Comparison of Al¹⁸F- and ⁶⁸Ga-labeled NOTA-PEG₄-LLP2A for PET imaging of very late antigen-4 in melanoma

Abstract

Malignant melanoma is an aggressive cancer with poor prognosis. Very late antigen-4 (VLA-4) is overexpressed in melanoma and many other tumors, making it an attractive target for developing molecular diagnostic and therapeutic agents. We compared Al¹⁸F- and ⁶⁸Ga-labeled LLP2A peptides for PET imaging of VLA-4 expression in melanoma. The peptidomimetic ligand LLP2A was modified with chelator 2-S-(4-isothiocyanatobenzyl)-1,4,7-triazacyclononane-1,4,7-triacetic acid (p-SCN-Bn-NOTA), and the resulting NOTA-PEG₄-LLP2A peptide was then radiolabeled with Al¹⁸F or ⁶⁸Ga. The two labeled peptides were assayed for in vitro and in vivo VLA-4 targeting efficiency. Good Al¹⁸F and ⁶⁸Ga radiolabeling yields were achieved, and the resulting PET tracers showed good serum stability. In the in vivo evaluation of the B16F10 xenograft mouse model, both tracers exhibited high accumulation with good contrast in static PET images. Compared with ⁶⁸Ga-NOTA-PEG₄-LLP2A, Al¹⁸F-NOTA-PEG₄-LLP2A resulted in relatively higher background, including higher liver uptake (1 h: 20.1 ± 2.6 vs. 15.3 ± 1.7 %ID/g, P < 0.05; 2 h: 11.0 ± 1.2 vs. 8.0 ± 0.8 %ID/g, P < 0.05) and lower tumor-to-blood ratios (2.5 ± 0.4 vs 3.3 ± 0.5 at 1 h, P<0.05; 5.1 ± 0.9 vs 7.3 ± 0.6 at 2 h, P<0.01) at some time points. The results obtained from the mice blocked with unlabeled peptides and VLA-4 negative A375 xenografts groups confirmed the high specificity of the developed tracers. Despite the relatively high liver uptake, both Al¹⁸F-NOTA-PEG₄-LLP2A and ⁶⁸Ga-NOTA-PEG₄-LLP2A exhibited high VLA-4 targeting efficacy with comparable in vivo performance, rendering them promising candidates for imaging tumors that overexpress VLA-4.

Graphical Abstract

Introduction

Melanoma is a very aggressive malignant tumor with an increasing incidence, especially among Caucasians, and has become one of the major health problems in Western countries [1]. It is one of the most devastating cancers, and is resistant to most treatments. Despite recent advances in the treatment of malignant melanomas, the overall 5-year-survival rate is < 10% [2]. Early diagnosis and accurate staging are important for the choice of appropriate therapeutic interventions to improve patient outcomes.

Noninvasive imaging techniques, including ¹⁸F-fluorodeoxyglucose-based positron emission tomography/computed tomography (PET/CT), play an increasingly critical role in the staging of melanoma and for assessing the response to treatment [3–5]. However, limitations of ¹⁸F-FDG-PET have been described in detecting distant small metastatic lesions, such as sentinel lymph node deposits in stages I and II of this cancer [6, 7]. In addition, the high accumulation of ¹⁸F-FDG in infectious and inflammatory lesions may mimic tumor deposits [8]. With the rapid growth of molecular imaging and precision medicine in the last decades, many radiopharmaceuticals that bind receptors have been developed for more specific imaging of melanoma [9]. Some, including radiolabeled benzamides that target melanin [10–13], α-melanocyte-stimulating hormone analogues that bind to the melanocortin-1 receptor (MC1R) [14–16], and peptides that bind to the integrin family [17–19], have exhibited promising results.

Very late antigen-4 (VLA-4, also called integrin α₄β₁) is a non-covalent, heterodimeric transmembrane receptor. It is overexpressed in melanoma, and plays a vital role in tumor progression, angiogenesis and metastasis by promoting adhesion and migration of tumor cells [20–22]. Enhanced VLA-4 expression has also been found in lymphomas, leukemias, osteosarcoma, and multiple myeloma [23]. Peptidomimetic LLP2A is a recently discovered VLA-4 targeting ligand with a very high affinity (IC₅₀ = 2pM) [24]. Many radionuclides, including ⁶⁴Cu [25–27], ⁶⁸Ga [27, 28], ¹⁸F [29, 30], ¹¹¹In [31], ^99mTc [32], and ¹⁷⁷Lu [33] have been successfully attached to this ligand for diagnosis and/or therapy of VLA-4-abundant tumors. Good images with high tumor-to-background contrasts have been obtained in different animal models [25–33]. In a mouse melanoma model, the ¹⁷⁷Lu-LLP2A bioconjugate showed promising therapeutic efficacy, especially when combined with immune checkpoint inhibitor therapy [34]. It also has shown great therapeutic potential in the treatment of melanoma metastases [33]. There has been clinical success with theranostic combinations of ⁶⁸Ga- and ¹⁷⁷Lu-labeled somatostatin ligands and prostate-specific membrane antigen (PSMA)-targeted ligands [35–39]. LLP2A-based diagnostic and therapeutic agents have also shown great potential in the treatment of VLA-4-overexpressing melanomas.

