Aluminum Fluoride-18 Labeled Mannosylated Dextran: Radiosynthesis and Initial Preclinical Positron Emission Tomography Studies

Purpose In addition to being expressed on liver sinusoidal endothelial cells, mannose receptors are also found on antigen-presenting cells, including macrophages, which are mainly involved in the inflammation process. Dextran derivatives of various sizes containing cysteine and mannose moieties have previously been labeled with 99mTc and used for single-photon emission computed tomography imaging of sentinel lymph nodes. In this study, we radiolabeled 21.3-kDa D10CM with positron-emitting 18F for initial positron emission tomography (PET) studies in rats. Procedures D10CM was conjugated with 1,4,7-triazacyclononane-1,4,7-triacetic acid (NOTA) chelator and radiolabeled with the aluminum fluoride-18 method. The whole-body distribution kinetics and stability of the intravenously administered tracer were studied in healthy male Sprague-Dawley rats by in vivo PET/CT imaging, ex vivo gamma counting, and high-performance liquid chromatography analysis. Results Al[18F]F-NOTA-D10CM was obtained with a radiochemical purity of >99% and molar activity of 9.9 GBq/μmol. At 60 minutes after injection, an average of 84% of the intact tracer was found in the blood, indicating excellent in vivo stability. The highest radioactivity concentration was seen in the liver, spleen, and bone marrow, in which mannose receptors are highly expressed under physiological conditions. The uptake specificity was confirmed with in vivo blocking experiments. Conclusions Our results imply that Al[18F]F-NOTA-D10CM is a suitable tracer for PET imaging. Further studies in disease models with mannose receptor CD206-positive macrophages are warranted to clarify the tracer’s potential for imaging of inflammation. Graphical abstract Supplementary Information The online version contains supplementary material available at 10.1007/s11307-023-01816-7.


Introduction
New carbohydrate-derived radiopharmaceuticals have recently reached major clinical breakthroughs.For example, 99m Tc-Tilmanocept (Lymphoseek®) has been approved for clinical use by the U.S. Food and Drug Administration and the European Medicines Agency.99m Tc-Tilmanocept (17 kDa) is a mannosylated dextran developed for imaging of sentinel lymph nodes with single-photon emission computed tomography [1].Indeed, dextran is a platform molecule that has been used in many types of pharmaceutical applications, and mannose is useful for decoration of drug molecules to achieve specific targeting.The attachment of mannose moieties to the dextran backbone may direct the radiotracer to the carbohydrate-recognition domain of C-type lectin mannose receptor (cluster of differentiation 206, CD206) on the surface of macrophages and dendritic cells.A study by Núñez and coworkers showed that uptake of mannosylated dextran in lymph nodes was four times higher than uptake of corresponding radiolabeled dextran without mannosylation [2].In a recent report, Pirmettis and coworkers labeled a mannosylated dextran with 99m Tc(CO) 3 using a cysteine moiety as a chelator.The generated compound, 99m Tc(CO) 3 -DCM20, has been studied for the detection of sentinel lymph nodes in preclinical settings, and promising results have been obtained on single-photon emission computed tomography [3].Our research focuses on the development of radiopharmaceuticals for positron emission tomography (PET) imaging, and we have radiolabeled many types of molecules, including oligosaccharides for mannose receptor targeting [4].In this study, the aluminum fluoride-18 (Al[ 18 F]F) labeling technique was used to radiosynthesize a 21.3 kDa dextran containing cysteine and mannose moieties (D10CM), which we refer to as Al[ 18 F]F-NOTA-D10CM, where NOTA denotes the chelator 1,4,7-triazacyclononane-1,4,7-triacetic acid used for attaching Al[ 18 F]F to the structure.In addition, we report a preclinical evaluation study of Al[ 18 F]F-NOTA-D10CM in healthy rats.

Preparation of NOTA-D10CM
NOTA-NHS ester (117 mg) dissolved in 2 mL of dimethyl sulfoxide was added to a solution of D10CM (100 mg) in 4 mL of borate buffer (0.1 M, pH 9.0).The reaction mixture was stirred at room temperature for 18 hours and then transferred to an ultrafiltration cell (model 8050, Millipore Corp, Bedford, MA) fitted with an ultrafiltration membrane (YM03).The volume was adjusted to 50 mL with deionized water, and the solution was then concentrated to 5 mL by applying nitrogen gas under pressure directly to the ultrafiltration cell.The procedure was repeated four times, and the residue was finally lyophilized to give the product as a white solid (100 mg).

