Review 29 July 2020 , , , , and 1 Biomedical Chemistry, Department of Clinical Radiology and Nuclear Medicine, Medical Faculty Mannheim of Heidelberg University, 68167 Mannheim, Germany 2 Molecular Imaging and Radiochemistry, Department of Clinical Radiology and Nuclear Medicine, Medical Faculty Mannheim of Heidelberg University, 68167 Mannheim, Germany 3 Department of Oncology, Division of Oncological Imaging, University of Alberta, Edmonton, AB T6G 1Z2, Canada 4 Institute of Pharmacy and Molecular Biotechnology, University of Heidelberg, 69120 Heidelberg, Germany
Over the past few years, an approach emerged that combines different receptor-specific peptide radioligands able to bind different target structures on tumor cells concomitantly or separately. The reason for the growing interest in this special field of radiopharmaceutical development is rooted in the fact that bispecific peptide heterodimers can exhibit a strongly increased target cell avidity and specificity compared to their corresponding monospecific counterparts by being able to bind to two different target structures that are overexpressed on the cell surface of several malignancies. This increase of avidity is most pronounced in the case of concomitant binding of both peptides to their respective targets but is also observed in cases of heterogeneously expressed receptors within a tumor entity. Furthermore, the application of a radiolabeled heterobivalent agent can solve the ubiquitous problem of limited tumor visualization sensitivity caused by differential receptor expression on different tumor lesions. In this article, the concept of heterobivalent targeting and the general advantages of using radiolabeled bispecific peptidic ligands for tumor imaging or therapy as well as the influence of molecular design and the receptors on the tumor cell surface are explained, and an overview is given of the radiolabeled heterobivalent peptides described thus far.
Whole-body imaging techniques are indispensable tools for the characterization of physiological as well as pathological conditions in daily clinical patient care. In particular, molecular imaging, comprising the nuclear medicine imaging modalities positron emission tomography (PET) and single photon emission computed tomography (SPECT), offers the advantage of tissue characterization on a functional level, thus enabling the detection and characterization of functional changes before morphological alterations can be detected using magnetic resonance imaging (MRI) and computed tomography (CT). The most precise and meaningful information can be obtained by combining functional and morphological imaging modalities (e.g., PET/CT, SPECT/CT, or PET/MRI), representing an important basis for sensitive and specific clinical diagnosis of different diseases.
As functional imaging with PET and SPECT plays such a pivotal role for the detection and characterization of neurologic, cardiologic, and especially oncologic pathologies, the number of radiotracers able to address specific pathologic functional changes has been steadily growing over the last decades. As a result, the research for new radioligands, enabling the visualization target structures with higher sensitivity and specificity continues unabatedly. In general, every bioactive compound accumulating in target cells and tissues can form—when radiolabeled with a suitable β+- or γ-emitter—a valuable imaging agent for PET or SPECT. Thus, different compound classes have been used for the development of new imaging agents for PET and SPECT, such as small molecules, peptides, peptide mimetics, RNAs, and antibodies.
The imaging of malignant diseases with high specificity mostly utilizes radiolabeled peptides as imaging agents as this substance class is—besides antibodies—able to bind to cell surface receptors that are overexpressed by the respective malignancy with high affinity and specificity, thus allowing for the accurate discrimination between benign and malignantly transformed tissue. Furthermore, peptides usually exhibit low toxicities and immunogenicities, and are readily synthesized and chemically modified to produce homogeneous products with tailored properties. Moreover, they show—in contrast to also highly target-specific and -affine antibodies—a favorably fast tissue penetration, target accumulation, and non-target tissue clearance and thus highly advantageous pharmacokinetic properties. Thus far, numerous radiolabeled peptide drugs have been developed for both diagnostic imaging of peptide receptor expression or peptide receptor radionuclide therapy (PRRT) [1,2,3,4,5,6,7,8].
Over the last years, a large number of radiolabeled peptides were developed for diagnostic imaging and therapy, including multivalent peptides consisting of more than one copy of the targeting peptide. This peptide homomultimer concept bears several advantages. First, the metabolic stability of peptide multimers is commonly increased compared to the respective monomers. This can be attributed to the usually contained artificial structure elements hampering a degradation by endogenous peptidases and to the steric demand of the constructs, resulting in steric hindrance towards peptidase attacks [9,10,11,12]. This results in a longer bioavailability of the compound and thus increased probability of multimer receptor binding. Second, higher target avidities (avidity is the affinity of a ligand being able to bind with more than one target-affine binding unit) are generally obtained, resulting in tighter target binding, higher tumor uptakes combined with higher tumor-to-background ratios and a prolonged tumor retention [13,14,15,16,17,18].
The increased avidity of peptide multimers arises from different factors. One is the possibility of the multimer to concomitantly bind the target cell with more than one peptide copy, resulting in tighter target binding. Such concomitant binding also decreases the possibility of complete ligand dissociation. Even if one interaction is lost, other binding interactions remain and ensure target association. In this case, the effect termed “forced proximity” becomes particularly significant—the unbound peptide copies of the multimer stay near the cell surface due to the one ligand bound to its receptor, forcing the unbound peptides of the multimer to stay in reach of other free target receptors. This increases the probability of unbound peptides of the multimer to interact with other free receptors on the tumor cell surface, while increasing the probability of re-binding in case of ligand dissociation. These effects are particularly pronounced when receptor clustering is triggered by peptide binding [19], further increasing the local concentration of receptors near the site of initial peptide binding, and therefore the probability of further binding events.
Figure 1. Schematic depiction of the “forced proximity” effect. Considering a peptide multimer having already bound with one peptide to a receptor on the cell surface, unbound peptide copies of the same multimer are located near the cell in close proximity of other free target receptors, increasing the probability of further peptide–receptor interactions or re-binding upon ligand dissociation.
