Preclinical Evaluation of a New Format of 68Ga- and 111In-Labeled Affibody Molecule ZIGF-1R:4551 for the Visualization of IGF-1R Expression in Malignant Tumors Using PET and SPECT

The Insulin-like growth factor-1 receptor (IGF-1R) is a molecular target for several monoclonal antibodies undergoing clinical evaluation as anticancer therapeutics. The non-invasive detection of IGF-1R expression in tumors might enable stratification of patients for specific treatment and improve the outcome of both clinical trials and routine treatment. The affibody molecule ZIGF-1R:4551 binds specifically to IGF-1R with subnanomolar affinity. The goal of this study was to evaluate the 68Ga and 111In-labeled affibody construct NODAGA-(HE)3-ZIGF-1R:4551 for the imaging of IGF-1R expression, using PET and SPECT. The labeling was efficient and provided stable coupling of both radionuclides. The two imaging probes, [68Ga]Ga-NODAGA-(HE)3-ZIGF-1R:4551 and [111In]In-NODAGA-(HE)3-ZIGF-1R:4551, demonstrated specific binding to IGF-1R-expressing human cancer cell lines in vitro and to IGF-1R-expressing xenografts in mice. Preclinical PET and SPECT/CT imaging demonstrated visualization of IGF-1R-expressing xenografts already one hour after injection. The tumor-to-blood ratios at 3 h after injection were 7.8 ± 0.2 and 8.0 ± 0.6 for [68Ga]Ga-NODAGA-(HE)3-ZIGF-1R:4551 and [111In]In-NODAGA-(HE)3-ZIGF-1R:4551, respectively. In conclusion, a molecular design of the ZIGF-1R:4551 affibody molecule, including placement of a (HE)3-tag on the N-terminus and site-specific coupling of a NODAGA chelator on the C-terminus, provides a tracer with improved imaging properties for visualization of IGF-1R in malignant tumors, using PET and SPECT.


Introduction
Insulin-like growth factor-1 receptor (IGF-1R) overexpression is involved in the neoplastic transformation of cells, contributing to a malignant phenotype by an enhanced proliferation rate and suppressed apoptosis [1,2]. The involvement of IGF-1R overexpression in the development, progression, metastasis, and therapy resistance of several malignancies, such as breast [3,4], prostate [5,6], pancreatic [7], and ovarian [8] cancers, is well documented in both preclinical and clinical studies. Understanding the importance of IGF-1R signaling has prompted the development and clinical evaluation of several potential anticancer therapeutics, targeting signaling through the IGF axis [9]. These drug candidates are, for example, monoclonal antibodies, directly interacting with IGF-1R, and they include ganitumab [10][11][12], figitumumab [13,14], cixutumumab [15,16], and dalotuzumab [17,18]. The results from clinical trials including these mAbs has unfortunately only shown a modest clinical benefit for the patients. However, the clinical trials were performed on unselected patient groups, and a more prominent response was found for a subset of the ovarian [10], pancreatic [19], and prostate [20] cancer patients. Moreover, the combination of the standard regimen with such antibodies was sometimes associated with an increased rate of adverse effects [11,13]. It has therefore been suggested that a subset of patients might benefit from IGF-1R-targeted therapy on the precondition of identification of the appropriate predictive biomarkers [11,21,22]. Both preclinical and clinical data suggest that the antiproliferative effect of anti-IGF-1R antibodies correlates with the level of IGF-1R expression [23][24][25]. Accordingly, identification of the tumors overexpressing IGF-1R might enable stratification of the patients that would benefit the most from such therapies and help to harness the full potential of IGF-1R-targeting pharmaceutics.
The most straightforward approach to measure the expression level would be to analyze biopsy samples. However, the invasive nature of biopsies limits the number of samples that can be collected and does not address the possible heterogeneity of expression of IGF-1R in the primary tumor and the metastases, or the variation of the expression over time. Apparently, the development of a non-invasive methodology for the determination of the IGF-1R status would allow for repeated measurements and would facilitate both clinical development and the routine use of IGF-1R-targeted pharmaceutics. A possible solution might be PET (positron emission tomography) or SPECT (single photon emission computed tomography) visualization of the expression by radiopharmaceuticals specifically binding to IGF-1R in vivo.