¹⁸F is the most commonly used radionuclide for PET with a favorable half-life (109.78 min), high imaging resolution, and good scalability to GBq levels via on-demand cyclotron production. It can be labeled to ligands via the attached 1,4,7-triazacyclononane-1,4,7-triacetic acid (NOTA) chelator by Al¹⁸F chelating. In addition to ¹⁸F, ⁶⁸Gallium (half-life = 67.71 min) has been increasingly considered as a promising radionuclide for routine clinical PET imaging. It is conveniently produced from an in-house ⁶⁸Ge/⁶⁸Ga generator, and has a very simple radiolabeling procedure. Both ¹⁸F and ⁶⁸Ga have their own intrinsic advantages and attract most of the interest in constructing peptide-based tracers. Few attempts have been made to develop ⁶⁸Ga- and ¹⁸F-BF₃-labeled LLP2A peptide for melanoma imaging [27–30], and there has been no direct comparison between ⁶⁸Ga- and ¹⁸F-labeled LLP2A. We prepared a new Al¹⁸F-labeled LLP2A using a one-step Al¹⁸F-chelation method, and assessed its performance in B16F10 (VLA-4-positive) and A375 (VLA-4-negative) mouse models. The results were compared with those obtained from ⁶⁸Ga-NOTA-PEG₄-LLP2A. Characterization and biological evaluations of the labeled peptides included LogP values, serum stability, cell uptake, cell internalization, PET imaging, and ex vivo biodistribution.

Materials and Methods

Reagents and Instruments.

All chemicals were obtained from J&K Chemicals (Beijing, China), or Sigma-Aldrich (St. Louis, MO, USA) and were used without further purification. The mass spectrometry spectra were recorded on Applied Biosystems Voyager DE-STR MALDI-TOF (USA). Radioactivity was quantified using a dose calibrator (CRC-15R, Capintec, Ramsey, NJ, USA) or gamma counter (2470 WIZARD; PerkinElmer, Waltham, MA, USA). For purification of peptides and analysis of the radiolabeled conjugates using two elution buffers (0.1 v% trifluoroacetic acid (TFA) in deionized water as elution buffer A and 0.1 v% TFA in acetonitrile as elution buffer B), high-performance liquid chromatography (HPLC) was performed on an LC-20AT system (Shimadzu Corporation, Tokyo, Japan) equipped with an SPD-20A UV/vis detector (Shimadzu) and a flow count radiation detector (Bioscan, Washington, DC, USA).

Peptide synthesis

LLP2A was prepared via standard 9-fluorenylmethoxycarbonyl (Fmoc)-based solid phase synthesis as reported previously [24]. Briefly, Fmoc-Lys(Dde)-OH was first attached to Rink amide resin followed by Fmoc-PEG₄-CH₂COOH, Fmoc-Ach-OH, Fmoc-Aad(tBu)-OH, N-α-Fmoc-Lys(Alloc)-OH, 2-(4–3-o-tolylureido)phenylacetic acid, and Trans-3-(3-pyridyl)acrylic acid. All coupling was performed in dimethylformamide (DMF) using O-(7-azabenzotriazol-1-yl)-N,N,N′,N′-tetramethyluronium hexafluorophosphate (HATU) (3 equivalents), 1-hydroxybenzotriazole (HOBt) (3 equivalents), and N,N-Diisopropylethylamine (DIEA, 9 equivalents) as coupling reagents. After all the coupling, the Dde protecting group was cleaved with 2% hydrazine in DMF (3 × 20 min), washed with DMF and dichloromethane (DCM), and cleaved from the resin by twice treating with TFA/Phenol/H₂O/triisopropylsilane (92.5/2.5/2.5/2.5) for 2 h at room temperature. The solution was filtered and the peptide was precipitated in cold diethyl ester. The crude peptide was purified on a semi-preparative C18 column (Sepax GP-C18, 250 × 10 mm) using HPLC. Fractions from 12.5 to 15 min were collected and lyophilized to obtain pure PEG₄-LLP2A. HPLC method: 0–3 min, 25% buffer B; 3–30 min, from 25% buffer B to 40% buffer B; flow rate, 4 mL/min.

