Development of Dual Receptor Enhanced Pre-Targeting Strategy—A Novel Promising Technology for Immuno-Positron Emission Tomography Imaging

Abstract

PET imaging has become an important diagnostic tool in the era of precise medicine. Various pre-targeting systems have been reported to address limitations associated with traditional immuno-PET. However, the application of these mono-receptor based pre-targeting (MRPT) strategies is limited to non-internalizable antibodies, and the tumor uptake is usually much lower than that in the corresponding immuno-PET. To circumvent these limitations, we develop the first Dual-Receptor Pre-Targeting (DRPT) system through entrapping the tumor-receptor-specific radioligand by the pre-administered antibody. Besides the similar ligation pathway happens in MRPT, incorporation of a tumor-receptor-specific peptide into the radioligand in DRPT enhances both concentration and retention of the radioligand on tumor, promoting its ligation with pre-administered mAb on cell-surface and/or internalized into tumor-cells. In this study, ⁶⁴Cu based DRPT shows superior performance over corresponding MRPT and immuno-PET using internalizable antibodies. Besides, the compatibility of DRPT with short-lived and generator-produced ⁶⁸Ga is demonstrated, leveraging its advantage in reducing radio-dose exposure. Furthermore, the feasibility of reducing the amount of the pre-administered antibody is confirmed, indicating the cost saving potential of DRPT. In summary, synergizing advantages of dual-receptor targeting and pre-targeting, we expect that this DRPT strategy can become a breakthrough technology in the field of antibody-based molecular imaging.

Graphical Abstract

Based on a pair of designed chemical tools namely TCO-mAb (monoclonal antibody) and Tz-RTL (Radiolabeled Targeting Ligand), a novel dual receptor enhanced pre-targeting (DRPT) strategy for PET imaging is developed. Our DRPT increases the tumor uptake and specificity as compared to traditional mono-receptor pre-targeting (MRPT) strategies while improving the tumor to background contrast as compared to the traditional immuno-PET, rendering it a novel breakthrough technology for antibody-based molecular imaging.

1. Introduction

In the current age of precision medicine, Positron Emission Tomography (PET) imaging plays an important role in the field of oncology.^[1-4] PET imaging serves as a critical diagnostic tool^[5] by providing specific information regarding tumor receptor expression levels.^[6-8] This clinically vital data can facilitate the selection of proper targeting therapies for subsequent treatment. Owing to the extraordinary specificity and affinity, monoclonal antibodies (mAbs) are considered ideal tumor-targeting vectors for oncologic purposes,^[9-12] especially when low molecular weight ligands with high affinity towards targets of interest are not available. However, the slow clearance of mAbs from the blood as well as their extremely high uptake in non-target organs such as the liver^[13,14] generally result in low signal to noise ratios. In addition, due to the relatively long physiological half-life of mAbs, radioisotopes with a comparable long half-life (such as ⁸⁹Zr with a t_1/2 of 78.4 hrs) are usually required when undertaking traditional immuno-PET^[12]. This necessarily imparts radiation-safety concerns for patients; as well as logistical issues for the long-term housing of radioactive animals (such as monkey) in the preclinical research setting.

To overcome these drawbacks of traditional immuno-PET, various pre-targeting strategies have been proposed, featured by in vivo labeling pre-administered slow-clearing mAbs with fast-clearing radioactive-molecule (RM) via covalent or non-covalent interactions.^[15-19] Among these pre-targeting strategies, one promising representative relies on the trans-cyclooctene (TCO)-tetrazine (Tz) ligation^[20], which is an inverse electron-demand Diels-Alder (IEDDA) reaction with ultra-fast kinetics compared to other biorthogonal reactions. In particular, the administration of Tz-RM is usually performed days after the administration of TCO-mAb to ensure the sufficient accumulation of TCO-mAb in tumor as well as its sufficient clearance from normal organs. This approach may improve the signal to noise ratio when compared to traditional immuno-PET.^[21-23] Furthermore, pre-targeting strategies also have the advantage of utilizing shorter half-life radioisotopes (such as, ⁶⁴Cu with a t_1/2 of 12.7 hrs, ¹⁸F with a t_1/2 of 110 mins, and ⁶⁸Ga with a t_1/2 of 68 mins)^[21,24,25], owing to the fast pharmacokinetics of its second reagent Tz-RM. Therefore, the overall radiation exposure associated with pre-targeting strategies have the potential to be much less than that observed with traditional immuno-PET. Despite their advantages, intrinsic limitations for these mono-receptor based pre-targeting (MRPT) strategies have also been noted. First, MRPT may not be suitable for internalizable mAbs^[26]. The internalization of mAbs can significantly decrease the amount of mAbs available on the cell surface which are required to trap the fast-clearing Tz-RM. Second, MRPT may result in relatively low radionuclide uptake in tumor. This can be attributed to insufficient ligation caused by low concentration and short retention of Tz-RM within the tumor environment, due to its lack of tumor-receptor targeting moiety. For example, the trastuzumab based traditional immuno-PET and MRPT showed tumor uptakes of 37%ID/g and 1.5%ID/g, respectively.^[27]