The use of radionuclide visualization of molecular target expression for subsequent radionuclide therapy (theranostics) is an established practice in the treatment of thyroid, prostate, neuroendocrine, and hematologic malignancies [26]. Currently, theranostic approaches are also finding use in regimens including monoclonal antibodies and antibody-drug conjugates [27]. In the case of IGF-1R, the development of imaging probes is challenging because of its noticeable expression in a number of healthy tissues, e.g., the lung, liver, intestines, and salivary gland (Human Protein Atlas: https://www.proteinatlas.org/ENSG00000140443-IGF1R/tissue, accessed on 15 June 2022). Furthermore, the overexpression in tumors is often modest, and only a relatively low expression level (10,000-30,000 receptors per cell) is necessary for the tumor to respond to IGF-1R-targeting antibodies [23,24]. This puts high demands on the imaging properties of the probes for IGF-1R visualization.
Several classes of radiolabeled probes for visualization of IGF-1R have been preclinically evaluated [28]. For example, the small-molecule IGF-1R tyrosine kinase inhibitor BMS-754807 was radiofluorinated with an aim to be used for PET imaging [29]. [ 18 F]F-BMS-754807 demonstrated specific binding to glioblastoma, breast cancer, and pancreatic tumor sections in vitro [29], and its biodistribution was evaluated in rodents [30]. However, no tumor imaging using [ 18 F]F-BMS-754807 has yet been reported.
Radiolabeled therapeutic anti-IGF-1R antibodies have been evaluated for imaging of IGF-1R in preclinical models [31][32][33]. R1507, a fully human monoclonal antibody against IGF-1R, has been labeled with 111 In for SPECT, using benzyl-isothiocyanatediethylenetriaminepentaacetic acid (DTPA). It has been shown that the uptake of [ 111 In]In-R1507 in tumors correlates with the IGF-1R expression level [31]. However, the contrast of imaging was low due to the long residence time of IgG in blood, a consequence of its large size (molecular weight 150 kDa), and interaction with the neonatal Fc receptor (FcRn), which protects IgG from degradation by cells in contact with blood [34]. Seven days were required to reach a tumor-to-blood ratio of 8.1 ± 2.5 for tumors with a high expression level of IGF-1R. To utilize the imaging power of PET, the long-lived positron emitter 89 Zr was used to label another anti-IGF-1R antibody, 1A2G11 [33]. In this case, the tumor-to-blood ratio was around 1 even 120 h after injection.
The use of smaller imaging agents improves the contrast [35]. Indeed, the use of a smaller F(ab') 2 -fragment (molecular weight of 110 kDa, and lacking interaction with FcRn) of the R1507 antibody resulted in a significantly increased tumor-to-blood ratio [32,36]. A tumor-to-blood ratio of 7.7 was achieved for a SUM149 breast cancer xenograft mouse model at 24 h after injection [32]. This indicated that the use of smaller protein-based imaging probes might improve the IGF-1R imaging contrast. IGF-1(E3R), a synthetic analogue of a natural ligand IGF-1, which is not recognized by IGF binding protein-3 (IGFBP-3), has been conjugated with DTPA by using its cyclic anhydride and labeled with 111 In [37]. This small (7.5 kDa) tracer demonstrated a strong correlation between its uptake and the IGF-1R expression level in human tumor xenografts in mice. The uptake in tumors with high IGF-1R expression (MCF-7/HER2-18) was 2.5 ± 0.4% ID/g, and the tumor-to-blood ratio was 5.8 at 4 h after injection. A possible disadvantage of this tracer was the stochastic attachment of the label due to conjugation of the chelator to any one of the four amino groups in the protein. Thus, the labeled product was a mixture of several variants with different positions of the chelator, which might affect the biodistribution and performance of the tracer.
