Characterization and evaluation of DOTA-conjugated Bombesin/RGD-antagonists for prostate cancer tumor imaging and therapy

Abstract

Introduction

Here we present the metallation, characterization, in vivo and in vitro evaluations of dual-targeting, peptide-based radiopharmaceuticals with utility for imaging and potentially treating prostate tumors by virtue of their ability to target the α_Vβ₃ integrin or the gastrin releasing peptide receptor (GRPr).

Methods

[RGD-Glu-6Ahx-RM2] (RGD: Arg-Gly-Asp; Glu: glutamic acid; 6-Ahx: 6-amino hexanoic acid; RM2: (D-Phe-Gln-Trp-Ala-Val-Gly-His-Sta-Leu-NH2)) was conjugated to a DOTA (1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid) bifunctional chelator (BFCA) purified via reversed-phase high-performance liquid chromatography (RP-HPLC), characterized by electrospray ionization-mass spectrometry (ESI-MS), and radiolabeled with ¹¹¹In or ¹⁷⁷Lu. Natural-metallated compounds were assessed for binding affinity for the α_Vβ₃ integrin or GRPr in human glioblastoma U87-MG and prostate PC-3 cell lines and stability prior to in vivo evaluation in normal CF-1 mice and SCID mice xenografted with PC-3 cells.

Results

Competitive displacement binding assays with PC-3 and U87-MG cells revealed high to moderate binding affinity for the GRPr or the α_Vβ₃ integrin (IC₅₀ range of 5.39 ± 1.37 nM to 9.26 ± 0.00 nM in PC-3 cells, and a range of 255 ± 47 nM to 321 ± 85 nM in U87-MG cells). Biodistribution studies indicated high tumor uptake in PC-3 tumor-bearing mice (average of 7.40 ± 0.53% ID/g at 1 h post-intravenous injection) and prolonged retention of tracer (mean of 4.41 ± 0.91% ID/g at 24 h post-intravenous injection). Blocking assays corroborated the specificity of radioconjugates for each target. Micro-single photon emission computed tomography (microSPECT) confirmed favorable radiouptake profiles in xenografted mice at 20 h post-injection.

Conclusions

[RGD-Glu-[¹¹¹In-DO3A]-6-Ahx-RM2] and [RGD-Glu-[¹⁷⁷Lu- DO3A]-6-Ahx-RM2] show favorable pharmacokinetic and radiouptake profiles, meriting continued evaluation for molecular imaging in murine U87-MG/PC-3 xenograft models and radiotherapy studies with ¹⁷⁷Lu and ⁹⁰Y conjugates.

Advances in Knowledge and Implications for Patient Care

These heterovalent, peptide-targeting ligands perform comparably with many mono- and multivalent conjugates with the potential benefit of increased sensitivity for detecting cancer cells exhibiting differential expression of target receptors.

Introduction

Despite decades of research, prostate cancer remains an important disease; in 2014, diagnoses are expected to represent 27% of all new cancer cases in American men, exceeding new diagnoses for all other types of male cancers. In addition, it continues to fall behind only lung cancer with regard to cancer-related mortality, with nearly 30,000 directly attributable deaths estimated for 2014 [1].

Although direct mortality from non-metastatic prostate cancer is relatively low, the grave prognosis, excruciating pain, and increasing costs of palliative therapy associated with chronic and metastatic stages [2], [3] motivate the pursuit of innovative and more effective methods for earlier and accurate detection and therapy, as well as approaches for differentiating indolent from aggressive forms of the disease. Surgery and various radiation- and chemically-based modalities often fail to prevent the disease from eventually becoming refractory to hormone-centered therapy and achieving the metastasis which accounts for the majority of the complications and mortalities associated with prostate cancer [4], [5]. Furthermore, treatments themselves may cause side effects due to the inherently imprecise nature in which they are delivered, and diagnostic methods such as serum prostate specific antigen (PSA) and needle biopsy historically have been controversial in their utility for accurate, early detection; analysts contend that these methods result in missed diagnoses in addition to misdiagnoses and unnecessary treatments; conversely, others maintain that they have demonstrated a beneficial impact on overall mortality [6], [7], [8], [9], [10], [11], [12]. Although the recently FDA-approved Xofigo™, has been shown to improve overall survival by controlling metastatic bone lesions [13], it is not useful as a diagnostic agent. However, it and other radiopharmaceuticals provide evidence that peptide radioligands can be very selective and demonstrate the required affinity for use in imaging. The selective, over-expression of peptide receptors in human tumors is often exploited in targeting strategies for imaging and treating various tumors with radionuclide therapy [14], [15]. It is well-established that gastrin releasing peptide is a potent epithelial mitogen and that its receptor, the GRPr—also known as subtype 2 of the bombesin (BBN) receptor superfamily—is one such biomarker greatly over-expressed in a variety of human cancers, including prostate cancer (although not in benign prostatic hyperplasia). Peptide-based targeting radiopharmaceuticals can be easily synthesized and tailored to specifically target biomarkers that are expressed predominantly on tumor cells of certain human cancers and negligibly on normal tissue cells. Furthermore, they have enhanced ability to penetrate tumors and tumor vasculature and, with appropriate modifications to their structure, are quickly cleared from the blood and non-target tissues without appreciable immunogenicity [14], [15], [16].

