Chelating the Alpha Therapy Radionuclides ²²⁵Ac³⁺ and ²¹³Bi³⁺ with 18-Membered Macrocyclic Ligands Macrodipa and Py-Macrodipa

Abstract

The radionuclides ²²⁵Ac³⁺ and ²¹³Bi³⁺ possess favorable physical properties for targeted alpha therapy (TAT), a therapeutic approach that leverages α radiation to treat cancers. A chelator that effectively binds and retains these radionuclides is required for this application. The development of ligands that can be used for this purpose, however, is challenging because the large ionic radii and charge-diffuse nature of these metal ions give rise to weaker metal-ligand interactions. In this study, we evaluated two 18-membered macrocyclic chelators, macrodipa and py-macrodipa, for their ability to complex ²²⁵Ac³⁺ and ²¹³Bi³⁺. Their coordination chemistry with Ac³⁺ was probed computationally and with Bi³⁺ experimentally via NMR spectroscopy and X-ray crystallography. Furthermore, radiolabeling studies were conducted, revealing the efficient incorporation of both ²²⁵Ac³⁺ and ²¹³Bi³⁺ by py-macrodipa that matches or surpasses the well-known chelators macropa and DOTA. Incubation in human serum at 37 °C showed that ~90% of the ²²⁵Ac³⁺–py-macrodipa complex dissociates after 1 d. The Bi³⁺–py-macrodipa complex possesses remarkable kinetic inertness in an EDTA transchelation challenge study, surpassing that of Bi³⁺–macropa. This work establishes py-macrodipa as a valuable candidate for ²¹³Bi³⁺ TAT, providing further motivation for its implementation within new radiopharmaceutical agents.

Graphical Abstract

Targeted alpha therapy (TAT) is a promising therapeutic strategy that leverages α-particle-emitting radionuclides to annihilate tumor cells. Compared to conventional internal radiotherapy using β-particle emitters, the implementation of significantly more massive α particles, which deposit their energy over much shorter distances, provides key advantages. The short range of α radiation can yield enhanced selectivity for targeted cancer cells, while minimizing damage to surrounding healthy cells. Moreover, the very large linear energy transfer (LET) of α particles is significantly more effective in causing lethal DNA double strand breaks that kill cancer cells in a more efficacious manner compared to the lower-LET β particles.^1-6

To date, over eight radionuclides have been identified as potential candidates for use in TAT based on their decay properties and production routes.⁷ Among these nuclides, ²²⁵Ac³⁺ and ²¹³Bi³⁺ have received considerable attention that has manifested in clinical studies.^{8-10 225}Ac (t_1/2 = 9.9 d) emits four α particles through its decay chain, a property that confers it with high cytotoxic potency. Its 9.9-day half-life is also well matched with the in vivo circulation timescales of macromolecular targeting vectors like antibodies.^{11,12 213}Bi (t_1/2 = 45.6 min), a daughter of ²²⁵Ac³⁺, emits one α particle through its decay chain and can be conveniently obtained from ²²⁵Ac/²¹³Bi generators.¹³ Its shorter half-life can be optimally matched to small-molecule targeting vectors, rendering it useful for different systems than those used for ²²⁵Ac³⁺.^14,15

To convert these promising radionuclides into useful radiotherapeutic agents, a chelator that efficiently binds and stably retains them is required.^16,17 The development of chelators for large metal ions like Ac³⁺ and Bi³⁺, however, is challenging, partly because their low charge density weakens electrostatic interactions with ligand donor atoms.

We recently reported a new ligand called macrodipa¹⁸ and its second-generation analogue py-macrodipa¹⁹ (Chart 1). These “macrodipa-type” chelators feature a unique “dual size selectivity”, characterized by their good affinities for both the large and small rare-earth metal ions (Ln³⁺). This unusual selectivity profile arises from a significant conformational toggle that occurs when they form complexes with Ln³⁺ ions of different sizes. Large Ln³⁺ form 10-coordinate, nearly C₂-symmetric complexes (Conformation A), whereas an 8-coordinate, asymmetric complex arises for small Ln³⁺ (Conformation B).^18,19 We have further demonstrated that this property makes py-macrodipa a valuable candidate for nuclear medicine applications with both ¹³⁵La³⁺ and ⁴⁴Sc³⁺, Ln³⁺ radiometal ions with the largest and smallest ionic radii within this series.¹⁹

Chart 1.

Structures of Chelators Discussed in This Work.

Based on this successful application of macrodipa and py-macrodipa for the Ln³⁺ ions, we sought to evaluate these ligands with biomedically relevant ions beyond the Ln³⁺ series, namely Ac³⁺ and Bi³⁺. The potentials of both chelators for TAT applications using their radioisotopes ²²⁵Ac³⁺ and ²¹³Bi³⁺ were determined and benchmarked to those of the well-known chelators macropa and DOTA (Chart 1), which have established precedence for nuclear medicine applications with these radiometals.^20-23

We assessed the coordination chemistry of these ligands with stable Bi³⁺. The ¹H and ¹³C{¹H} NMR spectra of their Bi³⁺ complexes (Bi³⁺–macrodipa and Bi³⁺–py-macrodipa) were acquired in D₂O (Figures 1 and S1-S4). These spectra reveal the presence of a single, well-resolved species that lacks symmetry for both complexes. Thus, Bi³⁺–macrodipa and Bi³⁺–py-macrodipa most likely attain the asymmetric Conformation B, which is the preferred binding mode of these ligands for small Ln³⁺ (Figure S5-S6).

