_Y2, BapA_Ms, BapA_Ylip, BapA_BcC1, and BapA_BcA5 was completed using two chromogenic substrates, H-βhGly-pNA and H-β³hLeu-pNA, at concentrations ranging from 0.19 mM to 20 mM (Table 1). Lyophilized enzymes were used for analyses.

Kinetic analysis showed that BapA_Ylip had the highest k_cat and the lowest K_m, indicating that it had the highest activity and affinity for the substrate H-βhGly-pNA (k_cat/K_m value of 1.9 × 10⁹ M⁻¹ · s⁻¹). BapA_3-2W4 had the highest k_cat and the lowest K_m with H-β³hLeu-pNA as the substrate (k_cat/K_m of 1.27 × 10⁷ M⁻¹ · s⁻¹). The k_cat and K_m values for each enzyme are shown in Table 1. The k_cat/K_m values for BapA_Ylip, BapA_Ms, BapA_BcA5, and BapA_BcC1 with H-βhGly-pNA as the substrate were 245-, 1.65-, 6.71-, and 3.87-fold higher, respectively, than the k_cat/K_m for DmpA with the same substrate (62, 68). The calculated k_cat/K_m value was the highest for BapA_3-2W4 with H-β³hLeu-pNA as the substrate; however, BapA_Ylip, BapA_BcA5, and BapA_BcC1 also had high k_cat/K_m values with this substrate (Table 1).

BapA_Ylip had the highest specific activity of 36.7 U mg⁻¹ with the substrate 2.5 mM H-βhGly-pNA, and BapA_Ms had the lowest specific activity of 0.01 U mg⁻¹ with the substrate H-β³hLeu-pNA. BapA_BcA5 had the highest specific activity of 2.8 U mg⁻¹ with the substrate H-β³hLeu-pNA (Table 2). The specific activities of BapA_BcA5 and BapA_BcC1 were 14.4 and 10.7 U mg⁻¹ with H-βhGly-pNA. BapA_Y2 had specific activities below 1 U mg⁻¹ with both H-βhGly-pNA and H-β³hLeu-pNA (Table 2).

While the enzyme preparations did show some activity with α-amino acid substrates (Fig. 5), this activity cannot be conclusively attributed to the β-aminopeptidases, due to the possibility that protein contaminants with aminopeptidase activity may hydrolyze these substrates. There was no release of 4-nitroaniline from the chromogenic α-amino acid substrate H-d-Val-Leu-Lys-pNA by any enzyme preparation (Table 2 and Fig. 5).

FIG 5.

Hydrolysis of pNA-linked substrates by β-aminopeptidases. Absorbance of 4-nitrolaniline released from substrate hydrolysis by the different β-aminopeptidases was measured spectrophotometrically at 405 nm after 24 h of incubation with mixing. As some of the enzymes hydrolyzed some substrates very slowly, this measurement was performed to determine whether the enzymes could hydrolyze substrates over an extended period. Error bars represent the standard error of the mean (n = 3). BC_C1, BapA from Burkholderia sp. strain C1; BC_A5, BapA from Burkholderia sp. strain A5; YL, BapA from Y. lipolytica; MS, BapA from M. smegmatis MC2 155; 3-2W4, BapA from S. xenopeptidilytica strain 3-2W4; Y2, BapA from S. microcystinivorans strain Y2.

Thermostability and pH optima.

The thermostability and optimal pH for activity were determined for each of the β-aminopeptidases. The thermostability of β-aminopeptidase enzymes was determined by measuring the residual activity of each enzyme after preincubation across a range of temperatures (ranging from 25 to 70°C) in universal buffer (pH 8.0) for 30 min. These activity measurements were carried out at 25°C using equal amounts of enzyme based on the protein concentration. The positive control, BapA_3-2W4, was stable up to 70°C, which is consistent with previous findings of Geueke et al. (56). BapA_Ylip was stable up to 45°C (Table 3). BapA_BcA5 and BapA_BcC1 were stable up to 50°C. BapA_Ylip, BapA_BcA5, and BapA_BcC1 were inactivated at 55°C. BapA_Ms was stable up to 60°C and inactivated at 70°C (Table 3).

TABLE 3.

Thermostability of β-aminopeptidases^a

Enzyme	T50 (°C)
BapAYlip	51
BapAMs	66
BapABcA5	52
BapABcC1	52
BapA3-2W4	>70

^aT₅₀ is defined as the temperature at which the enzyme retains 50% of its activity after a 15-min incubation compared to incubation at 25°C (BapA_Ylip and BapA_Ms) or 30°C (BapA_BcA5 and BapA_BcC1). BapA_3-2W4 retained greater than 50% activity at the maximum temperature tested (70°C).

