Introduction

The basis of peptide receptor radionuclide scintigraphy (PRRS) is the over expression of peptide receptors in pathologic conditions, e.g., as cancer. As specific molecular probes used for receptor imaging and tumour targeting, radiolabelled peptides have a huge impact on the diagnosis and treatment of diseases. In the future, more effective selective imaging and therapeutic agents based on radiopeptide radiopharmaceuticals are expected [1].

For imaging studies, peptides can be labelled with several radionuclides such as 99mTc, 111In, 68Ga, 64Cu, 123/124I. For routine clinical SPECT imaging, 99mTc is the radionuclide of choice. This is because it has good radiation physical characteristics [IT, 140 keV (90), 6 h] and is available through an inexpensive 99Mo/99mTc generator. Furthermore, it is well developed and has varied chemistry [2]. For routine clinical PET imaging, 18F is the radionuclide of choice due to its low energy of positron (0.69 MeV, 97 % β+, 3 % EC). Additionally, it has the highest potential resolution for PET imaging, a low tissue radiation dose and a half life of 109.8 min. This makes it possible for a more complex synthesis with sufficient quantities to be delivered to centres far from its production facility [3].

The radiolabeling of peptides with 99mTc can be directly achieved through amino acids. Alternatively, it can be indirectly achieved through BFCAs, e.g., HYNIC. Since HYNIC-peptide conjugate is easily labelled with 99mTc, using a number of co-ligands with high stability, HYNIC is one of the most widely used BFCAs for the labelling of peptides with 99mTc [4–6].

Direct fluorination of peptides with 18F is not possible. This is because of the harsh reaction conditions such as high temperature, pH, solvents and a large amount of peptide. Thus, indirect methods, which utilize fluorinated prosthetic groups, have been developed. The preparation of most fluorinated prosthetic groups is a multi-step procedure and is time-consuming. Furthermore, the final product usually needs to be purified by HPLC with a low yield. In addition, the reaction between peptide and fluorinated prosthetic group is not selective and a peptide precursor should be protected [7–15]. To avoid side reactions, chemoselective reactions, such as oxime and hydrazone bond formation, were introduced. Previous studies show that unprotected peptide precursors, functionalized with aminooxy or hydrazine, form a stable oxime or hydrazone bond with18F-fluorinated aldehyde as a prosthetic group in aqueous media [16–18]. In recent years, Human serum albumin (HAS) and unprotected peptides functionalized with HYNIC, such as RGD, octreotide and substance p, were fluorinated with 18Fluorobenzaldehyde with high stability via hydrazone formation [19–21]. For oxime bond formation, the reagent Eei-Aoa-NHS was introduced to incorporate aminooxy at N-terminal end of peptide for oxime bond formation. In a simple one step method, 18FDG was applied to fluorinate aminooxy peptides, through oxime bound formation [22–24].

The future approach is a fast, easy and efficient method for large scale production of 18F labelled peptides for clinical routine application at nuclear medicine centres. 18FDG has high potential as 18F-fluorinated prosthetic group. This is due to the availability of 18FDG in most PET centres and the presence of aldehyde group in acyclic form of 18FDG, as a result of mutarotation in aqueous solutions.

Our aim is to functionalize peptide for labelling with 18FDG and 99mTc for PET and SPECT imaging studies. With regard to high reactivity and chemoselectivity of HYNIC toward carbonyl for hydrazone bond formation and its stable complexes with 99mTc, HYNIC-conjugated peptides was considered for the labelling of peptides with 18FDG and 99mTc. In this study, LIKKPF as a model peptide was synthesized and conjugated with HYNIC. The HYNIC conjugated peptide was labelled with 99mTc using Na99mTcO4 and 18F using 18FDG. Here, we report the direct labelling of a HYNIC functionalized peptide with 18FDG. Moreover, the stability, labelling efficiency and radiochemical purity of 99mTc and 18F radiolabeled peptide were determined and compared.

