Live-cell imaging with Aspergillus fumigatus-specific fluorescent siderophore conjugates

Live-cell imaging allows the in vivo analysis of subcellular localisation dynamics of physiological processes with high spatial–temporal resolution. However, only few fluorescent dyes have been custom-designed to facilitate species-specific live-cell imaging approaches in filamentous fungi to date. Therefore, we developed fluorescent dye conjugates based on the sophisticated iron acquisition system of Aspergillus fumigatus by chemical modification of the siderophore triacetylfusarinine C (TAFC). Various fluorophores (FITC, NBD, Ocean Blue, BODIPY 630/650, SiR, TAMRA and Cy5) were conjugated to diacetylfusarinine C (DAFC). Gallium-68 labelling enabled in vitro and in vivo characterisations. LogD, uptake assays and growth assays were performed and complemented by live-cell imaging in different Aspergillus species. Siderophore conjugates were specifically recognised by the TAFC transporter MirB and utilized as an iron source in growth assays. Fluorescence microscopy revealed uptake dynamics and differential subcellular accumulation patterns of all compounds inside fungal hyphae.[Fe]DAFC-NBD and -Ocean Blue accumulated in vacuoles, whereas [Fe]DAFC-BODIPY, -SiR and -Cy5 localised to mitochondria. [Fe]DAFC -FITC showed a uniform cytoplasmic distribution, whereas [Fe]DAFC-TAMRA was not internalised at all. Co-staining experiments with commercially available fluorescent dyes confirmed these findings. Overall, we developed a new class of fluorescent dyes that vary in intracellular fungal targeting , thereby providing novel tools for live-cell imaging applications for Aspergillus fumigatus.

Live-cell imaging is the key technology to analyse physiological processes inside filamentous fungi, including the uptake and distribution dynamics of fluorescent dyes. The ability to visualize subcellular structures and follow developmental processes in vivo presents a major advantage over fixed samples. Confocal laser scanning microscopy is the most widely available and thus routinely used technology to image and quantify physiological activities with high spatial and high temporal resolution 1 . The combination of genetically encoded fluorescent fusion proteins with organelle-specific fluorescent dyes grants experimental access to virtually any cellular process. Also rapid advances in computer hard-and software now allow the processing of vast amounts of data, particularly those produced during multi-colour 3D-time-lapse recordings 2,3 .
Although a considerable number of organelle-specific fluorescent probes are available for the visualisation of different cellular compartments in filamentous fungi 4,5 , none are species-specific and therefore they are not applicable for identification or targeted treatment in situ.
To overcome this limitation, we exploited the iron acquiring siderophore system of filamentous fungi. Iron is essential for most prokaryotes and all eukaryotes to grow and reproduce. Therefore, iron acquisition has top priority for any organism but is limited by its bioavailability. Atmospheric oxygen rapidly oxidizes the watersoluble ferrous form (Fe(II)) into ferric hydroxide (Fe(III)). Microorganisms have developed different strategies to overcome this problem. Aspergillus fumigatus (A. fumigatus), an airborne human pathogen that causes Scientific RepoRtS | (2020) 10:15519 | https://doi.org/10.1038/s41598-020-72452-2 www.nature.com/scientificreports/ life-threatening invasive pulmonary aspergillosis (IPA) with a mortality rate of up to 90% 6,7 , has two high-affinity iron acquisition systems: reductive iron assimilation and siderophore-assisted iron uptake 8 . Siderophores are low molecular organic molecules that bind ferric iron and fulfill either an iron acquisition or internal storage function from the fungal environment 9 . A. fumigatus produces the hydroxamate-type siderophores desferri-fusarinine C (FsC) and desferri-N,N' ,N''-triacetylfusarinine C (TAFC) which are secreted in response to iron starvation, as well as desferri-ferricrocin (FC) and desferri-hydroxyferricrocin (HFC) which are used for internal iron handling in hyphae and conidia, respectively 9,10 . After secretion, TAFC binds ferric iron with particularly high binding affinity (pM = 31.8, TAFC 11 ) and is subsequently reabsorbed via the specific energy-dependent major facilitator transporter MirB 12 . Uptake of [Fe]TAFC by MirB is characteristic for A. fumigatus as it is not recognized by most other fungi or bacteria 13,14 . Therefore, modification of TAFC is a highly promising approach for the generation of novel and accurate diagnostic biomarkers for the detection of A. fumigatus infections.