The optimal pH for enzyme activity was determined by incubating each enzyme in various universal buffers with a pH range of 4 to 11 for 15 min and then measuring the enzyme activity. The activity of the nondenatured enzymes was compared to that of the heat-denatured enzymes under the same conditions to ensure that any enzyme activity variations were due solely to pH effects. BapA_Ms had activity of 20 to 25 U ml⁻¹ across a pH range of pH 7 to 11, with the highest activity (25 U ml⁻¹) at pH 11 (Fig. 6). BapA_Ylip had activity ranging from 30 to 73 U ml⁻¹ over a pH range of pH 5 to 11, with the highest activity of 73 U ml⁻¹ at pH 7 (Fig. 6). BapA_BcA5 and BapA_BcC1 had optimal activities of 31 U ml⁻¹ and 45 U ml^−1, respectively, at pH 6 and with activity reduced considerably at higher or lower pH values (Fig. 6).

FIG 6.

Optimal pH for β-aminopeptidase activity. The enzymes were preincubated for 15 min in different universal buffers prepared at various pH values (4 to 11), and the enzyme activities were measured at 405 nm and 25°C. The results shown are an average of the activities determined in triplicate, and error bars represent the standard error of the mean (n = 3). Symbols: circles, BapA_Ylip; diamonds, BapA_Ms; triangles, BapA_BcA5; squares, BapA_BcC1.

Enzyme inhibition.

Further characterization was carried out by determining the effect of a range of protease inhibitors on the β-aminopeptidase activity of the enzymes. The inhibitors tested included ampicillin (5 and 10 mM), Pefabloc SC (serine protease inhibitor, 5 and 10 mM), EDTA (5 and 10 mM), CaCl₂ (1 and 10 mM), ZnCl₂ (1 and 5 mM), and α-leucine-dipeptide (5 and 10 mM). ZnCl₂ reduced the activity of all four enzymes, with BapA_BcC1 and BapA_Ms almost completely inhibited (Fig. 7). No inhibition of BapA_BcA5 activity was observed with ampicillin, Pefabloc SC, EDTA, CaCl₂, and α-leucine-dipeptide; instead, addition of all compounds (except ZnCl₂) resulted in increased activity. BapA_BcC1 was inhibited by ampicillin but not by Pefabloc SC, EDTA, or CaCl₂. BapA_BcC1 was activated by the addition of Pefabloc SC (Fig. 7). BapA_Ms was not inhibited by ampicillin or CaCl₂ but was significantly inhibited by Pefabloc SC and to a lesser extent by EDTA. BapA_Ylip was not inhibited by ampicillin, Pefabloc SC, or EDTA but was inhibited by CaCl₂, α-Leu-dipeptide, and ZnCl₂ (Fig. 7).

FIG 7.

Effects of various inhibitors and metal ions on the activity of the β-aminopeptidases. The measured activity was expressed as the percentage activity remaining after incubation for 30 min at 25°C with the potential inhibitor or metal ion. The enzyme activity was also measured in the absence of the inhibitors or metal ions, and this measured activity was taken as 100%. The mean results of three independent experiments are represented in the graph. The enzyme activity was measured at 405 nm at 25°C. Error bars represent the standard error of the mean (n = 3). BcA5, BapA from Burkholderia sp. strain A5; BcC1, BapA from Burkholderia sp. strain C1; Ms, BapA from M. smegmatis MC2 155; Ylip, BapA from Y. lipolytica.

Effect of BapA overexpression.

The biological and physiological roles of β-aminopeptidases are presently unknown. To investigate the function of β-aminopeptidases, the wild type and a bapF mutant of Pseudomonas aeruginosa MPAO1 were used for phenotypic analysis. There was no significant difference in biofilm and pyocyanin production between the wild-type MPAO1, the bapF MPAO1 mutant, and vector-only controls. Whole cells of the wild-type P. aeruginosa MPAO1 strain did not hydrolyze H-βhGly-pNA, nor did cells from the bapF MPAO1 mutant strain. Overexpression of the bapF gene from P. aeruginosa MPAO1 in the wild-type MPAO1 and complementation of the bapF MPAO1 mutant with the MPAO1 bapF gene resulted in both strains rapidly hydrolyzing H-βhGly-pNA (Fig. 8A). Expression of BapF in the wild-type strain resulted in increased biofilm formation (optical density at 600 nm [OD₆₀₀], 6.45 ± 0.317) compared to that of the vector-only control (OD₆₀₀, 4.51 ± 0.404) (P < 0.0001) (Fig. 8B). Pyocyanin production was also higher (9.32 μg/ml ± 0.20) in the wild-type strain overexpressing BapF than in the vector-only control (5.93 μg ml⁻¹ ± 1.97) (Fig. 8C).

FIG 8.

