[68Ga]Ga-Schizokinen, a Potential Radiotracer for Selective Bacterial Infection Imaging

Gallium-68-labeled siderophores as radiotracers have gained interest for the development of in situ infection-specific imaging diagnostics. Here, we report radiolabeling, in vitro screening, and in vivo pharmacokinetics (PK) of gallium-68-labeled schizokinen ([68Ga]Ga-SKN) as a new potential radiotracer for imaging bacterial infections. We radiolabeled SKN with ≥95% radiochemical purity. Our in vitro studies demonstrated its hydrophilic characteristics, neutral pH stability, and short-term stability in human serum and toward transchelation. In vitro uptake of [68Ga]Ga-SKN by Escherichia coli, Pseudomonas aeruginosa, Staphylococcus aureus, and S. epidermidis, but no uptake by Candida glabrata, C. albicans, or Aspergillus fumigatus, demonstrated its specificity to bacterial species. Whole-body [68Ga]Ga-SKN positron emission tomography (PET) combined with computerized tomography (CT) in healthy mice showed rapid renal excretion with no or minimal organ uptake. The subsequent ex vivo biodistribution resembled this fast PK with rapid renal excretion with minimal blood retention and no major organ uptake and showed some dissociation of the tracer in the urine after 60 min postinjection. These findings warrant further evaluation of [68Ga]Ga-SKN as a bacteria-specific radiotracer for infection imaging.

T he diagnosis and treatment of bacterial infections are challenging because of the difficulties in isolating and identifying pathogens early, determining pathogen sensitivity to drugs, selecting the most effective treatment options, and monitoring the success of treatment over an appropriate duration.Treatment outcomes worsen when pathogens become resistant to commonly prescribed antimicrobial agents. 1 In 2019, bacterial pathogens caused an estimated 7.7 million deaths globally, 1.27 million of which were directly linked to antimicrobial resistance (AMR). 2 Escherichia coli, Acinetobacter baumannii, Staphylococcus aureus, Streptococcus pneumoniae, Klebsiella pneumoniae, and Pseudomonas aeruginosa were the leading causes of death.Although the use of rapid molecular and genome-based diagnostic techniques is increasing, 3 infection-specific in situ imaging as a diagnostic tool remains an unmet clinical need.This approach is particularly relevant for the early detection of invasive infections in immunocompromised patients, for locating and detecting microbial infections in the body (for example, infection of implanted prosthetics 4−6 ), and for monitoring treatment efficacy.−9 None of these radiotracers can directly image the infection in the body but rather capture secondary host responses against infections. 10esearch has been ongoing to identify microbial targets for developing sensitive and specific radiotracers for imaging infections. 1 1 − 1 3 Among them, sorbitol analogue 2-[ 18 F]fluorodeoxysorbitol (2-[ 18 F]FDS) targeting the sugar metabolic pathway, 14 radiolabeled D-amino acids (DDAs) targeting cellular components, 15 and gallium-68 ( 68 Ga)-labeled siderophores targeting the iron transport pathway 4 are the front runners for bacteria-specific radiotracer developments.However, both 2-[ 18 F]FDS and DDAs have recently shown preclinical imaging of fungal infections, 14,16 thus rendering them broad-spectrum infection imaging agents.Interestingly, siderophores, which transport the essential transition metal iron(III) (Fe 3+ ) into microbial cells from the surrounding environment, can be designed as bacterial-and fungal-specific radiotracers.Typically, Fe 3+ transport is mediated by the recognition of siderophores via specific cell-surface receptors and the internalization of siderophore-bound iron by the active transportation system of the membrane, Siderophore-Iron-Transporters (SITs).−19 During infection, pathogens increase the production of these low molecular weight siderophores with high affinity for iron to obtain Fe 3+ because host proteins such as transferrin, lactoferrin, and ferritin tightly regulate the availability of iron in the host environment. 20Siderophore-mediated iron transport in microbes has been discussed in detail elsewhere. 21 68 a is a PET radioisotope that chemically resembles Fe 3+ .This allows microbes to import siderophores radiolabeled with 68 Ga into microbial cells instead of Fe 3+ -siderophores.Different 68 Ga-siderophores have been investigated preclinically for imaging infection, and PET/CT imaging in animal infection models has proven the utility of 68 Ga-siderophores as diagnostic tools. 22Furthermore, the utilization of xenosiderophores offers the opportunity to develop both broadspectrum and narrow-spectrum radiotracers that could be useful in relevant clinical cases, such as polymicrobial infections and vascular graft infections due to fistulae with the enteric or respiratory tract.
