Putative lantibiotic-like gene cluster of P. Elgii B69
B69 was subjected to whole-genome shotgun sequencing, yielding 7.9 Mb of sequence on 278 assembled contigs [23
]. Data mining for the LanC homolog amidst the genomic data of P. elgii
B69, using the SpaC sequence of P. polymyxa
E681 as a driver, resulted in the identification of a lantibiotic-like gene cluster containing five probable open reading frames (ORFs), designated elgT1
, and elgA
(Figure ). All genes, except elgT1
, were transcribed in the same orientation. The amino acid sequences of the four products of the elg
) showed high levels of identity (31%-38%) with those of homologous proteins from several type AI lantibiotic gene clusters (Table ).
Figure 1 Elg gene cluster, ElgA amino acid sequence and sequence alignment with type AI prelantibiotics. A, The biosynthetic gene cluster of P. elgii B69 consists of five ORFs, elgT1, elgC, elgT2, elgB, and elgA. The number of amino acids encoded by each gene (more ...)
Deduced peptides and proteins derived from the elg gene cluster
ElgT1 (596 amino acids (a.a.)) and ElgT2 (625 a.a.) showed high-level identity with numerous adenosine-5'-triphosphate (ATP)-binding cassette (ABC) transporter proteins. ElgT1 shared 31% identity with SpaT, a protein responsible for the transportation of the ericins A and S of B. subtilis
A1/3 [GenBank: AAL15565
], and 31% identity with EtnT, which is responsible for the export of the entianin of B. subtilis
]. Similarly, ElgT2 showed strong homology (38% identity) with the subtilin-transport protein of B. subtilis
ATCC 6633 [GenBank: P33116
], and was homologous to NisT of Lactococcus lactis
N8 [GenBank: CAA79469
] and NsuT of Streptococcus uberis
42 [GenBank: ABA00880
] (34% identity in both cases). These proteins are responsible for the transportation of nisin Z and nisin U, respectively [26
]. The two proteins ElgT1 and ElgT2 also shared 28% sequence identity with each other, suggesting that they have similar functions in the processes of transportation and secretion of elgicins. ElgT1 and ElgT2 may serve as a two-component ABC transporter, similar to MibTU and CinTU, which are probably involved in the export of microbisporicin and cinnamycin [28
]; however this function is uncommon in the maturation of lantibiotics.
encodes a protein containing 454 amino acids, which shows strong homology to the lantibiotic cyclase, MibC, of Microbispora corallina
NRRL 30420 (33% identity) [GenBank: ADK32556
]. MibC is involved in the formation of (Me)Lan bridges in microbisporicin [28
]. The amino acid sequences encoded by the lanC
genes have some conserved structural motifs, including GXAHG, WCXG, and CHG, in which the cysteine and histidine residues are highly conserved [30
]. The alignment of ElgC with several type AI lantibiotic synthetases showed that ElgC contains several conserved sequences, such as GVSHG (positions 244-248), WCYG (positions 316-319), and CHG (positions 366-368), wherein His247, Cys317, Cys366, and His367 are strictly conserved. These observations indicate that ElgC is a lantibiotic synthetase that catalyzes the synthesis of Lan and MeLan residues.
A large ORF upstream and overlapping elgT2
by 4 bp encodes a protein of 1037 amino acids. The putative protein ElgB shares 31% identity with MibB of M. corallina
NRRL 30420 [GenBank: ADK32555
] and 30% identity with SpaB of B. subtilis
ATCC 6633 [GenBank: P39774
]. The proteins MibB and SpaB are responsible for the dehydration of serine and threonine residues in the propeptide to form the unsaturated amino acids of microbisporicin and subtilin, respectively [28
]. Thus, ElgB appears to be a dehydratase involved in the process of maturation.
encodes the prepeptide of the elgicins, with a length of 64 amino acids. No lantibiotics reported thus far share homology with ElgA, suggesting that the mature proteins derived from ElgA are novel lantibiotics. The alignment of the putative leader peptide of ElgA with those of other lantibiotics revealed the existence of a possibly conserved motif "FDLD" (Figure ), which resembles the "FDLN" motif in the leader peptide of type AI lantibiotics [32
]. Considering that the elg
gene cluster contains the lanB
genes encoding the modified enzymes, it could be concluded that the elgicins are type AI lantibiotics.
gene cluster lacks the immunity genes lanI
. LanEFG acts as an ABC transporter for lantibiotic immunity; for example, NisEFG expels lantibiotic molecules that have entered the cytoplasmic membrane into the extracellular environment [33
]. Considering the mechanism of LanEFG-imparted immunity, ElgT1T2 is likely to play a role in self-protection, in addition to that of secretion and transportation of the elgicins.