Synthesis of NOTA-PEG₄-LLP2A

p-SCN-Bn-NOTA (1.9 mg) and DIEA (10 μL) were added to a solution of LLP2A-PEG₄ (2.0 mg) in DMF (0.2 mL). After reaction at room temperature for 4 h, 0.8 mL water was added, and followed by HPLC purification. 2.1 mg product was obtained after lyophilization (yield, ~71%). HPLC method: Sepax GP-C18 (250 × 4.6 mm), 0–3 min, from 10% buffer B to 20% buffer B; 3–23 min, from 20% buffer B to 60% buffer B; flow rate 1 mL/min. Retention time: 16.1 min. HRMS (Bruker SolariX 7.0T): calcd. for C₇₉H₁₁₁N₁₅O₂₀S: m/z [M+2H]²⁺ 811.9003; observed, [M + H]²⁺ = 811.9000.

Synthesis of non-radioactive Ga- and AlF-NOTA-PEG₄-LLP2A

For Ga-NOTA-PEG₄-LLP2A: NOTA-PEG₄-LLP2A (20 μL, 5 mM) and GaCl₃ (2 uL, 100 mM) were added to sodium acetate buffer (50 μL, 0.25 M, pH 4.5). The mixture was heated at 100°C for 10 min and the pH was monitored and kept at 4–5. Then, HPLC purification was performed and the peak at 16.0 min was collected. HRMS (Bruker SolariX 7.0T): calcd. for C₇₉H₁₀₈GaN₁₅O₂₀S: m/z [M+2H]²⁺ 844.8514; observed, [M + H]²⁺ = 844.8493. The same HPLC method as NOTA-PEG₄-LLP2A was used for this chemical.

For AlF-NOTA-PEG₄-LLP2A: AlCl₃ (100 μL, 200 mM) and NaF (100 μL, 200 mM) were mixed and shook for 10 mins at room temperature to obtain AlF²⁺ solution. NOTA-PEG₄-LLP2A (20 μL, 5 mM) and AlF²⁺ solution (2 uL, 100 mM) were added to sodium acetate buffer (50 μL, 0.25 M, pH 4.5). The mixture was heated at 100°C for 10 min and the pH was monitored and kept at 4–5. Then, HPLC purification was performed and the peak at 15.9 min was collected. HRMS (Bruker SolariX 7.0T): calcd. for C₇₉H₁₀₈AlFN₁₅O₂₀S: m/z [M+2H]²⁺ 833.3786; observed, [M + H]²⁺ = 833.3889. The same HPLC method as NOTA-PEG₄-LLP2A was used for this chemical.

Radiolabeling

Al¹⁸F labeling: no-carrier-added ¹⁸F-fluoride was used, which was produced by a GE MINItrace cyclotron via the ¹⁸O(p,n)¹⁸F reaction. Briefly, the target was washed with 4 mL water to collect residual activity after routine clinical ¹⁸FDG production. The target water was passed through a Sep-Pak QMA light cartridge (Waters), which was preconditioned with 5 mL of 0.5 M potassium bicarbonate. The ¹⁸F was eluted with 300 μL saline and the fractions were collected into 6 tubes (each 50 μL). To the tube containing the highest activity (500~740 MBq), AlCl₃ (2 μL, 5 mM in sodium acetate buffer, pH 4, 0.1 M), sodium acetate buffer (10 μL, pH 4, 1M), acetonitrile (50 μL) and NOTA-PEG₄-LLP2A (4 μL, 5 mM) were added. The solution was heated at 100°C for 15 min, and then diluted with 1 mL water. The resulting mixture passed through a C18 light cartridge (Waters), followed by washing the cartridge with 5 mL water to removed unreacted ¹⁸F, and then the desired Al¹⁸F-NOTA-PEG₄-LLP2A was eluted with 0.5 mL ethanol. The product was reconstituted in saline and passed through a 0.22μm Millipore filter into a sterile vial, and its radiochemical purity was determined by radio-HPLC.

⁶⁸Ga labeling: ⁶⁸Ga³⁺ in 0.05 M HCl was obtained from a ⁶⁸Ge/⁶⁸Ga generator (Isotope Technologies Graching, Germany). Briefly, 4 mL 0.05 M HCl was passed through the column and the fractions were collected into 8 tubes (each 0.5 mL). Sodium acetate buffer (150 μL, 0.25 M, pH 4.5) and NOTA-PEG₄-LLP2A (2 μL, 5 mM) were added to the tube containing the highest activity (370~500 MBq). The pH of the final radiolabeling solution was 4~5. Then, the mixture was heated at 100°C for 10 min. The radiolabeling yield of the product ⁶⁸Ga-NOTA-PEG₄-LLP2A was determined by radio-HPLC.

