Pretargeting molecules TF2 and RDC018
The bsAb TF2 contains two CEACAM5 and one HSG-binding sites  and was produced using the Dock-and-Lock method as described previously . A schematic representation of the pretargeting agents is provided by Schoffelen et al. . RDC018 is a peptide-hapten derived from IMP-288, a 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid (DOTA)-conjugated d-Tyr-d-Lys-d-Glu-d-Lys tetrapeptide, in which both lysine residues are substituted with a HSG moiety via their ε-amino group (Additional file 1: Figure S1) . In addition to the DOTA chelate for radiolabeling, RDC018 is C-terminally conjugated with the fluorescent DyLightTM 800 NHS ester. TF2  and RDC018 were kind gifts from Immunomedics Inc.
All labeling procedures were performed under metal-free conditions. Briefly, [111In]InCl3 (Mallinckrodt Medical BV/Curium, Petten, the Netherlands) was added to IMP-288 or RDC018 in two volumes of 0.1 M 2-(N-morpholino)ethanesulfonic acid (MES), pH 5.5. After 20 min of incubation at 95 °C, 50 mM ethylenediaminetetraacetic acid (EDTA) was added to the labeling reaction to a final concentration of 1 mM EDTA to chelate unincorporated 111In. Lastly, Tween80 (Sigma-Aldrich, Saint Louis, MO, USA) was added to the labeling product in a final concentration of 0.01%. The labeling efficiency was determined by instant thin-layer chromatography on Varian silica gel strips (ITLC-SG; Agilent Technologies, Amstelveen, the Netherlands) using 0.1 mM ammonium acetate (NH4Ac) buffer with 0.1 M EDTA (pH 5.5) as the mobile phase. If labeling efficiency was below 95%, labeled products were purified using solid-phase extraction on an HLB cartridge (Waters Chromatography B.V., Etten-Leur, the Netherlands) with 100% EtOH as mobile phase. Final radiochemical purity was > 95% for all compounds.
The antibody hMN-14 was conjugated to IRDye800CW (fluorophore:antibody substitution ratio 1.4) and diethylenetriaminepentaacetic acid (DTPA) which was labeled with [111In]InCl3 at a specific activity of 0.78 MBq μg−1, as previously described .
CEA-expressing human colon adenocarcinoma cells LS174T and CEA-negative human renal cell carcinoma cells SK-RC-52 were obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA). Cells were tested for mycoplasma negativity. Cells were cultured in sterile conditions using RPMI-1640 medium supplemented with 10% heat-inactivated fetal calf serum and 2 mM l-glutamine, without antibiotic additive. Cells were cultured in T150 tissue culture flasks in a humidified incubator with an atmosphere of 95% air and 5% carbon dioxide at 37 °C. All cells were harvested with trypsin/EDTA.
Xenograft mouse models
All animal experiments were approved by the Institutional Animal Welfare Committee of the Radboud University Medical Center and were conducted in accordance to the guidelines of the Revised Dutch Act on Animal Experimentation (2014).
Female BALB/cAnNRj-Foxn1nu/nu nude mice (7–9 weeks-old, 18–22-g body weight; Janvier), housed in individually ventilated cages (5 mice per cage) under non-sterile standard conditions with free access to standard animal chow and water, were adapted to laboratory conditions for 1 week before experimental use. For the biodistribution and microSPECT/CT experiments, mice were subcutaneously inoculated with 2 × 106 LS174T cells (left flank) and 2 × 106 SK-RC-52 cells (right flank) both suspended in 200 μL RPMI-1640 medium. For the image-guided resection experiment, intraperitoneal tumor growth was induced by an intraperitoneal injection of 3 × 105 LS174T cells suspended in 200 μL RPMI-1640 medium. Tail vein injections were performed for intravenous administration of antibodies and peptides.