While SPECT cameras are widely available, PET imaging provides better spatial resolution and higher quantification accuracy. Therefore, the development of a PET imaging probe for IGF-1R would be desirable. For this purpose, Su and co-workers evaluated the affibody-based tracer NOTA-Z IGF-1R:4:40 labeled with a long-lived positron emitter, 64 Cu ( Figure 1D) [44]. This tracer provided a tumor-to-blood ratio of 4.1 ± 0.6 at 24 h after injection. Although the use of NOTA provides stable coupling of 64 Cu [45], earlier studies on 64 Cu-labeled anti-HER2 affibody molecules have shown that the renal metabolism of such tracers after reabsorption in the proximal tubuli causes a release of 64 Cu into the blood stream, decreasing the tumor-to-blood ratio [46]. An alternative positron-emitting nuclide for labeling of anti-IGF-1R affibody molecules is 68 Ga (T 1/2 = 68 min). To investigate the performance of a 68 Ga-labeled variant of Z IGF-1R:4551 , a new design was applied in this study ( Figure 1E). A (HE) 3 -tag was introduced at the N-terminus since it has previously been found that the use of this tag can favorably influence the biodistribution of affibody molecules [47]. For coupling of the radionuclide, a maleimido derivative of the macrocyclic triaza-chelator NODAGA (1,4,7-triazacyclononane-1-glutaric acid-4,7-diacetic acid) was conjugated to a unique cysteine placed in the Cterminus of the affibody construct. NODAGA provides stable chelation with 68 Ga [48]. It has to be noted that it is desirable to have the possibility of SPECT imaging of IGF-1R expression, since SPECT scanners are more common than PET cameras in the medical community. The use of NODAGA in NODAGA-(HE) 3 -Z IGF-1R:4551 provides the possibility for labeling with the single-photon emitter 111 In [49].
The goal of this study was to perform a preclinical evaluation of NODAGA-(HE) 3 -Z IGF-1R:4551 , labeled with 68 Ga or 111 In, for imaging of IGF-1R expressing tumors by using PET and SPECT.
The IGF-1R-expressing cell line DU145 (prostate cancer) and the ovarian cancer cell line SKOV3 were obtained from the American Type Culture Collection (ATCC). The cells were cultured in Roswell Park Memorial Institute (RPMI) 1640 medium (Sigma-Aldrich, St. Louis, MO, USA) supplemented with 10% fetal calf serum, 2 mM L-glutamine, 100 IU/mL penicillin, and 100 mg/mL streptomycin. The data on the in vitro studies and biodistribution were analyzed by using GraphPad Prism (version 8.0.1 for Windows; GraphPad Software, La Jolla, CA, USA) to determine significant differences (p < 0.05, t-test).

Protein Production
The gene encoding (HE) 3 -Z IGF-1R:4551 with a C-terminally placed cysteine was expressed from the pET-21a(+) plasmid (Novagen, Madison, WI, USA), under control of the T7-promoter, essentially as earlier described for a different affibody construct [50]. Protein production was carried out overnight at 25 • C, after which the cells were harvested by centrifugation and lysed in a French Press. After heat treatment at 70 • C for 10 min, the (HE) 3 -Z IGF-1R:4551 protein was purified by anion-exchange chromatography on a Q-sepharose column, followed by reversed-phase chromatography on a Resources RPC column (GE Healthcare, Uppsala, Sweden), using an ÄKTA system (GE Healthcare). The fractions containing (HE) 3 -Z IGF-1R:4551 were pooled and lyophilized.