Radio-HPLC method: 0–4 min, 5% buffer B; 4–5 min, from 5% buffer B to 90% buffer B; 5–10 min, 90% buffer B; 10–15 min, 5% buffer B; flow rate, 1.0 mL/min.

Serum stability

Al¹⁸F-NOTA-PEG₄-LLP2A and ⁶⁸Ga-NOTA-PEG₄-LLP2A (~3.7 MBq) were each mixed with 0.5 mL human serum, and the two mixtures were incubated at 37°C for 4 h (Al¹⁸F-tracer) and 2 h (⁶⁸Ga-tracer), respectively. In order to remove plasma protein, 0.5 mL acetonitrile was added to the mixtures, followed by 5 min centrifugation at 14000 RPM. The supernatants were separated and 90% of the radioactivity was recovered. Approximately 37 kBq of the supernatants were taken for radioactive purity determination using radio-HPLC. Free ⁶⁴Cu/⁶⁸Ga came out at ~ 3.5 min and the radiolabeled conjugates came out at 11 min.

Cell culture

Cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM), supplemented with 10% fetal bovine serum, penicillin (100 units/mL), streptomycin (100 μg/mL) L-glutamine (300 μg/mL), sodium pyruvate (100 mg/mL), and glucose (4.5 g/L) and maintained at 37°C, 5% CO₂.

Cell uptake and internalization

B16F10 cells were seeded in 12-well plates (200,000 cells per well) 24 h prior to the experiment. The cells were then washed twice with 1 mL Hank’s Balanced Salt Solution (HBSS) and 1 mL media (DMEM with 0.1% BSA and 1 mM Mn²⁺) was added to each well. The cells were then incubated with Al¹⁸F-NOTA-PEG₄-LLP2A/⁶⁸Ga-NOTA-PEG₄-LLP2A (~74 kBq per well) with and without the addition of excess LLP2A-PEG₄ (10 μg per well) to determine nonspecific uptake. At different time points (0.5, 1, and 2 h for ⁶⁸Ga-tracer; 0.5, 1, 2, and 4 h for ¹⁸F-tracer), the radioactive media were aspirated, and the plate was washed twice with PBS buffer (pH 7.2). The total cell uptakes were measured by counting the radioactivity on a gamma counter. To collect the surface-bound fraction, each well was treated with 20 mM sodium acetate-HBSS (pH 4.0) and was incubated at 37°C for 10 min. After removal of the surface-bound fraction, cell pellets were dissolved in 0.5% SDS. All of the fractions were counted in a well counter (2470 Automatic Gamma Counter WIZARD, PerkinElmer, Norwalk CT, USA). The amount internalized was the amount of activity in the final cell pellet, corrected for activity in the blocked fractions and background activity. The percentages of the radiotracers internalized into the cell were calculated based on the equation: (the gamma counts obtained from cell-internalized radioactivity) / (the gamma counts obtained from the total cell associated radioactivity).

Animal Model

Four-week-old female C57BL/6 and BALB/c nude mice purchased from HFK Bioscience (Beijing, China) were used in this study. All mice were maintained under specific pathogen-free conditions, within facilities approved by the Laboratory Animal Care of Huazhong University of Science and Technology, and in compliance with the regulations and standards of the Institutional Animal Care and Use Committee of Tongji Medical College of Huazhong University of Science and Technology. For B16F10 and A375 tumor xenografts, C57BL/6 and BALB/c nude mice were injected subcutaneously in the right shoulder with one million cells in PBS buffer, respectively.

Micro-PET imaging

Mice bearing B16F10/A375 xenografts (4 mice per group) were injected intravenously with Al¹⁸F-NOTA-PEG₄-LLP2A or ⁶⁸Ga-NOTA-PEG₄-LLP2A (5.5–7.4 MBq, 100–200 pmol per mouse) with or without a blocking dose (100 μg unlabeled LLP2A-PEG₄). At 0.5 h, 1 h and 2 h (and 4 h for ¹⁸F) post injection (p.i.), mice were anesthetized with 2% isoflurane and PET images were acquired. All PET scans were performed with a small-animal PET scanner (BioCaliburn LH Raycan Technology Co., Ltd.™, Suzhou, China). The intrinsic spatial resolution was 1.0 mm, the time resolution was 1.50 ns full-width at half-maximum, the energy resolution was 13.0% at 511 keV and the axial field of view was 5.3 cm.