To directly address these limitations, we developed a novel dual-receptor enhanced pre-targeting (DRPT) strategy utilizing a pair of chemical tools--TCO-mAb and Tz-RTL (Herein, RTL represents radioactive tumor targeting ligands such as radioactive peptides or small molecule; Scheme 1). In addition to inheriting advantages of MRPT over traditional immuno-PET, DRPT possesses several unique features which properly address limitations of MRPT. First, DPRT can utilize internalizbale mAb. With the incorporation of internalizable RTL, Tz-RTL can ligate with TCO-mAb not only on the cell surface but also inside the cell. This biological feature makes DRPT a strategy suitable for both non-internalizable and internalizable mAbs. Second, DPRT imparts relatively increased tumor uptake. Benefiting from the tumor targeting property, the tumor concentration and retention of Tz-RTL in DRPT will be increased as compared to those of Tz-RM in MRPT, leading to an enhanced ligation efficiency and further resulting in increased tumor uptake. Third, DPRT may provide higher tumor specificity. The utilization of two concurrent tumor specific targets affords DRPT higher tumor specificity when compared to MRPT, in which only one tumor specific biomarker is targeted. Given these three proposed unique features, we hypothesized that DRPT can improve upon the scientific rigor of current pre-targeting strategies and result in a wider adoption of immuno-PET imaging in clinical oncology.

Scheme 1.

Differential Mechanisms of DRPT and MRPT: Both strategies start with the pre-administration of TCO-mAbs and their concomitant slow clearance from non-tumor organs in days. Subsequently, Tz-radioactive tumor targeting ligands (Tz-RTL) were injected for DRPT (orange); while Tz-radioactivemolecule (Tz-RM) was injected for MRPT (blue). Tz-RTL in DRPT could accumulate at the tumor site and be internalized into tumor cells, ligating with pre-administered TCO-mAbs either on cell surface or inside the cell. In contrast, Tz-RM in MRPT neither accumulate at the tumor site nor being internalized into tumor cells, resulting in only limited surface ligation. Therefore, DRPT leads to greatly enhanced tumor uptake (as compared to MRPT), and is applicable to internalizable mAbs.

In this proof-of-concept study, we investigated the in vivo performance of DRPT through targeting epidermal growth factor receptor (EGFR) and integrin α_Vβ₃. These two receptors were logically chosen as they are simultaneously overexpressed by a variety of human cancers. A pair of DRPT reagents namely Cetuximab-TCO (serving as TCO-mAb) and MeTz-NOTA-RGD (serving as Tz-RTL), were likewise designed and prepared for this in vivo evaluation. Herein, cetuximab is a monoclonal antibody that specifically binds to human EGFR and could be internalized into tumor cells upon forming the immunocomplex with the antigen.^[28,29] Even after a comprehensive optimization, cetuximab-based MRPT showed ~7X less tumor uptake as compared to traditional immuno-PET using the directly radiolabeled cetuximab.^[27] Benefiting from the internalization of MeTz-NOTA-RGD which targets integrin α_Vβ₃^[30], it is expected that our proposed DRPT strategy may overcome limitations observed in cetuximab-based MRPT.

Besides ⁶⁴Cu based in vivo evaluation, we also investigate the possibility of performing DRPT using ⁶⁸Ga, which is a radioisotope not applicable to traditional immuno-PET as a result of its short half-life. The compatibility of DRPT with ⁶⁸Ga is highly valuable because of the expected reduction in dosimetry and the ease of ⁶⁸Ga production by a ⁶⁸Ge/⁶⁸Ga generator. This is clinically important because either ⁶⁴Cu or ⁸⁹Zr requires the services of a local cyclotron for productio^[31], which is not widely available in most hospitals.

2. Results and Discussion

2.1. Preparation of chemical tools

A pair of DRPT reagents, Cetuximab-TCO and MeTz-NOTA-RGD, were successfully prepared for in vivo evaluation. In particular, Cetuximab-TCO was prepared via a widely used primary amine based antibody modification approach,^[16] and the recovery yield after the dialysis purification was above 95%. The average number of TCO on each antibody was determined by a FPLC based quantification method, and the result indicated that there were 6 TCO per antibody on average. The in vitro immunoreactivity of the TCO functionalized cetuximab was then measured via the reported Lindmo method^[32]. Specifically, prepared Cetuximab-TCO was radiolabeled with MeTz-NOTA(⁶⁴Cu). Upon the desalting column purification, recovered antibodies were incubated with different numbers of U87MG cells. Ratios of total/bound activity were plotted against 1/cell number (Figure S1), and the immunoreactive fraction was found to be above 90%, indicating the prepared Cetuximab-TCO maintained immunoreactivity against the human EGFR receptor.