(A) Hydrolysis of chromogenic β-hGly-pNA by P. aeruginosa MPAO1 strains. The cuvettes contain supernatant after 5 min of incubation with chromogenic substrate and the specified strain as follows: 5 mM βGly-pNA plus 50 mM Tris-HCl (pH 8.0) alone (cuvette 1), wild-type (WT) MPAO1 (cuvette 2), WT MPAO1(pMMB67EH) (cuvette 3), WT MPAO1(pDLL316) (cuvette 4), WT MPAO1 (bapF [cloned into SalI site]) (pMMB67EH) (cuvette 5), bapF MPAO1 mutant (cuvette 6), bapF MPAO1 mutant (pMMB67EH) (cuvette 7), bapF MPAO1 mutant (pDLL316) (cuvette 8), and bapF MPAO1 mutant (bapF [cloned into SalI site]) (pMMB67EH) (cuvette 9). (B) Biofilm formation under dynamic conditions by P. aeruginosa strains WT MPAO1(pMMB67EH) and WT MPAO1(pDLL316) (overexpression construct; wild-type + BapF in figure), measured spectrophotometrically at OD₆₀₀ before and after crystal violet staining. Biofilm formation was measured in technical quadruplicates and biological triplicates. The difference between biofilm formation of the WT MPAO1(pMMB67EH) and that of WT MPAO1(pDLL316) was statistically significant (P < 0.0001, R² = 0.944). (C) Pyocyanin formation by P. aeruginosa strains WT MPAO1(pMMB67EH) and WT MPAO1(pDLL316) (overexpression construct; wild-type + BapF in figure), measured spectrophotometrically at OD₅₂₀ before and after crystal violet staining. Pyocyanin concentration was measured in technical quadruplicates and biological triplicates. The difference between biofilm formation of WT MPAO1(pMMB67EH) and that of WT MPAO1(pDLL316) was statistically significant (P < 0.05).

DISCUSSION

Early studies seeking to understand the origin of β-peptide-degrading ability in mixed cultures taken from soil samples, lake sediments, and wastewater treatment plants resulted in the isolation of enzymes exhibiting β-aminopeptidase activity from microorganisms belonging to the Ochrobactrum, Sphingosinicella, and Pseudomonas genera (7, 56, 57, 63, 70). Since these initial discoveries, genetic analyses have shown that sequences encoding similar enzymes are present in almost all forms of life except for animals and viruses (4, 66, 68, 83), but the biological role of the translated protein remains elusive. In this study, β-aminopeptidases from M. smegmatis MC2 155, Y. lipolytica, and two Burkholderia sp. strains (denoted A5 and C1) were characterized to provide further insights into the distribution and biological function of what currently appears to be an evolutionary oddity. To this end, microorganisms originating from diverse environments (deep sea hydrothermal vents, activated sludge, human clinical samples, food media, and oil fields) and different points on the evolutionary time scale (Archaea, Proteobacteria, and Eukaryota) were specifically chosen as enzyme sources (91–95).

Like the well-studied β-aminopeptidases DmpA, BapA_Y2, BapA_3-2W4, BapA_Ps, and BapF, the β-aminopeptidases isolated in this study were encoded by a single gene. The enzymes all shared similar properties, including becoming active as a heterodimer, hydrolysis (albeit with differing activities) of chromogenic β-peptide substrates both in vitro and in vivo, and self-cleavage at the nucleophilic serine residue (i.e., the residue at the start of the β-subunit) (Fig. 2). Characterization of the β-aminopeptidases DmpA and BapA_3-2W4 has shown that the residues Ser250, Ser288, and Glu290 (DmpA numbering) are important for self-processing and catalysis (58, 68). The important Ser-Ser-Glu residues appear to be conserved across all β-aminopeptidases, indicating self-processing and catalytic mechanisms similar to those of the DmpA family of Ntn hydrolases (7, 63, 70, 96). A curiosity was the β-aminopeptidase from the hyperthermophilic archaeon P. horikoshii OT3, which despite clear evidence of self-processing did not catalyze hydrolysis of either of the chromogenic substrates tested. This suggests that either the reaction conditions were not suitable or the enzyme has a somewhat different substrate preference that may be related to its origin or evolutionary position.