Ex vivo Bio-distribution

Mice bearing B16F10/A375 xenografts (5 mice per group) were injected intravenously with Al¹⁸F-NOTA-PEG₄-LLP2A or ⁶⁸Ga-NOTA-PEG₄-LLP2A (5.5–7.4 MBq, 100–200 pmol per mouse) with or without a blocking dose (100 μg unlabeled LLP2A-PEG₄). The tumor-bearing mice were sacrificed at 0.5 h, 1 h, and 2 h, (and 4 h for ¹⁸F-tracer) p.i.. The organs and tumors were harvested, weighed, and counted in a gamma counter. The radioactivity in each tissue was normalized as the percentage injected dose per gram of tissue (%ID/g).

Statistics analysis

Significance analysis of the data was performed by unpaired t-testing using Prism software (GraphPad Software Inc., San Diego, CA). Multiple comparisons were performed by multiple t-tests together with the Holm-Sidak method for correction. A value of P < 0.05 was considered significant and P < 0.01 was considered highly significant.

Results

Synthesis and radiolabeling

A good isolated yield (~ 50%) of NH₂-PEG₄-LLP2A was obtained from the traditional Fmoc-based solid phase synthesis. p-SCN-Bn-NOTA was attached to the prepared NH₂-PEG₄-LLP2A with a 71% yield; after HPLC purification, the purity of the NOTA-PEG₄-LLP2A product was > 95% (see Figure 1a for the structures).

Figure 1.

(a) Structures of Al¹⁸F-NOTA-PEG₄-LLP2A and ⁶⁸Ga-NOTA-PEG₄-LLP2A. (b) Stability of Al¹⁸F-NOTA-PEG₄-LLP2A and ⁶⁸Ga-NOTA-PEG₄-LLP2A after incubation in serum at 37°C for 4h and 2h, respectively. (c, e) Cell internalization of Al¹⁸F-NOTA-PEG₄-LLP2A and ⁶⁸Ga-NOTA-PEG₄-LLP2A in B16F10 cells. The radioactivity that was removed from the cells by treatment with 20 mM sodium acetate-HBSS (pH 4.0) was considered as the membrane-bound fraction. The radioactivity measured in the cell lysates was considered as the internalized fraction. Total cell-associated activity equals the activity of internalized fraction plus membrane-bound fraction. (d, f) B16F10 cell uptake of the two tracers and after blocking with co-incubation of 10 μg unlabeled LLP2A-PEG₄. (***, p<0.001; n = 3)

The Al¹⁸F radiolabeling of the conjugate was performed at 100°C in 0.1 M sodium acetate buffer (pH 4) for 15 min. Radiochemical yields ranging from 38%–42% were obtained with high radiochemical purity (> 95%). The ⁶⁸Ga radiolabeling of the conjugate was performed at 100°C in a solution of 0.5 mL ⁶⁸GaCl₃ in HCl (0.05M) and 0.15 mL sodium acetate (0.25 M) for 15 min. A greater than 95% radiolabeling yield was achieved. Nonradioactive Ga-NOTA-PEG₄-LLP2A and AlF-NOTA-PEG₄-LLP2A standards were prepared to confirm the successful ⁶⁸Ga and Al¹⁸F radiolabeling. The molar activity of ⁶⁸Ga-NOTA-PEG₄-LLP2A (37~74 MBq/nmol) was slightly higher than that of Al¹⁸F-NOTA-PEG₄-LLP2A (20~37 MBq/nmol). Both Al¹⁸F²⁺ and ⁶⁸Ga³⁺ labeled tracers exhibited good stability, and less than 5% disassociation of ¹⁸F and ⁶⁸Ga were observed after being incubated in human serum at 37° (Figure 1b). The LogP values of Al¹⁸F-NOTA-PEG₄-LLP2A/⁶⁸Ga-NOTA-PEG₄-LLP2A were −1.80 ± 0.03 and −1.92 ± 0.05, respectively, indicating that both tracers were hydrophilic.

In vitro cell study

To determine the specific binding and internalization profiles of Al¹⁸F-NOTA-PEG₄-LLP2A and ⁶⁸Ga-NOTA-PEG₄-LLP2A, cell uptake studies were performed using the VLA-4 overexpressing mouse B16F10 melanoma cells. 4 h and 2 h cell uptake studies were conducted for ¹⁸F- and ⁶⁸Ga-tracers respectively because of their different half-life times. As shown in Figure 1c and 1e, the uptake of Al¹⁸F-NOTA-PEG₄-LLP2A increased within 1 h and then remained at a plateau up to 4 h, while the uptake of ⁶⁸Ga-NOTA-PEG₄-LLP2A gradually increased over 2 h. The uptake values of ⁶⁸Ga-NOTA-PEG₄-LLP2A were significantly higher than those of Al¹⁸F-NOTA-PEG₄-LLP2A at the same time point (P < 0.01). A total of 32.0%–35.5% (Al¹⁸F-tracer) and 26.9%–36.5% (⁶⁸Ga-tracer) of the cell bonded tracers (1%–1.4% and 2.2%–2.4% of total added dose, respectively) were internalized into the cells, while most of them remained on the cell surface. In the groups receiving the blocking dose, the uptake of Al¹⁸F-NOTA-PEG₄-LLP2A and ⁶⁸Ga-NOTA-PEG₄-LLP2A were reduced significantly by 61.7% and 79.2% respectively, validating their high specificity for VLA-4.