Another chemical tool MeTz-NOTA-RGD was prepared using our previously developed bifunctional NOTA chelator^[33] via a multiple step synthetic scheme (Scheme S1). Briefly, using commercially available NH₂-PEG₃-NH(Boc), N₃-NOTA(tBu)-COOH was first converted to N₃-NOTA(tBu)-PEG₃-NH(Boc), which was then conjugated with RGD-PEG₄-BCN to give RGD-PEG₄-NOTA(tBu)-PEG₃-NH(Boc). After removing all protection groups by 95% TFA, the released primary NH₂ subsequently reacted with the MeTz-NHS to generate the desired chemical tool MeTz-NOTA-RGD. The internalization of the prepared chemical tool MeTz-NOTA-RGD was confirmed via in vitro internalization study (Figure S2) using the integrin α_Vβ₃ expressed cell line.

Except the two DRPT reagents mentioned above, the other two chemical tools MeTz-NOTA-RAD and MeTz-NOTA were also prepared for control experiments. Preparation details of all utilized chemical tools are described in the supporting information.

2.2. PET imaging by DRPT

PET imaging studies were then performed to evaluate DRPT in vivo performance. Mice bearing 4T1 subcutaneous allograft (mice breast cancer cell line, human EGFR negative^[34]) and U87MG subcutaneous xenograft (human glioblastoma cell line, human EGFR positive^[35]) were administered with 100μg Cetuximab-TCO (targeting human EGFR) via the tail vain injection. The antibody was allowed to accumulate at the tumor site as well as clear from the blood pool for 24h. Subsequently, 11.1MBq MeTz-(⁶⁴Cu)NOTA-RGD (targeting integrin α_Vβ₃) at a specific activity of 18.5MBq/nmole was administered via the tail vain injection, and PET images were taken at 4h, 24h & 48h post injections. It was found that the human EGFR positive U87MG tumor could be visualized at all three examined time points with better tumor to background ratios at the two later time points (24h & 48h post injection), while the human EGFR negative 4T1 tumor showed minimal accumulation (Figure 1a). Quantitative PET imaging region of interest (ROI) analysis revealed consistent results (Figure 1b). Specifically, U87MG exhibited an uptake of 2.40±0.27%ID/g at the 4h time point, which increased to 5.03±0.42%ID/g at the 24h time point and maintained at this level for the next 24h (4.10±0.87%ID/g at the 48h time point). While 4T1 showed much lower uptakes at the same examined time points (1.50±0.27%ID/g, 1.20±0.25%ID/g, and 1.17±0.06%ID/g for 4h, 24h, and 48h time points, respectively), demonstrating the in vivo specificity of Cetuximab-TCO as well as the biological specificity of DRPT to distinguish EGFR positive tumor from the EGFR negative tumor. Noted that the optimal contrast was achieved at late time points, which may hinder the translation of DRPT from bench sides to clinical applications. This could be attributed to the time required for the sufficient in vivo ligation, and applying a click chemistry group with a faster reaction rate (eg. replacing methyltetrazine with tetrazine) may shorten the optimal imaging time point.

Figure 1.

PET imaging by DRPT on mice bearing U87MG & 4T1 tumors: (a). PET images recorded at 4h, 24h and 48h post injection time points; (b) ROI based biodistribution results at 4h, 24h and 48h post injection time points.

The specificity of the second DRPT reagent (MeTz-NOTA-RGD) towards its target was also assessed through a control experiment where RGD in this second DRPT reagent was replaced by its inactive analogue RAD, which did not bind with the antigen but possessed similar physical chemical properties. Results showed that with MeTz-NOTA-RAD, the uptake by U87MG tumor dropped to 1.67±0.32%ID/g based on the ROI analysis (Figure 2a&2b), demonstrating the specificity of MeTz-NOTA-RGD towards the integrin α_vβ₃ in DRPT. Another control experiment using blocking dose of RGD (10mg/kg) was also performed, and the PET imaging result (Figure S3) revealed an obviously decrease in the tumor uptake, further validating the specificity of the MeTz-NOTA-RGD. Moreover, since MeTz-NOTA-RGD is a radiolabeled tumor targeting peptide, it is necessary to confirm that the tumor uptake observed in DRPT was not solely a property of this second reagent itself. A control experiment was performed, and it was found that the injection of the same amount of MeTz-NOTA-RGD alone (without pre-administration of Cetuximab-TCO) resulted in markedly diminished tumor accumulation when compared to the DRPT (Figure 2a). A tumor uptake value of 1.75±0.21%ID/g was noted based on the subsequent ROI analysis (Figure 2b) which was much lower than that by DRPT (5.03±0.42%ID/g). These results suggest that the observed tumor uptake in DRPT was primarily a factor of antibody mediated trapping of the MeTz-NOTA-RGD, rather than merely a factor of this second regent’s intrinsic biological specificity for the integrin α_vβ₃ antigen.

Figure 2.

Investigation of the specificity of MeTz-NOTA-RGD for the integrin α_vβ₃ as well as its contribution to the overall tumor uptake by this intrinsic specificity: (a) PET images recorded at the 24h post injection time point; (b) ROI based biodistribution results at the 24h post injection time point.

