Introduction

Iron is an essential cofactor for a variety of important cellular processes and, therefore, can be considered as a vital nutrient for virtually all forms of life [1, 2]. Most microorganisms use special mechanisms to acquire iron including production of siderophores [3]. Mainly in the iron-poor environments, microorganisms such as Aspergillus fumigatus produce large amounts of siderophores to scavenge iron (III) and enable its uptake into the organism [4]. It has recently been recognised that iron plays an essential role in infection in general [5] and in fungal infections in particular. It has been shown that in particular, the siderophore system is essential for the virulence of A. fumigatus [6].

Siderophores are low molecular weight, iron-chelating ligands synthesised by almost all microorganisms for iron acquisition and storage; however, not all aspects of siderophore utilisation by microorganisms are fully understood. Luckey et al. [7] reported that some bacterial strains lost the ability to synthesise siderophores but retained the ability to utilise siderophores released by other microorganisms. Similarly, Haas [8], Philpott et al. [9] and Heymann et al. [10] reported that several fungal species are able to utilise siderophores produced by other fungi, indicating the lack of specificity of siderophore systems for particular microorganisms.

A great variety of different siderophores are known today, the majority being of hydroxamate, catecholate or α-hydroxycarboxylate type, each having a high selectivity for iron (III) [11]. The chemistry of iron (III) and gallium (III) is very similar and was already widely exploited in the field of nuclear medicine in the use of 67Ga-citrate. 68Ga is short-lived, generator-produced isotope that has recently become the subject of great interest for molecular imaging applications using positron emission tomography (PET) [12]. We have recently shown that various siderophores can be labelled with 68Ga [13], and 68Ga-triacetylfusarinine C (TAFC) and 68Ga-ferrioxamine E (FOXE) are able to detect A. fumigatus infection in a rat infection model using PET imaging [14, 15].

TAFC is a common trihydroxamate-type siderophore of many fungal species (Aspergillus sp., Fusarium sp., etc.). Although many different forms of fusarinines have been detected, the cyclic acetylated trimer is regarded as the product with the highest chemical stability [16]. Adjimani and Emery have even shown that TAFC is able to extract iron from other siderophores and thereby feed the producing microorganism with iron from exogenous siderophores [17].

Ferrioxamines were originally isolated and characterised as ferrioxamines A to H. All ferrioxamines are trihydroxamate-type siderophores and are either cyclic or linear. FOXE is a cyclic siderophore with high affinity to iron (III) mainly produced by actinomycetes and other bacteria [18].

Here, we report on the characterisation and comparison of in vitro and in vivo uptake of 68Ga-TAFC and 68Ga-FOXE in different microorganisms and human lung cancer cells to evaluate their specificity and sensitivity for A. fumigatus infection imaging.

Materials and Methods

Chemicals

All commercially obtained chemicals were of the highest available purity and were used without further purification. Siderophores were obtained from Genaxxon BioScience GmbH (Ulm, Germany), and 68Ge/68Ga generator, from Eckert & Ziegler Eurotope GmbH (Berlin, Germany).

Radiolabelling

Both siderophores were labelled with 68Ga using acetate buffer at room temperature (RT) (TAFC) or at 80 °C (FOXE) [13]. Radiochemical purity was determined using reverse-phase high-performance liquid chromatography gradient method and/or instant thin-layer chromatography on silica gel impregnated glass fibres, as described previously [1315].

Preparation of Microbial Cultures for In Vitro Studies

A. fumigatus ATCC46645, Aspergillus terreus DSM826, Aspergillus flavus ATCC9643, Rhizopus oryzae AS5 and Fusarium solani AS94 were cultured at 37 °C in liquid Aspergillus minimal media (AMM) [19] containing 1 % glucose and 20 mM glutamine as carbon and nitrogen source, respectively. Iron-containing media were supplemented with 30 μM FeSO4, whereas for iron-limiting cultures, iron was omitted. For all other microbial strains, the iron-replete and iron-limited main cultures were at first precultured for 18 h and inoculated with a single colony at 37 °C. Such an inoculum was subsequently used for the culturing of the main cultures. The preculture medium for Candida albicans ATCC90028, Klebsiella pneumoniae and Pseudomonas aeruginosa ATCC9027 was yeast peptone dextrose (YPD) + 0.5 % glucose; that for Mycobacterium smegmatis mc2155, YPD +  0.5 % Tween 80; and that for Staphylococcus aureus, Roswell Park Memorial Institute (RPMI) + 1 % casamino acids. With exception of P. aeruginosa, the iron-replete main culture medium for all these strains was the same as the preculture medium, and for iron-limiting conditions, the ferrous iron chelator dipyridyl was added to a final concentration of 200 μM. For P. aeruginosa, the main culture media were iron-replete and iron-limiting AMM (see above), respectively. Iron-deficient conditions were verified by detection of extracellular siderophores production, which is repressed by iron.

