A Pretargeted Imaging Strategy for EGFR-Positive Colorectal Carcinoma via Modulation of Tz-Radioligand Pharmacokinetics

Previously, we successfully developed a pretargeted imaging strategy (atezolizumab-TCO/[99mTc]HYNIC-PEG11-Tz) for evaluating programmed cell death ligand-1 (PD-L1) expression in xenograft mice. However, the surplus unclicked [99mTc]HYNIC-PEG11-Tz is cleared somewhat sluggishly through the intestines, which is not ideal for colorectal cancer (CRC) imaging. To shift the excretion of the Tz-radioligand to the renal system, we developed a novel Tz-radioligand by adding a polypeptide linker between HYNIC and PEG11. Pretargeted molecular probes [99mTc]HYNIC-polypeptide-PEG11-Tz and cetuximab-TCO were synthesized. [99mTc]HYNIC-polypeptide-PEG11-Tz was evaluated for in vitro stability and in vivo blood pharmacokinetics. In vitro ligation reactivity of [99mTc]HYNIC-polypeptide-PEG11-Tz towards cetuximab-TCO was also tested. Biodistribution assay and imaging of [99mTc]HYNIC-polypeptide-PEG11-Tz were performed to observe its excretion pathway. Pretargeted biodistribution was measured at three different accumulation intervals to determine the optimal pretargeted interval time. Pretargeted (cetuximab-TCO 48 h/[99mTc]HYNIC-PEG11-Tz 6 h) and (cetuximab-TCO 48 h/[99mTc]HYNIC-Polypeptide-PEG11-Tz 6 h) imagings were compared to examine the effect of the excretion pathway on tumor imaging. [99mTc]HYNIC-polypeptide-PEG11-Tz showed favorable in vitro stability and rapid blood clearance in mice. SEC-HPLC revealed almost complete reaction between cetuximab-TCO and [99mTc]HYNIC-polypeptide-PEG11-Tz in vitro, with the 8:1 Tz-to-mAb reaction providing a conversion yield of 87.83 ± 3.27 %. Biodistribution and imaging analyses showed that the Tz-radioligand was cleared through the kidneys. After 24, 48, and 72 h of accumulation in HCT116 tumor, the tumor-to-blood ratio of cetuximab-TCO was 0.83 ± 0.13, 1.40 ± 0.31, and 1.15 ± 0.21, respectively. Both pretargeted (cetuximab-TCO 48 h/[99mTc]HYNIC-PEG11-Tz 6 h) and (cetuximab-TCO 48 h/[99mTc]HYNIC-polypeptide-PEG11-Tz 6 h) clearly delineated HCT116 tumor. Pretargeted imaging strategy using cetuximab-TCO/[99mTc]HYNIC-polypeptide-PEG11-Tz could be used for diagnosing CRC, as the surplus unclicked [99mTc]HYNIC-polypeptide-PEG11-Tz was cleared through the urinary system, leading to low abdominal uptake background. Our novel pretargeted imaging strategy (cetuximab-TCO/[99mTc]HYNIC-polypeptide-PEG11-Tz) was useful for imaging CRC, broadening the application scope of pretargeted imaging strategy. The pretargeted imaging strategy clearly delineated HCT116 tumor, showing that its use could be extended to selection of internalizing antibodies.


Introduction
Colorectal cancer (CRC) is a common malignant tumor of the gastrointestinal tract that poses a serious threat to human health, with its morbidity ranking third among all malignant tumors and its mortality ranking fourth in the world [1]. In recent years, with lifestyle changes, the incidence of CRC has increased year by year. As early diagnosis and treatment will greatly improve the prognosis of CRC, it is important to develop a specific targeted molecular probe for CRC. Anatomical imaging methods, such as endoscopic ultrasonography, computed tomography (CT), and magnetic resonance imaging (MRI), have obvious advantages over invasive biopsy and endoscopy in the evaluation of tumor infiltration to surrounding tissues as well as lymph node and distant metastasis. Nevertheless, the imaging techniques are not specific for CRC diagnosis. 2-Deoxy-2-[ 18 F]fluoro-Dglucose ([ 18 F]FDG) positron emission tomography/ computed tomography (PET/CT) is a functional imaging technology that can reflect tumor glucose metabolism. However, the most commonly used positron radioactive tracer, 18 F-FDG, is still not a CRC-specific molecular probe.