2.3. Comparison of DRPT and other imaging strategies via ⁶⁴Cu based studies

DRPT was then compared to MRPT and immuno-PET using the same mice model bearing both U87MG and 4T1 tumors. Based on PET images recorded at 24h post injection, the U87MG tumor could also be visualized by MRPT (Figure 3a) but with a much lower uptake value than by DRPT according to the ROI analysis (Figure 3b, 1.57±0.42%ID/g by MRPT vs 5.03±0.42%ID/g by DRPT). The higher tumor uptake observed in DRPT than in MRPT could be attributed to two major factors. One factor may be that in DRPT, MeTz-NOTA-RGD could be internalized by tumor cells upon its binding to the integrin α_Vβ₃ and then ligate with the internalized Cetuximab-TCO; while in MRPT, MeTz-NOTA could not be internalized thus it was not able to ligate with those internalized Cetuximab-TCO. The other factor may be that the tumor concentration of MeTz-NOTA-RGD in DRPT could be higher than that of MeTz-NOTA in MRPT (owing to the tumor-receptor targeting property of RGD), resulting in an improved ligation efficiency. In addition to the enhanced tumor uptake, DRPT also provided a better tumor to background contrast as compared to MRPT. In particular, the tumor to blood ratio in DRPT was 6.43±1.67 while that in MRPT was 3.22±0.12, and the tumor to kidney ratio in DRPT was 5.99±1.32 while that in MRPT was 2.27±0.03. (Figure 3c) Therefore, DRPT demonstrated a better tumor specificity as compared to MRPT, which could be attributed to that both reagents applied in DRPT (Cetuximab-TCO & MeTz-NOTA-RGD) can target the tumor while in MRPT, the second reagent (MeTz-NOTA) lacked a tumor targeting group. Besides MRPT, DRPT was also found to be superior to traditional immuno-PET using the intact mAb. Although a higher tumor uptake was observed in traditional immuno-PET, high background noise caused by the slow in vivo clearance of intact mAb was also observed; especially in the liver (Figure 3a). This may seriously hinder its clinical application for tumor detection of cancers within major organs such as liver, spleen and pancreas. Further ROI analysis revealed a better tumor to background contrast by DRPT than by immuno-PET. This was featured by the enhanced tumor to blood ratio (Figure 3c, 6.43±1.67 by DRPT vs 2.04±0.08 by immuno-PET). In summary, DRPT successfully exhibited a more promising in vivo performance than either MRPT or immuno-PET based on this PET imaging study.

Figure 3.

Comparison of DRPT, MRPT and immuno-PET: (a) PET imaging results of mice bearing U87MG & 4T1 tumors by DRPT, MRPT and immuno-PET (Cetuximab (⁶⁴Cu)); (b) ROI based biodistribution results of DRPT, MRPT and immuno-PET (Cetuximab (⁶⁴Cu)); (c) Tumor to organ ratios by DRPT, MRPT and immuno-PET (Cetuximab (⁶⁴Cu)). Two sided t-test was used to compare DRPT and MRPT for tumor to kidney ratios (n=3, p=0.008), tumor to liver ratios (n=3, p=0.037), and tumor to blood ratios (n=3, p=0.029). Two sided t-test was used to compare DRPT and Immuno-PET for tumor to blood ratios (n=3, p=0.039). *, p < 0.05; **, p < 0.01.

As the proposed DRPT strategy involved the administration of radiolabeled peptides as the second reagent, it is of interest to have it compared to traditional peptide-based imaging. To this end, we compared the PET imaging result of DRPT with our previously reported result^[33] from a monomeric RGD tracer. Considering the fast in vivo clearance of peptides, imaging of the same U87MG tumor at 1h post injection time point by the RGD monomer was used for the comparison since 1h was deemed as a more optimal imaging time point than 24h for peptide-based tracers. With the same imaging scale bar, the U87MG tumor could be clearly visualized by DRPT while it was difficult to be identified by RGD monomeric tracer (Figure 4a). Subsequent ROI comparison (Figure 4b) revealed that DRPT could provide a much higher tumor uptake (5.03±0.42%ID/g) than the corresponding peptide based monomeric tracer (1.50±0.42%ID/g), along with a significant improvement in the tumor to kidney ratio (Figure 4c, 6.43±1.67 by DRPT vs 0.43±0.23 by the RGD monomer). All these results suggested another potential indication of DRPT for being applied in the peptide-based targeting radiotherapy, where the high tumor uptake is desired for causing the sufficient radio therapeutic effects while kidney is usually considered as the dose limiting organ. Specifically, for those peptides labeled with therapeutic radionuclides such as ⁹⁰Y or ¹⁷⁷Lu for targeting radiotherapy, it is possible to further enhance their therapeutic efficacy by applying the DRPT strategy through modifying these peptides with tetrazine and pairing them with TCO modified antibodies targeting another receptor available on the same cell surface.