These characteristics provide strong motive for the continued development and refinement of this category of diagnostic and therapeutic modalities, and there is much opportunity for investigating the numerous, unexplored combinations of their structural components to favorably alter their in vivo pharmacokinetic profiles. While there are novel diagnostic and therapeutic modalities being developed, it is important that the progress and investment made thus far be built upon, as there may yet be much to attain. It is encouraging to note that small differences in otherwise very similar compound structures can lead to surprising differential capability for binding specificity, affinity and overall in vivo efficacy. For instance, Lane et al. and multiple other investigators have shown that minor differences in linking moieties can manifest profound variability in a single parent compound's tumor-to-normal tissue uptake, as well as its mode and timing of excretion [17], [18].

Researchers continue to employ radiolabeled, bombesin-based agonist and antagonist ligands to target cancers for imaging and therapy because of their high binding affinity for the GRPr, accounting for the abundance of literature pertaining to the subject [19], [20], [21], [22], [23], [24], [25], [26], [27], [28], [29], [30], [31]. BBN-based molecules that target the GRPr in an agonistic fashion mimic the endogenous ligand known as gastrin-releasing peptide and are among the most utilized ligands for targeting human cancers. RM2 is an antagonist analog of BBN that has recently been shown to have improved uptake and retention in tumors as compared to agonistic-type GRPr-targeting ligands [32]. More recent targets of interest include members of the integrin receptor family, among them α_Vβ₃, a suspect in the development of metastasis due to roles in cell attachment, cell-to-extracellular matrix interactions and angiogenesis [33], [34], [35], and the fact that it is expressed in many tumors, including malignant melanoma, glioblastoma, and breast as well as prostate. However, integrins are only moderately expressed on the surfaces of some prostate tumors. For example, the GRPr density per PC-3 cell averages approximately 2.5 × 10⁵ [36], whereas in comparison, the α_Vβ₃ integrin receptor density per cell has been found to be only 2.8 × 10³ on PC-3 prostate cancer cells in in vitro assays [37]. It is important to acknowledge, however, that these receptor numbers may not accurately reflect the clinical presentations of human prostate cancers.