Materials and methods

For this study, the amino acids and resin were obtained from Bachem (Bubendorf, Switzerland). Coupling reagents, HOBT and DIC were purchased from Sigma-Aldrich (St. Luis, MO, USA). Cold FDG [19FDG] and Succinimidyl-N-Boc-HYNIC were purchased from ABX advanced Biochemical compounds GmbH (Radeberg, Germany). All of the chemicals, solvents and reagents were of analytical quality and used without further purification. Silica gel 60 F254 pre-coated aluminium sheets from Merck were used for TLC. Using normal serum, 99mTcO4 − was eluted from a 99Mo/99mTc generator (Pars-isotope, Tehran, Iran). Furthermore, 18FDG was provided from a routine in-house synthesis at the PET/CT Unit, Ferdous Nuclear Medicine centre, Dr Masih Daneshvari Hospital, Shahid Beheshti University of Medical Sciences (Tehran, Iran). Additionally, 18FDG was prepared by TracerLab MXFDG (GE Medical Systems Benelux s.a) with specification according to the European Pharmacopeia. SepPak Plus C-18 was purchased from Waters Corporation (Milford, USA). The distribution of radioactivity on TLC was determined using a TLC Scanner Mini-Scan, MS.1000. This was equipped with flow count B-FC-1000 and gamma detector MS3200 (Bioscan, Washington, USA). Mass-Spectra was recorded on LC–MS Triple Quad 6410 Agilent Technologies using series 1200 HPLC system (Tokoyo, Japan) column: C-18, 250 × 4.6 mm, 5 μm, mobile phase: A: H2O + 0.1 % TFA, B: acetonitrile, flow rate: 1 mL/min, 20 μL, total run time: 40 min. A NaI well counter (Triathler multilabel tester, Hidex, Finland) and a dose calibrator (Atomlab 100, Biodex, NY) were used to measure low and high levels of radioactivity, respectively.

Peptide synthesis

Peptide LIKKPF (Leu-Ile-Lye–Lye-Pro-Phe) was manually synthesized on solid phase using standard Fmoc strategy, based on the Merrifield method [25]. Briefly, the peptide sequence Leu-Ile-Lye–Lye-Pro-Phe was assembled on Wang Resin with two equiv of N-α-Fmoc-protected amino acid and two equiv HOBt and DIC as a coupling reagent in six steps. A solution of two equiv of Succinimidyl-N-Boc-HYNIC and two equiv of DIPEA in dry DMF was added to the resin. After shaking for 45 min at room temperature (RT), the solution was removed and the resin was washed with DMF and CH2Cl2, respectively. The completeness of the coupling reaction was checked by a Kaiser Test. The cleavage of peptide from resin was checked using cocktail TFA/TIS/H2O (95:2.5:2.5) for 45 min [26]. The solvents evaporated and peptide was precipitated with diethyl ether. The identity of peptide was confirmed by LC–MS.

Labelling studies

Labelling of HYNIC-LIKKPF with cold FDG (19FDG)

The HYNIC functionalized peptide (2 mg) was reacted with 2 mg of 19FDG in 96 % ethanol in saline (200 µL) pH = 2–3 for 30 min at 100 °C. The reaction mixture was immediately cooled diluted with water to a final volume of 1 mL. Purification was achieved by a Sep-Pak C18 cartridge. The cartridge was activated using 10 mL ethanol, followed by 10 mL of water and 20 mL of air. The reaction mixture was passed through a cartridge, followed by 5 mL of H2O to remove un-reacted FDG and finally, 2 mL of 96 % ethanol to obtain peptide [24]. The identity of 19FDG-HYNIC-LIKKPF was confirmed by LC-MS.

Radiolabeling of HYNIC-LIKKPF with 18FDG

Different concentrations of peptide, (0.1–5) mg in 40–400 µL of 96 % ethanol, pH (2–2.5, 5–6, 8–8.5) were incubated with 18FDG (1–5 mCi/200–250 µL) at 25, 80, 100 and 120 °C for 30 min. The pH was adjusted by TFA or NaOH. Reaction mixtures were purified by a Sep-Pak C18 cartridge. On the collected fractions, TLC chromatography was performed on TLC silica gel 60 F254, acetonitrile:water (95:5) as a mobile phase, over a path of 8 cm [24, 27]. For HPLC radiochromatogram, 20 µL of purified mixture was applied to LC–MS. The outlet of LC–MS column was disconnected from mass and connected to a fraction collector. Forthy fractions (1 mL/min) were collected and activity was measured.

Radiolabelling of HYNIC-LIKKPF with 99mTc

HYNIC-peptide was labelled with 99mTc using SnCl2 as a reducing reagent and EDDA and tricine as co-ligand. Briefly, (15–1000) µg peptide was mixed with 10 mg tricine, 5 mg EDDA, 7 µg SnCl2 in a final volume 550 µL at pH 5–6. The labelling was initiated with the addition of 5–10 mCi 99mTc. The incubation was completed at 100 °C for 30 min. The radiochemical purity was determined by Radio-TLC, using TLC-SG as a stationary phase and MEK, sodium citrate 0.1 M pH 5, and methanol: ammonium acetate 1 M (1:1) as mobile phases, over a path of 8 cm. HPLC radiochromatogram was obtained as mentioned on “Radiolabeling of HYNIC-LIKKPF with 18FDG“ section.