Small animal PET imaging

The static images of Al¹⁸F-NOTA-PEG₄-LLP2A in mice bearing B16F10 tumors afforded excellent tumor visualization with high tumor retention and tumor-to-background contrast at 0.5 h, 1 h, 2 h and 4 h after intravenous injection. The activity accumulation in liver was high and minimal activity accumulations were observed in other normal organs, except bladder.

PET images obtained from the group injected with ⁶⁸Ga-NOTA-PEG₄-LLP2A tracers were similar to those injected with Al¹⁸F-NOTA-PEG₄-LLP2A, and all the tumors were clearly visualized with highest tumor uptake achieved at an early p.i. time point (0.5 h). Compared to Al¹⁸F-NOTA-PEG₄-LLP2A, ⁶⁸Ga-NOTA-PEG₄-LLP2A showed faster clearance and also slightly less liver uptake, with higher bladder uptake due to the renal excretion of injected radioactivity.

To validate in vivo specificity, blocking doses were applied to the mice injected with Al¹⁸F-NOTA-PEG₄-LLP2A and ⁶⁸Ga-NOTA-PEG₄-LLP2A. As shown in Figure 2(c, d), the activity accumulation in B16F10 xenografts was significantly reduced by the blocking agent, showing good specificity. To further confirm their VLA-4 targeting specificity, mice bearing VLA-4-negtive A375 xenografts were injected with ⁶⁸Ga-NOTA-PEG₄-LLP2A and PET imaging was conducted at the same time points. The resulting PET images did not show specific accumulation in the A375 xenografts (Figure 3).

Figure 2.

Representative static PET images of the B16F10 xenograft at 0.5 h, 1 h, 2 h, and 4 h p.i. of 5.5–7.4 MBq (100–200 pmol) Al¹⁸F-NOTA-PEG₄-LLP2A (a) and ⁶⁸Ga-NOTA-PEG₄-LLP2A (b), and the blocked group co-injected with 100 μg unlabeled LLP2A-PEG₄ (c and d, respectively). Arrows indicate tumors.

Figure 3.

Representative static PET images of C57BL/6 mice bearing B16F10 tumor xenografts (a) and BALB/c nude mice bearing A375 tumor xenografts (b) at 0.5 h, 1 h, and 2 h p.i. of 5.5–7.4 MBq (100–200 pmol) ⁶⁸Ga-NOTA-PEG₄-LLP2A. Arrows indicate tumors.

Ex vivo Biodistribution

Biodistribution data are summarized in Figures 4 and 5 and Tables S1, S2 (See electronic supplementary material). Both Al¹⁸F-NOTA-PEG₄-LLP2A and ⁶⁸Ga-NOTA-PEG₄-LLP2A exhibited similar in vivo distribution and excretion profiles, such as high uptake in tumor, liver, spleen, and kidneys at the same time point p.i.. Both tracers cleared from the kidneys and liver; however, the liver clearance was significantly higher for Al¹⁸F-NOTA-PEG₄-LLP2A than for ⁶⁸Ga-NOTA-PEG₄-LLP2A at 1 h and 2 h p.i. (1 h: 20.1 ± 2.6 vs. 15.3 ± 1.7 %ID/g, P < 0.05; 2 h: 11.0 ± 1.2 vs. 8.0 ± 0.8 %ID/g, P < 0.05). Significantly lower uptake in the colon was observed for ⁶⁸Ga-NOTA-PEG₄-LLP2A than for Al¹⁸F-NOTA-PEG₄-LLP2A at 1 h and 2 h p.i. (1 h: 4.8 ± 0.9 vs 2.4 ± 0.3 %ID/g, P < 0.01; 2h: 2.8 ± 0.2 vs. 2.1 ± 0.2 %ID/g, P < 0.01). The kidney uptake was also significantly lower for ⁶⁸Ga-NOTA-PEG₄-LLP2A than for Al¹⁸F-NOTA-PEG₄-LLP2A at 2 h p.i. (5.9 ± 0.5 vs. 4.5 ± 0.3 %ID/g, P < 0.01). The biodistribution comparisons of Al¹⁸F-NOTA-PEG₄-LLP2A and ⁶⁸Ga-NOTA-PEG₄-LLP2A are shown in Figure S1.

