A Glimpse at Siderophores Production by Anabaena flos-aquae UTEX 1444

In this study, a strain of Anabaena flos-aquae UTEX 1444 was cultivated in six different concentrations of iron (III). Cultures were extracted with organic solvents and analyzed using our dereplication strategy, based on the combined use of high-resolution tandem mass spectrometry and molecular networking. The analysis showed the presence of the siderophores’ family, named synechobactins, only in the zero iron (III) treatment culture. Seven unknown synechobactin variants were present in the extract, and their structures have been determined by a careful HRMS/MS analysis. This study unveils the capability of Anabaena flos-aquae UTEX 1444 to produce a large array of siderophores and may be a suitable model organism for a sustainable scale-up exploitation of such bioactive molecules, for the bioremediation of contaminated ecosystems, as well as in drug discovery.


Introduction
Living organisms require several kinds of metals for their wellness. In virtue of its unique coordination and redox chemistry, iron is, among the metal ions, involved in vital metabolic functions in plants, animals, and microorganisms. The two most frequent oxidation states of iron are +2 and +3, which are also referred to as ferrous and ferric, respectively. Because of its oxidation states, iron plays crucial roles in biocatalysis and electron transport chains, in several biological systems [1]. Microbes use different methods to collect iron from the environment; among others, they produce Fe 3+ -chelating molecules, named siderophores [2]. Siderophores are low-molecular-weight molecules (500-1500 Da), endowed with and named after their high affinity for iron (III) (Kf > 1030). Pathogens adapt their own iron-uptake strategy, in response to the type of infection (acute or chronic) and to the availability of iron [3] of the host. Acting as ferric ion scavengers, siderophores contribute significantly to the virulence of pathogenic microbes. Fe 3+ is transported inside the cell as a hexadentate octahedral complex with the siderophores. Depending on the primary oxygen-donating ligands that bind the iron, siderophores are classified as follows: hydroxamates, catecholates, carboxylates, and mixed-types siderophores [4].
Siderophores are powerful molecules that have applications in ecological research; for the removal of petroleum hydrocarbons from oceans and as algal bloom biocontrollers, in agriculture as soil bioremediators, and in drug discovery as iron chelation therapy, antibiotic carriers, and fish pathogen biocontrollers [5][6][7]. The most intriguing application is their use as ligands for antibiotics, the so-called "Trojan horse" strategy [8]. The potential Mar. Drugs 2022, 20, 256 2 of 10 chemical space of siderophore-antibiotic conjugates is a great opportunity to explore an untapped natural resource and design a new class of molecules to face drug-resistant pathogens [6,9].
Cyanobacteria are a class of photosynthetic microorganism, spread out in a large array of environments, from tropical areas to extremely cold waters. They are present in marine waters, as well as in freshwaters. This unique class of microorganisms is well known to produce several different classes of secondary metabolites, either toxins, named cyanotoxins, or bioactive natural products with interesting pharmacological properties [10]. Cyanobacteria rely on ferric iron to survive. While iron is one of the most abundant elements on Earth, bioavailable iron in freshwater and marine environments is limited, falling in the picomolar to a low nanomolar range. To survive in such an iron-depleted environment, cyanobacteria produce siderophores. Recently, it has been further speculated that cyanobacteria use siderophores as antimicrobial agents and as shields, to protect themselves from heavy-metal toxicity [11]. Cyanobacterial siderophores ( Figure 1) include mainly either hydroxamates or catecholates [12]. The first evidence of presence and types of siderophores in cyanobacteria date back to the early 1980s, when iron uptake in Anabaena sp. in iron starvation conditions was demonstrated to be mediated by the hydroxamate-type siderophore schizokinen [13]. On the other hand, cyanobacterial genomes are known to harbor a rich variety of gene clusters with unknown function. Recently, we reported on the awakening of a widely distributed class of silent gene clusters by iron starvation that yielded cyanochelin's production, β-hydroxy aspartate lipopeptides involved in iron uptake [14].
It is well known that a single strain produces different molecules when grown under different environmental conditions. This concept is the basis of the OSMAC ("One Strain, Many Compounds") strategy [15], a cultivation-based approach that consists in altering cultivation parameters, such as nutrient content, metal ions, rate of aeration, temperature, Mar. Drugs 2022, 20, 256 3 of 10 in order to trigger the productions of compounds of biomedical interest and to activate silent biosynthetic gene clusters.
In this context, in the frame of our ongoing research on the potentiality of cyanobacteria as incubators of bioactive molecules, we investigated the metabolome of the strain of Anabaena flos-aquae (UTEX 1444), when cultivated in the condition of iron deficiency and iron overload. Our study revealed the presence of a suite of synechobactins (seven of which are new variants), only in the extracts from the "zero-Iron (III) treatment" culture.
In this context, in the frame of our ongoing research on the potentiality of cyanobacteria as incubators of bioactive molecules, we investigated the metabolome of the strain of Anabaena flos-aquae (UTEX 1444), when cultivated in the condition of iron deficiency and iron overload. Our study revealed the presence of a suite of synechobactins (seven of which are new variants), only in the extracts from the "zero-Iron (III) treatment" culture.