Although the enzymes were quite similar in fundamental structural features and active in both hydrolytic and synthetic directions, there were clear differences in substrate specificity, reaction rates, pH optima, and thermal resistance. The differences are most likely due to amino acid sequence percentage identities of less than 51% (see Table S2 in the supplemental material). Variation in substrate preference is often linked to an enzyme’s substrate binding region. Structural analyses of DmpA (97), BapA_3-2W4 (68), and BapA_BcA5 (98) have shown that substrate specificity of β-aminopeptidases is directly linked to the width of the substrate binding pocket and is important in determining substrate specificity (68). These pockets are formed by 15 to 20 amino acid residues that are found upstream of Glu144 (DmpA numbering), and the residues are highly variable between each β-aminopeptidase (Fig. S2) (68). The enzymes BapA_3-2W4 and BapA_Y2 have wide substrate binding pockets, whereas the DmpA pocket has a Trp residue, which partially occludes the substrate binding pocket, hence sterically limiting the size of the N-terminal amino acids that can effectively fit in the active site (68). Amino acid sequence alignments showed that a Trp residue presumably corresponding to that in DmpA is present in BapA_BcA5, BapA_BcC1, and BapF, suggesting a substrate preference profile similar to that of DmpA. BapA_BcA5 and BapA_BcC1 displayed high specific activities with the smaller amino acid substrates used in this study. There is clear diversity in the active site, which results in variation in substrate specificity, and differences in other structural components result in changes to enzyme properties (e.g., pH optima, thermal resistance).

A role in metabolism or detoxification of peptide-based metabolites is possible. This function would require substrate flexibility, which is apparent in some instances. For example, some of the β-aminopeptidases hydrolyzed substrates with N-terminal α-amino acids. BapA_Ylip, BapA_BcA5, and BapA_BcC1 all hydrolyzed H-l-Ala-pNA, showing substrate preferences similar to those of DmpA, BapA_Ps, and BapF (57, 70). We also supported the findings of Geueke et al. (7) that BapA_3-2W4 and BapA_Y2 hydrolyze H-βhGly-pNA at a low rate. This is because of the short backbone length of α-amino acids with the single carbon atom between the NH₂ and CO groups. The reduced distance between NH₂ and CO in α-peptides restricts the positioning of the substrate in the active site of the β-aminopeptidase (99).

Further evidence suggesting a specific role in each organism was the response to protease inhibitors. Other than ZnCl₂, which inhibited all the enzymes tested, the effect of protease inhibitors was variable and largely unique to each enzyme. The serine protease inhibitor Pefabloc SC actually increased the activity of BapA_BcA5, BapA_BcC1, and BapA_Ylip. This phenomenon was also observed when DamA was tested with the serine protease inhibitors TLCK (Nα-p-tosyl-l-lysine chloromethyl ketone) and leupeptin (64). As penicillin acylases and β-aminopeptidases both belong to the Ntn hydrolase superfamily and have structural and catalytic similarities but not high sequence similarity, the effect of ampicillin on BapA_3-2W4 was tested (99, 100). BapA_3-2W4 was inhibited at high ampicillin concentrations, and biochemical and structural analyses showed that ampicillin binds but is not converted by the enzyme (99). BapA_BcC1 and BapA_Ms were partially inhibited by ampicillin, and the activities of the other enzymes were not inhibited by ampicillin. One hypothesis is that β-aminopeptidases have an evolutionary link to penicillin acylases, again suggesting a metabolic role (99). The almost complete inhibition by zinc cations of all enzymes fits with the enzyme structure. Zinc ions act as inhibitors by interacting with the side chains of the residues Asp, Glu, Cys, and His (101). In β-aminopeptidases, the residues Glu144 and Asp290 (DmpA numbering) are conserved across all enzymes (Fig. 2) and play a crucial role in enzyme self-processing and substrate catalysis (68).

Characterization of heterologously expressed β-aminopeptidases provided limited information about the cellular role of these proteins because of the absence of a clearly identifiable natural substrate. The absence of signal sequences indicated that the enzymes are located in the cytoplasm and are active on compounds that are either produced by cells or transported into cells. The genes flanking the β-aminopeptidase-encoding region were analyzed using sequence data from the GenBank (NCBI) database (90). The flanking genes did not appear to have any function that may relate to a potential role of β-aminopeptidases. Phenotypic analysis of a bapF mutant of P. aeruginosa MPAO1 was used to provide more information on the role of BapF in this microorganism. Although phenotypic differences between the wild-type MPAO1 and the bapF MPAO1 mutant were not detected, overexpression of BapF in the two strains resulted in notable phenotypic differences. These included increases in biofilm formation and pyocyanin production. The PseudoCAP database has manually assigned BapF association with amino acid biosynthesis and metabolism (102). Enzymes such as BapF may have a role in β-amino acid metabolism in which the resulting products contribute to, or are components of, biofilm. Polyamines, products of β-alanine metabolism, have been shown to be important for biofilm formation in some organisms (103). Similarly, β-alanine is a key metabolite in chorismate synthesis. Chorismate is a key metabolite required for production of pyocyanin, which is an important virulence factor of P. aeruginosa PAO1 (104). The results presented here suggest that the β-aminopeptidase produced by P. aeruginosa MPAO1 may play a metabolic role that supports biofilm formation and/or pyocyanin production.

