Article Cite This: J. Nat. Prod. XXXX, XXX, XXX−XXX
pubs.acs.org/jnp
Myxochelin- and Pseudochelin-Derived Lipoxygenase Inhibitors from a Genetically Engineered Myxococcus xanthus Strain Angela Sester,†,‡ Lea Winand,† Simona Pace,§ Wolf Hiller,⊥ Oliver Werz,*,§ and Markus Nett*,†
Downloaded via MACQUARIE UNIV on August 29, 2019 at 20:44:29 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
†
Department of Biochemical and Chemical Engineering and ⊥Department of Chemistry and Chemical Biology, TU Dortmund University, Emil-Figge-Straße 50, 44227 Dortmund, Germany ‡ Leibniz Institute for Natural Product Research and Infection Biology e.V., Hans-Knöll-Institute, Beutenbergstraße 11a, 07745 Jena, Germany § Chair of Pharmaceutical and Medicinal Chemistry, Institute of Pharmacy, Friedrich-Schiller-University, Philosophenweg 14, 07743 Jena, Germany S Supporting Information *
ABSTRACT: Precursor-directed biosynthesis was used to introduce selected aryl carboxylic acids into the pseudochelin pathway, which had recently been assembled in Myxococcus xanthus. Overall, 14 previously undescribed analogues of the natural products myxochelin B and pseudochelin A were generated and structurally characterized. A subset of 10 derivatives together with their parental molecules were evaluated for their activity toward human 5-lipoxygenase. This testing revealed pseudochelin A as the most potent 5lipoxygenase inhibitor among the naturally occurring compounds, whereas myxochelin A is the least active. Replacement of the catechol moieties in myxochelin B and pseudochelin A affected the bioactivity to different degrees.
T
line synthase gene mxcM, leading to the additional formation of 3.12 Feeding of such a recombinant M. xanthus strain with different aryl carboxylic acids hence appeared as a promising approach to gain access to 2- and 3-derived myxochelins. Here, we report on the outcome of this experiment as well as on the biological evaluation of the generated compounds.
he myxochelins (1, 2) and their putative biosynthetic derivatives, namely, pseudochelin A (3) and the hyalachelins (4−6), are siderophores that feature a lysinederived core structure and two catechol groups.1−3 Based upon the occurrence in different bacterial lineages, 1 represents the most widely distributed member of this natural product family.4−6 Pharmacological testing of 1 revealed promising antileukemic properties, which could be correlated to inhibition of the enzyme human 5-lipoxygenase (5-LO).7 Subsequently, several analogues of 1 with modified aryl moieties were prepared by precursor-directed biosynthesis in the myxobacterium Pyxidicoccus fallax, and their inhibitory activities toward 5-LO were determined in a cell-free assay.8 While the results of this study clearly showed that the presence of both catechol units is not essential for 5-LO inhibition, the importance of the lysinol motif for the biological activity remained unclear. This was due to the fact that the biotechnological production of myxochelins exhibiting different core scaffolds is not feasible in P. fallax. Not only does the bacterium lack the genetic prerequisites for the production of the derivatives 2−6, but the myxochelin biosynthesis enzymes are also very restrictive regarding the utilization of amino acids other than L-lysine.9 In order to increase the structural diversity of the naturally made myxochelins, we thus decided to resort to Myxococcus xanthus. This bacterium is a native producer of both 1 and 2,9−11 and its myxochelin pathway can be even further expanded by heterologous expression of the imidazo© XXXX American Chemical Society and American Society of Pharmacognosy
■
RESULTS AND DISCUSSION In consideration of the substrate incorporation previously observed in P. fallax,8 three compounds, i.e., benzoic acid, salicylic acid, and 6-fluorosalicylic acid, were selected for the feeding studies in M. xanthus. Since 2 is the direct biosynthetic precursor of 3,12 the generation of the latter inevitably affects the yield of the former and its derivatives. In order to secure sufficient material for biological testing, the 2-derived myxochelins were thus produced in the wild-type strain of M. xanthus, which lacks the necessary mxcM gene for the formation of 3. Moreover, every M. xanthus culture (wild-type and recombinant) was supplemented with the adsorber resin XAD7HP to improve the myxochelin and pseudochelin yield, respectively.8 After 6 days of incubation, the resin was recovered from the culture broth by filtration and then extracted with methanol. The presence of myxochelins and pseudochelins was probed by LC/MS analysis. Due to the Received: May 1, 2019
A
DOI: 10.1021/acs.jnatprod.9b00403 J. Nat. Prod. XXXX, XXX, XXX−XXX
Journal of Natural Products
Article