Figure 4.

Biodistribution in mice bearing B16F10 tumors at 0.5 h, 1 h, 2 h and 4 h p.i. of Al¹⁸F-NOTA-PEG₄-LLP2A (5.5–7.4 MBq, 100–200 pmol) and 4 h after blocking by coinjection of 100 μg unlabeled LLP2A-PEG₄ with the Al¹⁸F-NOTA-PEG₄-LLP2A). (a); Tumor-to-blood ratios (b); Tumor-to-muscle radios (c). (n = 5, ** P < 0.001)

Figure 5.

Biodistribution of ⁶⁸Ga-NOTA-PEG₄-LLP2A (5.5–7.4 MBq, 100–200 pmol) in mice bearing B16F10 tumors at 0.5 h, 1 h and 2 h p.i., 2 h blocked (co-injected with 100 μg unlabeled LLP2A-PEG₄ with the ⁶⁸Ga-NOTA-PEG₄-LLP2A), and in mice bearing A375 tumors at 2 h p.i. (a); Tumor-to-blood ratios (b) and tumor-to-muscle ratios (c) of ⁶⁸Ga-NOTA-PEG₄-LLP2A in mice bearing B16F10 tumors at 0.5, 1 and 2 h post-injection (n = 5, ** P < 0.001)

Discussion

The aim of this study was to develop LLP2A-based PET tracers using same peptidomimetic precursor, with different positron emitters (¹⁸F and ⁶⁸Ga), and compare their in vivo performance for PET imaging of melanoma. A commonly-used chelator, p-SCN-Bn-NOTA, served as the key coordinating group to form stable complexes with Al¹⁸F and ⁶⁸Ga. To our knowledge, this is the first report of an Al¹⁸F-labeled LLP2A tracer. Previously, there have been only a few reports of ¹⁸F-labeled LLP2A using the ¹⁸F-organotrifluoroborate method [29, 30]. In their study, PEG₂-LLP2A-trfluroroborate conjugates were synthesized and labeled with ¹⁸F by the isotope exchange method with sufficient molar activity but relatively lower radiochemical yields of 3.7% and 11.3%, respectively.

Stability is extremely important for radiotracers, especially for ¹⁸F radiolabeled tracers, since free ¹⁸F⁻ binds to bone with high affinity. Thus, we conducted an in vitro stability test, and both prepared tracers remained intact after several hours’ incubation in serum and PBS. For Al¹⁸F-NOTA-PEG₄-LLP2A, relatively high bone uptake (4.47 ± 0.37 %ID/g at 0.5 h) was determined by an ex vivo biodistribution study. The high bone uptake usually indicates defluorination and low in vivo stability of ¹⁸F-based tracers. In this case, defluorination might also occur for Al¹⁸F-NOTA-PEG₄-LLP2A, however, the majority was most likely contributed by marrow uptake because of the natural expression of VLA-4 in the bone marrow [40]. Moreover, higher bone uptake (5.75 ± 2.04%ID/g at 0.5 h) was also observed in mice receiving ⁶⁸Ga-NOTA-PEG₄-LLP2A, which confirmed specific uptake in bone marrow. The bone uptake was significantly reduced for both tracers when cold PEG₄-LLP2A was coinjected, which further validated the stability.