Currently, targeted molecular probes for CRC mainly include antibody [2][3][4], polypeptide [5,6], and nanoparticle [7,8] probes targeting epithelial growth factor receptor (EGFR) or vascular endothelial growth factor (VEGF). Antibody probes have exquisite affinity and selectivity for molecular targets, such as EGFR or VEGF overexpressed by CRC. However, owing to their slow pharmacokinetics, the use of antibodies as tracers requires labeling with isotopes with long half-lives (e.g., 111 In, 64 Cu, or 131 I), which significantly increases radiation dose to non-target tissues. Polypeptide probes have several advantages over antibody probes, including low molecular weight, reduced circulation time, easy access to target sites, and passivity to the immune system with little or no immunogenicity. However, stability and affinity to target sites vary among different polypeptide tracers. Nanoparticle probes are equipped with different targeting units for different receptors, but this type of probe poses potential threats of immunogenicity and renal toxicity.
EGFR has become a therapeutic target for CRC owing to the close association of EGFR expression with disease progression and metastasis [9]. The therapeutic antibody cetuximab has been approved by the US FDA for CRC treatment. As an EGFR inhibitor, cetuximab specifically targets the extracellular domain of EGFR and blocks intracellular tyrosine kinase activity. Moreover, molecular tracers incorporating cetuximab labeled with long-lived radionuclides (such as 111 In, 89 Zr, 64 Cu, and 124 I) have been developed to detect EGFR expression and evaluate therapeutic responses to EGFR blockade [10][11][12][13][14]. However, the extended circulation time of radiolabeled antibodies leads to prohibitively high radiation dose to healthy organs and low tumor/background imaging contrast. In this study, we developed a novel pretargeted imaging strategy for evaluating EGFR expression in CRC.
Previously, we developed a pretargeted single-photon emission computed tomography (SPECT) imaging strategy for evaluating immune checkpoint ligand PD-L1 expression in tumors based on bioorthogonal Diels-Alder click chemistry [15]. The molecular probe mainly includes two components: TCO-modified atezolizumab (atezolizumab-TCO), which can target PD-L1, and 99m Tc-labeled Tzradioligand ([ 99m Tc]HYNIC-PEG 11 -Tz). Pretargeted imaging of (atezolizumab-TCO/[ 99m Tc]HYNIC-PEG 11 -Tz) involves the following steps: (1) atezolizumab-TCO is first injected into the bloodstream; (2) atezolizumab-TCO binds specifically to PD-L1 overexpressed by tumors and is then gradually cleared from blood; (3) [ 99m Tc]HYNIC-PEG 11 -Tz is injected; (4) [ 99m Tc]HYNIC-PEG 11 -Tz binds to the pretargeted atezolizumab-TCO via in vivo click ligation, followed by rapid clearance of excess Tz-radioligand. The pretargeted imaging strategy clearly delineates PD-L1expressing H1975 human lung cancer xenografts with high imaging contrast and significantly reduces background radiation dose to non-target organs. However, the surplus unclicked [ 99m Tc]HYNIC-PEG 11 -Tz is cleared somewhat sluggishly through the intestines. This is, of course, not an ideal situation for imaging of abdominal tumors, especially CRC. Therefore, we further developed a novel Tzradioligand with a more favorable pharmacokinetic profile for CRC imaging. Regarding the research design, we added a polypeptide chain containing hydrophilic amino acids between the HYNIC and PEG 11 to shift the excretion of the Tz-derivative to the renal system, thereby reducing abdominal uptake background and facilitating CRC imaging. Technetium ( 99m Tc) is the most widely used radionuclide for diagnosis in SPECT imaging; it is easily obtained from 99 Mo/ 99m Tc and presents an appropriate half-life (6.02 h) and gamma emission energy (140 keV). Thus, 99m Tc was chosen as a radionuclide in the present study. A challenging EGFR-targeting antibody, cetuximab, was selected because cetuximab is known to be internalized over time once bound to the membrane EGFR [16,17]; this will verify whether the pretargeted imaging strategy can be extended to a wider selection of antibodies. Ultimately, we developed another pretargeted imaging strategy using a combination of a TCOmodified internalizing antibody (cetuximab-TCO) and polypeptide-added Tz-radioligand ([ 99m Tc]HYNIC-polypeptide-PEG 11 -Tz) for detecting EGFR expression in CRC based on the renal excretion of the novel Tz-radioligand; this strategy will further broaden the application scope of pretargeted imaging strategy.