Molecular Networking and Synechobactins' Identification
All extracts from cultures #1-6 were analyzed by LC-HRMS using an LTQ Orbitrap instrument. Data-dependent acquisition was used to trigger MS 2 scans of the ten most intense ions detected in the full MS scan. The raw LC-MS data were pre-processed using the MZmine program 2.53 [19], which allowed us to remove isotopic peaks, to identify adducts and to perform quantitation. The preliminary analysis (data not shown) of MZmine data revealed that the most significant extract for each culture was that of butanol, as the most representative of each entire metabolome. Therefore, the subsequent data processing and molecular networking were performed on the butanol extracts,

Molecular Networking and Synechobactins' Identification
All extracts from cultures #1-6 were analyzed by LC-HRMS using an LTQ Orbitrap instrument. Data-dependent acquisition was used to trigger MS 2 scans of the ten most intense ions detected in the full MS scan. The raw LC-MS data were pre-processed using the MZmine program 2.53 [19], which allowed us to remove isotopic peaks, to identify adducts and to perform quantitation. The preliminary analysis (data not shown) of MZmine data revealed that the most significant extract for each culture was that of butanol, as the most representative of each entire metabolome. Therefore, the subsequent data processing and molecular networking were performed on the butanol extracts, originating from #1-6 cultures. The global .mgf file, containing the MS 2 data from #1-6 butanol extract and the quantitation table obtained by mzMine were submitted to the online platform at the Global Natural Products Social Molecular Networking website [20], where a Feature-Based Molecular Network (FBMN) was generated https://gnps.ucsd.edu/ProteoSAFe/status. jsp?task=664d019a6b52486d880014bb1e16bd69 (accessed on 12 November 2021) [21]. The resulting molecular network was then visualized using the Cytoscape program 3.9.0 [22] (see also Figures S1-S13, Supplementary Material).
The comprehensive quali-quantitative network of #1-6 butanol extracts (Figure 3a) contains 165 features, grouped into 10 clusters. In the network, each node is represented as a pie chart showing the amount of the compound in the source cultures containing different iron concentrations (Figure 3b). One of the ten clusters was composed of 23 nodes, 13 of which were derived solely from the zero iron (III) culture (Figure 3c, nodes in red). These nodes did not match any known compounds in GNPS' library. The ESI-HRMS spectrum of each of these compounds displayed, in addition to the [M + H] + ion (apo form, not Fe-bound), an ion at 52.9117 amu difference [M + 56 Fe − 3H] + , corresponding to their Fe-adduct, thus, suggesting the presence of Fe-chelating compounds. Six nodes were identified as synechobactins (Table 1), i.e., the hydroxamate-type siderophores. Synechobactins have a backbone consisting of citric acid, linked to two 1,3-diaminopropane units. The first diaminopropane unit is N-acetylated and N-hydroxylated, forming the hydroxamate group. The second diaminopropane unit is N-hydroxylated and N-acylated, with a fatty acid tail that varies in length and degree of unsaturation [23]. The molecular mass of 6, corresponding to the molecular formula of C 30 H 55 O 9 N 4 + , was 2 amu (2 hydrogen atoms) less than that of synechobactin C 16 (5), suggesting the presence of a double bond. The additional unsaturation was allocated to the fatty acid tail, the peak at m/z 361.1706 in the MS 2 spectrum of 6 resulting from the neutral loss of the unsaturated fatty acid C 16 H 30 O 2 (254.2235 amu). The position and the configuration of the double bond remained unassigned. To the best of our knowledge, the only reported synechobactin structure in which an unsaturated acyl chain is present is rhizobactin (C 10 ) [25]; therefore, we propose that synechobactin C 16 Compounds (1)(2)(3)(4)(5)(6)(7)(8) share the same dihydroxamate-citrate moiety, being different only in the structure of the fatty acid residue; compounds 9-12 are modified in the polar part. Both the cleavages from the pseudomolecular ions' peaks at m/z 517.3212 (9) and 545.3525 (10) of dodecanoic and tetradecanoic acid, respectively, are observed for synechobactin A and C 14 , generating the fragment ion C 12 H 21 O 6 N 4 + (m/z 317.1459), in consistence with the non-acetylated hydroxamate/citrate backbone. This was confirmed by the presence of the peak at m/z 429.2601, in the MS 2 spectrum of 9, due to the loss of the 3-(hydroxyamino)propan-1-amine (88.0634 amu, C 3 H 8 ON 2 ) unit, which was also present in the MS 2 spectrum of synechobactin A, due to the loss of the N-(3-aminopropyl)-Nhydroxyacetamide residue (132.0894 amu). The 15.9948 amu difference between synechobactin C 14 (4) and the ion at m/z 573.3839 (compound 11) accounts for an oxygen atom. The loss of tetradecanoic acid from the pseudomolecular ion of 11 leads to the peak at m/z 345.1772 (C 14 H 25 O 6 N 4 ), indicating that the oxygen atom is missing in the dihydroxamatecitrate portion of the molecule. The peak at m/z 328.1508 (C 14 H 22 O 6 N 3 ), in both the MS 2 spectra of synechobactin C 14 (4) and deoxyC 14 (11), corresponds to the fragment (citryl)-N-(3-aminopropyl)-N-hydroxyacetamide, pointing to the lack of a hydroxyl group on the nitrogen of the other hydroxamate. Likewise, in compound 12, the dehydrated form of deoxy synechobactin C 14 , dehydration occurs on the second hydroxamate unit. The thirteenth node of the synechobactins' cluster (m/z 658.2852, C 28 H 50 O 10 N 4 Fe + ) is the iron-bound ion of synechobactin oxyC 14 (7).

