Chemical and biological characterization of new Re(CO)₃/[^99mTc](CO)₃ bombesin analogues

Abstract

Introduction

Bombesin, a neuropeptide with potential for breast and prostate tumor targeting, is rapidly metabolized in vivo, and as a result, uptake in tumor xenografts in mice is poor. An improvement can be expected from the introduction of nonnatural amino acids and spacers. Leu¹³ was replaced by cyclohexylalanine and Met¹⁴ by norleucine. Two spacers, -βAla–βAla- and 3,6-dioxa-8-aminooctanoic acid, were inserted between the receptor-binding amino acid sequence (7–14) of bombesin (BBS) and the retroN^α-carboxymethyl histidine chelator used for labeling with the [^99mTc](CO)₃ core and the rhenium (Re) congener.

Methods

The biological characterization of the new compounds was performed both in vitro on prostate carcinoma PC-3 cells (binding affinity, internalization/externalization) and in vivo (biodistribution in nude mice with tumor xenografts). The stability was also investigated in human plasma. The Re analogues were prepared for chemical characterization.

Results

The nonnatural amino acids led to markedly slower degradation in human plasma and PC-3 cell cultures. The receptor affinity of the new technetium 99m ([^99mTc])-labeled BBS analogues was similar to the unmodified compound with K_d<1 nM. Uptake in the pancreas and in PC-3 tumor xenografts in nude mice was blocked by unlabeled BBS. The best target-to-nontarget uptake ratio was clearly due to the presence of the more polar spacer, -βAla–βAla-.

Conclusions

The different spacers did not have a significant effect on stability or receptor affinity but had a clear influence on the uptake in healthy organs and tumors. Uptake in the kidneys was lower than in the liver, which is likely to be due to the lipophilicity of the compounds. A specific, high uptake was also observed in the gastrin-releasing peptide receptor-rich pancreas. Thus, with the introduction of spacers the in vivo properties of the compounds can be improved while leaving the affinity unaffected.

Introduction

Many neuropeptide receptors have been shown to be overexpressed at high concentrations in a variety of human tumors, and they may therefore be used as potential targets for imaging and therapy with radiolabeled neuropeptides [1], [2], [3], [4], [5]. In fact, the targeting of overexpressed peptide receptors in tumors by small peptides has become a very strong focus of interest for nuclear medicine [1]. The variability of receptors being overexpressed (breast cancer [6] as an example) led to the conclusion that not only one but many peptides must be developed. Peptides have favorable properties for use as radiolabeled tracers, particularly high receptor affinity and rapid plasma clearance. However, under physiological conditions, they are rapidly metabolized in plasma by endogenous peptidases to terminate their biological effect. One interesting peptide, viz., bombesin (BBS), was selected for this study. BBS is a tetradecapeptide [pGlu-Gln-Arg-Leu-Gly-Asn-Gln-Trp-Ala-Val-Gly-His-Leu-Met-NH₂, the carboxyterminal sequence BBS(8–14) is responsible for the binding] originally isolated from the skin of the amphibian Bombina orientalis [7]. It shows high affinity to the gastrin-releasing peptide (GRP) receptor, a G protein-coupled receptor characterized by the typical configuration of seven transmembrane domains [8]. BBS and its mammalian counterpart GRP produces a wide spectrum of biological responses in the central nervous system as well as in peripheral tissues [9], [10]. In addition to the physiological effects, it has been established that GRP/BBS promotes proliferation in normal and human cancer cell lines. For example, it was shown that BBS can stimulate cell proliferation in the androgen-independent human prostate carcinoma cell line PC-3 [11], [12]. Overexpression of the GRP receptor has been demonstrated in a large number of tumors, such as prostate and breast cancers, which are among the leading causes of cancer death [13], [14], [15], [16], [17]. Radiolabeled BBS peptides might therefore be useful for GRP receptor scintigraphy. High uptake of the diagnostic tracer in nonoperable tumors is the most important criterion for subsequent radionuclide therapy. Some analogues have been previously studied for labeling with rhenium 188 ([¹⁸⁸Re]) and technetium 99m ([^99mTc]) [18], [19], [20], [21]. As mentioned above, the rapid in vivo degradation can restrict their potential use as radiopharmaceuticals. The problem of a poor in vivo stability may be solved by modifying the sequence, and many known modifications of small peptides are available. However, the design of an analogue with a high tumor uptake compared to healthy tissues is much more difficult. High in vivo stability is a necessary prerequisite but it is not sufficient to reach a good target to nontarget ratio. It seems that this ratio can be improved by insertion of a spacer between the chelator used for labeling and the binding sequence of BBS. It is not yet clear whether the properties of the spacer group and the properties of the radiolabeled complex independently influence GRP receptor binding and uptake in cancer cells. The optimal length of the spacer was determined by Smith et al [22] who investigated the pharmacokinetics and pancreatic accumulation of different analogues with a series of aliphatic spacer groups ranging from 0 to 11 carbon atoms. Analogues in which the aliphatic spacer group ranges from five to eight carbon atoms appear to be the most promising for targeting [22]. Since one objective of our work was the development of a radiopharmaceutical labeled with fac-[^99mTc(H₂O)₃(CO)₃]⁺ [23], [24], we selected the chelator retroN^α-carboxymethyl histidine [(N^αHis)Ac] and obtained the neutral complex [^99mTc(CO)₃((N^αHis)Ac)]. Many complexes used by other groups are either positively or negatively charged. We observed solubility problems with our BBS analogues and therefore decided to introduce more polar spacers than the aliphatic spacers used previously.

New BBS analogues with different modifications have been synthesized and labeled with fac-[^99mTc(H₂O)₃(CO)₃]⁺. BBS-IV, BBS-IX, BBS-XXX have no spacer, whereas BBS-38 and BBS-39 have -βAla–βAla- and 3,6-dioxa-8-aminooctanoic acid as spacers, respectively (Table 1). In order to improve metabolic stability, two natural amino acids of the sequence have been replaced by nonnatural amino acids. These modifications consisted of the replacement of Leu¹³ by cyclohexylalanine (Cha) and Met¹⁴ by norleucine (Nle) (BBS-IV and BBS-IX, respectively) or both modifications in the same molecule (BBS-XXX, BBS-38 and BBS-39). Here, we present the synthesis of the new analogues as well as their chemical and pharmacological characterization. This includes specifically the metabolic stability in vitro, binding properties, internalization, externalization and the in vivo biodistribution in nude mice bearing human PC-3 tumor xenografts. [^99mTc] was used for the pharmacological investigation as it has a low radiation burden and is readily available. However, this radionuclide is not useful (even with a very high amount of radioactivity) for chemical analysis. Therefore, the characterization by mass spectrometry (MS) was done with the cold Re-labeled compounds. Our analogues are more lipophilic than BBS-II, the unchanged analogue, and show an insufficient solubility in water, which made obtaining nuclear magnetic resonance (NMR) data difficult. We decided to include the NMR data for BBS-II for comparison. Another reason for the use of the cold Re-labeled compounds for characterization is that our ultimate goal is radiotherapy with the [¹⁸⁸Re]- or [¹⁸⁶Re]-labeled analogues.