mAb-TCO Modification and Characterization
Briefly, 0.2 M NaHCO 3 was added to cetuximab in phosphate buffer solution (PBS) (817 μl, 12.12 mg/ml) to adjust the final pH of the buffer to 8.6. To this solution was added an appropriate volume of TCO-NHS in N,N-dimethylacetamide. Finally, the protein concentration was 3.0 mg/ml, N,N-dimethylacetamide content was 10 %, and TCO/mAb reaction stoichiometry was 35:1. The resulting solution was incubated with gentle shaking for 3 h at room temperature in the dark. Subsequently, a 40 KD Zeba™ spin desalting column (Thermo Fisher Scientific) was used for buffer exchange and purification of the final product. Aggregation and purity of the product were determined using a SEC-HPLC system. Protein concentration of this product was determined by using a spectrophotometer (Thermo Fisher Nanodrop 2000) at an absorbance of 280 nm. The final product was stored in PBS (pH = 7.4).
We used reversed-phase HPLC (RP-HPLC) to determine the rough average number of TCO attached to each cetuximab. A concentration gradient of HYNIC-polypeptide-PEG 11 -Tz (10-500 μg/ml) was prepared for generating a standard curve. Briefly, 3.2 μl N,N-dimethylformamide and 2.8 μl 8 mg/ml HYNIC-polypeptide-PEG 11 -Tz solution in N,N-dimethylformamide were added to 24 μl of the product obtained in the above step, resulting in a total volume of 30 μl and a Tz/cetuximab molar ratio of 20. The reaction solution was incubated with gentle shaking at 37°C for 20 min in the dark, and then the consumption of HYNICpolypeptide-PEG 11 -Tz was determined with RP-HPLC to calculate the average number of TCO conjugated to each cetuximab. The average number of TCO conjugated to each cetuximab was determined using the equation below:  11 -Tz (3.6-4.0 μg, 16.65-18.50 MBq in 100 μl NS) via intravenous tail vein injection. Micro-SPECT/CT scans were conducted on a Nano SPECT/CT scanner (BioScan, Washington DC, USA) at 30 min, 2 h, and 6 h after Tzradioligand injection (n = 3, a total of three mice were used for imaging study). The mice were then anesthetized by 2 % isoflurane/oxygen gas inhalation, and anesthesia was maintained using 1 % isoflurane/oxygen gas during SPECT/CT scan. CT was performed with the following parameters: frame resolution, 256 × 512; current, 0.15 mA; tube voltage, 45 kVp. SPECT was then performed in the same bed position as CT scan, with the following parameters: four high-resolution conical collimators with nine-pinhole plates; matrix, 256 × 256; resolution, 1 mm/pixel; energy peak, 140 keV; window width, 10 %; and scan time, 30 s/ projection for 30 min and 2 h, 50 s/projection for 6 h posttetrazine injection, 24 projections in total. 3D ordered subset expectation maximization images were reconstructed using HiSPECT algorithm. InVivoScope (Version 1.43; Bioscan) was used for post-processing of reconstructed SPECT/CT data. Biodistribution was analyzed at the same time point after intravenous injection of Tz-radioligand (3.6-4.0 μg, 16.65-18.50 MBq in 100 μl NS via the tail vein; n = 3, a total of nine mice were used for the biodistribution assay) with SPECT imaging (the mice used for biodistribution assay were separate from those used for imaging). The mice were anesthetized with isoflurane and euthanized by cervical dislocation. Blood was withdrawn by heart puncture, and each organ and tissue of interest was harvested, blotted dry, and weighed. The radioactivity of each sample was measured using a γ-counter along with standards to determine the %ID/g for each sample of interest.