In a recent study, we developed strategies to synthesize different TAFC derivatives by subsequently substituting the free amines of the FsC molecule 15,16 . Uptake assays demonstrated that various modifications are possible without losing recognition of TAFC by MirB 17 . During the development of novel hybrid imaging applications for fungal infection diagnostics, combining positron emission tomography (PET) with optical imaging by introduction of a near-infrared (NIR) dye 18 , we successfully visualized A. fumigatus infections. However, we also observed loss of uptake depending on the fluorescent label used.
To investigate this effect, we broadened our scope to include additional fluorescent dyes that cover a wider spectrum of excitation wavelengths (Fig. 1), and characterized their individual uptake behaviour. The aim was to identify optimized compounds that are selectively recognized by A. fumigatus for applications including livecell imaging microscopy.
Starting from the diacetylated form (diacetylfusarinine C, DAFC) 18 we coupled a series of different fluorescent dyes and characterized their recognition by MirB.

Methods
Chemicals. All chemicals were purchased from commercial sources as reagent grade and used without further purification unless stated otherwise. Fluorescent dyes were obtained in their carboxylic acid form, as N-hydroxysuccinimid-esters or isothiocyanates for coupling and were used without further purification: Fluorescein isothiocyanate (FITC) (Sigma Aldrich, Vienna, Austria)) 6- Synthesis of the fluorescent conjugates and radiolabelling. [Fe]FsC was used as a starting material to produce [Fe]DAFC as previously described 16,20 . Fluorescent dyes were coupled depending on their different functional moiety. Carboxylic acid derivatives were activated with O-(7-azabenzotriazol-1-yl)-1,1,3,3-tetramethyluronium-hexafluorophosphate (HATU) for conjugation with the free amine of [Fe]DAFC in dimethylformamide (DMF), or already pre-activated dyes were used (isothiocyanate, NHS-ester). After reaction at room temperature and under light exclusion, the product was purified by preparative RP-HPLC to give a coloured www.nature.com/scientificreports/ solid powder after freeze drying. Identity was confirmed by MALDI-TOF MS (Bruker Daltonics, Bremen, Germany).
To obtain the iron free siderophores, the product was treated with a 1,000-fold excess of ethylenediaminetetraacetic acid (EDTA) at pH 4 and purified by preparative RP-HPLC, as described earlier 18 . 68 Ga-compounds were prepared as previously described 21 using gallium-68 from a commercial 68 Ge/ 68 Ga-generator (IGG100, Eckert & Ziegler Isotope products, Berlin, Germany) and incubating with iron free siderophore in acetate buffer at pH 4.5. The radiolabelled conjugates were used without further purification for LogD determination and uptake/ competition assays, respectively. For all other experiments [Fe]-complexes were used.
More detailed chemical information and analytical conditions are provided as supplementary material.
Uptake and competition assay. Uptake assays were performed as previously described 17,18  Growth promotion assay. Growth promotion assays were performed as previously described 17,22 using a mutant strain (ΔsidA/ΔftrA) of A. fumigatus that lacks sidA and ftrA which have siderophore production and reductive iron assimilation functions. Spores were point inoculated (104 conidia) in 24-well plates, containing 0.5 mL of Aspergillus minimal medium agar and an increasing concentration of iron containing siderophore ranging from 0.1-50 μM. Plates were incubated for 48 h at 37 °C in a humidity chamber and visually assessed 18 . Without siderophore supplementation, no growth of this mutant strain was observed.