Because both of the aforementioned receptors are expressed only minimally on normal human cells, they are potentially ideal biomarkers for which we can design radiolabeled BBN/RGD derivatives for molecular imaging and targeted radionuclide therapy. Monomeric and multimeric RGD peptides have shown the ability to specifically target the α_Vβ₃ integrin [35], [38], [39], [40], and a number of RGD-based targeting vectors have been radiolabeled with ^99mTc, ¹¹¹In, ⁶⁸Ga, ¹⁸ F and ⁶⁴Cu radionuclides and evaluated for targeting tumors that express it [38], [39], [41], [42], [43], [44], [45], [46], [47], [48], [49]. The emerging concept of simultaneous, multi-receptor targeting with a single radiocompound brings fresh opportunities for advancement in targeted radiopeptide theranostics. Monovalent radioligands for prostate cancer imaging and therapy have been investigated by multiple research groups, and have traditionally been modeled after a single, endogenous peptide hormone, small peptide, or antibody, which begets the tendency for them to bind only to those cancer cells which adequately express the target molecule. However, results of new heterodimeric peptide conjugates have shown improved molecular imaging properties when compared to monomeric peptide constructs [48], [50], [51], [52], [53], [54], presumably a consequence of an effective increase in the total number of receptors available to which the heterodimer may bind. For example, the total number of receptors available for binding a heterodimeric radioligand would effectively be the sum of all GRPr and α_Vβ₃ receptors on cell surfaces and should be higher than that of either monovalent GRPr-/α_Vβ₃ single-receptor targeting cells of interest [55]. Heterodimers targeting α_Vβ₃ and GRPr have been shown to be superior to monovalent GRPr/α_Vβ₃-targeting ligands, which can be limited by one or the other molecular targets/biomarkers being expressed either in very low numbers or not at all [51], [52], [53], [54]. Differential target receptor expression of cancer cells among individual patients (or even within the same patient) poses an obstacle with which we must contend; this variable receptor expression may also be a function of disease stage (e.g., early versus chronic or metastatic). To meet this challenge, it would be ideal to develop compounds capable of binding to more than one relevant target with very good specificity and affinity. Such a feature should lead to enhanced accumulation and retention that results in greater sensitivity and also helps mitigate stage-variable expression. This is not only a solid approach for managing the long-standing problems of late and missed diagnoses, but also provides more accurate diagnostic profiling for individual patients that could perhaps be altered via radiometal-chelator modifications for use as an efficient, normal tissue-sparing therapeutic option.

The PC-3 cell line is a well-characterized, androgen-independent, human prostate cancer cell line that co-expresses GRPr and integrin α_Vβ₃ [35] (although the latter to a much lesser extent than the former), which is our motivation for its use in the current study. Because the human glioblastoma cell line U87-MG is known to express the integrin target in far greater numbers, it is also used to evaluate our dual-targeting compounds. In the present study, we have designed and synthesized two, radiolabeled RGD-RM2 heterodimeric peptide antagonists for dual receptor targeting, referred to as [RGD-Glu-[¹¹¹In-DO3A]-6-Ahx-RM2] and [RGD-Glu-[¹⁷⁷Lu-DO3A]-6-Ahx-RM2].

¹⁷⁷Lu- and ¹¹¹In-radiolabeled radiopharmaceuticals have been of interest to our group and many others due to their ideal nuclear characteristics; each has a half-life, decay mode and emission profile suitable for imaging, and the beta emission of ¹⁷⁷Lu also renders it useful as a radiotherapeutic. ¹¹¹In is produced via cyclotron in a (p,2n) reaction and decays primarily by electron capture, producing two gamma photons (171 keV and 245 keV) which fall into an ideal range for imaging purposes. A physical half-life of 2.8d also makes ¹¹¹In ideal for molecular imaging. For example, since it is readily available, it is easily procured within a reasonable time to allow for drug preparation, quality control, drug delivery and biodistribution. ¹¹¹In has been extensively used to label peptide- and antibody-based compounds for many years and is, perhaps, one of the more widely-used radionuclides for diagnostic imaging—most notably in the case of ¹¹¹In-DTPA-Octreotide (Octreoscan®) [56]. ¹⁷⁷Lu has also become a favorite radionuclide for use in nuclear medicine for the labeling and improvement of both monovalent and multivalent compounds [57], [58], [59], [60], [61], [62], [63], [64], [65], [66]. As a reactor-produced product, ¹⁷⁷Lu can be obtained in a high specific activity/carrier-free state from a ¹⁷⁶Yb target via indirect neutron capture (n,gamma) [67] as to avoid the presence of long-lived, high-energy, metastable isotopes. However, it is most frequently synthesized in moderate specific activity [68] via direct neutron capture using a ¹⁷⁶Lu-enriched target. In addition to its two gamma rays that are suitable for imaging (113 keV and 208 keV), ¹⁷⁷Lu emits a beta particle with an energy of 498 keV that can achieve a tissue penetration depth of ~ 2 mm, making it excellent for use with smaller size tumors [61], [69]. In addition, the 6.7d half-life makes it conducive for use as a diagnostic or therapeutic agent. Both ¹¹¹In and ¹⁷⁷Lu lend themselves well to successful, inert chelation within the DOTA bifunctional chelator as a result of their inherent chemical properties, such as 3 + oxidation states, hard metal centers and larger-sized ionic radii.