Reagents for astatination of biomolecules. 5. Evaluation of hydrazone linkers in (211)At- and (125)I-labeled closo-decaborate(2-) conjugates of Fab' as a means of decreasing kidney retention.

Evaluation of monoclonal antibody (mAb) fragments (e.g., Fab', Fab, or engineered fragments) as cancer-targeting reagents for therapy with the α-particle emitting radionuclide astatine-211 ((211)At) has been hampered by low in vivo stability of the label and a propensity of these proteins localize to kidneys. Fortunately, our group has shown that the low stability of the (211)At label, generally a meta- or para-[(211)At]astatobenzoyl conjugate, on mAb Fab' fragments can be dramatically improved by the use of closo-decaborate(2-) conjugates. However, the higher stability of radiolabeled mAb Fab' conjugates appears to result in retention of radioactivity in the kidneys. This investigation was conducted to evaluate whether the retention of radioactivity in kidney might be decreased by the use of an acid-cleavable hydrazone between the Fab' and the radiolabeled closo-decaborate(2-) moiety. Five conjugation reagents containing sulfhydryl-reactive maleimide groups, a hydrazone functionality, and a closo-decaborate(2-) moiety were prepared. In four of the five conjugation reagents, a discrete poly(ethylene glycol) (PEG) linker was used, and one substituent adjacent to the hydrazone was varied (phenyl, benzoate, anisole, or methyl) to provide varying acid sensitivity. In the initial studies, the five maleimido-closo-decaborate(2-) conjugation reagents were radioiodinated ((125)I or (131)I), then conjugated with an anti-PSMA Fab' (107-1A4 Fab'). Biodistributions of the five radioiodinated Fab' conjugates were obtained in nude mice at 1, 4, and 24 h post injection (pi). In contrast to closo-decaborate(2-) conjugated to 107-1A4 Fab' through a noncleavable linker, two conjugates containing either a benzoate or a methyl substituent on the hydrazone functionality displayed clearance rates from kidney, liver, and spleen that were similar to those obtained with directly radioiodinated Fab' (i.e., no conjugate). The maleimido-closo-decaborate(2-) conjugation reagent containing a benzoate substituent on the hydrazone was chosen for study with (211)At. That reagent was conjugated with 107-1A4 Fab', then labeled (separately) with (125)I and (211)At. The radiolabeled Fab' conjugates were coinjected into nude mice bearing LNCaP human tumor xenografts, and biodistribution data were obtained at 1, 4, and 24 h pi. Tumor targeting was achieved with both (125)I- and (211)At-labeled Fab', but the (211)At-labeled Fab' reached a higher concentration (25.56 ± 11.20 vs 11.97 ± 1.31%ID/g). Surprisingly, while the (125)I-labeled Fab' was cleared from kidney similar to earlier studies, the (211)At-labeled Fab'was not (i.e., kidney conc. for (125)I vs (211)At; 4 h, 13.14 ± 2.03 ID/g vs 42.28 ± 16.38%D/g; 24 h, 4.23 ± 1.57 ID/g vs 39.52 ± 15.87%ID/g). Since the Fab' conjugate is identical in both cases except for the radionuclide, it seems likely that the difference in tissue clearance seen is due to an effect that (211)At has on either the hydrazone cleavage or on the retention of a metabolite. Results from other studies in our laboratory suggest that the latter case is most likely. The hydrazone linkers tested do not provide the tissue clearance sought for (211)At, so additional hydrazones linkers will be evaluated. However, the results support the use of hydrazone linkers when Fab' conjugated with closo-decaborate(2-) reagents are radioiodinated.