Pretargeted Biodistribution Study
Male athymic nude mice bearing subcutaneous HCT116 xenografts were administered 100 μg (approximately 0.69 nmol) cetuximab-TCO in 100 μl PBS via intravenous tail vein injection (the injected dose of 100 μg mAb-TCO was in accordance with the dose in published pretargeted tumor imaging studies [15,[18][19][20][21][22][23][24]). To validate the optimal pretargeted interval time, three different accumulation intervals of 24, 48, and 72 h were used in the pretargeted biodistribution study (n = 3). After three pretargeted accumulation periods, the same mice were then administered [ 99m Tc]HYNIC-polypeptide-PEG 11 -Tz (3.6-4.0 μg, 16.65-18.50 MBq in 100 μl NS), also via tail vein injection. At 6 h after Tz-radioligand administration, the animals were anesthetized with isoflurane, and blood was withdrawn by heart puncture (the time point of 6 h was selected according to the results of our previous investigation [15]). Next, the mice were euthanized by cervical dislocation, and samples including the tumor, heart, lung, liver, spleen, stomach, small intestine, large intestine, kidney, muscle, bone, skin, brain, and thyroid tissues as well as feces (in distal colon) and urine were harvested, blotted dry, and weighed. The radioactivity of all samples was measured in a γ-counter to determine the %ID/g.

Pretargeted Biodistribution and Imaging Comparison
Male nude mice bearing subcutaneous HCT116 xenografts were administered 100 μg cetuximab-TCO via intravenous tail vein injection (n = 3). After 48-h accumulation interval (the highest tumor/blood ratio), which was appropriate according to the results of pretargeted biodistribution study, the mice were administered [ 99m Tc]HYNIC-PEG 11

Statistical Analysis
GraphPad Prism version 5.02 was used for statistical calculation and biodistribution data analysis. Data were analyzed by the unpaired, two-tailed Student's t test and one-way analysis of variance (ANOVA) for comparison between two or three groups. Group variation is described as mean ± standard deviation. Differences with P G 0.05 were considered significant.

Results
Synthesis, Characterization, Radionuclide Labeling, Radiochemical Purity, and In Vitro Reactivity of Molecular Probes Modification of cetuximab by TCO-NHS ester was conducted in an alkaline environment (pH = 8.6) for 3 h. Next, cetuximab-TCO was purified by using a 40 KD Zeba™ spin desalting column. The aggregation and purity of the product were 0.73 % and 99.27 %, respectively, as shown by SEC-HPLC (Suppl. Fig. 1a, see ESM). As shown in Suppl. Fig.  1b and c, a standard curve for determination of HYNICpolypeptide-PEG 11 -Tz concentration was obtained from RP-HPLC of a concentration gradient of HYNIC-polypeptide-PEG 11 -Tz (10-500 μg/ml). After the reaction of purified cetuximab-TCO with 20-fold molar equivalents of HYNICpolypeptide-PEG 11 -Tz was completed, the consumption of HYNIC-polypeptide-PEG 11 -Tz was used for calculating the average number of TCO conjugated to each cetuximab. Finally, the effective average TCO moieties attached to each cetuximab was calculated to be 8.23.