Figure 4.

Comparison of DRPT and the RGD monomeric tracer: (a) PET imaging results of mice bearing U87MG & 4T1 tumors by DRPT and the RGD monomeric tracer; (b) ROI based biodistribution results of DRPT and the RGD monomeric tracer. Two sided t-test was used to compare DRPT and RGD monomer for tumor to kidney ratios (n=3, p=0.0004). ***, p < 0.001.

2.4. Intracellular ligation between internalized DRPT reagents

We desired to validate that the internalized mAb and peptide can ligate via click chemistry at a close proximity. However, direct monitoring the ligation between internalized DRPT reagents at the cellular level is difficult because it is hard to distinguish ligation products formed inside cells from those formed outside and subsequently internalized. As the major concern here should be the proximity of cetuximab and RGD inside the cell instead of the intracellular TCO-Tz ligation which has already been reported by other research groups^[36], we designed and performed an in vitro experiment based on a reported photo-triggered click reaction^[37]. This approach provided specificity for detecting only the click reaction occurring inside of the cell between internalized reagents^[38]. In particular, cetuximab functionalized with the coumarin-N₃ and RGD functionalized with the photo-ODIBO were prepared^[39] to mimic the two DRPT reagents -- Cetuximab-TCO and MeTz-NOTA-RGD, respectively. Herein, photo-ODIBO is a photo-triggered metal-free click chemistry moiety which can ligate with N₃ upon the UV irradiation at 365nm,^[40] thus it allowed us to initiate the click reaction after the internalization process ceased. The fluorescence of coumarin will only be activated upon the cycloaddition due to the localization of the lone pair of electrons from the internal nitrogen atom of the azide to the triazole ring^[41], thus providing the feasibility to monitor the ligation inside cells via a fluorescence microscopy’s FITC channel. The procedure of this experiment (Figure 5a) consists of three steps: 1. Incubate cells with both Cetuximab-coumarin-N₃ and RGD-photo-ODIBO for their internalization; 2. After the incubation, block the further biomolecule exchange across the cell membrane via the colchicine treatment^[42] and trigger the ligation with a UV irradiation; 3. Monitor the click reaction via a fluorescence microscopy. U87MG cells which expressed both EGFR and integrin α_Vβ₃ were used in this experiment, and results showed that strong fluorescence signal in the cytoplasm was observed under the FITC channel (Figure 5b). Since the exchange of biomolecule across the cell membrane was blocked by the colchicine treatment, the observed fluorescence signal in the cytoplasm could be attributed to the ligation happening inside the cell. Another control group was set up in which cells were treated with colchicine before incubated with RGD-photo-ODIBO. After triggering the click reaction, it was found that no significant fluorescence signal could be observed in this control group (Figure 5c), further demonstrating that the previously observed fluorescence signal in the experimental group was resulted from the click reaction inside the cell between internalized RGD-photo-ODIBO and Cetuximab-coumarin-N₃. Given the fact that the reaction rate between ODIBO and azide (k ~ 45M⁻¹s⁻¹)^[40] is much slower than that between methyl tetrazine and TCO (k ~ 800M⁻¹s⁻¹)^[43], it is possible that the internalized MeTz-NOTA-RGD can ligate with internalized Cetuximab-TCO.

Figure 5.

Investigation of click reactions between internalized chemical tools by an in vitro fluorescence based assay: (a) workflow of the assay; (b) experimental group monitored under the fluorescence microscopy, cells were treated with colchicine after the incubation of cetuximab-coumarin-N₃ and RGD-photo-ODIBO; (c) control group monitored under the fluorescence microscopy, cells were treated with colchicine before the incubation of RGD-photo-ODIBO.

2.5. DRPT on BxPC3

To verify that DRPT is a universal strategy which can be applied to other types of cancer, we further evaluated its performance on mice bearing BxPC3 xenograft (human pancreatic cancer cell line, human EGFR positive^[44]). PET images recorded at 24h post injection showed that BxPC3 tumor could be clearly visualized by DRPT while the second reagent itself resulted in a lower concentration of the radio activity at the tumor site (Figure S4a). Subsequent ROI analysis provided consistent results (Figure S4b) that the tumor uptake in the DRPT group was 2X higher than that in the second reagent control group, indicating a successful antibody mediated dose enhancement. Therefore, in addition to the U87MG model, the successful application of DRPT on the BxPC3 model was also demonstrated, suggesting that DRPT can serve as a universal strategy for imaging different types of cancer.

2.6. Investigation of ⁶⁸Ga based DRPT

Although ⁶⁴Cu based DRPT showed promising results, it could be estimated that the inconvenience of the ⁶⁴Cu production^[45] may hinder its wide clinical application, and thus a radioisotope with a more convenient producing procedure would be highly desirable. Besides, since one of the major clinical advantages of using the pre-targeting strategy is the significantly reduced radiation dose exposure compared to traditional ⁸⁹Zr based immuno-PET imaging, it is of interest to explore the possibility of using a radioisotope possessing a half-life time that is even shorter than ⁶⁴Cu to further leverage this advantage. ⁶⁸Ga, rendered by its ease of the generator based production^[46] and a shorter half life time than ⁶⁴Cu, was thought to be a better choice of radionuclide. Therefore, we further investigated the ⁶⁸Ga based DRPT in mice bearing subcutaneous BxPC3 xenograft with the same pair of chemical tools.