To date, though no siderophore-based radiotracers have been developed clinically, the FDA approved the siderophoreconjugated β-lactam cefiderocol to treat Gram-negative bacterial infections. 23Furthermore, the natural hydroxamate siderophore desferrioxamine B (DFO-B; Figure 1A) is already an approved drug for iron overload treatment.
Researchers have repurposed DFO-B by investigating its potential for imaging infections.It has turned out to be a broad-spectrum radiotracer for imaging bacterial and fungal infections. 24,25A phase I/IIa clinical trial using [ 68 Ga]Ga-DFO-B for imaging the upper/lower respiratory tract or orthopedic bacterial infections is ongoing in Europe (EudraCT no.2020-002868-31).Another small observational clinical study with [ 68 Ga]Ga-DFO-B is being conducted to image vascular graft infections in the UK (ID: NCT05285072).
Schizokinen (SKN) is a citrate-based hydroxamate-siderophore produced by the Gram-positive bacterium Bacillus megaterium (Figure 1B).SKN is a low-molecular-weight (420 Da) negatively charged compound.SKN has been reported to possess a higher affinity for Fe 3+ than DFO-B (pFe 3+ value for SKN, 26.8 > DFO-B, 25.0). 26,27In vivo studies of SKN revealed low bioavailability via the oral route, suggesting that SKN might not be a suitable candidate for iron overload treatment, 26 similar to DFO-B.An earlier study characterized the coordination chemistry of gallium(III)-SKN and revealed the structural resemblance of Ga(III)-SKN to the iron analog Fe(III)-SKN. 28Therefore, owing to its comparable affinity for Fe 3+ , SKN deserves evaluation as an infection-specific PET tracer.
In this study, we established radiolabeling of SKN with 68 Ga and evaluated the in vitro uptake of this complex by Grampositive (S. epidermidis, S. aureus) and Gram-negative (P.aeruginosa, E. coli) bacteria and fungi (Candida albicans, C. glabrata, and Aspergillus f umigatus).Subsequently, we performed PET/CT imaging and determined the in vivo pharmacokinetics in a healthy animal model.As far as we are aware, our study is the first report on [ 68 Ga]Ga-SKN in terms of radiolabeling and characterization, microbial uptake specificity, in vivo animal imaging by PET/CT, and ex vivo biodistribution (BioD).
We also assessed the stability of [ 68 Ga]Ga-SKN in terms of transchelation in the presence of an excess of the competitive chelator diethylenetriamine pentaacetate (DTPA; 6 mM).The results show that after 5 min, the radiotracer was ∼90% intact, glabrata, and A. f umigatus after 45 min of incubation (as %AD).Each experiment was performed in triplicate (mean ± SD).P value < 0.05 was significant "*," and "ns" means nonsignificant.and after 60 min, it was ∼50% intact (Figure S5).Our findings are similar to those reported for [ 68 Ga]Ga-Ornibactin, 35 which was found 80% intact after 30 min and 70% intact after 60 min, and [ 68 Ga]Ga-DFO-B, 24 which was 94% intact after 30 min and 85.3% intact after 60 min in the presence of 6 mM DTPA.However, the pH of the reaction for our DTPA study was not adjusted to a neutral value, resulting in an acidic environment.This is particularly important for [ 68 Ga]Ga-SKN because SKN is pH-sensitive. 29,36Nevertheless, the transchelation study shows the complex is not inert to ligand exchange, although in vivo, the main competing ligand would be transferrin, not DTPA. 32Future work should investigate the incubation of the tracer at neutral pH in the presence of DTPA and at different time points in both DTPA and human serum.Furthermore, any breakdown products or metabolites of SKN will also need to be investigated, as well as their potential toxicity.
We evaluated the uptake of [ 68 Ga]Ga-SKN by bacterial and fungal strains.We confirmed the stability of [ 68 Ga]Ga-SKN in media by iTLC (alike Figure S1) before the uptake experiment.Previous studies have predicted the utilization of xenosiderophore SKN in S. aureus based on the presence of siderophore receptors. 37Accordingly, our study confirmed the high uptake of [ 68 Ga]Ga-SKN by both S. aureus and S. epidermidis.The literature shows that E. coli K-12 is devoid of SKN-specific outer membrane receptors and that there is no in vitro utilization of SKN by this strain. 38However, high similarity in amino acid sequences between the Fe 3+ -aerobactin IutA transporter protein present in E. coli and the outer membrane transport protein Fe 3+ -SKN SchT in Anabaena sp.PCC 7120 has been previously reported. 39Thus, E. coli harboring aerobactin transporter proteins (E. coli may produce multiple siderophores, including enterobactin, aerobactin, yersiniabactin, and salmochelin) 40 may also transport SKN.