In this study, we identified β-aminopeptidases by screening microorganisms from the Australian biota and scanning genomic DNA databases to identify β-aminopeptidases. The genes were from a diverse range of microorganisms, including a Gram-positive bacterium and a yeast, and encode functional enzymes (BapA_Ms, BapA_Ylip, BapA_BcA5, and BapA_BcC1) which have features typical of the DmpA family of Ntn hydrolases. The enzymes had β-aminopeptidase activity, demonstrating that this activity is widespread in different types of microorganisms despite the absence of a naturally occurring substrate. Despite belonging to same family, the enzymes showed clear differences in biochemical properties, substrate preferences, and enzyme kinetics. We propose that this group of enzymes has an important metabolic role potentially associated with amino acid metabolism involving hydrolysis of β-amino acids incorporated into the structure of natural compounds or with amino acid recycling, which may be unique to each different organism and its typical environment.

While the biological role of these enzymes is of interest, application in biotechnology may also be of significant value, as laboratory-scale testing has shown that β-aminopeptidases hydrolyze racemic mixtures of β-amino acid amides to enantiopure β-amino acids and can be used in the synthesis of β-peptides (72, 77–80, 82). Enzyme diversity and robustness will be important in realizing these applications. This study has provided new insights into the cellular role of β-aminopeptidases, with further research into this enzyme group required to elucidate their natural role and to enable their use in biotechnology.

MATERIALS AND METHODS

Materials.

Fmoc (9-fluorenylmethoxy arbonyl)-protected β-alanine was obtained from Sigma-Aldrich, Australia. H-βhGly-pNA and H-l-β³hLeu-pNA were obtained from Mimotopes Pty Ltd., Australia. The β-peptides were classified as described by Seebach et al. (1), and β-Ala is referred to as H-β-hGly when following this nomenclature. The chromogenic α-amino acid substrates H-l-Ala-pNA, H-l-Leu-pNA, H-l-Arg-pNA, and H-d-Val-Leu-Lys-pNA were obtained from Sigma-Aldrich, Australia.

β-Peptide synthesis.

β-Alanine peptides were synthesized via standard solid-phase peptide synthesis using Fmoc–β-alanine according to the general method outlined by Guichard et al. (74) and Guichard and Seebach (105). The product was purified by reverse-phase high-pressure liquid chromatography to >95% purity.

Medium and growth conditions.

E. coli strains were grown in 2YT medium (16 g liter⁻¹ tryptone, 10 g liter⁻¹ yeast extract, and 5 g liter⁻¹ NaCl with 1.5% [wt/vol] agar for solid medium) supplemented with the appropriate antibiotics at 37°C. Liquid cultures were shaken at 200 rpm.

The strains isolated from activated sludge were grown aerobically at 30°C with shaking at 200 rpm. Activated sludge from an aeration tank from the Lilydale wastewater treatment plant, Victoria, Australia, was used as the source of microorganisms. One hundred microliters of activated sludge was sequentially passaged, eight times every 7 days, in 461S (Straffon) minimal medium (106) with 0.5 g liter⁻¹ of poly-β-alanine as the sole carbon, nitrogen, and energy source and enriched for poly-β-peptide-degrading microorganisms. The liquid cultures were then streaked onto solid 461S minimal medium containing 0.5 g liter⁻¹ poly-β-alanine as the sole carbon, nitrogen, and energy source, and individual colonies were then streak plated until pure cultures were obtained.

DNA manipulation and analysis.

The β-aminopeptidase genes from Y. lipolytica, M. smegmatis MC2 155, and P. horikoshii OT3 were selected from BLAST searches using BapA_3-2W4, BapA_Y2, and DmpA as queries. The β-aminopeptidase genes from Y. lipolytica, M. smegmatis MC2 155, P. horikoshii OT3, S. xenopeptidilytica 3-2W4, and S. microcystinivorans Y2 were codon optimized for E. coli, synthesized, and cloned into pUC57-Kan by GenScript (Piscataway, NJ). The pUC57-Kan vectors harboring the β-aminopeptidase genes were digested with NdeI and BamHI, and the genes were cloned into the T7-based expression vector pET28a(+) (Novagen) so that the resultant proteins contained an N-terminal 6-histidine tag (His₆ tag). A start codon was introduced into the 5′ N terminus of the gene. A stop codon was introduced after the gene to prevent transcription of the C-terminal His tag.

The two strains isolated from activated sludge and purified from enrichment cultures were designated A5 and C1. PCR amplification was carried out using primer pair bak4/bak11w and the 16S rRNA universal1/universal2 primers to amplify 800- and 1,000-bp products, respectively (107). The amplified products were then purified using a PCR purification kit (Qiagen, Germany) and sequenced using the same primers used for amplification. Based on RefSeq and BLASTn analysis of data from GenBank, National Center for Biotechnology Information (NCBI), using the 16S rRNA sequencing data, the two strains with the highest E value matches (E value of 0.0) and complete sequence matches were selected as references for strains A5 and C1 (108); these were Burkholderia species.