Given the 10-fold reduction in half-life, a major concern of replacing ⁶⁴Cu with ⁶⁸Ga is whether the reaction kinetic of the Tz-TCO conjugation could be fast enough to ensure the sufficient in vivo ligation. A preliminary in vivo imaging study was conducted to demonstrate the feasibility of using ⁶⁸Ga in DRPT. Similar to the ⁶⁴Cu based DRPT, each mouse was administered with 100 μg Cetuximab-TCO 24h prior to the injection of the second reagent MeTz-(⁶⁸Ga)NOTA-RGD. Considering the half-life time of ⁶⁸Ga, imaging was performed at 1.5h and 3h post injection time points. It was observed that the tumor could be clearly visualized at both time points (Figure 6) with uptake values 2X (for 1.5h) and 3X (for 3h) higher than those obtained by administering the second reagent alone (Figure S5). This finding suggested the successful in vivo ligation in the ⁶⁸Ga-DRPT group. Besides, the tumor uptake in the ⁶⁸Ga-DRPT group remained similar from 1.5h to 3h while in the control group dosed with only the second reagent, the tumor uptake reduced by half during the same period. This provided further evidence of the in vivo ligation in the ⁶⁸Ga-DRPT group. Therefore, the feasibility of using ⁶⁸Ga in DRPT was demonstrated.

Figure 6.

Investigation of the feasibility of using ⁶⁸Ga in DRPT by imaging mice xenografted with the BxPC3 tumor. PET images were recorded at 1.5h & 3h post injection time points.

Consuming expensive antibodies at a larger amount is deemed to be one drawback of the pre-targeting strategy compared to traditional immuno-PET, which may lead to the potential high cost in the clinical application.^[47] Therefore, a trial of reducing the amount of the pre-administered antibody from 100 μg to 50 μg, an amount widely used in traditional immuno-PET, was undertaken. Specifically, each mouse bearing the BxPC3 xenograft was sequentially administered with 50 μg Cetuximab-TCO and MeTz-(⁶⁸Ga)NOTA-RGD with a lag time of 24h. Tumor could be clearly visualized in both images taken at 1.5h and 3h post injection time points (Figure 7). Subsequent ROI analysis revealed that tumor uptake values, despite lower than those obtained by DRPT with 100 μg Cetuximab-TCO, were still 1.7X (for 1.5h) and 2.5X (for 3h) higher than those obtained by treating with the second reagent alone (Figure S6), indicating a successful in vivo ligation and validating the feasibility of this optimization with reduced pre-administered antibody.

Figure 7.

Investigation of ⁶⁸Ga based DRPT with reduced amounts of mAb (50 μg Cetuximab-TCO). PET images were recorded at 1.5h & 3h post injection time points.

3. Conclusion

We investigated whether a dual-receptor enhanced pre-targeting strategy improves upon MRPT and traditional immuno-PET approaches. Our results suggested that the DRPT approach retained the intrinsic advantages of MRPT while simultaneously addressing its well-known limitations. Moreover, we have demonstrated the feasibility of performing DRPT with ⁶⁸Ga. This may allow for the reduction in radiation dose during PET imaging and may make the use of the DRPT approach more clinically applicable due to the ease of ⁶⁸Ga production as compared to other radioisotopes such as ⁶⁴Cu or ⁸⁹Zr. Finally, we demonstrated that the DRPT approach is capable of achieving a 50% reduction in mAb dosage compared to MRPT, which may eventually translate to cost savings. Taken together, the DRPT strategy presented here has been demonstrated as a novel promising PET imaging technology which may advance the field of oncology research and practice. Featured by an ideal balance between the tumor uptake and tumor to non-tumor ratios, DRPT also possesses the potential to be applied in the molecularly targeted radionuclide therapy, which will be explored in our future study. A major limitation associated with this strategy would be the requirement of identifying two receptors that can be targeted at the same time. For those scenarios in which only one receptor is available, it would be necessary to find a pair of an antibody and a peptide targeting different binding sites of that specific receptor if DRPT needs to be applied.

4. Experimental Section

Labeling of MeTz-NOTA-RGD with ⁶⁴Cu & ⁶⁸Ga:

The ⁶⁴Cu labeling of MeTz-NOTA-RGD was conducted by incubating 1nmole MeTz-NOTA-RGD and 18.5MBq ⁶⁴Cu in 100μl 0.1 M NH₄OAc buffer (pH ~ 6. 8) at 37 °C for 30 minutes. Labeling yield was above 95% based on the HPLC monitoring.