NODAGA Conjugation
(HE) 3 -Z IGF-1R:4551 was dissolved to a concentration of 1 mg/mL in PBS buffer (1 mL). To reduce potentially oxidized cysteines, tris (2-carboxyethyl) phosphine (TCEP) was added to a final concentration of 5 mM, followed by incubation for 30 min at 37 • C. The pH was adjusted to 6.5, using a 1 M HCl solution, after which the NODAGA chelator (20 mM in DMSO) was added to a 3:1 (chelator:protein) ratio. The conjugation reaction was allowed to proceed for 16 h at room temperature. The reaction mixture was purified (aliquot by aliquot) by reversed-phase high-performance liquid chromatography (RP-HPLC) on Agilent 1200 HPLC system (Agilent Technologies, Santa Clara, CA, USA), using a Zorbax 300SB-C18 column (9.4 × 250 mm, 5 µm particle size, Agilent Technologies, Santa Clara, CA, USA) with a 20 min gradient of 20-65% B (A = 0.1% trifluoroacetic acid (TFA) in H 2 O; B = 0.1% TFA in CH 3 CN), at a flow rate of 1.5 mL/min. An analysis of the purified product was performed on the same system, using a Zorbax CB300-C18 column (4.6 × 150 mm, 3. The fractions containing NODAGA-(HE) 3 -Z IGF-1R:4551 were pooled, lyophilized, and stored at −20 • C until labeling and biologic evaluation. Immediately before the evaluation, NODAGA-(HE) 3 -Z IGF-1R:4551 was redissolved in PBS to a concentration 3.4 mg/mL. Aliquots containing 20 µg in 5.9 µL PBS were prepared and stored at −20 • C.

Radiolabeling
For labeling with Ga-68, an aliquot of 20 µg of NODAGA-(HE) 3 -Z IGF-1R:4551 in PBS was mixed with 100 µL of 1.25 M sodium acetate, pH 3.6. The mixture was incubated with 80 µL of gallium-68 eluate (29)(30) MBq) at 50 • C for 15 min. Thereafter, the labeled compound was incubated with a 500-fold molar excess of EDTA (Ethylenediaminetetraacetic acid) for 5 min at 50 • C. After incubation, 1 µL samples were collected for measurement of the radiochemical yield using ITLC (see below). The labeled conjugates were purified by using NAP-5 columns (GE Healthcare) pre-equilibrated with PBS containing 1% BSA. The radiochemical purity of the conjugates was determined by ITLC.
To evaluate its stability, the purified conjugate was incubated with a 1000-fold excess of EDTA for 1 h at room temperature. The control samples were incubated in the same conditions but without the addition of EDTA. The protein-bound activity was analyzed by using ITLC. The stability test was performed in triplicates.
For labeling with In-111, an aliquot of 20 µg of NODAGA-(HE) 3 -Z IGF-1R:4551 in PBS was mixed with 50 µL of 0.2 M sodium acetate, pH 6.0. The mixture was incubated with 30 µL of 111 In solution in 0.1 M HCl (10 MBq) at 60 • C for 30 min. Thereafter, the labeled compound was incubated with a 500-fold molar excess of EDTA for 5 min at 60 • C. After incubation, 1 µL samples were collected for measurement of the radiochemical yield, using ITLC (see below). No purification was required since the radiochemical purity was over 95%.
The stability under EDTA challenge was determined in the same way as for the 68 Ga-labeled variant, but the incubation time was 6 h.
Instant thin-layer chromatography (ITLC) was performed by using glass microfiber chromatography paper impregnated with a silica gel (Agilent, Santa Clara, CA, USA). The ITLC strips were developed by 0.2 M citric acid, pH 2.0. In this system, the radiolabeled affibody molecules remain at the application point (R f = 0.0), while free radiometals (both 68 Ga and 111 In) and their complexes with EDTA migrate with the solvent front (R f = 1.0).

In Vitro Studies
The cells were seeded one day prior to the experiments in 3 cm petri dishes, with a density of 10 6 cells/dish. Each experiment was performed in triplicate for each data point.