To enable isolation of the β-aminopeptidase genes from these new strains, primers were designed using the available genomic DNA sequences of strains identified by BLASTn in NCBI that matched the sequences of the isolated strains best. Due to high sequence similarity in the flanking region of the putative β-aminopeptidase gene from the two isolated strains, only one set of primers was designed. The forward primer was 5′-GAATTGCCATATGCGCACGAGGGATCTCGG-3′, and the reverse primer was 5′-CGCTTGTCGACTCAGGCGCCGCGCC-3′, with incorporated NdeI and SalI sites (in bold) for subsequent construction of T7-based expression vectors containing the cloned β-aminopeptidase genes. A 1,170-bp PCR fragment was amplified by PCR using Phusion high-fidelity DNA polymerase (Thermo Fisher Scientific, Australia), per the manufacturer’s protocol, from both Burkholderia sp. isolates using single colonies of each. The PCR product was cloned into the pCR-BluntII-TOPO Zero Blunt vector (Life Technologies, USA), per the manufacturer’s instructions. The TOPO vector was transformed into RbCl₂-competent DH5α cells (109). The transformation mixture was then incubated in 2YT broth supplemented with 50 μg ml⁻¹ kanamycin, with shaking at 37°C for 16 h. Plasmids were extracted using the QIAprep spin miniprep kit (Qiagen, Germany). The PCR products were digested from the TOPO vector and then ligated into the pET28a(+) vector.

All β-aminopeptidase expression vectors containing β-aminopeptidase genes were transformed into chemically competent E. coli DH5α cells. The inserted genes were then sequenced to confirm the correct gene insertion (Micromon, Monash University, Australia), and the resultant plasmids were designated pGDBapA_Ylip, pGDBapA_Ms, pGDBapA_BcA5, pGDBapA_BcC1, and pGDBapA_PH0078 for plasmids harboring β-aminopeptidases from Y. lipolytica, M. smegmatis, Burkholderia sp. A5 and C1, and P. horikoshii OT3.

Expression of β-aminopeptidases.

The T7-based expression vectors harboring the different β-aminopeptidase genes were transformed into RbCl₂-competent E. coli BL21(DE3)/pLysS for expression of the encoded proteins. The vector only was also transformed as a control. The 100-ml seed cultures, containing 2YT medium supplemented with 50 μg ml⁻¹ kanamycin and 34 μg ml⁻¹ chloramphenicol, were inoculated with the recombinant E. coli cells harboring the different expression vectors and grown overnight at 30°C with shaking at 200 rpm. The seed cultures were used to inoculate 500 ml of 2YT broth, supplemented with 50 μg ml⁻¹ kanamycin and 34 μg ml⁻¹ chloramphenicol to an OD₆₀₀ of 0.1. The cultures were grown at 30°C with shaking, until an OD₆₀₀ of 0.5 to 0.6 was reached. β-Aminopeptidase gene expression was induced with a final concentration of 2 mM isopropyl-thio-β-galactoside (IPTG) (Sigma-Aldrich, Australia), and the culture was incubated for a further 4 h at 30°C with shaking at 200 rpm. The cells were then harvested by centrifugation at 8,000 × g for 20 min at 4°C. The cell pellet was washed by resuspension in 50 mM Tris-HCl (pH 8.0) and centrifuged at 8,000 × g for 5 min, and the washing was repeated. The whole cells were then resuspended in 50 mM Tris-HCl (pH 8.0) and used in an enzyme assay to determine function, or the washed cell pellet was frozen at −80°C and used for protein purification.

Western blot analysis of crude cell lysates expressing the β-aminopeptidases was performed to confirm expression of His₆-tagged enzymes. After SDS-PAGE analysis, proteins were transferred to a nitrocellulose membrane (Whatman, Protran; pore size, 0.45 μm). The membrane was washed with 1× Tris-buffered saline (TBS) and 0.1% (vol/vol) Tween 20 buffer twice for 10 min. The membranes were then incubated with penta-His monoclonal antibody (1:2,000) (Qiagen, Germany) for 1 h, followed by goat anti-mouse IgG (H+L)–horseradish peroxidase (HRP) conjugate (Millipore) (1:3,000). The membrane was visualized using Western Lightning Plus ECL enhanced chemiluminescence substrate (Perkin Elmer, USA), according to the manufacturer’s instructions.

N-terminal amino acid sequencing.

The protein bands were cut out from the SDS-PAGE gel and washed with dithiothreitol (DTT)-iodoacetamide and digested with trypsin overnight in 50 μl of 20 mM sodium bicarbonate buffer. The digests were analyzed by liquid chromatography-tandem mass spectrometry (LC-MS/MS) at a gradient of 95% buffer A (0.1% formic acid) to 70% buffer B (80% acetonitrile–0.1% formic acid), at a flow rate of 300 nl/min over 30 min using an UltiMate 3000 nano high-performance liquid chromatograph (HPLC) (ThermoFisher, Scientific, USA) and an HCT Ultra 3D ion trap mass spectrometer (Bruker Daltonics, USA). The LC-MS/MS data were then searched against an in-house curated database using the MASCOT search engine by the Monash Biomedical Protein Facility for N-terminal amino acid sequence identification.