Radiolabelling of HYNIC-LIKKPF with 18FDG in presence of SnCl2, tricine and EDDA

The radiolabelling was achieved as mentioned on “Radiolabeling of HYNIC-LIKKPF with 18FDG“ section.

Stability studies

The stability of 99mTc-HYNIC-peptide and 18FDG-HYNIC-peptide were studied at room temperature for 12 and 4 h, respectively. At different time points, radiochemical purity of complexes was checked by a Radio-TLC. In another study, approximately 10 µg of radiolabelled peptides were individually added to 0.5 mL of human serum plasma and incubated at 37 °C. At different time points (maximum 12 h for 99mTc and 4 h for 18FDG), plasma proteins were precipitated out by reacting with 0.5 mL acetonitrile. The activity bound to the plasma protein was measured by counting the activity associated with the precipitate. The supernatant was analysed by Radio-TLC.

Results

LIKKPF was successfully synthesized via the standard Fmoc method, functionalized with HYNIC at the N-terminal and analysed by LC-MS. HYNIC-peptide: calculated for C45H69N11O8: 879.53; found m/z = 880 [M + H]+. Analytical RP-HPLC: t R = 6.1 min, 20 % A: 80 % B. At this MW (879.53 g/mol) 1 HYNIC molecule covalently attached to each molecule of peptide (Fig. 1).

Fig. 1
figure 1

LC–MS chromatogram of HYNIC-LIKKPF

Full size image

The HYNIC-LIKKPF was conjugated with cold FDG (19FDG). The yield of labelling was 51 %. The LC–MS analysis of purified labelled peptide (purification by C18 Sep Pack cartridge) showed a single mass peak in 1044 [M + H] + which corresponds to FDG-HYNIC-LIKKPF (Fig. 2) calculated for C50H78FN11O12: 1043.5. RP-HPLC: t R = 5.5 min, 20 % A: 80 % B (Fig. 3).

Fig. 2
figure 2

LC–MS chromatogram of 19FDG-HYNIC-LIKKPF

Full size image
Fig. 3
figure 3

Hydrazone bond formation between 18FDG and HYNIC-LIKKPF

Full size image

The radiolabelling of HYNIC-LIKKPF with 99mTc was examined at temperatures (25, 80, 100, 120 °C), incubation times (5, 10, 15, 20, 25, 30 min), peptide (15, 100, 500, 1000 µg), and pH (2–2.5, 5–6, 8–8.5). Results are shown in Figs. 4, 5 and 6. HPLC radiochromatogram is shown in Fig. 7. The RCP was determined via Radio-TLC. TLC-SG, MEK: 99mTcO4 − (R f = 0.9–1.0), 99mTcO2, 99mTc-HYNIC-peptide, and co-ligand (R f = 0). TLC-SG, sodium citrate 0.1 M, pH 5: 99mTcO4 −, co-ligand (R f = 0.9–1.0), 99mTcO2, 99mTc-HYNIC-peptide (R f = 0). TLC-SG, methanol:ammonium acetate 1 M (1:1): 99mTcO4 −, 99mTc-HYNIC-peptide, and co-ligand (R f = 0.9–1.0), 99mTcO2 (R f  = 0). The RCP over 90 % was achieved with (15–1000) μg peptide, 5 mg EDDA, 10 mg tricine, 7 μg SnCl2, pH 5–6, at 100 °C for 30 min.

Fig. 4
figure 4

RCP of 99mTc-HYNIC-LIKKPF as a function of time. The pH used 2–2.5, 5–6, and 8–8.5 with 100 µg of peptide, 5 mCi Na99mTcO4, at 100 °C

Full size image
Fig. 5
figure 5

RCP of 99mTc-HYNIC-LIKKPF as a function of time. The temperatures used 25, 80, 100, 120 °C with 100 µg of peptide, 5 mCi Na99mTcO4, pH 5–6

Full size image
Fig. 6
figure 6

RCP of 99mTc-HYNIC-LIKKPF as a function of time. The amounts of peptide used 15, 100, 500, 1000 µg at 100 °C, pH 5–6, 5 mCi Na99mTcO4