The binding specificity of [ 68 Ga]Ga-NODAGA-(HE) 3 -Z IGF-1R:4551 and [ 111 In]In-NODAGA-(HE) 3 -Z IGF-1R:4551 to DU145 and SKOV3 cells was tested by incubating the cells with 1 nM of labeled conjugates for 30 min at 37 • C. To saturate the IGF-1 receptors, the cells were incubated with a 1000-fold excess of unlabeled (HE) 3 -Z IGF-1R:4551 for 20 min before adding the labeled compound in one set of dishes. Afterward, the cells were washed and detached by trypsin, and the radioactivity in cells was measured by using an automated gamma-spectrometer with a NaI (TI) detector (2480 Wizard, Wallac, Finland) to calculate the percent of cell-bound radioactivity.
To study cellular processing of [ 111 In]In-NODAGA-(HE) 3 -Z IGF-1R:4551 , a modified acid wash method [51] was used. Briefly, DU145 and SKOV3 cells were continuously incubated with 3 nM labeled conjugate at 37 • C. At predetermined time points (1, 2, 4, 6, and 24 h after incubation start), a group of dishes (n = 3) was removed from the incubator, and the incubation medium was collected. To separate the membrane-bound radioactivity, the cells were treated with 0.2 M glycine buffer containing 4 M urea, pH 2.5, for 5 min on ice, and the solution was collected. To isolate the internalized fraction of the radioconjugates, the cells were detached by treatment with 1 M NaOH, at 37 • C, for 30 min and collected. The activities of the incubation medium, the acidic buffer containing the membrane-bound conjugate, and the cells with the internalized fraction were measured to determine the membrane-bound and the internalized fractions.

In Vivo Studies
The animal experiments were planned and performed in accordance with national legislation on laboratory animal protection, and the study was approved by the local Ethics Committee for Animal Research in Uppsala (Permit 4C/16).
To establish IGF-1R-positive and IGF-1R-negative xenografts, 5 × 10 6 DU145 cells (in Matrigel, BD Biosciences) or Ramos cells (IGF-1R negative) were injected subcutaneously in the hind legs of female BALB/c nu/nu mice. The xenografts were allowed to grow for 2 weeks. In the biodistribution experiments, groups of four mice were used. At the time of the experiment, the average weight of mice with DU145 xenografts was 21.9 ± 1.0 g and 23.0 ± 0.3 g for mice bearing Ramos xenografts, respectively. The average tumor weight was 80 ± 37 mg and 46 ± 29 mg for mice bearing DU145 and Ramos xenografts, respectively.
The targeting properties of the 111 In-and 68 Ga-labeled NODAGA-(HE) 3  In-labeled probe was used. The labeled conjugates were formulated for co-injection based on a total injected protein mass of 1 µg per mouse. At each time point, a group of mice was sacrificed by heart puncture after intraperitoneal injection of a mixture of ketamine (250 mg/kg) and xylazine (25 mg/kg). Samples of blood, salivary glands, lung, liver, spleen, pancreas, stomach, large intestine, kidneys, tumor, muscle, bone, and the remaining carcass were collected. Organs and tissue samples were weighed, and their activity was measured by using a gamma-spectrometer separately for 68 Ga and 111 In, as described earlier [52]. These values were used to calculate the uptake of 111 In-and 68 Ga-labeled NODAGA-(HE) 3 -Z IGF-1R:4551 as a percentage of injected dose per gram of tissue (%ID/g).
To test the in vivo specificity, a group of mice with IGF-1R-negative Ramos xenografts were injected with a mixture of 700 kBq [ 68 Ga]Ga-NODAGA-(HE) 3 -Z IGF-1R:4551 and 10 kBq 111 In-labeled NODAGA-(HE) 3 -Z IGF-1R:4551 . Tumors and blood samples were collected at 1 h p.i., and their activity was measured as described above.