The leader peptides of type AI lantibiotics are usually processed by a serine protease encoded by lanP
, which is not found in the elg
gene cluster. The leader peptide of ElgA may instead be processed by an intrinsic B69 serine protease. This elgicin maturation process might therefore resemble that of subtilin, wherein the leader peptide of presubtilin is removed by an intrinsic B. subtilis
subtilisin-like protease [13
Isolation and purification of elgicins
Genomic analysis of P. elgii B69 revealed the presence of a new lantibiotic-like gene cluster. To express this elg gene cluster, P. elgii B69 was grown aerobically at 30°C for 120 h in different fermentation media designed for the production of active substances. At harvest, extractions of B69 fermentation broths were achieved using column chromatographic fractionation on AB-8 macroporous resin (Haiguang Chemical Ltd., Tianjin, China). The KL medium fraction (5 g/L glucose, 4 g/L (NH4)2SO4, 2.6 g/L K2HPO4, 4 g/L MgSO4, 2 g/L NaCl, 2 g/L CaCl2, 2 mg/L FeSO4·7H2O, 2 mg/L ZnSO4·7H2O, and 1.5 mg/L MnSO4·H2O, pH 7.2) eluted by 80% methanol showed activity against the indicator strain P. ehimensis. This fraction was then applied to the solid-phase extraction (SPE) column. The fraction with activity against the indicator strain was eluted with 50% methanol and further separated by analytical reverse-phase high-performance liquid chromatography (RP-HPLC). Aided by the presence of several tyrosine residues within the precursor peptide ElgA, its ultraviolet (UV) absorption was measured at 280 nm during analytical HPLC. The fractions corresponding to the retention time of 21.290-22.036 min were isolated, and they showed activity against P. ehimensis.
Large-scale fermentation of P. elgii B69 was carried out in KL medium for the production of active substances. The target compounds were then isolated by a simple three-step purification procedure consisting of AB-8 resin fractionation, SPE, and preparative RP-HPLC, as described in the "Methods" section. In the preparative RP-HPLC profile, the three peaks corresponding to retention times of 34.21, 35.43, and 36.53 min (Figure ) were pooled and designated elgicin A, B, and C, respectively, of which elgicin B was the major component. These fractions were lyophilized and subjected to electrospray ionization-mass spectrometry (ESI-MS) for molecular analyses.
Figure 2 Reverse-phase HPLC profile of crude SPE-extraction. UV absorption was measured at 280 nm. MV, millivolt. Peak 1, with retention time of 34.21 min, corresponds to elgicins AI and AII. Peaks 2 and 3, with retention times of 35.43 and 36.53 min, correspond (more ...)
ESI-MS analyses of elgicins
To determine the molecular masses of elgicins, the lyophilized elgicins A, B, and C were dissolved in sterile water and subjected to ESI-MS. The MS spectrum of HPLC-purified elgicin A revealed four signals at the mass-to-charge ratios (m/z) 1135.07 [M + 4H]4+, 1512.89 [M + 3H]3+, 1149.31 [M + 4H]4+, and 1532.58 [M + 3H]3+ (Figure ). The molecular weight calculated from the two former signals was 4536 Da, and the others corresponded to a molecular weight of 4593 Da. These findings suggest that Peak 1 contained two compounds, designated elgicin AI and elgicin AII. The molecular weight of elgicin AII was 57 Da larger than that of elgicin AI; this difference corresponds to the molecular weight of a single glycine residue. In the case of Peak 2, the mass spectrum showed the presence of two strong signals at m/z values of 1177.72 [M + 4H]4+ and 1569.89 [M + 3H]3+, corresponding to a molecular mass of 4706 Da (Figure ). The molecular weight of elgicin B was 113 Da larger than that of AII; this difference corresponds to the molecular mass of a single leucine residue, as deduced from the prepeptide of ElgA that lacks an isoleucine residue (Figure ). Compound elgicin C, with a retention time of 36.53 min, had a molecular mass of 4820 Da, consistent with the two signals at m/z 1206.14 [M + 4H]4+ and 1608.30 [M + 3H]3+ (Figure ). The molecular mass of elgicin C was 114 Da larger than that of elgicin B; this difference is consistent with the molecular mass of a single asparagine residue.