Anabaena flos-aquae UTEX 1444: Strain, Cultivation and Extraction
The genus Anabaena BORY ex BORN. et FLAH. 1888 encompasses filamentous, nonbranched species, which differ in the form of cells, coiling of trichomes, shape and size of akinetes [26]. Anabaena species exhibit a wide range of strain-specific variability, and environmental conditions mainly affect cell dimensions and size of heterocysts [27].
The species A. flos-aquae Brébisson ex Bornet et Flahault 1888, is a free-floating organism, whose coiled filaments can be solitary or irregularly assembled, with narrow mucilaginous sheath. Cells are usually spherical to barrel-shaped, whereas akinetes can be cylindrical or elliptical and heterocysts usually spherical [28]. A. flos-aquae is diffused in freshwater environments worldwide and is responsible of blooms that can have noxious effects on other organisms, due to the excretion of toxic metabolites, as anatoxins [29].
From a taxonomical point of view, A. flos-aquae is a combined morphological group, in that an alternation of regularly and irregularly coiled filaments is observed for the same strain in laboratory cultures; moreover, vegetative cell shape and size, along with heterocysts and akinetes dimensions, were largely variable among different populations [30].

LC-HRMS Analyses and Molecular Networking
LC-HRMS experiments were performed using a Thermo LTQ Orbitrap XL highresolution ESI mass spectrometer (Thermo Fisher Scientific Spa, Rodano, Italy) coupled to an Agilent model 1100 LC system (Agilent Technologies, Cernusco sul Naviglio, Italy), which included a solvent reservoir, in-line degasser, binary pump, and refrigerated autosampler. A 5-µm Kinetex C18 column (50 × 2.10 mm), maintained at room temperature, was eluted at 200 mL min −1 with H 2 O (supplemented with 0.1% HCOOH) and MeOH, using gradient elution. The gradient program was as follows: 10% MeOH for 3 min, 10%→100% MeOH for 30 min, 100% MeOH for 7 min. Mass spectra were acquired in positive ion detection mode. MS parameters were a spray voltage of 4.8 kV, a capillary temperature of 285 • C, a sheath gas rate of 32 units N 2 (ca. 150 mL/min), and an auxiliary gas rate of 15 units N 2 (ca. 50 mL/min). Data were collected in the data-dependent acquisition (DDA) mode, in which the first, the second, up until the tenth most intense ions of a full-scan mass spectrum were subjected to high-resolution tandem mass spectrometry (HRMS/MS) analysis. The m/z range for data-dependent acquisition was set between 100 and 2000 amu. HRMS/MS scans were obtained for selected ions with CID fragmentation, isolation width of 2.0, normalized collision energy of 35, Activation Q of 0.250, and activation time of 30 ms. Data were analyzed using Thermo Xcalibur software.
Raw files were imported into MZmine 2.53 open-source software [14]. The mass detection was performed on raw data and exact masses with mass level 1 and centroided masses with mass level 2, by keeping the noise level at 1000. Chromatograms were built using an ADAP module with a minimum height of 1000, and m/z tolerance of 0.001 (or 5 ppm). Peak alignment was performed using the Join aligner algorithm (m/z tolerance at 0.005 (or 5 ppm), absolute RT tolerance at 0. were filtered out by setting the maximum relative height at 100%. Peaks without associated MS/MS spectra were finally filtered out from the peak list. Clustered data were then exported to .mgf file for GNPS, while chromatographic data including retention times, peak areas, and peak heights were exported to a .csv file.