Live-cell imaging. Fluorescence microscopy was performed on a Leica TCS SP5 II inverted confocal laser scanning microscope equipped with eight excitation laser lines between 405 and 633 nm, a four-channel filter-free AOBS (Acousto-Optical Beam Splitter) and three photo-multiplier tubes and one Leica HyD detector. Liquid cultures of fungal germling were prepared in μ-Slide 8 Well chambered coverslips (cat.no. 80821, ibidi GmbH, Martinsried, Germany). Each well was inoculated with 5 × 10 3 Spores in 200 μL minimal medium and incubated at 37 °C in a humidified chamber. A. fumigatus (ATCC 46,645) was cultivated for 14 h and A. terreus (ATCC 3,633) for 48 h to obtain well developed germlings and young hyphae without extensive cell fusion. For microscopy, fluorescent dyes were used at a final concentration of 10 μM and incubated for 5-20 min. For costaining experiments, FM1-43 (10 µM), CFW (10 µM) or DFFDA (10 µM) were added simultaneously with the siderophore conjugate. Blocking experiments with NaN 3 (final concentration of 1 mM) and [Fe]TAFC (final concentration of 1 mM) were performed by pre-incubation of the blocking substance for 15 min before adding the fluorophore conjugate. Excitation laser intensity during image acquisition was kept to a minimum to reduce photobleaching and phototoxic effects to the cells while still achieving good signal-to-noise ratios. The precise image acquisition settings are shown for each conjugate in Table S1. Images were recorded with a maximum resolution of 1024 × 1024 pixels and saved as PNG. Z-stack acquisition is indicated in the image description where applicable. Apart from brightness and contrast adjustments and cropping using the ImageJ 1.52a open source software platform (Wayne Rasband, NIH, Bethesda, MD, USA), images were not subjected to further manipulation 18 .

Results
Synthesis of fluorescent conjugates and radiolabelling. All fluorescent conjugates were synthesized with an excellent chemical purity (> 95%, monitored by analytical RP-HPLC, UV absorption at λ = 220 nm) and yields ranging from 20-40%. Corresponding mass analysis was in good agreement with the calculated values. Radiolabelling was achieved in 10 min at room temperature, with almost quantitative radiochemical yields (> 95%) (representative HPLC radiochromatograms are shown in Figure S1).
In vitro characterization. LogD  [Fe]DAFC-BODIPY and -SiR supported growth at 1 µM but did not induce sporulation, even when availability was raised to 50 µM.[Fe] DAFC-Cy5 promoted some growth at 1 µM but led to complete growth arrest above 50 μM, suggesting a concentration dependent inhibitory effect 18 . Interestingly, [Fe]DAFC-TAMRA did not support appreciable growth even at the highest possible concentration. Taken together these data indicate that all siderophore conjugates except [Fe]DAFC-TAMRA can be efficiently utilized as iron carriers by A. fumigatus.
Fluorescence microscopy. Live-cell imaging revealed heterogeneous results not only in terms of overall uptake of the [Fe]DAFC conjugates and dye-dependent uptake dynamics, but also in relation to the subcellular distribution pattern of the internalised molecules. Since [Fe]TAFC requires the MirB transporter to cross the plasma membrane, experiments with A. terreus which lacks MirB orthologs 17 , can be used as a control for potential unspecific, MirB-independent uptake. In addition, application of the free, non-conjugated fluorescent dye alone was used as control for siderophore-independent dye uptake (Fig. 4).
All fluorescent conjugates were efficiently internalised by A. fumigatus germlings and showed distinct subcellular localisation patterns, except [Fe]DAFC-TAMRA which remained outside under all tested conditions.
[Fe]DAFC-NBD and -Ocean Blue localised into big circular structures, most likely vacuoles.