Preparation of AlCl 3 Stock Solution in Acetate Buffer
The stock solution of AlCl 3 (2 mM) was prepared in acetate buffer (pH 4.0, 1 M).The acetate buffer was prepared by dissolving sodium acetate (8.2 g, 100 mmol) in TraceSELECT water.The pH was carefully adjusted to 4.0 with acetic acid and magnetic stirring, and the final volume was adjusted to 100 mL.Aluminum chloride hydrate (AlCl 3 6.5H 2 O, 50.1 mg, 200 μmol) was added to the acetate buffer.The stock solution was divided into aliquots and stored at 4°C.

Quality Analysis of Al[ 18 F]F-NOTA-D10CM
The radiochemical purity of the end product sample (0.3-0.8 MBq) was determined using an HPLC system (Hitachi; Merck, Darmstadt, Germany) equipped with a Jupiter Proteo column (reversed-phase C18, 150 × 4.6 mm, 300 Å; Phenomenex) and a radioactivity detector (Radiomatic 150TR flow-through, Packard, Meriden, CT).Solvent A was water containing 0.1% TFA, and solvent B was acetonitrile containing 0.1% TFA.In the elution protocol, the content of solvent B was increased from 8% to 40% over 0-9 minutes and then maintained at 40% over 9-10 minutes at a flow rate of 1 mL/minute.The stability of Al[ 18 F]F-NOTA-D10CM in the final product formulation buffer was assessed by HPLC analysis of aliquots taken at 1, 2, 3, and 4 hours after the end of synthesis (EOS).

LogD Measurement
PBS (600 μL) and 1-octanol (600 μL) were added to a 1.5 mL Eppendorf tube followed by the addition of 5 kBq of Al[ 18 F]F-NOTA-D10CM.The mixture was vortexed for 5 minutes, and then the water and organic phases were separated by centrifugation at 14,100 ×g for 3 minutes at room temperature.A 400 μL aliquot was taken from both phases for gamma counting (Wizard 3″, PerkinElmer/Wallac, Turku, Finland).The measurements were performed in triplicate.The LogD value was calculated using the formula: .

Determination of Molar Activity
The molar activity was determined by means of HPLC analysis using a wavelength of 220 nm under UV detection.Five NOTA-D10CM samples at concentrations of 1, 2, 4, 6, and 8 μM were analyzed in triplicate and used to generate a calibration curve.The HPLC method was the same as that described above for quality control.Then, 50 μL samples of radioactivity-decayed final product were analyzed by HPLC, and the concentration (c, nmol/ mL) of the product was calculated using the calibration curve.Finally, the molar activity at EOS was calculated as follows:

PET Studies
Healthy male Sprague-Dawley rats (n = 6, body weight 268.3 ± 31.8 g, age 7-8 weeks) and smaller sized rats of the same strain for the in vivo blocking study (n = 4, body weight 94.9 ± 3.7 g, age 4 weeks) were obtained from the Central Animal Laboratory, University of Turku, Turku, Finland.The PET/computed tomography (CT) imaging was acquired using an Inveon Multimodality scanner (Siemens Medical Solutions, Knoxville, TN).Rats were anesthetized with isoflurane (4-5% for induction and 1.5-2% for maintenance), and the tail vein was cannulated.A CT scan was first performed for attenuation correction and anatomical reference, and then a 60 minute PET acquisition was started at the time of intravenous injection of Al[ 18 F]F-NOTA-D10CM (51.6 ± 1.8 MBq, 431.1 ± 233.4 μL).To perform the blocking experiment, 100-fold molar excess of mannan from Saccharomyces cerevisiae (Sigma-Aldrich M7504)) in saline (400 μL) was injected 15 minutes before the Al[ 18 F]F-NOTA-D10CM injection.PET data obtained in a list-mode were reconstructed with an ordered subsets expectation maximization 3-dimensional (OSEM-3D) algorithm into 30 × 3 s, 9 × 10 s, 4 × 30 s, 5 × 60 s, and 10 × 300 s time frames.After the PET/CT imaging, animals were euthanized, and various tissues were collected, weighed, and measured with a gamma counter (Triathler 3″, Hidex, Turku, Finland).The radioactivity of the excised tissues was expressed as a percentage of the injected radioactivity dose per gram of tissue (%ID/g), corrected for the radioactivity remaining in the cannula and tail.
PET/CT image analysis was performed using Carimas 2.10 software (Turku PET Centre, Turku, Finland, www.turku petce ntre.fi/ carim as/).Regions of interest (ROIs) were manually defined in main organs using CT for anatomical reference.Time-activity curves were extracted from the 60 minute PET data and expressed as standardized uptake value (SUV) versus time after injection.

