On the origin of amphi-enterobactin fragments produced by Vibrio campbellii species

Amphi-enterobactin is an amphiphilic siderophore isolated from a variety of microbial Vibrio species. Like enterobactin, amphi-enterobactin is a triscatecholate siderophore; however, it is framed on an expanded tetralactone core comprised of four l-Ser residues, of which one l-Ser is appended by a fatty acid and the remaining l-Ser residues are appended by 2,3-dihydroxybenzoate (DHB). Fragments of amphi-enterobactin composed of 2-Ser-1-DHB-FA and 3-Ser-2-DHB-FA have been identified in the supernatant of Vibrio campbellii species. The origin of these fragments has not been determined, although two distinct isomers could exist for 2-Ser-1-DHB-FA and three distinct isomers could exist for 3-Ser-2-DHB-FA. The fragments of amphi-enterobactin could originate from hydrolysis of the amphi-enterobactin macrolactone, or from premature release due to an inefficient biosynthetic pathway. Unique masses in the tandem MS analysis establish that certain fragments isolated from the culture supernatant must originate from hydrolysis of the amphi-enterobactin macrolactone, while others cannot be distinguished from premature release during biosynthesis or hydrolysis of amphi-enterobactin. Graphical abstract Supplementary Information The online version contains supplementary material available at 10.1007/s00775-022-01949-0.


Introduction
Iron is a cofactor required by many enzymes involved in essential cellular processes. However, obtaining iron becomes challenging due to the low solubility of iron (III).
One strategy that bacteria have evolved to obtain iron is the biosynthesis of siderophores, low molecular weight organic compounds that bind Fe(III) with high affinity. These Fe(III)-siderophore complexes are taken up by the cell through outer membrane receptor proteins.
V. campbellii BAA-1116 contains a set of genes homologous to the biosynthetic gene cluster (BGC) of enterobactin, entA-F (Fig. 1A), yet instead of enterobactin, the strain produces amphi-enterobactin (Fig. 1B, C) [1]. In addition to the amphi-enterobactin aebA-F genes, the gene aebG encoding a long-chain fatty acid Co-A ligase (FACL) is located nearby this BGC [1]. The biosynthesis of 2,3-dihydroxybenzoic acid (2,3-DHBA) is carried out by AebABCE. Zane et al. [1] established that the biosynthesis of amphi-enterobactin begins by appending an AebG-activated fatty acid to l-Ser loaded on AebF (Fig. 1C). FACL enzymes are known to activate fatty acids to fatty acyl-CoA thioesters before integrating with the nonribosomal peptides [5,6]. Thus, this FACL initiates the biosynthetic process of amphi-enterobactin by appending the FA to the first loaded l-Ser residue on AebF NRPS. AebF continues its bifunctional activity of catalyzing the formation of amide bonds between DHB and another l-Ser, respectively. The thioesterase domain of AebF ultimately catalyzes the release of amphi-enterobactin through intramolecular cyclization, generating the macrolactone and releasing amphi-enterobactin from the NRPS [1].
Several bacterial strains, V. campbellii BAA-1116, Burkholderia cepacian K56-2, and V. vulnificus MO6-24/O, have been shown to engage in quorum-sensing regulation of siderophore production, where high cell density leads to an accumulation of quorum sensing molecules, which The aebABCDEF biosynthetic gene cluster first identified in Vibrio harveyi BAA-1116 [1]. Genes involved in siderophore biosynthesis and transport are represented by blue and orange arrows, respectively. White arrows represent hypothetical proteins whose function has not yet been determined. c Biosynthesis of amphi-enterobactin catalyzed by NRPS AebF. The potential points of pre-release of fragments in the biosynthesis of amphi-enterobactin are indicated (blue arrows). Each potential early release product has the fatty acid appended to the amine of a C-terminal l-Ser. C condensation domain; A adenylation domain; T thiolation domain; TE thioesterase domain with the Fe(II)-Fur complex decreases siderophore production [2,4,7,8]. A recent study reported the presence of amphi-enterobactin-related soluble fragments, particularly 2,3-dihydroxybenzoic acid (DHBA) and 2,3-dihydroxybenzoyl-l-serine (DHB-Ser), along with linearized amphi-enterobactin fragments mass spectrometry [4]. DHBA and DHB-Ser were found to be more abundant in comparison with amphi-enterobactin. McRose et al. [4] propose two possible sources of DHBA and DHB-Ser: premature release from the biosynthetic pathway or degradation of amphi-enterobactins. Because of the accumulation of DHBA and DHB-Ser found in the supernatant of V. campbellii BAA-1116, the study suggested an inefficient amphi-enterobactin biosynthetic process [4].
Amphi-enterobactin hydrolysis products composed of two l-Ser residues, one 2,3-dihydroxybenzoate (2,3-DHB) group and a fatty acid, have been reported previously [1,4]. In this report, we use a shorthand notation for these fragments, based on a binary code [9,10], where the number [1] depicts l-Ser appended by the fatty acid, and We report herein a mass fragmentation analysis that establishes these amphi-enterobactin hydrolysis fragments arise from the full siderophore, although we cannot rule out premature release. The amphi-enterobactin macrolactone siderophore is in fact produced as supported by the tandem MS analysis of the hydrolysis products.

