Site-selective labelling of peptides with fluorine-18 in solution and on solid phase using succinimidyl 4-[¹⁸F]fluorobenzoate ([¹⁸F]SFB) – a comparative study

Objectives:

The growing interest in peptides as probes for the molecular imaging of physiological and pathological processes by positron emissions tomography (PET) is accompanied with the increasing need for generally applicable methods that allow a selective labelling of these biomolecules [1]. The site-selective introduction of the well-established 4-[¹⁸F]fluorobenzoyl prosthetic group into peptides containing multiple amino acids with nucleophilic entities located in the side chains is very often challenging. This is especially valid for lysine-containing peptides. Sometimes, a selective labelling can be achieved by taking advantage of the slight difference between the pK_a values of the N-terminus and the ε-amino group of lysine [2]. However, the success of this strategy is moderate. Therefore, the potential to label various lysine containing peptides with [¹⁸F]SFB at defined sites by solid-phase synthesis was evaluated in this study in comparison to ¹⁸F-fluorobenzoylation in solution.
Methods:
Four different peptides were chosen to comparatively evaluate the solid phase approach and labelling in solution (Figure 1). The selection included the N-terminal telopeptide of collagen α1(I) 1 [3] and its derivative 2 [4] both containing a lysine residue important for collagen crosslinking, the k7 sequence of a cell-penetrating peptide 3 [5] and a derivative of a so-called SNEW peptide 4. The latter is a ligand for the EphB2 receptor whose binding affinity is critically dependent on a free N-terminus [6]. The aim was to label 1-3 at their N-terminus whereas 4 was to be labelled selectively at its C-terminal lysine residue. The precursors as well as the corresponding non-radioactive fluorobenzoylated peptides were synthesised by microwave-assisted solid-phase peptide synthesis employing the Fmoc/tBu strategy. For labelling in solution the fully deprotected peptides were reacted with [¹⁸F]SFB at pH 7 and 9. Labelling on solid phase was carried out by suspending the preswollen peptide-loaded resin in a mixture of DMF and phosphate buffer containing [¹⁸F]SFB followed by a washing step and final acidic deprotection and cleavage from the resin. For every labelling experiment the labelling yields were determined by radio-HPLC. In addition, peptides 1 and 2 were labelled in solution by using derivatives with Alloc/All-protected side chains followed by Pd-catalysed deprotection and solid phase extraction.

1 *H-Leu-Ser-Tyr-Gly-Tyr-Asp-Glu-Lys-Ser-Thr-Gly-Ile-Ser-Val-Pro-NH₂
2 *H-Gly-Gly-Gly-Asp-Pro-Lys-Gly-Gly-Gly-Gly-Gly-NH₂
3 *H-Lys-Lys-Arg-Lys-Ala-Pro-Lys-Lys-Lys-Arg-Lys-Phe-Ala-NH₂
4 H-Ser-Asn-Glu-Trp-Ile-Leu-Pro-Arg-Leu-Pro-Gln-His-Val-*Lys-NH₂

Figure 1. Sequences of the peptides studied in the labelling experiments. The desired labelling position is indicated by an asterisk.

Results:

Labelling in solution always resulted in product mixtures independent of the pH that were often difficult to purify. In contrast, labelling on solid-phases yielded ¹⁸F-fluorobenzoylated peptides in radiochemical purities of 95-99%. The optimised procedure allowed to prepare the labelled peptides in activities ranging from 30-100 MBq (8-15%, d.c.) with a specific activity of 10-12 GBq/μmol and a reaction time of 135-140 min (without [¹⁸F]SFB production). The labelling of the Alloc/All-protected peptides in solution was less successful due to the presence of unreacted [¹⁸F]SFB during the deprotection step.
Conclusions:
The selective labelling [¹⁸F]SFB on solid phase is a useful tool in the development of radiotracers based on peptides, especially if the lysine side chain or the N-terminus is functionally important for target binding.
References:
[1] Tolmachev & Stone-Elander (2010) Biochim. Biophys. Acta 1800, 487-510,
[2] Wester et al. (1995) J. Label. Compd. Radiopharm. 37, S513-S515,
[3] Helseth et al. (1979) Biopolymers 18, 3005-3014,
[4] Nagan & Kagan (1994) J. Biol. Chem. 269, 22366-22371,
[5] Rennert et al. (2008) ChemMedChem 3, 241-253
[6] Chrencik et al. (2007) J. Biol. Chem. 282, 36505-36513