Figure 1.

¹H NMR spectra of Bi³⁺–macrodipa and Bi³⁺–py-macrodipa (500 MHz, D₂O, pD 5, 25 °C).

As further validation, we characterized Bi³⁺–macrodipa and Bi³⁺–py-macrodipa by X-ray crystallography (Figure 2). The crystal structures of these complexes confirm that they attain the asymmetric Conformation B, consistent with our observations from NMR spectroscopy. Like their NMR spectra, these Bi³⁺ structures are comparable to those of the small Ln³⁺ analogues, Lu³⁺–macrodipa and Sc³⁺–py-macrodipa, with respect to the orientation of the picolinate donors and the lack of full engagement of all six macrocycle donor atoms.^18,19 A key difference between these Ln³⁺ and Bi³⁺ structures, however, is the absence of a coordinated water molecule in the latter. This void is most likely a consequence of the stereochemical activity^24,25 of the Bi³⁺ 6s² lone pair. These observations that Bi³⁺–macrodipa and Bi³⁺–py-macrodipa attain the asymmetric Conformation B rather than the symmetric Conformation A is somewhat surprising based on the similar ionic radii of Bi³⁺ and La³⁺,^26,27 a representative large Ln³⁺. This result suggests that the stereochemical activity of the 6s² lone pair plays a pronounced role in mediating the preferred conformations of these Bi³⁺ complexes.

Figure 2.

Crystal structures of (a) [Bi(macrodipa)]⁺ and (b) [Bi(py-macrodipa)]⁺. Thermal ellipsoids are drawn at the 50% probability level. Solvent and counterions are omitted for clarity.

Experimental characterization of Ac³⁺ complexes is challenging due to the high radioactivity and extremely limited availability of its longest-lived isotope ²²⁷Ac (t_1/2 = 21.8 y).²⁸ Thus instead, we probed the structures of Ac³⁺–macrodipa and Ac³⁺–py-macrodipa computationally using density functional theory (DFT) with Gaussian 16.²⁹ The hybrid TPSSh functional,³⁰ which has been validated for studying Ac³⁺ chemistry,^31,32 was adopted. A large-core relativistic effective core potential (LCRECP) and the associated basis set was assigned to the Ac³⁺ center,^33-35 whereas the 6-31G(d,p) basis set^36,37 was applied to all other lighter atoms. Aqueous solvation effects were accounted for with the SMD solvation model.³⁸

Because the ionic radii and coordination chemistry of Ac³⁺ and La³⁺ are similar,²⁸ we optimized Ac³⁺–macrodipa and Ac³⁺–py-macrodipa starting from the geometries of the corresponding La³⁺ complexes, which attain the symmetric Conformation A.^18,19 Within these structures (Figure 3), the Ac–O interatomic distances are 2.45–2.48 Å for negatively charged O and 2.70–2.79 Å for neutral O, whereas the Ac–N interactions range from 2.76–2.92 Å. These calculated distances are in expectation with experimentally measured Ac–O and Ac–N interatomic distances.^39-43 Additionally, we optimized both complexes in Conformation B. Consistent with our expectations, Conformation B is energetically unfavored for both complexes (Table S2).

Figure 3.

DFT-optimized structures of (a) [Ac(macrodipa)]⁺ and (b) [Ac(py-macrodipa)]⁺. Hydrogen atoms are omitted for clarity. Green: Ac, grey: C, blue: N, red: O.

Having established the coordination chemistry of these ligands, we next carried out radiolabeling studies to evaluate their potential value for ²²⁵Ac³⁺ and ²¹³Bi³⁺ TAT in comparison to the state-of-the-art chelators macropa and DOTA. These radionuclides were produced and purified according to previously-described protocols.^44-46

Different concentrations of macrodipa, py-macrodipa, macropa, and DOTA were combined with pH 5.5–6 buffered solutions containing either 20–40 or 30–300 kBq of ²²⁵Ac³⁺ and ²¹³Bi³⁺ at ambient or elevated temperature, and the radiochemical yields (RCYs) were determined by radio-TLC. The concentration-dependent RCYs for these four chelators are summarized in Figure 4. For both radionuclides, py-macrodipa is able to achieve significantly higher RCYs than its analogue macrodipa and the conventional chelator DOTA, which also required high temperatures for radiolabeling. RCYs of approximately 75% and 65% are obtained when using low py-macrodipa concentrations of 10⁻⁶ M and 10⁻⁸ M for ²²⁵Ac³⁺ and ²¹³Bi³⁺, respectively. With respect to ²²⁵Ac³⁺ chelation, py-macrodipa was slightly less effective than macropa, but was better at radiolabeling ²¹³Bi³⁺. We also performed ²²⁵Ac³⁺ radiolabeling with macrodipa and py-macrodipa at pH 7 (Table S3). Under this condition, both chelators were able to access greater RCYs, but still failed to surpass macropa. Overall, these studies show that py-macrodipa effectively radiolabels both ²²⁵Ac³⁺ and ²¹³Bi³⁺ under mild conditions.