Molecular Probe In Vitro Cell Experiments
Pretargeted cell binding assay was performed using high EGFR-expressing HCT116 cell line. The pretargeted experimental group showed high radioactivity retention in cells, with an average count per minute (CPM) value of 19,045 ± 557 (Fig. 3a). The cetuximab control group, blocking group, and IgG-TCO isotype control group cells showed significantly lower activity, with an average CPM value of 323 ± 52, 2326 ± 341, and 1519 ± 148, respectively (P G 0.05). The pretargeted cell binding assay confirmed the specific binding of cetuximab-TCO to HCT116 cells and verified preliminarily the feasibility of the molecular probes for use in in vivo pretargeted imaging. For saturation binding assay,  the specific binding (SB) curve of [ 99m Tc]HYNIC-cetuximab was obtained from TB curve and NSB curve. The SB curve showed a typical sigmoidal shape (Fig. 3b). The equilibrium binding constant (Kd) of [ 99m Tc]HYNIC-cetuximab for EGFR was 29.37 ± 4.82 nM, and the receptor density (Bmax) was 9.50 ± 0.53 × 10 −18 mol/cell.

Molecular Probe In Vivo Animal Experiments
[ 99m Tc]HYNIC-polypeptide-PEG 11 -Tz alone was subjected to pharmacokinetics, biodistribution, and SPECT imaging analyses to investigate its pharmacokinetic profile, tissue biodistribution, and excretion pathway, respectively. Our results showed that the Tz-radioligand was cleared quickly from the circulation and distributed rapidly into various tissues and organs. Its half-life (t 1/2 ) in blood distribution was 15.72 ± 1.83 min (Suppl. Fig. 3, see ESM). The short blood t 1/2 ensured low radioactivity exposure to healthy organs, and the background signal was relatively low when used for in vivo imaging. Biodistribution and imaging studies of [ 99m Tc]HYNIC-polypeptide-PEG 11 -Tz alone were analyzed at 30 min, 2 h, and 6 h after Tz-radioligand injection. In contrast to [ 99m Tc]HYNIC-PEG 11 -Tz [15], which showed a hepatobiliary elimination pathway, [ 99m Tc]HYNIC-polypeptide-PEG 11 -Tz was cleared completely through the urinary system, as indicated in Suppl. Table 1 and Fig. 4a. The radioactivity in the kidneys was 4.84 ± 1.89 %ID/g, 4.59 ± 0.82 %ID/g, and 2.98 ± 1.12 %ID/ g at 30 min, 2 h, and 6 h after the Tz-radioligand injection, respectively. The Tz-radioligand was excreted quickly into the bladder, and the %ID/g of urine was 59.01 ± 36.71, 21.11 ± 23.80, and 4.21 ± 3.16 at 30 min, 2 h, and 6 h after the Tz-radioligand injection, respectively. The accumulation and retention of the radiotracer in the liver, intestine, and feces remained very low (G 0.3 %ID/g at various time points after the Tz-radioligand injection), indicating no hepatobiliary elimination pathway for the radioligand. The low uptake (G 0.2 %ID/g) in HCT116 tumor at all time points suggested that the Tz-radioligand had no specific binding to high EGFR-expressing HCT116 tumor. Critically, its accumulation in the thyroid and stomach was particularly low (G 0.3 %ID/g), suggesting the high in vivo stability of the Tzradioligand. SPECT imaging at each time point clearly visualized the biodistribution of [ 99m Tc]HYNIC-polypeptide-PEG 11 -Tz in nude mice bearing subcutaneous HCT116 xenografts (Fig. 4b).