Figure 3 ESI-MS of RP-HPLC-purified elgicins AI, AII, B, and C isolated from fermentation medium. A, Peaks at 1512.89 [M + 3H]3+ and 1135.07 [M + 4H]4+ correspond to a mass of 4536 Da for elgicin AI. Peaks at 1532.58 [M + 3H]3+ and 1149.31 [M + 4H]4+ correspond (more ...)
Lantibiotics have small molecular weights (< 5 kDa) that usually range from 1700-4000 Da. Thus far, the molecular weights of only two lantibiotics, cytolysin LL
(isolated from the Enterococcus faecalis
strain FA2-2) and carnocin U149 (produced by Carnobacterium piscicola
U149), exceed 4 kDa (4164 and 4635 Da, respectively) [10
]. Our newly isolated four-component elgicins therefore have unusually large molecular weights of 4536 Da (elgicin AI), 4593 Da (elgicin AII), 4706 Da (elgicin B), and 4820 Da (elgicin C). To the best of our knowledge, no other lantibiotics have molecular weights greater than those of elgicins B and C.
Analysis of N-terminal amino acid sequence
To confirm whether the four-component antibacterial agents are derived from ElgA, HPLC-purified elgicin B was subjected to automated Edman degradation to determine its N-terminal amino acid sequence (Figure ). The first four amino acids were Leu-Gly-Asp-Tyr. The fifth residue was blocked completely, suggesting the presence of a dehydrated amino acid residue, a characteristic feature of lantibiotics. The Leu-Gly-Asp-Tyr sequence was consistent with the sequence of the propeptide that resulted from the removal of the leader peptide after cleavage at positions ranging between Asp21 and Leu22 of ElgA (Figure ). The observed molecular weight of elgicin B was 144 Da smaller than the calculated molecular weight of the unmodified propeptide, which can be explained by the loss of eight H2O molecules during posttranslational modification. Therefore, elgicin B is deduced to be the posttranslational modified product of ElgA.
Figure 4 Determination of N-terminal sequence of elgicin B using standard Edman degradation method. A, The 20 known amino acids served as standards. The peak representing the cysteine residue was not labeled. B-E, The first four amino acids in the N-terminal region (more ...)
The residue at position 21 of ElgA (Figure ) was asparagine and leucine was found at position 22. Considering the ESI-MS results, wherein the molecular weight of elgicin C was 114 Da larger and that of elgicin AII 113 Da smaller than that of elgicin B, the N-terminal amino acid sequences of the unmodified propeptides of elgicins C and AII could be Asp-Leu-Gly-Asp-Tyr and Gly-Asp-Tyr, respectively. Similarly, because the glycine residue was at position 23 of ElgA and the molecular weight of elgicin AI was 57 Da smaller than that of elgicin AII, the N-terminal amino acid sequence of the unmodified propeptide of elgicin AI could be Asp-Tyr. The observed molecular weights of these three peptides were 144 Da smaller than the calculated molecular weights of the respective predicted propeptides. This finding may be attributed to the loss of eight H2O molecules during maturation. Elgicins AI, AII, and C were thus confirmed to be the modified products of ElgA, that is, these four antibacterial agents possibly originated from the same prepeptide, ElgA, by peptide cleavage, followed by the removal of one amino acid at each N-terminal.
In the elg gene cluster, the presence of elgB, elgC, and the leader peptide of ElgA containing the motif "FDLD" confirmed that the elgicins are type AI lantibiotics. The origin of elgicins from identical pre-peptides by peptide cleavage and the removal of one amino acid from each corresponding N-terminus could be achieved in two ways. First, the serine protease could cleave at four cleavage sites of ElgA, that is, Ala20-Asp21, Asp21-Leu22, Leu22-Gly23, and Gly23-Asp24 (Figure ), resulting in the simultaneous production of these four peptides. Second, the Ala20-Asp21 could be cleaved by the serine protease to produce elgicin C, followed by the successive protease removal of Asp21, Leu22, or Gly23 residues from elgicin C to yield elgicins B, AII, and AI, respectively.
Antimicrobial activity of elgicins
Preparative RP-HPLC-purified elgicin compounds (150 μg) were pipetted onto a sterile paper disk and tested for antibacterial activity against various bacterial strains. As shown in Table , the active substances produced by P. elgii B69 showed inhibitory activity toward Staphylococcus epidermidis CMCC 26069, Staphylococcus aureus ATCC 43300, Pseudomonas aeruginosa ATCC 27853, Escherichia coli ATCC 35218, and Proteus vulgaris CMCC 49027. Other tested strains, namely, S. aureus ATCC 25923 and B. subtilis CGMCC 1.1470, were resistant to elgicins.
Antibacterial spectra of RP-HPLC-purified elgicin compounds