Preparation of Human Lung Cancer Cells for In Vitro Studies

H1299 non-small cell human lung cancer cells (ATCC) were maintained in tissue culture flasks (Cellstar, Greiner Bio-One, Kremsmuenster, Austria) in RPMI 1640, supplemented with 10 % (v/v) heat-inactivated FBS and 1 % (v/v) PSG at 37 °C with 5 % CO2 in a humidified atmosphere and grown in monolayer. On the day of the experiment, cells were removed with trypsin-EDTA and used at a density of 1 × 106 cells/ml for the uptake experiments.

In Vitro Uptake of 68Ga-Siderophores in Various Microbial Media

In vitro uptake was studied in A. fumigatus, A. flavus, A. terreus, C. albicans, R. oryzae, F. solani, P. aeruginosa, K. pneumoniae, S. aureus and M. smegmatis iron-deficient and iron-sufficient cultures. For the monitoring of uptake over time, 68Ga-siderophores were incubated in the microbial media for 10, 45 and 90 min at RT with or without blocking solution (Fe-TAFC or Fe-FOXE) in 96-well plates (Millipore, Billerica, MA, USA). The uptake was interrupted by filtration of the medium and rapid rinsing with ice-cold Tris buffer. The filters were collected and counted in a γ-counter.

Siderophore Utilisation Growth Assay

To exemplary confirm the ability or inability to take up TAFC, we developed a siderophore utilisation growth assay (Fig. 2). In this agar diffusion assay, the analysed species (107 conidia of A. fumigatus or 0.2 ml of K. pneumoniae preculture, respectively) was poured in 5 ml top agar (iron-limiting AMM medium + 0.7 % agar) on agar plates (iron-limiting AMM medium). Subsequently, 80 μl of 0.6 mM ligand-free TAFC solution was inoculated into a hole (5 mm diameter) punched into the middle of the plate. The plates were then incubated at 37 °C for 30 h. The ligand-free siderophore diffuses into the growth medium and chelates the present iron with the highest TAFC concentration in the vicinity of the hole. The growth of microorganisms unable to take up TAFC-iron is inhibited by high TAFC concentrations in this assay as TAFC iron is here the only iron source present.

In Vitro Uptake of 68Ga-Siderophores in Human Lung Cancer Cells

In vitro uptake was studied in human non-small cell lung cancer cells H1299 type. H1299 cells were seeded at a density of 1 × 106 cells per Eppendorf tube and incubated in triplicates with 68Ga-TAFC or 68Ga-FOXE at RT for 90 min. For positive and negative control, Eppendorf tubes containing iron-deficient and iron-sufficient cultures of A. fumigatus were incubated in triplicates with 68Ga-labelled siderophores at RT for 90 min in parallel. The incubation was interrupted by 2-min centrifugation at 5,000×g. The supernatant was collected and measured in a γ-counter. Cells sediment was disturbed by 1 ml of glycine and subsequent whirling. Aspergillus media were washed with 1 ml of ice-cold Tris buffer. All samples were centrifuged for 2 min, and supernatants were measured in γ-counter. Thereafter, 1 ml of 1 M NaOH was added to human lung cancer cells, mixed and centrifuged for 2 min. The supernatant was again collected and measured in a γ-counter.

Animal Experiments

All animal experiments were conducted in accordance with the regulations and guidelines of the Austrian and Dutch animal protection laws and with the approval of the Austrian Ministry of Science (66011/42-II/10b/2009) and Institutional Animal Welfare Committee of the Radboud University Medical Centre Nijmegen (revised Dutch Act on Animal Experimentation, 1997). Animal studies were performed using Lewis rats (Charles River Laboratories, Wilmington, MA)

In Vivo Imaging

In vivo uptake was studied in the A. fumigatus rat infection model [15, 16] and in rats with sterile (turpentine oil) and bacterial (S. aureus) intramuscular (i.m.) inflammation. An abscess was induced in the left calf muscle with approximately 1 × 109 colony-forming units of S. aureus in 0.1 ml 50:50 % suspension of autologous blood and normal saline. A sterile inflammation was induced in the right calf muscle by injecting 0.1 ml turpentine oil intramuscularly. During the procedure, animals were anaesthetised. After 24 h, when swelling of the muscle was apparent, the tracers were injected intravenously through the tail vein.

PET imaging was obtained using an Inveon animal PET/CT scanner (Siemens Preclinical Solutions, Knoxville, TN, USA) [15]. Static PET scans of 30 min were recorded at 30 min after i.v. injection of 68Ga-siderophore or 2-deoxy-2-[18F]fluoro-d-glucose (18F-FDG), the time established in the previous studies [15].