To validate the optimal pretargeted interval time, three different cetuximab-TCO accumulation intervals of 24, 48, and 72 h were tested in the pretargeted biodistribution study (Suppl. Table 2 and Fig. 5). Biodistribution experiments showed that the accumulation of cetuximab-TCO in the kidneys and urine was relatively high, with a %ID/g of more than 3.0 at each accumulation interval. After being filtered through the kidneys, the radiotracer was concentrated in the bladder and finally excreted out of the body through urine. In addition to the urinary system, high activity levels were observed in the tumor and blood (9 1 %ID/g at each accumulation interval), suggesting a great incidence of in vivo click reactions in blood in addition to ligations in the tumor. The tumor/blood ratio of cetuximab-TCO in HCT116 tumor was 0.83 ± 0.13, 1.40 ± 0.31, and 1.15 ± 0.21, respectively, after 24, 48, and 72 h of accumulation. One-way ANOVA revealed that the difference in tumor/ blood ratio for 24, 48, and 72 h pretargeted interval was significant (F = 5.357, P G 0.05). Pair-wise comparisons of any two pretargeted intervals using LSD test revealed that although the tumor/blood ratio at 48 and 72 h accumulation interval was not statistically different (P = 0.191), the difference in tumor/blood ratio between 24 and 48 h accumulation interval was significant (P G 0.05). Thus, it was clear that 48 h represented an appropriate accumulation interval between the administration of mAb-TCO and the subsequent injection of Tz-radioligand. The accumulation and retention of the radiotracer in other organs, including the hypervascular lung, liver, and spleen, remained generally very low. Statistical analysis revealed that the accumulation of the radiotracer in HCT116 tumor was significantly higher than that in each organ/tissue (including the heart, lung, liver, spleen, stomach, small intestine, large intestine, muscle, bone, skin, brain, and thyroid) for each pretargeted interval of 24, 48, and 72 h (all P G 0.05).

Discussion
T h e p r e t a r g e t e d s t r a t e g y ( a t e z o l i z u m a b -T C O / [ 99m Tc]HYNIC-PEG 11 -Tz) that we previously developed could be used to evaluate immune checkpoint ligand PD-L1 expression in tumors and simultaneously reduce the radiation dose to non-target organs [15]. However, the surplus unclicked [ 99m Tc]HYNIC-PEG 11 -Tz was cleared through the hepatobiliary system and remained in feces for a long time, which is not favorable for imaging abdominal tumor, especially for CRC. To further broaden the application scope of the pretargeted imaging strategy, we added a polypeptide chain between the HYNIC and Tz to shift the excretion of the Tz-radioligand to the renal system, thereby facilitating CRC imaging.
To improve the hydrophilicity of [ 99m Tc]HYNIC-PEG 11 -Tz to facilitate renal excretion, we chose the hydrophilic lysine, glutamate, and arginine in designing a polypeptide sequence. Ultimately, the polypeptide sequence was determined as Gly-Arg-Glu-Arg-Glu-Lys and synthesized successfully through the Fmoc-method. As to not increase the steric resistance of the click reaction between cetuximab-TCO and Tz-radioligand, we added the polypeptide chain between HYNIC and PEG 11 . Finally, the modified Tzderivative structure was identified as HYNIC-polypeptide-PEG 11 -Tz. HYNIC-polypeptide-PEG 11 -Tz was labeled with radionuclide 99m Tc in a similar manner to HYNIC-PEG 11 -Tz. The radiochemical purity of the final product, [ 99m Tc]HYNIC-polypeptide-PEG 11 -Tz, was more than 95 %, and the specific activity was 4.625 MBq/μg. The stability of [ 99m Tc]HYNIC-polypeptide-PEG 11 -Tz in NS, PBS, and FBS was similar to that of [ 99m Tc]HYNIC-PEG 11 -Tz [15].