Biodistribution of 111In-IMP-288 versus 111In-RDC018
In the first experiment, the biodistributions of 111In-IMP-288 and 111In-RDC018 were compared. Three different dose levels (0, 0.8, and 8 nmol) of TF2 or controls were tested with a 1:20 TF2:HSG-peptide ratio of each HSG-peptide in two subsets of 30 mice (5 mice per group). Mice at the zero dose level received the same amount of HSG-peptide (0.4 nmol) as mice at the highest dose level. Seventeen days after tumor inoculation, 200 μL TF2 in PBS-0.5% BSA or PBS-0.5% BSA was injected intravenously. Sixteen hours following TF2 administration, the radiolabeled HSG peptide (111In-IMP-288, 9.5 MBq μg−1 or 111In-RDC018, 5.8 MBq μg−1) was injected. Mice were euthanized by CO2/O2 asphyxiation, and the biodistribution of 111In-IMP-288 and 111In-RDC018 was determined 2 or 24 h after peptide injection. For this purpose, tissues of interest (tumor, muscle, lung, spleen, kidney, liver, pancreas, stomach, and duodenum) were dissected and weighed after which activity was measured in a shielded 3-in.-well-type γ-counter (Perkin-Elmer, Boston, MA, USA). Blood samples were obtained by heart puncture. For calculation of the uptake of activity in each tissue as a fraction of the injected activity, three aliquots of the injection dose were counted in the γ-counter simultaneously.
Biodistribution of 111In-RDC018 versus dual-labeled hMN-14
In the second experiment, the biodistribution profile of 111In-RDC018 in the pretargeted approach (TF2-RDC018) was compared to the dual-labeled humanized monoclonal antibody hMN-14 (reference compound) using the IVIS Lumina fluorescence camera (Xenogen VivoVision IVIS Lumina II, Caliper Life Sciences, Waltham, MA, USA) and MicroSPECT/CT (U-SPECT II; MILabs, Utrecht, the Netherlands). TF2 (1.4 nmol) and [111In]In-DTPA-hMN-14-IRdye800CW (32.2 μg, 0.78 MBq μg−1) were injected intravenously 17 days following subcutaneous tumor cell inoculation in two groups of 5 mice. Radiolabeled RDC018 (126 MBq μg−1, 0.18 μg per mouse, 22 MBq per mouse) was administered 16 h following TF2 injection. Mice which received TF2 and 111In-RDC018 were imaged at two time points (2 and 24 h post administration of the radiolabeled peptide). The reference group (n = 5) was imaged 24 and 48 h after dual-labeled hMN-14 injection.
MicroSPECT/CT and near-infrared fluorescence (NIRF) imaging
Mice with one CEA-positive and one CEA-negative tumor were scanned on a small-animal microSPECT/CT scanner with a 1.0-mm diameter pinhole collimator tube (acquisition time, 2 × 15 min) in prone position, followed by a CT scan (spatial resolution, 160 μm; 65 kV; 612 μA) for anatomical reference.
MicroSPECT/CT scans were reconstructed with MILabs reconstruction software, which uses an ordered-subset expectation maximalization algorithm.
NIRF images were acquired on the IVIS fluorescence imaging system (acquisition time, 5 min; binning, medium; Fstop, 2; excitation, 745 nm; excitation autofluorescence, 675 nm; emission, ICG; lamp level, high; FOV, D).
Image-guided (post mortem) resection
After the biodistribution experiments, an image-guided resection experiment was performed. Intraperitoneal tumors were induced in 3 groups of mice and after 3 weeks the targeting agents were administered. In the first group, 6 nmol of TF2 was administered to 5 mice and 16 h following TF2 injection, 0.3 nmol of radiolabeled RDC018 was administered. Two hours following 111In-RDC018 injection, mice were imaged with microSPECT/CT and NIRF imaging. Hereafter, image-guided resection using the IVIS fluorescence imaging system was performed. To confirm complete resection, additional optical imaging and SPECT/CT images were acquired. Next, animals were dissected to determine the biodistribution of the radiolabeled peptide as described above.
The two remaining groups of mice served as controls. In two mice, peritoneal tumors grew faster than expected and reached a humane endpoint before the start of the experiment. These mice were therefore euthanized prior to injection of the control compounds. As a result, each control group consisted of 4 mice. In the first group, (positive control) dual-labeled hMN-14  was administered to 4 mice with intraperitoneal LS174T tumors. Resection and analysis were performed 3 days after dual-labeled hMN-14 injection. As negative control, we used 4 mice with tumors pretargeted with the trivalent anti-CD20 bsAb TF4  in combination with 111In-RDC018 with the same dosing and timing as the TF2 group.
Statistical analyses were performed using GraphPad Prism software (version 5.03; GraphPad Software). Student’s t test was performed on the biodistribution studies IMP288 versus RDC018 (tumor, blood, and kidney), corrected for multiple testing (Bonferroni). A p value < 0.05 was used to reject the null hypothesis. Data are presented as mean and standard deviation.