All tumors could be clearly visualized in the PET images with high retention up to 4 h and 2 h p.i. of ¹⁸F- or ⁶⁸Ga-labeled tracers, respectively. The tumor uptake values of the two tracers were comparable at the same time points p.i. post injection, except a slightly higher tumor uptake of Al¹⁸F-NOTA-PEG₄-LLP2A (9.04 ± 2.30 %ID/g) at 0.5 h than ⁶⁸Ga-NOTA-PEG₄-LLP2A (6.77 ± 1.77 %ID/g) without statically significant. ⁶⁸Ga-NOTA-PEG₄-LLP2A had significantly higher tumor-to-blood ratios at 1 h (P < 0.05) and 2 h (P < 0.01), and significantly higher tumor-to-muscle ratio at 2 h (P < 0.05) than Al¹⁸F-NOTA-PEG₄-LLP2A. It should be noticed that the molar activity of ⁶⁸Ga-NOTA-PEG₄-LLP2A (37~74 MBq/nmol) was slightly higher than that of Al¹⁸F-NOTA-PEG₄-LLP2A (20~37 MBq/nmol), which might also contribute to the superior in vivo performance of ⁶⁸Ga-NOTA-PEG₄-LLP2A. The tumor uptake of ⁶⁸Ga-NOTA-PEG₄-LLP2A at 1 h was comparable to that of ⁶⁸Ga-DOTA-PEG₄-LLP2A [33] (9.1 ± 0.9 %ID/g) and ⁶⁸Ga-NODAGA-PEG₄-LLP2A [27] (8.7 ± 1.3 %ID/g) without statically significant. The tumor uptake of Al¹⁸F-NOTA-PEG₄-LLP2A at 1 h was comparable to that of ¹⁸F-DOTA-AMBF₃-PEG₂-LLP2A (9.46 ± 2.19 %ID/g) [30], but significantly higher than that of ¹⁸F-N-Pyr-p-BF₃⁻-PEG₂-LLP2A (4.41 ± 0.65 %ID/g, P=0.001) and ¹⁸F-AMBF₃-PEG₂-LLP2A (2.84 ± 0.34 %ID/g, P<0.0001) [29]. ⁶⁸Ga-NOTA-PEG₄-LLP2A exhibited relatively quicker clearance from nonspecific organs, which might due to slightly lower lipophilicity than Al¹⁸F-NOTA-PEG₄-LLP2A. It is consistent with their Log P profiles. The specificity of Al¹⁸F-NOTA-PEG₄-LLP2A and ⁶⁸Ga-NOTA-PEG₄-LLP2A were validated via blocking study, and further confirmed using a VLA-4-negative A375 tumor model.

Liver and bladder showed the most intense uptake, and the activity accumulation in liver decreased over time, indicating both tracers cleared predominantly via the hepatobiliary pathway and partly via renal clearance. Significantly higher liver uptake was observed at 1 h and 2 h p.i. of Al¹⁸F-NOTA-PEG₄-LLP2A compared to those of ⁶⁸Ga-NOTA-PEG₄-LLP2A (P < 0.01). The overall liver uptake of Al¹⁸F-NOTA-PEG₄-LLP2A and ⁶⁸Ga-NOTA-PEG₄-LLP2A were significantly higher than those of ⁶⁸Ga-NODAGA-PEG₄-LLP2A [27] and our previously reported ⁶⁸Ga-NE2P1A-PEG₄-LLP2A [28], which might due to the slightly higher lipophilicity of the chelator p-SCN-Bn-NOTA than those of NODAGA and p-SCN-PhPr-NE2P1A. Compared to the reported ¹⁸F-R-BF₃–PEG₂-LLP2A tracers, the liver uptake of Al¹⁸F-NOTA-PEG₄-LLP2A was relatively higher, but the colon uptake was much lower [29]. The high colonic uptake of ¹⁸F-R-BF₃–PEG₂-LLP2A was significantly reduced by introducing an appended DOTA moiety, which usually serves as a metal chelator, but acted as a hydrophilic appendant in their study [30]. The relatively high liver uptake of Al¹⁸F-NOTA-PEG₄-LLP2A and ⁶⁸Ga-NOTA-PEG₄-LLP2A may hinder their applications in the detection of liver metastatic lesions, which might also be solved by introducing a hydrophilic appendant.

Overall, ⁶⁸Ga-NOTA-PEG₄-LLP2A displayed clear superiority, compared with Al¹⁸F-NOTA-PEG₄-LLP2A, due to significantly lower liver uptake, and significantly higher tumor-to-muscle and tumor-to-blood ratios at some time points. Considering the wide availability of ¹⁸F⁻, development of ¹⁸F-labeled VLA-4 tracers may also be in demand in situations where the supply of ⁶⁸Ga (and/or ⁶⁸Ge/⁶⁸Ga generators) is limited. In addition, the residual activity of ¹⁸F after routine ¹⁸F-FDG production was collected by washing the target of the cyclotron and used for the production of Al¹⁸F-NOTA-PEG₄-LLP2A in this study. Much larger amounts of Al¹⁸F-NOTA-PEG₄-LLP2A can be produced with higher specific activity when the entire activity produced by the cyclotron is used. Thus, Al¹⁸F-NOTA-PEG₄-LLP2A is a good alternative VLA-4 imaging tracer for research and clinical translation.

Conclusion

In summary, two VLA-4 targeted radiotracers (Al¹⁸F-NOTA-PEG₄-LLP2A and ⁶⁸Ga-NOTA-PEG₄-LLP2A) have been successfully constructed for PET imaging of melanoma, and both tracers showed high in vivo VLA-4 targeting efficiency and good specificity. The favorable biodistribution and PET images make them promising candidates for noninvasively imaging VLA-4 overexpressing tumors.