Site specific discrete PEGylation of (124)I-labeled mCC49 Fab' fragments improves tumor MicroPET/CT imaging in mice

Abstract

The tumor-associated glycoprotein-72 (TAG-72) antigen is highly overexpressed in various human adenocarcinomas and anti-TAG-72 monoclonal antibodies, and fragments are therefore useful as pharmaceutical targeting vectors. In this study, we investigated the effects of site-specific PEGylation with MW 2-4 kDa discrete, branched PEGylation reagents on mCC49 Fab' (MW 50 kDa) via in vitro TAG72 binding, and in vivo blood clearance kinetics, biodistribution, and mouse tumor microPET/CT imaging. mCC49Fab' (Fab'-NEM) was conjugated at a hinge region cysteine with maleimide-dPEG 12-(dPEG24COOH)3 acid (Mal-dPEG-A), maleimide-dPEG12-(dPEG12COOH)3 acid (Mal-dPEG-B), or maleimide-dPEG12-(m-dPEG24)3 (Mal-dPEG-C), and then radiolabeled with iodine-124 ((124)I) in vitro radioligand binding assays and in vivo studies used TAG-72 expressing LS174T human colon carcinoma cells and xenograft mouse tumors. Conjugation of mCC49Fab' with Mal-dPEG-A (Fab'-A) reduced the binding affinity of the non PEGylated Fab' by 30%; however, in vivo, Fab'-A significantly lengthened the blood retention vs Fab'-NEM (47.5 vs 28.1%/ID at 1 h, 25.1 vs 8.4%/ID at 5 h, p < 0.01), showed excellent tumor to background, better microPET/CT images due to higher tumor accumulation, and increased tumor concentration in excised tissues at 72 h by 130% (5.09 ± 0.83 vs 3.83 ± 1.50%ID/g, p < 0.05). Despite the strong similarity of the three PEGylation reagents, PEGylation with Mal-dPEG-B or -C reduced the in vitro binding affinity of Fab'-NEM by 70%, blood retention, microPET/CT imaging tumor signal intensity, and residual 72 h tumor concentration by 49% (3.83 ± 1.50 vs 1.97 ± 0.29%ID/g, p < 0.05) and 63% (3.83 ± 1.50 vs 1.42 ± 0.35%ID/g, p < 0.05), respectively. We conclude that remarkably subtle changes in the structure of the PEGylation reagent can create significantly altered biologic behavior. Further study is warranted of conjugates of the triple branched, negatively charged Mal-dPEG-A.

Figure 1

Structure of the PEGylation reagents conjugated via the maleimide to the hinge region cysteine mCC49 Fab′. Mal-dPEG-A, n = 24, R = CH₂CH₂–COOH, MW = 4473.17 Da; Mal-dPEG-B, n = 12, R = CH₂–CH₂–COOH, MW = 2887.28 Da. Mal-dPEG-C: n = 24, R = H, MW = 4299.06 Da.

Figure 2

Time-radioactivity curves in mouse blood after administration of ¹²⁴I-Fab′, ¹²⁴I-Fab′-A, ¹²⁴I-Fab′-B, and ¹²⁴I-Fab′-C. Data were converted to %ID of whole blood pool calculated by assuming 78 mL/kg blood. Data are presented as mean ± SD N = 5. The inset table summarizes observed AUC_1–72h and blood clearance (CL).

Figure 3

microPET/CT images post injection of ¹²⁴I-Fab′-NEM (45 μCi), ¹²⁴I-Fab′-A (60 μCi), ¹²⁴I-Fab′-B (75 μCi), and ¹²⁴I-Fab′-C (64 μCi) in LS174T xenograft nude mice. Images are windowed identically, as described in Materials and Methods. (A) a representative mouse image from each group of N = 4 at 24 h; (B) a representative mouse from N = 4 injected with Fab′-A at 5, 24, 48, and 72 h.

Figure 4

Mouse blood radioactivity and LS174T tumor SUV from microPET/CT. Data from blood at 1 and 5 h show a similar pattern, and only 5 h blood is presented. Data are presented as Mean ± SD N = 5 except at 72 h n = 4 for B and N = 3 for C.

Figure 5

Correlation between blood radioactivity and tumor intensity in microPET/CT images. Blood radioactivity (%ID) of each individual mouse in all four groups was plotted against its normalized microPET tumor SUV. Upper panel, microPET/CT imaging at 5 h; bottom panel, microPET/CT imaging at 24 h.

Scheme 1