Results

Radiolabelling and In Vitro Stability

Both 68Ga-TAFC and 68Ga-FOXE showed high radiochemical purity and in vitro stability as described in the previous studies [1315]. Figure 1 displays chemical structures of studied 68Ga-siderophores.

Fig. 1
figure 1

Chemical structures of a 68Ga-TAFC and b 68Ga-FOXE.

In Vitro Uptake of 68Ga-Siderophores in Various Microbial Media

In vitro uptake of 68Ga-TAFC and 68Ga-FOXE was highly dependent on the mycelia iron load (Tables 1 and 2). Both compounds showed rapid uptake by iron-starved A. fumigatus cultures (Table 1), which could be partly blocked with excess of Fe-siderophore and significantly lower uptake by A. fumigatus grown under iron-sufficiency (Table 2). Tables 1 and 2 summarise the uptake values of both 68Ga-TAFC and 68Ga-FOXE in various microorganisms and H1299 cells. In bacterial (P. aeruginosa, K. pneumoniae, S. aureus, M. smegmatis) or yeast (C. albicans) cultures, 68Ga-TAFC revealed virtually no uptake in both iron-deficient and iron-sufficient media. 68Ga-FOXE showed similar results to 68Ga-TAFC except for S. aureus. In S. aureus iron-deficient media, clear uptake of 68Ga-FOXE was observed. In fungal cultures (A. flavus, A. terreus, R. oryzae, F. solani), both 68Ga-siderophores, especially 68Ga-FOXE, showed certain uptake in iron-deficient media, which was substantially lower in comparison with the uptake in A. fumigatus cultures.

Table 1 Comparison of 68Ga-FOXE and 68Ga-TAFC uptake in various microorganisms 10, 45 and 90 min after incubation in iron-deficient cultures and in human lung cancer cells (H1299)
Table 2 Comparison of 68Ga-FOXE and 68Ga-TAFC uptake in various microorganisms 10, 45 and 90 min after incubation in iron-sufficient cultures and in human lung cancer cells (H1299)

Siderophore Utilisation Growth Assay

In this assay (Fig. 2), K. pneumoniae displayed in contrast to A. fumigatus a clear growth inhibition zone, which is in perfect agreement with the in vitro 68Ga-TAFC uptake assay that indicated that K. pneumoniae lacks TAFC uptake (see Table 1). As a control, the plates were inoculated with ferric TAFC instead of ligand-free TAFC. In this case, iron traces that are not chelated by TAFC are available for growth. In this set-up, the growth of K. pneumoniae was not inhibited, which demonstrates that the growth inhibition is indeed due to iron chelation by TAFC and not by a potential iron-independent antibacterial activity of TAFC.

Fig. 2
figure 2

Siderophore utilisation assay demonstrating that growth of K. pneumoniae is in contrast to that of A. fumigatus inhibited by ligand-free TAFC.

In Vitro Uptake of 68Ga-Siderophores in Human Lung Cancer Cells

No uptake was observed in human lung cancer cells (H1299) for both 68Ga-siderophores (see Tables 1 and 2). Almost all the radioactivity was found in the supernatant containing cell media, and negligible radioactivity was observed in the glycine and NaOH supernatants for both compounds.

In Vivo Imaging

In vivo PET imaging in the A. fumigatus rat infection model showed rapid focal accumulation of 68Ga-siderophores in the lungs. Whereas no in vivo uptake in the lung region was detected in non-infected animals, and the only visible organs were the kidneys and bladder (see Fig. 3).

Fig. 3
figure 3

In vivo imaging of the A. fumigatus rat infection model (left-hand side images) and non-infected (right-hand side images) animals using a 68Ga-TAFC and b 68Ga-FOXE, 1 h. i.v. postinjection.

Both 68Ga-TAFC and 68Ga-FOXE revealed moderate clearly detectable uptake in the muscle with turpentine oil and no uptake in the S. aureus abscess, whereas 18F-FDG showed evident uptake in both sites of induced inflammation (see Fig. 4).

Fig. 4
figure 4

Comparison of a 18F-FDG, b 68Ga-TAFC and c 68Ga-FOXE uptake in the muscle of rats i.m. injected with turpentine oil and S. aureus abscess.

Discussion

Invasive aspergillosis (IA), which is mainly caused by A. fumigatus, is becoming one of the leading infective causes of morbidity and mortality in immunocompromised patients [2022]. A key factor for the patient survival is an early and accurate diagnosis of IA. Currently a number of tests and techniques are used in clinical practice, unfortunately lacking enough sensitivity and/or specificity that the diagnostics of IA continues to be highly challenging [23].

