The human epidermal growth factor receptor type 3 (HER3 or ErbB3) has recently attracted attention as a molecular target for anticancer therapy (1–3). HER3 has no tyrosine kinase activity of its own and participates in signaling by heterodimerization with other members of the HER family (4). One of the mechanisms behind acquired resistance of breast cancer patients to HER2-targeting therapies may be due to the upregulation of HER3 and hyperactivation of its signaling (5,6). In addition, HER3 has been shown to be involved in resistance to hormone therapy in breast cancer (7). Furthermore, HER3 expression in ovarian cancer is associated with more aggressive disease; in lung cancer, with resistance to tyrosine kinase-inhibiting therapy; and in prostate cancer HER3 expression may drive tumor cell growth in hormone refractory cancer (1). Thus, there is a strong rationale for the development of pharmaceuticals specifically targeting HER3. Besides monoclonal antibodies, HER3-targeting antisense oligonucleotides and scaffold proteins are under pre-clinical and early clinical development (3,8,9).

The clinical implementation of anti-HER3 therapy is a challenge. Treatment should be initiated at onset of HER3 overexpression to avoid overtreatment. As HER3 overexpression often occurs as a response to therapy, the samples from primary tumors would not be informative. Serial biopsies are undesirable due to invasiveness of the procedure and HER3 overexpression may not be synchronous in all the metastases that occur. The use of radionuclide molecular imaging is a useful strategy since it is non-invasive, repeatable and provides whole-body information.

Prerequisites for the successful application of molecular imaging are high specificity and sensitivity. Although radio-labeling of available therapeutic anti-HER3 monoclonal antibodies is an approach to the development of imaging probes, it is difficult to obtain a high sensitivity and specificity in this manner. Anti-HER3 antibody (10) and F(ab′)2 antibody fragments (11) were recently evaluated for positron emission tomography (PET) imaging in a preclinical setting. Slow clearance of antibodies from blood results in a high background even several days after injection, reducing imaging contrast and sensitivity. Additionally, bulky immunoglobulins with a molecular weight of >45 kDa tend to accumulate non-specifically in tumors due to an ‘enhanced permeability and retention’ (EPR) effect (12). This may cause false-positive findings, i.e., specificity of antibody-mediated imaging is low. A possible solution is the use of smaller imaging agents, which provide rapid extravasation and diffusion into the extracellular space, rapid clearance of unbound tracer, and exclude the EPR effect.

Previously, we demonstrated the feasibility of radio-nuclide imaging of HER3 expression in vivo (murine models) using a ^99mTc-labeled HER3-antagonistic affibody molecule ^99mTc(CO)₃-HEHEHE-Z08699 (13). Affibody molecules are small (7 kDa) scaffold proteins that are selected using molecular display techniques for high-affinity binding to different target proteins (14). Affibody molecules have demonstrated the ability of imaging of HER2 expression in breast cancer patients (15,16).

HER3-targeting affibody molecules Z08698 and Z08699 have been selected and demonstrate high affinities to HER3 (50 and 21 pM, respectively) (9). Such high affinities are important for the imaging of molecular targets with a relatively low tumor expression. Z08698 and Z08699 were also found to have high affinity to the murine counterpart of HER3, mErbM3, rendering mice an adequate model for the evaluation of imaging properties (13).

The aim of this study was the development of a HER3 imaging agent labeled with radiometals suitable for single photon emission computed tomography (SPECT) and PET.

Engineering of a unique cysteine into affibody molecules provides site-specific conjugation of the chelator, resulting in well-defined products with reproducible biodistribution properties. Findings regarding affibody molecules suggested that placement of a cysteine at the C-terminus is convenient, as this provides a maximum distance from a binding site and minimizes possible interference of the chelator with the molecular recognition (17,18). The use of the NOTA chelator (2,2′,2″-(1,4,7-triazonane-1,4,7-triyl)triacetic acid) provides stable labeling with ¹¹¹ In for SPECT, and with ⁶⁸Ga, ⁶⁴Cu or ¹⁸F (using aluminum monofluoride ¹⁸F-AlF chemistry) for PET. Previously, we showed that placement of a HEHEHE-tag (histidyl-glytamyl-histidyl-glytamyl-histidyl-glytamyl) at the N-terminus permits immobilized metalion affinity chromatography (IMAC) purification of affibody molecules and improves their biodistribution by reducing hepatic uptake (13,19). Bearing these considerations in mind, we generated two anti-HER3 affibody molecules with the HEHEHE-tag at the N-terminus and NOTA chelator at the C-terminus (i.e., HEHEHE-Z-NOTA) (Fig. 1).

Investigations on factors influencing imaging using HER2-targeting affibody molecules have demonstrated that low picomolar affinity is essential for the imaging of targets with low expression, such as HER3 (20). Previous findings have also demonstrated that small changes in amino acid composition can modify off-target interactions and affect blood clearance rate and hepatic uptake (20). Thus high affinity is not the only parameter to be considered for the selection of affibody molecules for imaging, and evaluation of several variants with approximately equal affinity may prove useful in selecting an agent with the most suitable in vivo properties.