The urinary excretion of the polypeptide-modified [ 99m Tc]HYNIC-polypeptide-PEG 11 -Tz was probably related to the increased hydrophilicity of the Tz-derivative, which may be caused by the hydrophilic amino acids and the presence of lysine residue in the polypeptide chain. It is known that the Tz-scaffold part in the Tz-radioligands is hydrophobic [25]. Therefore, in designing a Tz-derivative, it is necessary to add a hydrophilic structure to improve the hydrophilicity of the final product. In a study conducted by  Garcia et al. [18], [ 99m Tc]HYNIC-PEG 4 -Tz-Me was mainly cleared through the hepatobiliary system, and its relative absorption value by the liver and intestine was 9.91 ± 0.97 %ID/g and 23.35 ± 3.84 %ID/g, respectively, at 1 h after the Tz-radioligand injection. The authors further added a polypeptide sequence between HYNIC and PEG 4 to increase the hydrophilicity of the Tz-radioligand to logD of − 1.05 ± 0.02. The newly synthesized Tz-radioligand was excreted mainly through the kidneys. Its relative absorption value by urine was 81.92 ± 5.06 %ID at 1 h post-injection. In addition, Nichols et al. [26] and Devaraj et al. [27] successfully synthesized 68 Ga and 18 F-labeled Tz-coated polymer by introducing a well-established hydrophilic aminodextran backbone into a Tz-derivative, respectively. In our study, the noctanol/PBS distribution coefficient of [ 99m Tc]HYNIC-PEG 11  It has been shown that the introduction of positive charge to radiopharmaceuticals can increase their renal clearance and retention [28][29][30]. Similarly, the introduction of positive charge in the Tz-radioligands increased clearance through the kidneys, whereas Tz derivatives with no charge were mainly excreted via the hepatobiliary system. In a study performed by Zeglis et al., as 64 Cu-NOTA-Tz was cleared somewhat sluggishly through the gastrointestinal pathway [22], and the authors further created two novel Tzradioligands ( 64 Cu-NOTA-PEG 7 -Tz and 64 Cu-SarAr-Tz) to improve their pharmacokinetic profiles [23]. For 64 Cu-SarAr-Tz, the coordination environment changed from N 3 O 3 to N 6 , and more importantly, the overall charge of the Tz-radioligand was shifted from − 1 to + 2. Finally, the newly synthesized 64 Cu-SarAr-Tz was eliminated quickly and cleanly through the urinary tract [23]. In a pharmacokinetic profile study including 25 different Tz derivatives radiolabeled with either Al[ 18 F] or 68 Ga, Meyer et al. [31] observed that 68 Ga-NODA-Tz was excreted through the renal pathway, whereas Al[ 18 F]-NODA-Tz was excreted through the hepatic and intestinal pathway. The authors suggested that the different excretion pathways of the two Tz-radioligands resulted from the charge difference between 68 Ga-NODA-Tz (net charge: + 1) and Al[ 18 F]-NODA-Tz (net charge: 0). In addition, the authors found that uncharged Al[ 18 F]-NOTA-lysine-Tz was excreted mainly through the liver and intestines, whereas charged Al[ 18 F]-NOTA-(lysine) 2 -Tz (net charge: + 1) and Al[ 18 F]-NOTA-(lysine) 3 -Tz (net charge: + 2) were excreted mainly through the kidneys [31]. The two Tz-radioligands in our study showed no charge. The shift of the excretion route of the newly developed [ 99m Tc]HYNIC-polypeptide-PEG 11 -Tz was probably caused by the added polypeptide chain. Moreover, it has been reported that renal uptake and excretion may be related to the presence of lysine residues in the molecular structure [32,33]. There was a lysine residue in the polypeptide chain of [ 99m Tc]HYNIC-polypeptide-PEG 11 -Tz.
In the pretargeted (cetuximab-TCO 24, 48, or 72 h/ [ 99m Tc]HYNIC-polypeptide-PEG 11 -Tz 6 h) biodistribution experiments, we found that the tumor uptake and tumor/ blood ratios were generally relatively low after three accumulation intervals, which probably resulted from cetuximab-TCO internalization over time once bound to the membrane EGFR. The internalization reduced the amount of cetuximab-TCO on the cell surface that was available for clicking to the Tz-radioligand. Previously, the pretargeted strategy mainly focused on highly expressed membrane receptors, such as the carbohydrate antigen CA19.9 in pancreatic cancer [34,35], or antibodies that show high cell surface persistence