[ 68 Ga]Ga-SKN uptake by the E. coli strain may be due to this intrastrain variability.The utilization of SKN in P. aeruginosa via the inner membrane protein FoxB, which is responsible for ferrichrome, ferrioxamine B, and SKN transport, has already been reported. 38. aureus and S. epidermidis (Gram-positive) exhibited higher uptake (more than 100-fold in the case of S. epidermidis) than did E. coli and P. aeruginosa (Gram-negative) under irondepleted conditions (Figure 3A), showing Gram-positive species selectivity of [ 68 Ga]Ga-SKN.A previous study with [ 68 Ga]Ga-DFO-B showed its higher uptake in S. aureus and S. agalactica compared to P. aeruginosa, and the level of uptake was variable among strains of the same bacterial species due to the STs upregulation affected by inherent genetic determinants or environmental cues.24 Here, we observed higher in vitro uptake of [ 68 Ga]Ga-SKN in S. epidermidis than in S. aureus.
The uptake specificity of [ 68 Ga]Ga-SKN in both Gramnegative and Gram-positive bacteria was determined by adding a blocking agent at excess concentrations, such as Fe-ENT, Fe-PVD, or Fe-SKN (Figure 3A).It was determined that these blocking agents were able to block up to 23% of [ 68 Ga]Ga-SKN uptake in E. coli (Fe-ENT), 94% in P. aeruginosa (Fe-PVD), 30% in S. aureus (Fe-SKN), and 40% in S. epidermidis (Fe-SKN).It is speculated that this uptake could be blocked to a greater extent in E. coli and S. aureus with higher concentrations of the blocking agent in excess.
The cell viability after incubation with [ 68 Ga]Ga-SKN was minimally affected compared to the control for all bacterial strains except for S. epidermidis (Figure S6).For S. epidermidis, a 1.7-fold reduction was observed.This inhibition is likely to occur because of intracellular iron deprivation (due to ironlimited condition), which could otherwise be used in metabolism. 41Additionally, literature shows that coagulasenegative staphylococci (such as S. epidermidis) are more susceptible to iron deprivation. 41,42Interestingly, viability was improved (about a 2-fold increase) in the presence of excess cold Fe-SKN in S. epidermidis.This may be due to the partial rescue of iron-deprived cells.
As [ 68 Ga]Ga-SKN uptake was higher in P. aeruginosa and S. epidermidis, further investigations were performed to determine the role of the active transport system in this uptake.A previous study used NaN 3 (2 mM) to block [ 68 Ga]Ga−PVD-PAO1 uptake by P. aeruginosa. 33NaN 3 is an inhibitor of ATP synthesis and affects microbial growth and metabolic activities. 43In vitro uptake blockade of S. epidermidis and P. aeruginosa by the active transport blocker sodium azide (NaN 3 ,) compared to the blocking agents is shown in the inset of Figure 3A.The results showed that 30 mM NaN 3 blocked [ 68 Ga]-Ga-SKN uptake by up to 87% in P. aeruginosa and up to 50% in S. epidermidis.The total viable count and raw radioactivity (cpm) are shown in Figure S7.Our findings indicate that [ 68 Ga]Ga-SKN taken up by P. aeruginosa and S. epidermidis is an active transport-mediated uptake process by metabolically active cells.
In vitro uptake in 96-well plates by fungal strains, including C. albicans, C. glabrata, and A. f umigatus, was performed to assess the microbial specificity of [ 68 Ga]Ga-SKN.Although yeast and mold cell types are different, the uptake results (% AD) presented here were obtained with the same volume of 180 μL of cells.For C. glabrata and C. albicans, 180 μL of sample had ∼1.8 × 10 6 to 10 7 CFU.The results showed almost no uptake by C. glabrata, C. albicans, and A. f umigatus (Figure 3B).Recent literature makes it evident that A. fumigatus is unable to utilize xenosiderophore, SKN, and, so far, the absence of SITs for SKN in Candida species. 44Hence, our results indicated that [ 68 Ga]Ga-SKN uptake is specific to clinically relevant bacterial species, which could distinguish between clinically relevant bacterial and fungal infections.