Statistical Analysis
Results are expressed as mean ± standard deviation (SD).The significance of differences between the groups was determined with unpaired Student's t tests.P-values <0.05 were considered statistically significant.

Synthesis of NOTA-D10CM
The mannosylated dextran D10CM was prepared according to the published method [3,6].NOTA-D10CM was synthesized in a conjugation reaction between D10CM and the NOTA-NHS in borate buffer at pH 9. The reaction was accomplished over 18 hours at room temperature followed by product purification by ultrafiltration with a 3 kDa cutoff.NOTA-D10CM was obtained as a sound white powder after lyophilization and subjected to analysis with nuclear magnetic resonance.Results revealed that NOTA-D10CM had 7 free and 17 mannosylated S-derivatized cysteines with a minimum of one NOTA chelator for each D10CM molecule (Supplementary Fig. 1).

Preparation of Al[ 18 F]F-NOTA-D10CM
We produced [ 18 F]fluoride in an 18 O(p,n) 18 F nuclear reaction using H 2 [ 18 O]O as a target in the cyclotron.The radiosynthesis was performed with a remote-controlled radiosynthesis device [7].An anion-exchange cartridge was used to extract [ 18 F]fluoride from the target water, and physiological saline was used to elute [ 18 F]fluoride from the cartridge to the reaction vessel.We used saline instead of a base for the [ 18 F]fluoride elution to avoid the need to adjust the pH in the reaction mixture, which improved practicality.In typical cases, the reaction mixture contained 6.8 nmol NOTA-D10CM precursor in sodium acetate buffer at pH 4.0 with acetonitrile as cosolvent.After heating, the product was isolated by HPLC followed by solid phase extraction.The final product was formulated in PBS containing ascorbic acid, in which the ethanol content was less than 10% by volume.It was essential to have ascorbic acid in the formulation to prevent radiolysis.The total radiosynthesis time was 91.8 ± 15.7 minutes (n = 10) from the end of the bombardment.The radioactivity yield was 790.0 ± 339.9 MBq (12.9% ± 5.4) with radioactivity concentration of 312.2 ± 140.7 MBq/ mL (n = 10) starting from 6.1 ± 0.6 GBq of [ 18 F]fluoride, and the molar activity was 9.9 ± 4.6 GBq/μmol (n = 4) at the EOS.The decay-corrected radiochemical yield was 24.8% ± 13.1.Quality control of the final product was performed with HPLC, and the retention time of Al[ 18 F]F-NOTA-D10CM on a C18 reversed-phase column was 6.2 ± 0.4 minutes (Fig. 1b, c).
Al[ 18 F]F-NOTA-D10CM was found to be stable over a 4 hour shelf-life test, with consistent radiochemical purity of >99%.The LogD value was −3.10, indicating that the tracer was highly hydrophilic, which is in line with the nature of oligosaccharides.
To confirm the observations from the in vivo imaging, ex vivo gamma counting of the excised tissues was carried out.The ex vivo results were in line with the in vivo PET/ CT data, with %ID/g being the highest in liver (4.44 ± 1.32), spleen (2.40 ± 0.45), and bone marrow (1.83 ± 0.95) (Supplementary Table 1).To confirm the specificity of Al[ 18 F]F-NOTA-D10CM uptake, in vivo blocking experiments with mannan were performed.Mannan was chosen as a blocking agent on the basis of previous reports describing mannose receptor blockage experiments [12].According to in vivo PET/CT, the blocking resulted in significantly reduced Al[ 18 F]F-NOTA-D10CM uptake in CD206-rich tissues: an average reduction of 88.6% in liver, 35.1% in spleen, and 59.9% in bone marrow (Fig. 4, Supplementary Table 1).As expected, the dramatically reduced uptake in the liver, spleen, and bone marrow resulted in prolonged Al[ 18 F]F-NOTA-D10CM in the blood circulation and high renal radioactivity in the blocking experiments.However, despite renal excretion, portion of the excreted radioactivity was still observed in the liver.These changes in biodistribution were also clearly visible on whole-body PET/CT images (Fig. 2).
The in vivo stability of Al[ 18 F]F-NOTA-D10CM was studied in rats by taking serial blood samples up to 60 minutes after injection followed by HPLC analysis of protein-free plasma under radioactivity detection.The intact tracer accounted for 100% ± 0, 96.2% ± 1.8%, 91.6% ± 3.0%, 90.1% ± 5.3%, and 84.1% ± 3.7% of total plasma radioactivity at 5, 15, 30, 45, and 60 minutes after injection, respectively, indicating excellent in vivo stability (Fig. 5).We also measured the distribution of radioactivity among the different components of the blood: on average, 7.2% ± 4.0% of the blood radioactivity was found in cells and 92.8% ± 4.0% in plasma, while 66.3% ± 6.5% of plasma radioactivity was free-form, and 33.7% ± 6.5% was bound to proteins.