when bound to membrane receptors, such as A33 that targets the transmembrane glycoprotein huA33 in CRC [22,23,26,27,36,37]. The biodistribution of these pretargeted strategies suggested high tumor uptake and tumor/blood ratio. However, Houghton et al. and Keinänen et al. have both showed that it is possible to pretarget an antibody that is internalized [38,39], although the tumor uptake observed for pretargeted imaging with internalized mAbs was much lower than that of the corresponding antibodies directly radiolabeled with long half-lived radionuclides. Our biodistribution assay revealed that tumor uptake and tumor/blood ratio were relatively low after the three pretargeted intervals, which is in accordance with the results of Houghton and Keinänen. In addition, we found the tumor/blood ratio was highest (48 h: 1.40 ± 0.31 vs. 24 h: 0.83 ± 0.13 and 72 h: 1.15 ± 0.21) for allowing 48 h accumulation of cetuximab-TCO and subsequent [ 99m Tc]HYNIC-polypeptide-PEG 11 -Tz injection. Thus, 48 h represented an appropriate interval of cetuximab-TCO accumulation. At 24 h after cetuximab-TCO injection, the accumulation of the mAb-TCO in HCT116 tumor had not reached a saturation status owing to its large molecular weight and poor tissue infiltration. Simultaneously, there was a large amount of circulating cetuximab-TCO in blood, as most cetuximab-TCO had not been eliminated out of the body, which led to a great incidence of in vivo click reactions in blood in addition to ligations in the tumor after [ 99m Tc]HYNIC-polypeptide-PEG 11 -Tz injection. In fact, the biodistribution after 24 h of accumulation indicated that the uptake in blood (1.60 ± 0.48 %ID/g) was even higher than that in the tumor (1.32 ± 0.14 %ID/g). The tumor/ blood ratio for 48 and 72 h accumulation intervals was not statistically different.
Biodistribution assay and SPECT imaging of pretargeted cetuximab-TCO/[ 99m Tc]HYNIC-PEG 11 -Tz and cetuximab-TCO/[ 99m Tc]HYNIC-polypeptide-PEG 11 -Tz was conducted in subcutaneous HCT116 xenograft mice, with accumulation interval of 48 h. The results revealed that the surplus unclicked [ 99m Tc]HYNIC-PEG 11 -Tz was cleared rapidly through the intestines into feces and remained in it for a long time. The surplus unclicked [ 99m Tc]HYNIC-polypeptide-PEG 11 -Tz was cleared completely through the urinary system. The tumor/blood ratio of pretargeted cetuximab-TCO/[ 99m Tc]HYNIC-PEG 11 -Tz and cetuximab-TCO/ [ 99m Tc]HYNIC-polypeptide-PEG 11 -Tz was 1.61 ± 0.12 and 1.40 ± 0.31, respectively. Although the tumor/blood ratios were relatively low owing to cetuximab-TCO internalization, both pretargeted imaging strategies delineated the HCT116 tumor clearly. mAb-TCO internalization was shown not to be an absolute prerequisite for the development of a successful pretargeting method. The pretargeted cetuximab-TCO/[ 99m Tc]HYNIC-polypeptide-PEG 11 -Tz imaging could not detect the retention of the surplus unclicked Tz-radioligand in feces, which favors to specific targeting imaging for CRC. Shi et al. [40] previously used a pretargeted strategy (cetuximab-Tz or panitumumab-Tz/ Al[ 18 F]-NOTA-Reppe anhydride) for HCT116 tumor EGFR-targeting imaging. In the study by Shi et al., the tumor/blood ratio of pretargeted cetuximab-Tz and panitumumab-Tz/Al[ 18 F]-NOTA-Reppe anhydride was 13.02 ± 0.64 and 10.15 ± 1.56, respectively. Both pretargeted imaging strategies clearly delineated the HCT116 tumor. However, owing to the fact that Al[ 18 F]-NOTA-Reppe anhydride was excreted through the hepatobiliary system and remained in feces for a long time, the authors were unable to resolve the problem that imaging agent remaining in feces interfered with CRC imaging.

Conclusions
We successfully developed a novel pretargeted imaging strategy (cetuximab-TCO/[ 99m Tc]HYNIC-polypeptide-PEG 11 -Tz) for imaging of CRC, in which the surplus unclicked [ 99m Tc]HYNIC-polypeptide-PEG 11 -Tz was cleared through the urinary system, resulting in low abdominal uptake background, which further broadened the application scope of the pretargeted imaging strategy. We also concluded that mAb-TCO internalization was not an absolute prerequisite for the development of a successful pretargeted strategy.