extracts, a satisfactory separation was usually accomplished after two consecutive HPLC runs. The yields after purification ranged from 0.1 to 3.7 mg/L (Tables 1 and 2). In total, 14 derivatives (2a, 2b, 2d−2i, 3d−3i) could be fully structurally characterized by NMR and MS analyses. The 1H and 13C NMR spectra of these compounds (see Tables 3 and 4 and Supporting Information) were largely consistent with those of the parental molecules 2 and 3. The only differences occurred in the aromatic regions and could be readily attributed to the incorporation of non-native aryl carboxylic acids through the identification of characteristic mono-, di-, and trisubstituted benzene spin systems. The identity of every myxochelin and pseudochelin derivative was verified by HR-ESIMS measurements. Ten analogues (2d−2h, 3d−3h) were obtained in sufficient quantities to allow biological testing. After individual exposure of isolated human recombinant 5-LO to the test compounds including 1−3, the enzymatic conversion of arachidonic acid into all-trans derivatives of leukotriene B4 and 5-hydro(pero)xyeicosatetraenoic acid was quantified by HPLC. Consistent with our previous investigations,7,8 we observed a potent bioactivity for 1 against 5-LO (IC50 = 1.9 μM). For comparison, the FDA-approved 5-LO inhibitor zileuton possessed only slightly better efficiency in this cell-free assay, with an IC50 value of 0.5 μM.13 The two other natural products, 2 and 3, which were tested for the first time for their 5-LO inhibitory properties in this study, were found to be even more active than 1, with IC50 values of 1.4 and 0.9 μM, respectively. It is interesting that the observed increase in 5-LO inhibition of the bacterial metabolites is paralleling the biosynthetic reaction sequence, in which the formation of 2 and 3 requires additional enzymatic processing in comparison to 1.9,12 Whether this gradual improvement in bioactivity occurred by chance or by evolutionary constraints cannot be answered right now, though further analyses in this direction would be certainly warranted considering the recent discovery of lipoxygenases in myxobacteria.14−16 Of the tested myxochelin B derivatives, 2f and 2g failed to inhibit 5-LO product formation at concentrations lower than 10 μM. The analogues 2d and 2h showed considerably higher IC50 values than their parental molecule, whereas the replacement of the 2,3-dihydroxybenzoate moieties in 3 was surprisingly well tolerated. Except for 3f, which lacks both catechol moieties, the pseudochelin A derivatives retained most of their bioactivity (Tables 1 and 2). On the basis of these results, we could now further refine our previous structure−activity relationship (SAR) data.8 The molecular mode of action of 5-LO inhibition by the myxochelins and pseudochelins is still unclear. In general, 5-LO inhibitors are classified as redox-active compounds, active site iron chelators, substrate mimetics (competitive inhibitors), or noncompetitive inhibitors.17 While substrate competition is rather unlikely for 1−3 and their derivatives, it is possible that these compounds partially inhibit 5-LO due to the complexation of the active site iron. In summary, it has become evident that the presence of an imidazoline ring next to a catechol or phenolic moiety is beneficial for 5-LO inhibition. In contrast, the replacement of both catechol units by salicyl moieties will negatively affect the inhibitory properties. Furthermore, it turned out that the bioactivity of 2 and 3 cannot be further improved by introducing a single salicyl or 6-fluorosalicyl moiety into the biosynthesis. Since these two aryl carboxylic acids were
random incorporation of the 2,3-dihydroxybenzoate surrogates at two possible positions in 1−3, up to nine non-natural derivatives could be detected in the extracts of the recombinant strain, albeit at varying concentrations, after the feeding of a single precursor molecule (Figure 1). In general, salicylic acid and 6-fluorosalicylic acid were incorporated more efficiently than benzoic acid. Those myxochelins that derived from 1 and that had previously been described8 or that seemed to be present in only trace levels were not considered in the following purification efforts. Owing to the complexity of the
Figure 1. Base peak chromatogram from a crude culture extract of a recombinant mxcM-expressing M. xanthus strain, which had been fed with benzoic acid (A). Merged extracted ion chromatograms of pseudochelin A (3) and its derivatives (B), myxochelin B (2) and its derivatives 2a−2c (C), as well as myxochelin A (1) and its derivatives 1a−1c (D). The peak labeled with 3* could not be unequivocally assigned to 3a and/or 3b. B
DOI: 10.1021/acs.jnatprod.9b00403 J. Nat. Prod. XXXX, XXX, XXX−XXX
Journal of Natural Products
Article
Table 1. Production Yield and Inhibitory Effects (IC50 Values) of Myxochelin B (2) and Its Derivatives on 5-Lipoxygenase (5LO) in Comparison to Myxochelin A (1)
A (1) B (2) B1 (2a) B2 (2b) B3 (2c) B4 (2d) B5 (2e) B6 (2f) B7 (2g) B8 (2h) B9 (2i)
R1
R2
R3
R4
R5
R6
R7
5-LO IC50 [μM]a
yield [mg/L]b
OH OH H OH H OH OH OH OH OH OH
OH OH H OH H H OH H H OH H
H H H H H H H H F H F
OH OH OH H H OH OH OH OH OH OH
OH OH OH H H OH H H OH H H
H H H H H H H H H F F
OH NH2 NH2 NH2 NH2 NH2 NH2 NH2 NH2 NH2 NH2
1.9 ± 0.6 1.4 ± 0.5 n.d. n.d. n.d. 10 ± 6.2 5.6 ± 2.6 >10 >10 9.3 ± 4.5 n.d.
n.d. 6.1 0.9 0.9 10 1.1 ± 2.2 ± n.d.
0.3
0.9 0.7 0.5 0.6
yield [mg/L]b 15.9