The question of specificity of the diagnostic method is extremely important for subsequent therapy management. There are many different types of chemotherapeutic drugs that are used against specific pathogens or tumours. Many radiotracers used in nuclear medicine clinical practice for inflammation or infection imaging (e.g. 18F-FDG, 67Ga-citrate) are not specific, even though some experimental developments claim to be able to distinguish between inflammation and infection [24]. We have shown that the 68Ga-labelled siderophores TAFC and FOXE are able to image A. fumigatus infection in vivo at an early onset of infection [15]. The mechanism of 68Ga-siderophores action is based on the active transport via highly specific transporters, which are upregulated during A. fumigatus infection, resulting in rapid accumulation of the radiotracer at the site of infection, allowing early detection of infection having appropriate clinical impact. A specific uptake mechanism utilising the nutrition pathway of the microorganisms also potentially should allow uptake in infected tissues that are normally not accessible by hydrophilic radiotracers such as the brain. This specific mechanism also holds the promise that 68Ga-siderophores could be highly specific agents for A. fumigatus infection imaging, and we tried to address this in the presented paper.

TAFC is a common siderophore of many pathogenic fungal species [16]. It is interesting to note that transport systems for fusarinines in bacteria are unknown so far [3]. Although other fungal siderophores, such as ferrichromes and coprogen, are utilised by Escherichia coli, a fusarinine outer membrane receptor is still lacking. Whereas TAFC is produced and utilised by many fungal species, FOXE is a siderophore produced by bacteria (Streptomyces spp.) and used by many bacteria including S. aureus [25, 26], but also by several fungal species, as we shown here. We have tested the specificity of 68Ga-TAFC and 68Ga-FOXE in vitro on the set of representative microorganisms as well as in the human lung cancer cells. Both compounds showed no uptake in the human lung cancer cells. The distinction between infection and tumour is very important, especially considering that one of the major groups of immunocompromised patients suffer various malignancies including lung cancer. 68Ga-TAFC displayed no uptake in any tested bacterial cultures (P. aeruginosa, K. pneumoniae, S. aureus and M. smegmatis), whereas 68Ga-FOXE was clearly taken up by S. aureus. This finding was in contrast with subsequent in vivo tests using the i.m. rat bacterial (S. aureus) infection model. No in vitro uptake was observed for 68Ga-FOXE in the remaining tested bacterial cultures (P. aeruginosa, K. pneumoniae and M. smegmatis). C. albicans was chosen as widely occurring representative of yeasts which is potentially pathogenic in immunocompromised patients. Both compounds were not taken up by C. albicans at all. The last group microorganisms studied in vitro covered selected fungal species. 68Ga-TAFC showed certain uptake in R. oryzae and F. solani cultures, but surprisingly very low uptake in Aspergillus species (A. flavus, A. terreus) in relation to A. fumigatus. 68Ga-FOXE displayed evidently higher uptake in all tested fungal media. In summary, the in vitro uptake of both 68Ga-TAFC and 68Ga-FOXE in A. fumigatus cultures was significantly higher compared to all tested microorganisms and human lung cancer cells, and the in vitro studies indicated higher specificity of 68Ga-TAFC.

In vivo PET imaging was based on our previous studies [14, 15] and was extended by testing of 68Ga-siderophores and 18F-FDG in the rat model of sterile (turpentine oil) and bacterial (S. aureus) i.m. inflammation. Both 68Ga-siderophores displayed rapid and highly selective focal accumulation in the lungs of A. fumigatus infected rats, whereas in non-infected animals, no radioactive signal was detected in the lungs area. In vivo imaging in the sterile and bacterial inflammation model showed no in vivo uptake in the S. aureus abscess for both 68Ga-siderophores. This finding is in contrast to the in vitro data of 68Ga-FOXE in S. aureus cultures as mentioned before. 68Ga-TAFC and 68Ga-FOXE revealed detectable uptake in the area of induced sterile inflammation. This uptake might be due to the severity of the inflammation, causing non-specific extravasation of the tracers. As expected, 18F-FDG showed much higher uptake in both i.m. sites of inflammation. In vivo studies confirmed that 68Ga-siderophores can be used for Aspergillus infection imaging.

Conclusions

We have shown that both 68Ga-TAFC and 68Ga-FOXE can be used for imaging of IA. Moreover, 68Ga-TAFC showed high in vitro specificity towards A. fumigatus compared to other tested microorganisms and human lung cancer cells. Both 68Ga-siderophores showed highly selective accumulation in the infected lungs in the A. fumigatus rat infection model using PET imaging with 68Ga-FOXE being slightly superior in terms of sensitivity. The final answer of the usefulness and applicability of these compounds for early diagnosis of IA in patients lies now in the clinical studies.