Confirmation of enzyme functionality using whole cells.

Recombinant whole cells before and 4 h after induction were harvested, washed by resuspension, and diluted using 50 mM Tris-HCl, pH 8.0, to an OD₆₀₀ of 0.5. One milliliter of the whole cells was then incubated at 25°C with H-βhGly-pNA at a final concentration of 5 mM until the appearance of a yellow color was observed due to hydrolysis of the chromogenic substrate, indicating that the enzyme was active. The 1-ml sample was centrifuged for 5 min at 13,000 × g, and the absorbance at 405 nm was measured using a UV-visible spectrophotometer (Shimadzu, Japan).

Purification of β-aminopeptidases.

Three to four grams of the cell pellet harboring different β-aminopeptidases was weighed to ensure approximately equivalent protein concentrations and resuspended in buffer A (50 mM Tris-HCl, 150 mM NaCl, pH 8.0). The cells were disrupted with an EmulsiFlex-C5 homogenizer (Avestin, Canada). The lysate was centrifuged at 9,000 × g for 20 min at 4˚C. Imidazole was added to the clarified supernatant (soluble fraction) to a final concentration of 5 mM. The soluble fraction was gently mixed with 1 ml of Talon Superflow metal affinity resin charged with cobalt (Clontech) at 4°C for 1 h. The resin-lysate mixture was transferred to a 20-ml column and washed four times with 20 ml of buffer A supplemented with 5 mM imidazole. Bound proteins were then eluted sequentially in buffer A (5 ml at each step) containing 20 mM, 60 mM, 100 mM, and 200 mM imidazole. Fractions of 1 ml were collected. A 10-μl sample was then taken from each fraction and analyzed by SDS-PAGE using 10% Bis-Tris gels, which were stained with SimplyBlue SafeStain (Life Technologies, USA). The purest fractions were pooled, dialyzed against 0.5 mM Tris-HCl (pH 8.0) buffer, lyophilized, and stored at −80°C. The protein concentration was determined using Pierce BCA protein assay (Thermo Fisher Scientific, Australia).

Measurement of β-aminopeptidase activity.

Enzyme activity was determined by measuring the hydrolysis of 5 mM βhGly-pNA in 50 mM Tris-HCl, pH 8.0, and the formation of 4-nitroaniline (ε = 8,800 M⁻¹ · cm⁻¹) spectrophotometrically at OD₄₀₅ and 25°C, a method adapted for this study from a report by Geueke et al. (7).

The kinetics of the β-aminopeptidase enzymes with the substrates H-βhGly-pNA and H-β³hLeu-pNA (ε = 8,800 M⁻¹ · cm⁻¹) was determined using a concentration range from 0.19 mM to 20 mM. The reaction requiring H-β³hLeu-pNA contained 10% (vol/vol) dimethyl sulfoxide, 50 mM Tris-HCl (pH 8.0), and the enzyme. The V_max and K_m were determined using GraphPad Prism 6 (GraphPad Software, Inc., San Diego, CA, USA). The calculations were based on nonlinear regression of a Michaelis-Menten model. The k_cat value was calculated based on the theoretical molecular mass and assuming that all protein in the sample was active. The enzyme activity was also measured using the chromogenic α-amino acid substrates H-l-Ala-pNA, H-l-Leu-pNA, H-l-Arg-pNA, and H-d-Val-Leu-Lys-pNA. The absorbance measurements were subtracted from substrate-only absorbance values for all samples. The α-amino acid reaction mixtures contained 2.5 mM chromogenic substrate in 50 mM Tris-HCl (pH 8.0), 20% (vol/vol) dimethyl sulfoxide, and approximately 10 μg ml⁻¹ of enzyme. Activity was determined over 5 min, and if the activity was low (less than 0.5 U), the enzymes were left to incubate with shaking at 25°C for 24 h. After 24 h, the absorbance was determined at 405 nm spectrophotometrically. One unit is the amount of enzyme required to form 1 μM 4-nitroaniline per min, and measurements were carried out in triplicate.

Determination of pH optima and thermostability of enzymes.

The optimal pH for β-aminopeptidase activity was determined by incubating the enzymes in the presence of a universal buffer (pH range 4 to 11) (110). Nondenatured and heat-denatured (incubated at 99°C for 10 min) enzymes were preincubated in the respective universal buffer for 15 min at 25°C in triplicate. Heat-denatured enzymes were used as a control to ensure that variations in enzyme activity were due solely to pH change.