Encouraged by having shown that Ga 3+ is complexed by SKN, that the complex is stable at neutral pH, its short-term stability in the presence of excess DTPA and in human serum, that the complex is selectively taken up by certain microbial species, and the complex is expected to circulate in vivo before being renally excreted rapidly (based on the known behavior of other 68 Ga-siderophore complexes), 24,25,32 we further investigated in vivo pharmacokinetics of [ 68 Ga]Ga-SKN as a potential bacterial-specific radiotracer.MicroPET/CT imaging of healthy BALB/c mice injected with 68 Ga-labeled SKN revealed rapid clearance from the bloodstream, with major excretion by the renal system (Figure 4).This contrasts with the behavior of unchelated gallium, which shows prolonged circulation due to rapid binding to transferrin. 32Minimal retention was observed in the blood and other organs.Region of interest (ROI) data from PET imaging over 60 min are shown in Figure S8, which shows rapid circulation in the body and gradual accumulation of the radioactivity in the bladder.These data suggest that the biological half-life of [ 68 Ga]Ga-SKN is ∼7 min, which is similar to other published 68 Ga siderophores. 24,45he ex vivo BioD values were in accordance with the data obtained from PET/CT imaging 60 min postinjection (p.i.).Rapid bloodstream clearance of [ 68 Ga]Ga-SKN was observed, with the renal system accounting for the majority of its excretion (Figure 5).There was minimal retention in the blood and other organs at 60 min p.i.The highest activity after 60 min was found in the kidneys (excluding bladder as gradual accumulation occurred there), at 5.37 ± 0.85% of the injected dose per gram of organ (ID/g).Similar findings were also highlighted by the iron−SKN complex in vivo, where iron-SKN was rapidly cleared from the bloodstream. 26−35 Radio-HPLC and radio-iTLC analysis of urine showed that [ 68 Ga]Ga-SKN was excreted largely in the urine sample collected 60 min p.i., and there is some presence of unchelated gallium-68, which indicates it may have been dissociated from the [ 68 Ga]Ga-SKN over the time (Figure S9).This result is similar to [ 68 Ga]Ga-DFO-B, 32 which has been shown to be a broad-spectrum infection-specific radiotracer to detect Grampositive, Gram-negative, and fungal pathogens in relevant lung and muscle animal infection models.Therefore, we envisage that [ 68 Ga]Ga-SKN would be able to detect pathogens in in vivo infection models.
Direct imaging of live bacterial or fungal pathogens can reveal local infections in the body without the cumbersome and invasive direct sampling used in traditional slowthroughput infection diagnostics.In situ identification of causative pathogens will aid in distinguishing between bacterial and fungal pathogens, reduce the turnaround time to prescribe appropriate antimicrobial treatment, and improve patient outcomes.Moreover, monitoring the success of antimicrobial treatments will aid in addressing the AMR crisis.
We studied [ 68 Ga]Ga-SKN radiosynthesis, its PBS stability, transchelation, short-term stability in human serum, and in vitro uptake in bacterial and fungal cells, which demonstrated bacterial specificity.Further PET/CT imaging and ex vivo BioD in healthy animals demonstrated the rapid clearance of this radiotracer from the blood by the renal system with very little retention in the major organs and blood circulation.In vivo healthy animal urine analysis shows ∼22% stability of [ 68 Ga]Ga-SKN after 60 min p.i. collection, where free or unchelated 68 Ga-acetate is present, similar to other 68 Gasiderophore studies.The limitations of the study are the shortterm in vitro human serum stability test and ∼50% stability after 60 min in the presence of transchelator DTPA.However, based on published 68 Ga siderophores' in vitro and in vivo behaviors and their ability to detect pathogens, this brief period should be enough to reach the site of infection and accumulate inside microbial cells.Therefore, future work will include further investigation of its in vitro stability in human serum at different time points, a pH-adjusted transchelation assay, in vivo evaluation of [ 68 Ga]Ga-SKN to detect bacterial infections (and to distinguish between bacterial and fungal infections) in clinically relevant infection models in small animals, and whether further chemical modifications are needed to alter its PK.