Discussion
In this work, we describe the Al[ 18 F]F-labeling of NOTA-conjugated mannosylated dextran derivative and demonstrate its accumulation in mannose receptor-rich tissues by in vivo PET imaging.To our knowledge, this is the first time that the mannosylated dextran has been labeled with Al[ 18 F]F-labeling technique.Accordingly, facile incorporation of 18 F into NOTA-D10CM and promising first PET studies encourage further studies.
In a previous study by Pirmettis and coworkers, dextran (molecular weight 11.8 kDa) was used as a backbone, to which a mannose moiety for targeting and a cysteine moiety for chelation of 99m Tc were attached resulting DCM20 with 24 units of mannose per mol of dextran (22.3 kDa) [3].The 99m Tc-labeled DCM20 was intended for CD206targeted sentinel lymph node mapping by subcutaneous or intradermal injection.Further studies by Papasavva and coworkers compared a series of different size mannosylated dextrans (21.3-805.6 kDa) with 19-645 units of mannose.Their results revealed that 99m Tc-labeled D75CM with 74 mannose units (111.2kDa) was superior because of the highest sentinel lymph node accumulation via subcutaneous injection and accumulation in the liver after intravenous administration were observed [5].Due to excellent imaging performance and favorable biokinetics, we kept the molecule structure as similar as possible in this study.Accordingly, we did not omit the cysteine moiety, even though it is not needed for the purpose of 18 F-labeling in the current situation.In this study, we aimed to reveal the potential of smaller size D10CM (21.3 kDa) with 17 mannose units The precursor compound NOTA-D10CM was conveniently synthesized by N-acylation of the amino group of the cysteine residues under slightly alkaline conditions, and purified by ultrafiltration.The easy availability of the precursor facilitates its potential use in nuclear medicine, including PET imaging.Among the available 18 F-fluorination methods, Al[ 18 F]F-labeling is one of the To fix Al[ 18 F]F into a radiopharmaceutical, several good chelators, including NOTA, restrained complexing agent (RESCA), 1,4,7-triazacyclononane-1,4-diacetate (NODA), and 1,4,7-triazacyclononane,1-glutaric acid-4,7-acetic acid (NODAGA) have been developed [14].Among them, NOTA is probably the most often used chelator in the clinical setting.The Al[ 18 F]F-NOTA complex has shown excellent stability in vivo, not only in animals but also in humans [15,16].To our knowledge, there are no clinical safety issues associated with the moiety of Al[ 18 F]F-NOTA in the radiopharmaceutical used for imaging purposes.Therefore, we decided to use Al[ 18 F]F-NOTA chelation chemistry for the labeling of D10CM (Fig. 1).Indeed, the radiosynthesis protocol was rather reproducible and robust.Although the decay-corrected radiochemical yield of ~25% was only moderate, the radioactivity yield of ~800 MBq from a starting activity of 6 GBq of [ 18 F]fluoride was adequate for preclinical PET applications.
We noticed that it was important to store the 0.1 M acetate buffer in small aliquots, because frequent opening the vial caused change in pH due to evaporation, which contributed to the failure of the radiosynthesis.We also observed that a high concentration (>50%) of organic solvents (e.g., ethanol and acetonitrile) caused precipitation of Al[ 18 F]F-NOTA-D10CM.Therefore, we kept the concentration of organic solvents below 40% in chromatographic experiments, reaction mixtures, and solid phase extraction.For example, in the step involving elution of Al[ 18 F]F-NOTA-D10CM from the solid phase extraction tC18 cartridge to the end product vial, it was critical to use 30% ethanol instead of the 50-100% that we usually use for other radiopharmaceuticals.
Mannosylated dextran derivatives have also been successfully 18 F-labeled by covalent bond formation.For example, Ting and coworkers have applied boron-based 18 F-fluorination method to prepare a [ 18 F]BOMB Lymphoseek with the molar activity of up to 1.85 GBq/μmol [17].For comparison, the molar activity of our Al[ 18 F]F-NOTA-D10CM was 10 GBq/μmol, which is moderate.
Ultimately, the Al[ 18 F]F-NOTA-D10CM is intended for imaging of mannose receptor (CD206) expression.However, as a first step, we performed PET studies in healthy rats to evaluate its biodistribution and in vivo stability.Indeed, our results revealed that vast majority of Al[ 18 F]F-NOTA-D10CM had cleared from blood circulation within 20 minutes and accumulated in liver, spleen and bone marrow rich in mannose receptor-positive cells (Fig. 3 and 4, Supplementary Table 1).Furthermore, we found that the in vivo stability of Al[ 18 F]F-NOTA-D10CM was excellent, with approximately 84% of total plasma radioactivity originating from intact Al[ 18 F]F-NOTA-D10CM at 60 minutes after i.v.injection (Fig. 5).Interestingly, our results are in line with the study by Choi and coworkers, in which intravenously administered 68 Ga-labeled mannosylated human serum albumin showed a similar biodistribution trend especially in the liver, spleen and bone marrow [18], i.e. organs associated with the reticuloendothelial system.To confirm the specificity of Al[ 18 F]F-NOTA-D10CM uptake, an in vivo blocking study with molar excess of mannan was performed.Indeed, the uptake in liver, spleen and bone marrow was significantly reduced.The blocking effect was clearly seen not only in the ex vivo tissue radioactivity measurements, but also in in vivo PET images.When the uptake is blocked in the mannose receptor-rich organs, increased amount of Al[ 18 F]F-NOTA-D10CM becomes available in the blood circulation and increased uptake into other tissues, including muscle, is observed.To evaluate the in vivo stability of Al[ 18 F]F-NOTA-D10CM in rats, blood samples were taken at five time points within one hour and protein-free plasma was isolated for metabolite analysis by HPLC.In order to precipitate plasma proteins, 10% sulfosalicylic acid had to be used instead of the commonly used organic solvent acetonitrile.This was to avoid precipitation of Al[ 18 F]F-NOTA-D10CM with a high concentration of organic solvents in the sample matrix, as described in above.