Full size image
Fig. 7
figure 7

HPLC radiochromatogram of 99mTc-HYNIC-LIKKPF

Full size image

The radiolabelling of HYNIC-LIKKPF with 18FDG was examined at temperatures (25, 80, 100, 120 °C), incubation times (5, 10, 15, 20, 25, 30 min), and pH (2–2.5, 5–6, 8–8.5). Results are shown in Figs. 8 and 9. HPLC radiochromatogram is shown in Fig. 10. The RCP was determined via Radio-TLC. The fluorinated peptide remained at origin (R f = 0), while 18FDG moved up with solvent (R f = 0.45). Preliminary studies showed that the optimal reaction temperature, incubation time and pH were 100 °C, 30 min, 2–2.5, respectively. By decreasing the volume of 18FDG activity, the RCP increased. Considering the activity concentration of our daily synthesized 18FDG, 50 mCi/mL, in all of the experiments, the final volume of reaction mixture was adjusted to 200–250 μL. 18FDG (1–5 mCi) was used for the labelling of different amounts of peptide (100 °C, 30 min, pH 2–2.5). Since 18FDG solution contains glucose, which competes with 18FDG for hydrazone bond formation [24, 27], the amount of glucose was determined. This was achieved in a 18FDG solution using HPLC (Agilent 1260, USA) equipped with a flow count Radio-HPLC detector system (B-FC-1000, FC-3300, Bioscan) and pulse amperometric detector (RID), column (anion exchange resin, 0.25 m, 4.0 mm, 10 μm), mobile phase (0.1 N NaOH), flow rate (1 mL/min), run time (30 min). The results showed a glucose concentration of 20–250 μg/mL of 18FDG solution at different 18FDG production runs. Tables 1 and 2 present the results of RCP with different glucose concentration. The highest RCP (>95 %) was obtained with glucose concentration <50 μg/mL (Table 2). The significant reduction of RCP was obtained by either a higher amount of activity or a lesser amount of peptide. The fluorinated peptide was purified by passing through C18 Sep Pack cartridge with 95 % efficiency. At the end of purification, the RCY (decay corrected) based on amount of 18FDG activity was 40 ± 6.3 %. In vitro stability was checked via Radio-TLC at different time points. Results showed radiolabelled peptides were stable and no significant release of 99mTc or 18FDG were detected for at least 12 and 4 h in aqueous and human serum solutions, respectively (Figs. 11, 12). The RCP was >95 % at all time points and less than 5 % of activity was transferred to serum proteins.

Fig. 8
figure 8

RCP of 18FDG-HYNIC-LIKKPF as a function of time. The temperatures used 25, 80, 100, 120 °C with 500 µg of peptide, 1 mCi 18FDG, pH 5–6

Full size image
Fig. 9
figure 9

RCP of 18FDG-HYNIC-LIKKPF as a function of time. The pH used 2–2.5, 5–6, and 8–8.5 with 500 µg of peptide, 1 mCi 18FDG at 100 °C

Full size image
Fig. 10
figure 10

HPLC radiochromatogram of 18FDG-HYNIC-LIKKPF

Full size image
Table 1 RCP of HYNIC-LIKKPF with 18FDG (100 °C, 30 min, pH 2–2.5, glucose 50–250 μg/mL)
Full size table
Table 2 RCP of HYNIC-LIKKPF with 18FDG (100 °C, 30 min, pH 2–2.5, glucose <50 μg/mL)
Full size table
Fig. 11
figure 11