In vivo imaging was performed 1 h after injection to obtain a visual confirmation of the biodistribution data. Mice bearing DU145 xenografts were used for this purpose. One mouse was injected with 3.2 MBq (1 µg) [ 68 Ga]Ga-NODAGA-(HE) 3 -Z IGF-1R:4551 . Wholebody nanoPET images were acquired by using a nanoScan PET/MR (Mediso Medical Imaging Systems, Budapest, Hungary). The scan times were 45 min. A CT scan was performed immediately after the PET scan, using a nanoScan SPECT/CT (Mediso Medical Imaging Systems, Budapest, Hungary) with the same bed. The parameters for the CT scans were a 5 min acquisition time, an X-ray energy peak of 50 keV/670 µA, and 480 projections. A second mouse was injected with 1.2 MBq (1 µg) [ 111 In]In-NODAGA-(HE) 3 -Z IGF-1R:4551 . Whole-body SPECT/CT was performed by using nanoScan SPECT/CT (Mediso Medical Imaging Systems, Hungary). The acquisition time was 20 min. Gamma-peaks of 245 and 171 keV (window width of 20%) were used for acquisition. The CT scan was performed in the same way as in the case of [ 68 Ga]Ga-NODAGA-(HE) 3 -Z IGF-1R:4551 . A reconstruction of the scans was conducted by using the Tera-Tomo™ 3D reconstruction engine with decay correction at the injection administration time. The CT data were reconstructed by using filter back projection in Nucline 2.03 Software (Mediso Medical Imaging Systems Ltd., Budapest, Hungary). The PET and CT scans were fused by using InterView FUSION software (Mediso Medical Imaging Systems, Budapest, Hungary).

Production, Purification, and Conjugation
The (HE) 3 -Z IGF-1R:4551 protein was expressed in Escherichia coli and was purified by heat treatment, followed by anion-exchange and reversed-phase chromatography. The product was conjugated to NODAGA and thereafter purified by using reversed-phase chromatography. To analyze the purity of the conjugated protein, a sample was separated by analytical reversed-phase high-performance liquid chromatography (RP-HPLC, Figure 2A).
The protein was eluted as a single symmetrical peak. Determination of the area-undercurve shows that the protein was >99% pure, close to 100%. The molecular weight was also determined by mass spectrometry (Figure 2B), and the result differed less than 1 Da from the theoretical molecular weight of NODAGA-(HE) 3 -Z IGF-1R:4551 .
The specific activity of 1.5 MBq/µg (11.6 GBq/µmol) was reproducibly obtained for [ 68 Ga]Ga-NODAGA-(HE) 3 -Z IGF-1R:4551 . No measurable release was observed after the challenge with a large molar excess of EDTA (Table 1). Labeling with 111 In provided a radiochemical yield exceeding 98%, and no additional purification was required. The specific activity of 1.5 MBq/µg (11.6 GBq/µmol) was reproducibly obtained for [ 111 In]In-NODAGA-(HE) 3 -Z IGF-1R:4551 . In the EDTA challenge test, the difference between samples treated with EDTA and control samples was small and was within the accuracy of the method ( Table 1).

In Vitro Studies
The results of the qualitative receptor saturation experiment are presented in Figure 4.   Figure 5. The pattern of binding to both cell lines was similar. An initial rapid increase of cellassociated activity was recorded, followed by slower increase. The internalization rate by both cell lines was rather slow. The internalized fraction 24 h after the start of incubation was 15.9 ± 1.2% and 23.5 ± 5.5% of the cell-associated activity for DU145 and SKOV3 cells, respectively.
The blood concentration was further decreased with time, although at a slower rate. Another distinguished feature of both tracers was a high renal uptake. One hour after injection, the kidneys retained 67 ± 4% of the injected 68 Ga activity (uptake 250 ± 1% ID/g) and 62 ± 4% of the injected 111 In activity (uptake 231 ± 7% ID/g). The renal uptake was constant over time for both nuclides (no significant difference, p > 0.05, between the time points). A noticeable uptake of both tracers was also found in the lung, liver, stomach, and colon. There was an apparent radionuclide-based difference in the biodistribution. The uptake of [ 111 In]In-NODAGA-(HE) 3 -Z IGF-1R:4551 was significantly (p < 0.05 in a paired t-test) higher in the salivary gland, lung, pancreas, and colon at 1 h after injection, compared to the uptake of [ 68 Ga]Ga-NODAGA-(HE) 3 -Z IGF-1R:4551 .