General experimental procedures
A Varian Cary-Bio 300 UV-visible spectrophotometer was used to monitor microbial growth at 600 nm. Analytical HPLC was used to analyze both the supernatant and cell pellet extracts from V. campbellii CAIM 519T to identify the production of both the amphi-enterobactin macrolactone and hydrolysis products. Mass spectrometry analysis was carried out on a Waters Xevo G2-XS QTof with positive-mode electrospray ionization coupled to an ACQUITY UPLC H-Class system with a Waters BEH C18 column. The cultures were harvested by centrifugation (6000 rpm, 30 min, 4 °C). The supernatant was decanted, and the cell pellet was resuspended in methanol (25 mL/ pellet), transferred into 50-mL conical tubes, and shaken overnight at 180 rpm, 4 °C. The methanol extract was centrifuged (6000 rpm, 10 min, 4 °C), filtered through a 0.22-μm membrane, and concentrated under vacuum to one-third the original volume.

Cultivation of
Both the supernatant and the cell pellet extracts were purified with XAD resin. The cell pellet extract was diluted with four times the volume with doubly deionized water (Milli-Q IQ). The supernatant and cell pellet extract were incubated with Amberlite XAD-2 resin for 4 h at 120 rpm, 25 °C. After 4 h, the XAD resin was washed with 2 L of doubly deionized water. From the cell pellet extract, siderophores were eluted with 90% methanol. From the supernatant, siderophores were eluted with 80% methanol. The eluent was concentrated under vacuum to dryness and dissolved in 5 mL of methanol.

UPLC-MS and MS/MS analysis of extracts
Extracts were analyzed through positive ion mode ESI-MS on a Waters Xebo G2-XS QTof coupled to a Waters Acquity H-Class UPLC system. The extracts of the culture supernatant were analyzed with a linear gradient of 0-100% CH 3 CN (0.1% formic acid), while the cell pellet extracts were analyzed with a linear gradient of 50-100% CH 3 CN (0.1% formic acid) in ddH 2 O (0.1% formic acid) over 10 min. For MSMS analysis, a collision energy profile of 20, 25, 30 kEV was employed. Using MassLynx 4.1, chromatograms for masses of interest were generated and molecular ion peaks quantified by integration (ApexTrack algorithm).

Results and interpretation
Origin of the amphi-enterobactin fragments: premature release during biosynthesis or macrolactone ester hydrolysis While it has been established that Vibrio campbellii CAIM 519T produces a suite of amphi-enterobactins, with fatty acids ranging from C 10 to C 14, which are either saturated or monohydroxylated [3], fragments of these amphi-enterobactins are also present in the culture supernatant of V. campbelli CAIM 519T (Figs. 2, 3 and S2). We have turned to tandem MS to investigate whether selected fragments originate from hydrolysis of the amphi-enterobactin macrolactone siderophore.  [01], or the N-terminal l-Ser, depicted by [10] (Fig. 3).
The species eluting at 4.6 min and 5.4 min are associated with Peak B and Peak D, respectively (Fig. 2) however, due to the trace quantity produced, tandem MS characterization was not carried out. The complete set of isomers along with the associated binary nomenclature is shown in Fig. 3.
Biosynthesis of amphi-enterobactin is initiated during fatty acyl-CoA thioester acylation of l-Ser-S-P-pant-AebF [1]. Thus, the carboxyl group interacting with the thioesterase domain throughout the amphi-enterobactin biosynthesis will always be appended to the fatty acid that was loaded onto l-Ser. Premature release of amphi-enterobactin fragments along the biosynthetic pathway could potentially occur at the thioesterase domain of the NRPS, releasing a fragment with the fatty acid appended to the C-terminal Ser, as in [01], [001], or [0001] (Fig. 1).
Structural variation within fragments increases if hydrolysis products arise from the fully formed amphi-enterobactin macrolactone. While this set of fragments may contain the fatty acid appended to the C-terminal Ser, as in the premature release fragments [01], [001], or [0001], other fragments with the fatty acid appended at each of the other Ser residues in the oligoserine backbone may be formed as well. Depending on the site of macrolactone hydrolysis, all of the structures in Fig. 3 may be considered hydrolysis products from amphi-enterobactin.

Conclusions
In summary, tandem MS analysis of the hydrolysis fragments of amphi-enterobactin in the culture supernatant establish that isomers [10] and [100] must arise from hydrolysis of the macrolactone amphi-enterobactin siderophore as opposed to prerelease of di-Ser or tri-Ser fragments during biosynthesis. Evidence for the Future experiments involving an in vitro analysis of the biosynthesis proteins for amphi-enterobactin could provide insight into the potential for premature release of incomplete fragments along the biosynthetic pathway for amphi-enterobactin. Previous results from reconstructing enterobactin