Conclusion
To conclude, the chelator-conjugated mannosylated dextran derivative NOTA-D10CM was synthesized to achieve efficient and straightforward labeling with Al[ 18 F]F chelation chemistry.The new radiopharmaceutical, Al[ 18 F]F-NOTA-D10CM, was produced with a high quality and moderate yield, and presented excellent in vitro and in vivo stability in rats.Al[ 18 F]F-NOTA-D10CM was rapidly cleared from the blood circulation and showed the highest uptake in the macrophagerich organs of liver, spleen, and bone marrow.In vivo blocking experiments established that the uptake is specific and in line with high mannose receptor expression.On the basis of these favorable characteristics, we consider that further translational studies in animal disease models are warranted.
A m = Total radioactivity at EOS (MBq) c nmol mL × volume (mL) of final product at EOS All animal experiments were approved by the national Project Authorization Board in Finland (license number ESAVI/43134/2019) and carried out in compliance with the EU Directive 2010/EU/63 on the protection of animals used for scientific purposes.

Fig. 1 Fig. 2
Fig. 1 Chemical structure and quality control of Al[ 18 F]F-NOTA-D10CM.(a) In the structure, the dextran backbone is indicated in pink, cysteine in gray, NOTA chelator in blue, and mannose moiety in green.Representa-

Fig. 3
Fig.3Representative micrographs of liver, spleen and bone marrow from a healthy male Sprague-Dawley rat stained with anti-mannose receptor antibody.Arrows denote liver sinusoidal endothelial

Fig. 5
Fig. 5 In vivo stability of intravenously administered Al[ 18 F]F-NOTA-D10CM in rats.(a) Percentage of intact tracer from total plasma radioactivity at different time points after injection.Lines represent means, and error bars denote standard deviation (n = 3-4 for each time point).(b) Representative radio-HPLC chromatograms of intact Al[ 18 F]F-NOTA-D10CM and (c) rat plasma at 60 minutes postinjection