The ⁶⁸Ga labeling of MeTz-NOTA-RGD was conducted by incubating 1nmole MeTz-NOTA-RGD and 18.5MBq ⁶⁸Ga in 100μl 0.1 M NaOAc buffer (pH ~ 4. 0) at 90 °C for 10 minutes. Labeling yield was above 95% based on the HPLC monitoring.

Labeling of MeTz-NOTA-RAD with ⁶⁴Cu:

The ⁶⁴Cu labeling of MeTz-NOTA-RAD was conducted by incubating 1nmole MeTz-NOTA-RAD and 18.5MBq ⁶⁴Cu in 100μl 0.1 M NH₄OAc buffer (pH ~ 6. 8) at 37 °C for 30 minutes. Labeling yield was above 95% based on the HPLC monitoring.

Labeling of MeTz-NOTA with ⁶⁴Cu:

The ⁶⁴Cu labeling of MeTz-NOTA was conducted by incubating 1nmole MeTz-NOTA and 18.5MBq ⁶⁴Cu in 100μl 0.1 M NH₄OAc buffer (pH ~ 6. 8) at 37 °C for 30 minutes. Labeling yield was above 95% based on the HPLC monitoring.

Labeling of Cetuximab-TCO with ⁶⁴Cu:

1nmole Cetuximab-TCO in 200 μl PBS was added to 1nmole MeTz-(⁶⁴Cu)NOTA in 100μl 0.1M NH₄OAc (pH = 6.8). The reaction mixture was agitated at 37°C for 1h. Most MeTz-(⁶⁴Cu)NOTA was successfully conjugated with Cetuximab-TCO based on the FPLC monitoring, giving a specific activity of 0.111MBq/μg. Unconjugated MeTz-(⁶⁴Cu)NOTA was removed by the desalting column equilibrated with PBS to obtain Cetuximab (⁶⁴Cu). The radio purity of the prepared Cetuximab (⁶⁴Cu) was identified to be above 95% based on the FPLC monitoring.

Cell culture:

U87MG, 4T1, and BxPC3 cell lines were cultured in RPMI1640 supplemented with 1% penicillin G, 1% streptomycin, 1% glutamine and 10% fetal bovine serum (Invitrogen) at 37°C under 5% CO₂.

Mice model:

All animal studies were conducted under protocols approved by University of Pittsburgh or Oregon Health & Science University Institutional Animal Care and Use Committee.

For the U87MG & 4T1 model, each mouse was subcutaneously implanted with 3*10^6 U87MG cells in PBS:Matrigel = 1:1 on the right shoulder and 1*10^6 4T1 cells in PBS:Matrigel = 1:1 on the left shoulder. Tumors was allowed to grow for 10 days before imaging.

For the BxPC3 model, each mouse was subcutaneously implanted with 3*10^6 bxpc3 cells in PBS:Matrigel = 1:1 on the right shoulder. Tumor was allowed to grow for 10 days before imaging.

⁶⁴Cu based DRPT on mice bearing U87MG & 4T1 tumors:

Mice bearing 4T1 tumor (left shoulder) and U87MG tumor (right shoulder) were administrated with 100 μg Cetuximab-TCO via the tail vain injection. The antibody was allowed to accumulate at the tumor site as well as clear from blood for 24h. Subsequently, 11.1MBq MeTz-(⁶⁴Cu)NOTA-RGD at a specific activity of 18.5MBq/nmole was administrated via the tail vain injection, and PET images were taken at 4h, 24h & 48h post injection time points.

⁶⁴Cu based DRPT second reagent on mice bearing U87MG & 4T1 tumors:

Mice bearing 4T1 tumor (left shoulder) and U87MG tumor (right shoulder) were administrated with 11.1MBq MeTz-(⁶⁴Cu)NOTA-RGD at a specific activity of 18.5MBq/nmole via the tail vain injection, and PET images were taken at the 24h post injection time point.

⁶⁴Cu based DRPT with MeTz-NOTA-RAD on mice bearing U87MG & 4T1 tumors:

Mice bearing 4T1 tumor (left shoulder) and U87MG tumor (right shoulder) were administrated with 100 μg Cetuximab-TCO via the tail vain injection. The antibody was allowed to accumulate at the tumor site as well as clear from blood for 24h. Subsequently, 11.1MBq MeTz-(⁶⁴Cu)NOTA-RAD at a specific activity of 18.5MBq/nmole was administrated via the tail vain injection, and PET images were taken at the 24h post injection time point.

⁶⁴Cu based DRPT with blocking on mice bearing U87MG & 4T1 tumors:

Mice bearing 4T1 tumor (left shoulder) and U87MG tumor (right shoulder) were administrated with 100 μg Cetuximab-TCO via the tail vain injection. The antibody was allowed to accumulate at the tumor site as well as clear from blood for 24h. Subsequently, 11.1MBq MeTz-(⁶⁴Cu)NOTA-RGD at a specific activity of 18.5MBq/nmole was administrated together with 0.2mg RGD peptides via the tail vain injection, and PET images were taken at the 24h post injection time point.