The thermostability of the β-aminopeptidases was determined by incubating the enzymes at a range of temperatures (4 to 70°C) for 30 min and then equilibrating them to room temperature before measuring the enzyme activity. Each assay was repeated in triplicate at each temperature.

Enzyme inhibition.

The inhibitors tested were ampicillin (Sigma), Pefabloc SC (Fluka), EDTA, α-leucine dipeptide (Sigma), and the salts CaCl₂ and ZnCl₂. The purified enzymes (1 μg ml⁻¹) were incubated with the inhibitors at 25°C for 30 min in Tris-HCl buffer (pH 8.0), after which enzyme activity was measured spectrophotometrically at 405 nm using 5 mM H-βhGly-pNA as the substrate.

P. aeruginosa MPAO1 complementation and overexpression constructs.

The wild-type P. aeruginosa strain MPAO1 and the isogenic bapF MPAO1 mutant strain PW3678, genotype PA1486-G03::ISlacZ/hah, were purchased from the University of Washington, USA (111). The P. aeruginosa-E. coli shuttle vector pMMB67EH was used in the construction of a recombinant plasmid containing the MPAO1 bapF gene under the control of the P_tac promoter.

To construct this plasmid, the bapF gene was amplified wild-type MPAO1 using the bapF forward primer 5′CGCGAGAATTCCTACCGGATGCC-3′ and the bapF reverse primer 5′TGAGAAGCTTTCAGCGTCCCGGC-3′ and cloned into the HindIII restriction site of pMMB67EH. PCR amplification was used to confirm that the gene was inserted in the correct orientation, and the resulting recombinant plasmid was designated pDLL316. The empty vector pMMB67EH or pDLL316 was transferred into the wild-type and bapF mutant strains by RP4-mediated conjugation from an S17-1 E. coli donor strain (112). The wild type and the bapF mutant harboring pMMB67EH served as controls for subsequent experiments. To select for transconjugants, the bacteria were grown on LB agar containing 10 μg ml⁻¹ tetracycline, 150 μg ml⁻¹ carbenicillin, 30 μg ml⁻¹ kanamycin, and 0.1 mM IPTG for 48 h at 30°C.

Biofilm formation.

Biofilm formation was determined using overnight cultures grown in LB medium. The strains were subcultured and grown to an OD₆₀₀ of 0.5 and induced with 1 mM IPTG. A 0.2-ml volume of the culture was then aliquoted into a 96-well tray and incubated at 30°C with shaking at 100 rpm for 12 h. Biofilm formation was determined by measuring the OD₆₀₀ before and after 0.1% crystal violet staining, and measurements were carried out as biological quadruplicates with technical triplicates. Poststaining, the crystal violet was resolubilized from the cells in 96% ethanol and transferred to a new microtiter plate before measurement.

Pyocyanin extraction.

P. aeruginosa strains were subcultured in 125-ml conical flasks containing 20 ml of King A medium and induced with 1 mM IPTG at an OD₆₀₀ of 0.55, after which they were grown overnight at 37°C with shaking (113). Eight milliliters of bacterial cells was then pelleted by centrifugation at 10,000 rpm for 10 min. The supernatant was removed and filtered with a 0.22-μm filter. Then, 0.5 volume of chloroform was added, followed by vortexing and centrifugation. The bottom layer was transferred to a fresh tube, and 0.2 M HCl was added, which gave a pink to deep pink color because of the presence of pyocyanin, and the mixture was vortexed and then centrifuged for 2 min. The acidified culture supernatant was transferred to a 96-well tray, and the absorbance was measured at 520 nm and the pyocyanin concentration was determined according to the following equation: concentration of pyocyanin (μg ml⁻¹) = OD₅₂₀ × 17.072 (ε) (114).

Accession number(s).

All nucleotide sequences are available in GenBank, NCBI (https://www.ncbi.nlm.nih.gov/GenBank/index.html). Nucleotide sequences for the β-aminopeptidases in this study are available in GenBank under accession numbers YP_886452.1 for BapA_Ms, XP_504986.1 for BapA_Ylip, and WP_010884198.1 for PH0078. The 16S rRNA gene sequence used to classify Burkholderia sp. A5 is under accession numbers WP_014899455.1, NC_018514.1, and NC_008061.1. The 16S rRNA sequence to classify Burkholderia sp. C1 is available under accession numbers NC_008061.1 and NR_074687.1. The accession numbers of full genomic DNA sequences used to design primers to amplify the genes encoding BapA_BcA5 and BapA_BcC1 are available in GenBank. The primers designed from B. cepacia GG4 were from the NCBI reference sequence for chromosome 2, accession number NC_018514.1, and those designed from B. cenocepacia AU1054 were from the NCBI reference sequence for chromosome 2, accession number NC_008061.1.