Due to these favorable effects of peptide multimerization on target interaction and thus tumor targeting, this approach was adapted to different peptide systems such as RGD-based peptides (RGD = peptide with the sequence Arg-Gly-Asp) targeting integrin α v β 3 [9,10,15,17,20,21,22,23,24,25,26,27,28], bombesin analogs binding the gastrin-releasing peptide receptor (GRPR) [29,30,31,32], somatostatin derivatives binding somatostatin receptors (SSTRs) [33,34,35], analogs of α-melanocyte-stimulating hormone (α-MSH) addressing melanocortin receptors [36,37,38,39], minigastrin analogs binding to the cholecystokinin-2 receptor (CCK2R) [11,12], and neurotensin derivatives targeting neurotensin receptors [40,41,42].
However, a severe limitation for receptor targeting persists, irrespective of the valency of the applied peptide—not every tumor of a certain entity expresses or overexpresses a certain receptor and inter- and intra-individual differences can occur [43,44,45,46], limiting the sensitivity of tumor visualization and therapy response. In addition, the receptor status can change significantly upon disease progression (by tumor de-differentiation and metastasis [44,47,48] or induced by tumor therapy [49,50,51]), further limiting successful tumor targeting. To give an example, human breast cancer cells overexpress the GRPR and the neuropeptide Y receptor subtype 1 (NPY(Y 1)R) in about 75% and 66–85% of all cases, respectively [52,53]. Thus, a GRPR- or NPY(Y 1)R-monospecific peptide binder, irrelevant if monovalent or multivalent, can only address a portion of breast cancer lesions. Given the fact that the vast majority of these tumors (93% [54]) overexpresses at least one of them or even both, the direction for the development of highly specific and also sensitive radiolabeled imaging agents becomes clear. Combining two peptides that are able to specifically bind to the GRPR or the NPY(Y 1)R into one heterobivalent radioligand enables a considerably higher cancer targeting efficiency and sensitivity compared to the respective monospecific radioligands.
This concept is not only promising for human breast cancer but for several cancer types where concomitant receptor targeting is warranted. Of course, the heterobivalent or heteromultivalent agents exhibit the same superior properties as homobivalent or homomultivalent peptides compared to their respective peptide monomers and additionally are able to bind to different targets, thus increasing the overall probability of target binding and visualization.
In general, radiotracers being composed of two different target affine ligands do not necessarily need to address two different structures such as two different receptors (such substances are termed as “heterobivalent” ligands, Figure 2B), but may also contain two different agents addressing the same structure (e.g., one single receptor type) by at least one allosteric interaction. This latter substance class is termed as “bitopic” ligands (Figure 2A).
Figure 2. Schematic depiction of bitopic (A) vs. heterobivalent (B) binding. In the case of bitopic binding, two different binding sites (at least one being allosteric) on the same receptor are addressed by the radioligand, whereas in the case of heterobivalent binding, two orthosteric binding sites of two different receptors are bound.
Although bitopic ligands show tighter target binding compared with monomers and exhibit most of the advantages of homobivalent ligands (vide supra) [55], they cannot take full advantage of the benefits of the heterobivalency approach as they can address only one structure, for example, one tumor-specific receptor, on the target cell (even though it is bound with higher probability and avidity), not resulting in higher sensitivity of target visualization when receptor densities vary between lesions.
Thus, radiolabeled heterobivalent (or heteromultivalent) peptidic agents are most promising for the highly specific and sensitive imaging and therapy of malignant diseases and are thus in the focus of this review. Although only the terms “heterobivalent” or “heterodimeric” will be used in the following review, the same considerations apply to “heteromultivalent” or “heteromultimeric” systems. However, heteromultimers are very rarely described and most of the work was conducted on heterodimeric systems.
As mentioned above, heterobivalent peptidic ligands add several further advantages to those observed for peptide homomultimers (vide supra). In particular, they can bind to different receptors concomitantly or independently overexpressed on the target tissue. This is especially useful in tumor entities or lesions showing no homogenous receptor overexpression between individuals or even in the same patient, which has been shown systematically for several cancer types [54,56,57,58] and should be of general relevance for various other cancers.
Tumors with a non-uniform distribution of target receptors can also be visualized by monospecific peptidic agents but can show an insufficient tumor detection sensitivity due to the partial absence of the target receptor, usually resulting in faint accumulation, or worse, an incomplete detection of all lesions. In principle, it would be possible to address this issue by applying two different monospecific peptidic radioligands. However, this approach has some significant drawbacks. First, the regulatory hurdles were much higher for the examination of a patient using two different radiotracers—if they were injected concurrently, this would result in difficulties regarding dynamic analyses and clinical interpretation of the data as which tracer produced which signal cannot be discriminated. If the tracers were applied consecutively, this would mean two separate examinations, which would be uncomfortable for the patient and furthermore costly. In contrast, a heterobivalent tracer would in contrast have to be applied only once, and even more importantly, some of the intrinsic advantages of heterobivalent ligands such as the higher stability and overall avidity of the combined agents can be taken advantage of.
Furthermore, heterobivalent agents being able to address two different receptor types exhibit a higher overall target binding probability due to the absolute higher number of target receptors and thus a higher tumor visualization rate which is of special importance in cases wherein there is only a moderate receptor expression on the tumor tissue.
Another advantage that was proposed for peptide heterodimers—apart from target avidity—is their higher target specificity compared to two monospecific binders. In particular, when two peptides are combined wherein both exhibit a moderate target affinity, a higher tumor specificity can be achieved as benign tissues only express the target receptors in low density whereas the malign cells can express both receptors that can be bound by the hete