⁶⁴Cu based MRPT on mice bearing U87MG & 4T1 tumors:

Mice bearing 4T1 tumor (left shoulder) and U87MG tumor (right shoulder) were administrated with 100 μg Cetuximab-TCO via the tail vain injection. The antibody was allowed to accumulate at the tumor site as well as clear from blood for 24h. Subsequently, 11.1MBq MeTz-(⁶⁴Cu)NOTA at a specific activity of 18.5MBq/nmole was administrated via the tail vain injection, and PET images were taken at the 24h post injection time point.

⁶⁴Cu based Immuno-PET on mice bearing U87MG & 4T1 tumors:

Mice bearing 4T1 tumor (left shoulder) and U87MG tumor (right shoulder) were administrated with 5.55MBq Cetuximab (⁶⁴Cu) at a specific activity of 0.111MBq/μg via the tail vain injection. PET images were taken at the 24h post injection time point.

⁶⁴Cu based DRPT on mice bearing BxPC3 tumor:

Mice bearing BxPC3 tumor (right shoulder) were administrated with 100 μg Cetuximab-TCO via the tail vain injection. The antibody was allowed to accumulate at the tumor site as well as clear from blood for 24h. Subsequently, 11.1MBq MeTz-(⁶⁴Cu)NOTA-RGD at a specific activity of 18.5MBq/nmole was administrated via the tail vain injection, and PET images were taken at the 24h post injection time point.

⁶⁴Cu based DRPT second reagent on mice bearing BxPC3 tumor:

Mice bearing BxPC3 tumor (right shoulder) were administrated with 11.1MBq MeTz-(⁶⁴Cu)NOTA-RGD at a specific activity of 18.5MBq/nmole via the tail vain injection, and PET images were taken at the 24h post injection time point.

⁶⁸Ga based DRPT on mice bearing BxPC3 tumor:

Mice bearing BxPC3 tumor (right shoulder) were administrated with 100 μg Cetuximab-TCO via the tail vain injection. The antibody was allowed to accumulate at the tumor site as well as clear from blood for 24h. Subsequently, 11.1MBq MeTz-(⁶⁸Ga)NOTA-RGD at a specific activity of 18.5MBq/nmole was administrated via the tail vain injection, and PET images were taken at 1.5h & 3h post injection time points.

⁶⁸Ga based DRPT second reagent on mice bearing BxPC3 tumor:

Mice bearing BxPC3 tumor (right shoulder) were administrated with 11.1MBq MeTz-(⁶⁸Ga)NOTA-RGD at a specific activity of 18.5MBq/nmole via the tail vain injection, and PET images were taken at the 24h post injection time point.

⁶⁸Ga based DRPT with 50 μg Cetuximab-TCO:

Mice bearing BxPC3 tumor (right shoulder) were administrated with 50 μg Cetuximab-TCO via the tail vain injection. The antibody was allowed to accumulate at the tumor site as well as clear from blood for 24h. Subsequently, 11.1MBq MeTz-(⁶⁸Ga)NOTA-RGD at a specific activity of 18.5MBq/nmole was administrated via the tail vain injection, and PET images were taken at 1.5h & 3h post injection time points.

Small-animal PET/CT imaging studies:

At desired imaging time points, mice were anesthetized with 2% isoflurane, and static PET/CT imaging was performed with 10 min PET scanning followed by 5 min CT scanning, using an Inveon Preclinical Imaging Station. Images obtained were reconstructed with maximum a posteriori, 3D ordered-subsets expectation maximization, and 2D filtered back projection. PET and CT images were co-registered with Inveon Research Workstation (IRW) software for the region of interest (ROI) analysis.

Validation of intracellular ligation:

For the experimental group, at 37°C in 1%BSA medium, 5*10^4 cells seeded on slide chambers were incubated with 0.1μM Cetuximab- coumarin-N₃ for 1h followed by another 1h incubation with 1μM RGD-photo-ODIBO. After the incubation, free Cetuximab-coumarin-N₃ and RGD-photo-ODIBO were removed by PBS wash, and cells were further treated with 10μM colchicine for 20min at 4°C. Subsequently, cells were irradiated under UV360nm for 10min to activate the photo-ODIBO group and then incubated with 20mM sodium acetate in HBSS (pH = 4.0) for 20min to remove the surface bound fraction. Upon removal of the HBSS buffer, cells were fixed with 4% paraformaldehyde at room temperature for 10min and imaged under the FITC channel using the confocal fluorescent microscopy.

For the control group, same steps were performed except that the colchicine treatment was conducted before the incubation with RGD-photo-ODIBO.

Statistical analysis:

The results are expressed as mean ± SD (n=3). Two sided t-test was used to analyze two different groups. In all statistical analyses, p < 0.05 was considered statistically significant, and *, **, *** indicate P < 0.05, 0.01 and 0.001 respectively. GraphPad Prism (Version 6.01) was used for the statistical analysis.