Figure 4.

Radiochemical yields at different ligand concentrations. (a) RCYs of ²²⁵Ac³⁺ radiolabeling (25 °C for py-macrodipa, macropa, 40 °C for macrodipa, and 80 °C for DOTA; pH 5.5–6; 60 min reaction time). (b) RCYs of ²¹³Bi³⁺ labeling (25 °C for macrodipa, py-macrodipa, macropa and 95 °C for DOTA; pH 5.5–6; 6–8 min reaction time). Error bars represent the standard deviations. The ²¹³Bi³⁺ data with macropa and DOTA was taken from Ref 21.

We next assessed the kinetic inertness of ²²⁵Ac³⁺–py-macrodipa by incubating it in human serum at 37 °C (Table S5). These studies show that ²²⁵Ac³⁺–py-macrodipa is fairly labile, as ~90% of the complex dissociated after 1 d. By contrast, ²²⁵Ac³⁺–macropa remained 98% intact in human serum after 5 d. This excellent kinetic inertness is consistent to a previously reported serum challenge on ²²⁵Ac³⁺–macropa.²⁰ Hence, despite the efficient radiolabeling properties of py-macrodipa, it is not an optimal candidate for TAT applications with ²²⁵Ac³⁺.

Because ²¹³Bi³⁺ decays quickly (t_1/2 = 45.6 min), probing the ²¹³Bi³⁺ complex kinetic inertness by this serum challenge assay is impractical. Instead, we performed a transchelation challenge assay^19-21,47-49 on the macrodipa, py-macrodipa, and macropa complexes with stable Bi³⁺. The transchelation reactions of these Bi³⁺ complexes were monitored by UV–Vis spectroscopy in the presence of a 10-fold excess EDTA, a ligand with high affinity for Bi³⁺,^50,51 at pH 5.0 and 25 °C. Under this condition, the Bi³⁺ ion is transchelated by EDTA, following pseudo-first-order kinetics. The resulting half-lives (t_1/2) for this transchelation process, a comparative measure of complex kinetic inertness, are shown in Table 1. Bi³⁺–macrodipa is kinetically labile to this transchelation challenge. The kinetic inertness of Bi³⁺–py-macrodipa is remarkably enhanced, as reflected by a t_1/2 of 13 d. Moreover, its inertness is greater than that of Bi³⁺–macropa, indicating that py-macrodipa is a promising candidate for TAT applications with ²¹³Bi³⁺.

Table 1.

Half-lives of Bi³⁺ Complexes when Challenged with 10 Equivalents of EDTA.^a

	t 1/2
Bi3+–macrodipa	9.2 ± 0.1 min
Bi3+–py-macrodipa	13.2 ± 1.2 d
Bi3+–macropa	2.2 ± 0.2 d

^a[BiL] = 100 μM, pH 5.0, 25 °C.

In summary, we evaluated the viability of macrodipa and py-macrodipa as chelators for ²²⁵Ac³⁺ and ²¹³Bi³⁺. Their coordination chemistry with Ac³⁺ and Bi³⁺ were characterized computationally and experimentally, respectively. Our radiolabeling studies revealed that py-macrodipa is highly effective at radiolabeling both radiometals, outperfoming both macrodipa and DOTA. Although the lability of Ac³⁺–py-macrodipa precludes its use with ²²⁵Ac³⁺ in nuclear medicine, the efficient formation and high stability of Bi³⁺–py-macrodipa, which surpasses Bi³⁺–macropa, suggests that this ligand is a valuable candidate for ²¹³Bi³⁺ chelation. These results highlight that py-macrodipa joins other promising candidates for ²¹³Bi³⁺ chelation that have arisen in recent years.^21,52-60 Ongoing work is directed towards the synthesis of a bifunctional analogue of py-macrodipa to apply this chelator in TAT, as well as the development of “macrodipa-type” chelators with enhanced Ac³⁺ complex stabilities.

SYNOPSIS.

The α-emitting radionuclides ²²⁵Ac³⁺ and ²¹³Bi³⁺ are promising candidates for targeted alpha therapy (TAT), a form of nuclear medicine that harnesses α radiation to kill cancer cells. Here, we investigate the chelation of these radiometals with the ligands macrodipa and py-macrodipa to assess their suitability for TAT. In particular, py-macrodipa is demonstrated to be a promising candidate for ²¹³Bi³⁺ chelation, surpassing the current state-of-the art chelators macropa and DOTA.