[Fe]DAFC-FITC evenly distributed throughout the cytoplasm but also accumulated in circular structures. The control experiments with A. terreus confirmed that efficient uptake requires MirB because six of the seven siderophore conjugates were not internalised by germlings of this species. Only [Fe]DAFC-SiR produced weak intracellular signals suggesting unspecific uptake by an unknown passive mechanism. Application of fluorescent dyes alone showed that FITC, Ocean Blue, NBD and TAMRA did penetrate the cell wall matrix but did not enter the cell during the observation time. In contrast, unconjugated BODIPY, SiR and Cy5 internalised rapidly, most likely via endocytosis across the plasma membrane. There was no obvious distinction between A. fumigatus and A. terreus. Exemplary, NaN 3 and [Fe]TAFC blocked cellular accumulation of [Fe]DAFC-Ocean Blue, indicating energy-dependant uptake by a membrane transporter, most likely MirB (Fig. S3). www.nature.com/scientificreports/ Co-staining with organelle-specific fluorescent dyes was used to confirm the localisation of internalised [Fe] DAFC conjugates. For instance, the lipophilic plasma membrane marker FM1-43, becomes endocytosed and distributes into mitochondrial and vacuolar membranes over time 3 . The mitochondria of filamentous fungi are longitudinal organelles that tend to accumulate near the tip of actively growing hyphae 5 to support the high

Discussion
Several attempts have been made to target the siderophore iron acquisition system with fluorescent probes. Desferoxamin B (DFO B) has been labelled with nitrobenzofuran (NBD) to acquire novel insight into the siderophore methabolism of the important plant pathogenic fungus Ustilago maydis 24 . The same DFO B-NBD conjugate has also been used to investigate Plasmidium falciparum infections. This modification allowed to overcome major drawbacks in the therapy of infected erythrocytes and also allowed to measure cellular uptake of the siderophore 25 . Doyle and colleagues modified FsC with NBD to perform fluorescence microscopy on A. fumigatus in order to visualize siderophore uptake into hypha 26 . The current study focussed on the uptake properties of different fluorescent [Fe]DAFC conjugates that specifically target A. fumigatus due to its specific interaction with the MirB transporter.
Conjugation of the different fluorescent dyes resulted in high purity compounds with acceptable yields, depending on the conjugation strategy. As expected, they all showed a high dependency to size and charge of the fluorophore. Notably, conjugation increased water solubility of all compounds, compared to the dye alone, allowing a reduction of the amount of organic solvent used for live-cell imaging. Uptake assays also revealed different values depending on the dye. Based on previous findings we postulated that highly charged molecules have a tendency towards unspecific binding to the outer cell wall of the hyphae 17,18 . This phenomenon was indeed observed for [ 68 Ga]Ga-DAFC-BODIPY, -TAMRA, -Cy5 and -SiR, which showed no decrease when blocked with [Fe]TAFC or in iron-sufficient media. As they all are charged molecules from + 1 to -1, interaction with the charged fungal cell wall matrix is likely. On the other hand, all conjugates showed a decrease of [ 68 Ga]Ga-TAFC uptake in competition assays, indicating specific interaction with the MirB transporter.   Figure S2 shows Calcolfuor White/FM1-43 co-staining.  5,27,28 . Taken together, the results show that the modification of TAFC with a variety of fluorescent dyes differentially affects cellular uptake and storage behaviour of the individual conjugates. The fact that different fluorescent moieties resulted in different subcellular localization of siderophore conjugates demands cautious interpretation of the cellular fate of siderophores after uptake via fluorescent labelling. Our studies underline that it is very important to consider that different fluorophores show different uptake behaviour and should be chosen wisely 29 , especially in terms of unspecific binding of the fluorophore itself. Importantly, due to the high specificity of the MirB transporter for TAFC, these dyes are also selective for A. fumigatus. Consequently, these fluorescent probes allow specific labelling of A. fumigatus in mixed cultures, e.g. containing different microbial or mammalian cells. Moreover, these probes can be used to visualization iron starvation in A. fumigatus as their uptake is repressed by iron. Furthermore, these probes allow identification of compounds that interact with the siderophore transporter via competition assays.

Conclusion
Overall, this study demonstrates that fluorescent dyes can be coupled to the siderophore DAFC while retaining its recognition and uptake by the siderophore-transporter MirB in A. fumigatus. Furthermore, additional radiolabelling of the conjugates is possible under mild conditions, to facilitate in vitro and in vivo applications. Since TAFC is specific for the MirB transporter, the described compounds can be used for species-specific fluorescence labelling. Furthermore, by introducing various fluorescent dyes, different cellular compartments can be enlightened at different wavelengths. This structure-related information is also valuable for the development of antifungal siderophore conjugates that might be used for the treatment of A. fumigatus infections.