The tumor uptake of both tracers remained stable between 1 and 3 h after injection (no significant difference between time points), but the uptake of [ 111 In]In-NODAGA-(HE) 3 -Z IGF-1R:4551 decreased significantly (1.4-fold) by 24 h. There was a decrease of the uptake in normal organs and tissues between 1 and 3 h after injection, but it was less that 2-fold. Accordingly, only the increase of tumor-to-blood ratio between these time points was prominent for both tracers (p < 0.001), and the increase of tumor-to-muscle ratio was significant (p < 0.05) ( Table 3). Increasing the time between injection and measurement for [ 111 In]In-NODAGA-(HE) 3 -Z IGF-1R:4551 from 1 to 24 h resulted in a significant increase of tumor-to-organ ratios, but the increase was mainly (except from blood) less than two-fold.
Images of mice bearing IGF-1R-expresing DU145 xenografts are presented in Figure 7.  Figure 7B), were capable of visualization of the tumors. Besides the tumors, a high activity uptake was observed in the kidneys, livers, and salivary glands, and this result is in agreement with the biodistribution data. Overall, the PET imaging provided better tumor visualization compared to the SPECT imaging.

Discussion
Excellent imaging of the expression level of several cancer-associated molecular targets by using radiolabeled affibody molecules has been demonstrated in preclinical studies and in clinical trials [38]. The small size of these targeting proteins facilitates their prompt localization in tumors on the one hand, and a rapid clearance of an unbound tracer from blood on the other hand, enabling high contrast of imaging within a couple of hours after injection. Furthermore, a precondition for the high contrast imaging is a strong affinity of the affibody probe to the molecular target [53], and affibody molecules with strong sub-picomolar affinity to many different cancer-relevant targets have been identified [39]. The typical ability of affibody molecules to refold after denaturation permits the use of elevated temperatures and a broad range of pH values during labeling, thus expanding the repertoire of suitable labeling methods compared to those available to most mAbs.
An important feature of affibody molecules to consider is the strong influence of radionuclide and a chelator for its attachment on their biodistribution and targeting characteristics [38]. This is a double-edged sword: Selection of an unfortunate combination might result in a tracer with poor imaging properties. Conversely, a systematic optimization of the molecular design (including labeling chemistry, i.e., the selection of radionuclide, chelator, and position of the label) of the affibody-based tracer may appreciably increase the imaging contrast and sensitivity of the procedure [38]. This is particularly important in the case of imaging of IGF-1R, which usually only has a modest expression in tumors, accompanied by a noticeable expression in normal tissues.
Our previous imaging experience with HER3 expression using affibody molecules [54] showed that placement of a (HE) 3 -tag at the N-terminus and a NODAGA-chelator at the C-terminus of the affibody provided the best PET imaging when 68 Ga was used as a label. However, the compositions of amino acids of the binding surface of HER3-and IGF-1Rbinding affibody molecules are different, and this will likely influence the biodistribution profile of the tracers appreciably by contribution to off-target interactions in vivo [55][56][57]. Thus, the selection of an optimal molecular design should be performed for every new affibody-based imaging probe.