A Feature-Based Molecular Network was generated on GNPS' online platform [15], with the following parameters: the parent mass tolerance and MS/MS fragment ion tolerance were set both at 0.05 Da, the cosine score at above 0.5, and matched peaks above 5. Spectra were retained only if the nodes appeared in each other's respective top 10 most similar nodes. The spectra in the network were then searched against GNPS spectral libraries using a cosine score above 0.7 and at least 6 matched peaks. The molecular network was visualized using Cytoscape software [16].

Conclusions
It is well known that cyanobacteria have a unique adaptive character due to their metabolism capability to react to adverse environmental conditions. They represent a large group of bioagents, less explored as bioresources that produce high-value natural products with biotechnological and ecological relevance. This paper reports on our study on the cyanobacteria Anabaena flos-aquae strain UTEX 1444, as a promising sustainable bioresource of bioactive molecules, i.e., siderophores. We explored metabolome variations when UTEX 1444 was cultivated in the condition of iron deficiency and iron overload. As presumed, the cultivation of the strain in conditions of iron deficiency revealed its capability to produce siderophores. In our procedure, to obtain a fast dereplication of #1-6 organic extracts, molecular networking analyses of MS/MS data were successfully used. Comprehensive FBMN, featuring #1-6 metabolites' relative abundance comparison, immediately disclosed the presence of a cluster composed of metabolites, belonging only to the zero iron (III) culture and, therefore, allowed us to easily and quickly pinpoint the siderophore cluster; moreover, molecular networking allowed for a fast detection of synechobactins and new variants, thanks to their typical fragmentation pattern. HRMS/MS analysis and the fragmentation pattern of synechobactin A, C 16 and the new synechobactin C 16:1 are reported as a useful diagnostic way to detect them in any extract. The structure and the kind of cyanobacterial siderophores were determined in only a few previously published papers [7].
Together with the five known synechobactins, six new variants have been identified, showing unusual modifications occurring in the nature of the acyl chain (unsaturated, hydroxylated) and in the hydroxamate-citrate backbone. To our best knowledge, this is the first report unveiling the biosynthetic capability of UTEX 1444 to produce a large array of synechobactins. It is interesting to note that UTEX 1444 produced 11 variants, beside schizokinen, showing the ability to incorporate a variety of saturated and unsaturated fatty acids, likely as a final biosynthetic step. Modifications also occur in the hydroxamate-citrate backbone, extending the chemical variants of the synechobactins known so far.
Considering the potential high impact of siderophores production as ferric ion scavengers, as well as toxic metal chelators, and then, the large array of applications in many different fields, from ecological research to drug discovery [5][6][7], our future studies will be focused on making UTEX 1444 a suitable model organism for a scale-up production of siderophores. This report, encompassing methodological improvements in siderophore production, characterization and detection through the dereplication of extracts, using a powerful bioinformatic approach, is a sustainable contribution to quest natural molecules that improve environmental and human health. Following EU direction, this study follows the "Do no significant harm" requirement and overcomes the "supply" problem that is the main bottleneck in the study of novel bioactive molecules from marine organisms.