Stability of 99mTc-HYNIC-LIKKPF in aqueous solution and human serum plasma

Full size image
Fig. 12
figure 12

Stability of 18FDG-HYNIC-LIKKPF in aqueous solution and human serum plasma

Full size image

Discussion

The role and importance of peptide radiopharmaceuticals in imaging studies has encouraged vast research, with scholars working towards an easy, fast and efficient peptide radiolabelling method. 18F and 99mTc are the best radionuclides for PET and SPECT studies. At present, a small number of peptides labelled with 99mTc are in clinical routine studies [28, 29]. Radiolabelling of peptides with 18F is more complex than labelling with 99mTc. A small number of 18Fluorinated prosthetic groups have been developed. There is not yet a clinically acceptable 18fluorinated prosthetic group for the radiolabelling of peptides with 18F. Synthesis of prosthetic groups is time consuming, containing multi-steps with a low to moderate yield. In addition, the peptide precursor should be protected before conjugation to prosthetic group. In vivo studies have also revealed a high hepatic and intestinal uptake, low target to non-target ratio, as well as a low stability of 18Fluorinated peptide [12, 30, 31]. To overcome the mentioned problems, chemoselective oxime and hydrazone reactions have been introduced [16–18]. Mutarotation of sugars in aqueous solutions, from cyclic to acyclic forms, provides an aldehyde group for oxime or hydrazone bond formation. The acyclic form of 18FDG, which increases at 100 °C, contains aldehyde group [18]. Due to the availability of 18FDG in most PET centres, there is potential for 18FDG as 18F-fluorinated prosthetic group to be clinically accepted for the radiolabelling of peptides. Previous studies have revealed that the addition of carbohydrate into a peptide may improve the in vivo pharmacokinetics of peptide by decreasing lipophilicity [33, 34]. The hydrophilic nature of 18FDG is expected to increase renal excretion, compared with hepatobiliary excretion [23].

18FDG and 99mTc are the most available radiopharmaceuticals in nuclear medicine centres worldwide. Thus, the primary goal of this study was to prepare a peptide kit formulation to be easily radiolabelled with 99mTc or 18FDG for imaging studies using SPECT or PET systems. The peptide LIKKPF was isolated by Burtea et al. [32] with high affinity and specificity for phosphatidyl serine. Since our group has recently been working on the design, synthesis and radiolabelling of peptides for apoptosis imaging, we selected the LIKKPF as the model peptide. The peptide was conjugated to HYNIC and characterized by LC–MS. The HYNIC-LIKKPF was labelled with 99mTc, using EDDA and tricine as co-ligand with RCP over 90 % (15–1000 μg peptide, 5 mg EDDA, 10 mg tricine, 7 μg SnCl2, pH 5–6, at 100 °C for 30 min). In this study, we also evaluated the labelling of HYNIC-LIKKPF with 18FDG as a prosthetic group. Our results showed that hydrazone bond formation between HYNIC-LIKKPF and 18FDG is a fast, effective and chemoselective reaction, which performs in aqueous media in the presence of chemicals such as EDDA, tricine and SnCl2. The optimal pH for hydrazone bond formation is 2–2.5. The amount of activity had a strong influence on RCP of oxime bond formation, as was mentioned by Hultsch et al. [27]. There is a competition between glucose and 18FDG for hydrazone bond formation. In our experiments, the glucose was not removed from the 18FDG solution. The radiolabelling of peptide was performed in the presence of glucose. The amount of glucose increases by using more 18FDG activity, which results in low RCP. Meanwhile, a higher RCP is obtained by using a larger amount of peptide. The glucose concentration of our daily synthesized 18FDG was usually 50–250 μg/mL. Some days, the glucose was <50 μg/mL. At this glucose concentration, RCP >95 % was achieved with peptide as low as 0.2 mg and 1 mCi 18FDG. At glucose concentration of 50–250 μg/mL, RCP >90 % was achieved with peptide as low as 2 mg and 1 mCi 18FDG. It decreased significantly by using less peptide or more 18FDG activity. Since most of the days the glucose concentration was 50–250 μg/mL, with 1 mg HYNIC-LIKKPF and 1 mCi 18FDG, the RCP of at least 70 % was achieved. The labelled peptide was purified using C18 Sep Pack cartridge with 95 % efficiency for further studies.

In order to get RCP over 90 % without further purification, the glucose should be completely removed using HPLC, as suggested by Hultsch et al. [27]. Otherwise, 18fluorinated peptide should be purified before in vivo experiments to remove un-reacted 18FDG and other impurities. With 18FDG solution free of glucose, it is expected to get RCP >90 % with peptide amount as low as 10–100 μg and 18FDG activity >10 mCi. In these conditions, it is possible to have a bifunctional kit formulation for PET and SPECT imaging.

Conclusion

In this study, the HYNIC-LIKKPF was labelled for the first time with 18FDG and 99mTc. In biological studies, 99mTc-HYNIC-LIKKPF with RCP >90 % is used without further purification, while 18FDG-HYNIC-LIKKPF with RCP ≈ 70 % should be purified. It is an assumption that the higher RCP would be achieved with peptide as low as (10–100) µg and 18FDG activity >10 mCi by removing glucose from 18FDG solution. Since low amounts of peptide are usualy labeled with 99mTc, one peptide kit formulation would be probably used for labeling with 18FDG and 99mTc.