The results of this study confirmed the favorable features of NODAGA-(HE) 3 -Z IGF-1R:4551 for imaging of IGF-1R expression. The protein was efficiently labeled with both 68 Ga and 111 In, and the labels were stable (Table 1). Generally, affibody molecules are stable to proteolysis [38]. Particularly, earlier studies have demonstrated that Z IGF-1R:4551 is stable in murine serum for at least one hour [41,43], and affibody molecules are nearly completely cleared from blood by one hour after injection. High thermodynamic stability and kinetic inertness of NODAGA complexes with 68 Ga and 111 In have also been demonstrated earlier [49]. Still, there is always a possibility that a fraction of a radionuclide would be coupled not by a macrocyclic chelator but by an unspecific "chelating pocket", which is formed by electron-donating sidechains. There is a risk that this coupling would be unstable and result in a release of a radionuclide in vivo. Therefore, we challenged the radiolabeled affibody molecules with a 1000-fold excess of EDTA, which should remove nuclides from weak binding sites. The results of this challenge (Table 1) (Figure 4). It has to be noted that, due to low level of IGF-1R expression (B max = 3.6 × 10 4 receptors per cell for the cells line DU145 with the highest expression, DU-145 [41]), the absolute values of the target-bound activity were modest. The internalization of radiolabeled NODAGA-(HE) 3 -Z IGF-1R:4551 was quite slow, with less than 30% after 24 h incubation ( Figure 5). This pattern is typical for other IGF-1R binding affibody molecules (see, for example, Reference [43]). Accumulation of both [ 68 Ga]Ga-NODAGA-(HE) 3 -Z IGF-1R:4551 and [ 111 In]In-NODAGA-(HE) 3 -Z IGF-1R:4551 in the IGF-1R-positive DU145 xenografts was significantly (p < 0.05) higher than in the IGF-1R-negative Ramos xenografts ( Figure 6), thus suggesting IGF-1R-specific uptake in tumors in vivo. Experimental imaging (Figure 7) confirmed that the visualization of IGF-1R expression in tumors is possible by using both PET and SPECT already 1 h after injection.
The half-life of 68 Ga (T 1/2 = 67 min) permits imaging up to 3-4 h after injection. Our data suggest that an increase of the time between injection and imaging, from 1 to 3 h, would increase the tumor-to-blood ratio by approximately two-fold (Table 3), which can be expected to further improve the imaging contrast. 111 In is a more long-lived (T 1/2 = 2.8 d) nuclide than 68 Ga, which permits image acquisition at later time points. Extending the time from 1 h to 24 h would permit an increase of the tumor-to-blood ratio from 4.1 ± 0.8 to 25.1 ± 1.3. It has to be noted that the uptake in lung, pancreas, stomach, colon, and salivary gland is IGF-1R-specific [40,41,43], and the clearance from these tissues was slow. The increase of ratios of radioactivity concentrations in tumors and these organs with time was only modest.
To evaluate the effect of the molecular design on the imaging properties of the anti-IGR-1R affibody imaging probes, we summarized their tumor-to-blood ratios in Table 4. Since the bloodborne activity contributes to the background in any organ and tissue, the tumor-to-blood ratio is one of most universal characteristics for comparison of different  3 -Z IGF-1R:4551 is higher than the ratio provided by [ 111 In]In-IGF-1(E3R) at 4 h after injection (5.8) [37]. It is as high as the value provided by 111 In-labeled intact antibody R1507 at 7 days after injection (8.1 ± 2.5) or by its F(ab') 2 fragment at 24 h after injection (7.7) [32]. When PET is not available, 111 In could be used as a label for NODAGA-(HE) 3 -Z IGF-1R:4551 . The disadvantages of the SPECT detection might be alleviated by an increase of the interval between injection and imaging to 24 h. At this time point, [ 111 In]In-NODAGA-(HE) 3 -Z IGF-1R:4551 provides the highest tumor-to-blood ratio reported for any of the IGR-1R imaging probes reported on in the literature. Obviously, 99m Tc has more favorable imaging properties and is cheaper than 111 In. However, the development of imaging probes is very expensive, and it is not realistic to expect that two different tracers with different chelators, one for PET and one for SPECT, would be developed simultaneously. The development of NODAGA-(HE) 3 -Z IGF-1R:4551 , which might be used for both radionuclide imaging modalities, seems to be a more realistic choice. In conclusion, a molecular design of the Z IGF-1R:4551 affibody molecule that includes the placement of the (HE) 3 -tag at N-terminus and site-specific coupling of a NODAGA chelator at C-terminus provides a tracer with improved imaging properties for the visualization of IGF-1R expression in malignant tumors when using PET and SPECT modalities.