■ MATERIALS AND METHODS
All chemicals, culture media, and analytical grade reagents were purchased from Sigma−Aldrich (UK; unless otherwise stated).Siderophores, including desferrioxamine B (DFO-B), pyoverdine (PVD), and enterobactin (ENT), were purchased from Sigma−Aldrich (UK).Schizokinen (SKN) was isolated and purified from Bacillus megaterium 26 and/or obtained via chemical synthesis. 27icrobial strains were purchased commercially, except for Candida albicans (ATCC 90028), Candida glabrata (ATCC 90030), and Aspergillus f umigatus (ATCC 46640), which were kindly provided by Dr. Silke Schelenz, KCH Clinical Lead Infection Sciences, King's College Hospital, London, UK.All of the strains and their growth media and temperatures, assay media, and blocking agents are listed in Table S1.Methods for the microbial uptake assay are described in the Supplementary Methods.

■ ANALYTICAL METHODS
Instant Thin Layer Chromatography (iTLC) was performed with a glass microfiber paper strip that is impregnated with silica gel (SG; Agilent Technologies, UK; mobile phase: 10% ammonium acetate, 30:70% water/methanol, or 50% water/ methanol).Post-iTLC, strips were subjected to scanning using a Raytest Rita-Star TLC scanner with a positron (β+) detector  (LabLogic, UK) and analyzed using Laura software (LabLogic, UK) or a Cyclone Plus Phosphor Imager (PerkinElmer).An Agilent Eclipse XDB C18 (5 μm 4.6 × 150 mm in diameter) reversed phase (RP) column was used for high-performance liquid chromatography (HPLC) along with UV detection at 220 nm (Gina Star TM software version 5.8).Further analysis was performed using Laura software (LabLogic, UK).Details of the composition of the mobile phase and gradients for RP-HPLC are mentioned in the Supplementary Methods.Radioactivity was measured with a gamma counter (LKB Wallac 1282 CompuGamma Gamma Counter or a Perki-nElmer 3470 Wizard2 Gamma Counter).
■ RADIOSYNTHESIS OF [ 68 Ga]Ga-SKN A total of 5 mL of 0.1 M ultrapure HCl was introduced to a 68 Ge/ 68 Ga generator (Eckert and Ziegler Eurotope GmbH, Berlin, Germany) to elute 68 GaCl 3 in five tubes following the fractionated elution approach.
Radiolabeling with 68 Ga was first optimized with variable amounts (μg) of SKN under different reaction conditions (pH, temperature, etc.).Briefly, 10−50 μg of SKN (1−4 μg/μL in water) was mixed with 60 μL of 3.6 M sodium acetate and 100−200 μL of eluted 68 GaCl 3 (10−120 MBq) and incubated at room temperature for approximately 10−15 min.The pH of the reaction mixtures was then adjusted to 6−7 with the addition of extra sodium acetate.The radiochemical purity (RCP) of [ 68 Ga]Ga-SKN was confirmed using instant thinlayer chromatography (iTLC), as described in the Analytical Methods section.
■ CHARACTERIZATION OF [ 68 Ga]Ga-SKN The distribution coefficient was performed following the shake flask method. 32The PBS, human serum, and DTPA stability of [ 68 Ga]Ga-SKN were determined via RP-HPLC.All of these protocols are described in the Supplementary Methods.

■ ANIMAL EXPERIMENTAL DESIGN
An in vivo animal study was performed following the Animals (Scientific Procedures) Act, 1986.Appropriate project and personal licenses were approved by the UK Home Office.King's College London Animal Welfare and Ethical Body approved these animal experiment protocols.Healthy female Balb/b mice (10−11 weeks, 10−12 g) were purchased from Charles River UK Ltd.The experiment was designed with four mice to evaluate the in vivo PK and ex vivo BioD of the radiotracer.Mice were intravenously injected with [ 68 Ga]Ga-SKN (110−120 μL in PBS, 4−6 MBq, and 5.5−6.0 μg of SKN per animal) and filter sterilized with Millex-LG 0.20 μm prior to injection.

■ PET/CT IMAGING
The NanoScan PET/CT (Mediso Medical Imaging Systems) was employed for dynamic PET scanning and CT scanning. 32ll of the images were rebuilt by Tera-Tomo (Monte Carlobased full 3D iterative software).All of these rebuilt images were analyzed by VivoQuant 1.21 software (inviCRO), and quantification for regions of interest (ROIs) was performed for specific mouse tissues.Mice were anesthetized with isoflurane (1.0−1.5 L/min oxygen flow rate and 2−2.5% isoflurane) prior to cannulation of the tail vein, after which a CT scan was performed.Later, the radiotracer was injected as stated above, and a dynamic PET scan was performed for 1 h.

Figure 1 .
Figure 1.Chemical structure of desferrioxamine B and schizokinen.