Findings obtained showed that targeting of HER3 and other tyrosine kinases with an expression pattern similar to that of HER3 (endogenous expression in healthy tissue, e.g., IGF-1R) suggested that imaging contrast could be improved at later time-points (i.e., at 24 h p.i.) (13,21). Thus, ¹¹¹In was used as a label whose half-life permits studying biodistribution at the day of injection and during the following day. The targeting properties of ¹¹¹In-HEHEHE-Z08698-NOTA and ¹¹¹In-HEHEHE-Z08699-NOTA were compared in vitro and in vivo in nude mice bearing HER3-expressing BT-474 breast carcinoma xenografts.

High-quality Milli-Q water (resistance >18 MΩ/cm) was used to prepare the solutions. [¹¹¹In]-indium chloride was purchased from Covidien, Ltd. (Dublin, Ireland). Cells used during the in vitro experiments were detached using trypsin-EDTA solution (0.05% trypsin, 0.02% EDTA in buffer; Biochrom AG, Berlin, Germany). Radioactivity was measured using an automated gamma-counter with a 3-inch NaI(Tl) detector (1480 Wizard; Wallac Oy, Turku, Finland). The purity of radiolabeled affibody molecules was determined by radio ITLC (150-771 Dark Green, Tec-Control Chromatography strips from Biodex Medical Systems) and cross-validated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). The distribution of radioactivity along the thin layer chromatography strips was measured on a Cyclone Storage Phosphor System and analyzed using the OptiQuant image analysis software (Perkin-Elmer, Waltham, MA, USA).

For the in vitro and in vivo experiments, the HER3-expressing BT-474 breast carcinoma cell line [American Type Tissue Culture Collection (ATCC) via LGC Promochem, Borås, Sweden] was used. Animal experiments were planned and performed in accordance with national legislation on protection of laboratory animals and were approved by the Ethics Committee for Animal Research in Uppsala, Sweden.

Data on the cell uptake and biodistribution were assessed by an unpaired, two-tailed t-test using GraphPad Prism (version 4.00 for Windows GraphPad Software) to determine significant differences (P<0.05).

Elucidation of the role of HER3 in cancer resistance to different types of treatments has resulted in the development of HER3-targeting therapies. Imaging may be a valuable clinical tool for the stratification of patients for such therapies, making treatment more personalized. Previously, we demonstrated the feasibility of imaging of HER3-expressing xenografts using a ^99mTc-labeled anti-HER3 HEHEHE-Z08698 affibody molecule (13). That investigation initiated a series of studies focused on the development of affibody-based HER3-imaging agents suitable for PET and SPECT applications. To achieve that aim, a NOTA chelator, site-specifically conjugated to a C-terminal cysteine of the affibody molecule was selected. In a new HEHEHE-Z-NOTA format tested for two variants, Z08698 and Z08699, the affibody molecules retained their low picomolar affinity to HER3. However, the altered format was found to have some effect on the mutual affinities. The dissociation constant of HEHEHE-Z08698-NOTA (45 pM) was lower than that of HEHEHE-Z08699-NOTA (77 pM), while the opposite relationship was observed for the parental affibody molecules Z08698 (50 pM) and Z08699 (21 pM). The two conjugates retained a capacity for high-fidelity refolding after heating, which is essential as labeling using macrocyclic chelators requires relatively harsh temperature conditions. Labeling of conjugates was performed at elevated temperatures (85°C) that provided stable labeling of ¹¹¹In-HEHEHE-Z08698-NOTA and ¹¹¹In-HEHEHE-Z08699-NOTA. However, the two conjugates demonstrated preserved binding specificity to HER3-expressing BT-474 breast cancer cells after labeling.

The two conjugates demonstrated saturable binding to HER3-expressing breast cancer xenografts in vivo, which suggests target-specific uptake. Of note, saturable uptake was also demonstrated in mErbB3-expressing murine tissues (Fig. 3). Expression of a molecular target in normal tissues may appreciably affect imaging, and cross-reactivity of affibody molecules with the murine counterpart makes the animal model adequate for assessment of imaging properties.

Uptake of conjugates in tumor xenografts did not differ significantly between ¹¹¹In-HEHEHE-Z08698-NOTA and ¹¹¹In-HEHEHE-Z08699-NOTA at earlier time points (1–8 h p.i.). However, ¹¹¹In-HEHEHE-Z08698-NOTA had improved tumor retention of radioactivity at 24 h p.i. The main difference between the conjugates was in their uptake in normal organs and tissues. It is apparent that ¹¹¹ In-HEHEHE-Z08698-NOTA cleared more rapidly from blood but had a higher uptake in liver and spleen. Specifically, the two phenomena may be connected and elevated hepatic uptake may be a reason for more rapid conjugate elimination from blood. Furthermore, the HER3-binding site of the tested conjugates may be involved in off-target interactions and thereby influence their biodistribution profiles. The only differences between these affibody molecules in the binding sites were that Q11 and W25 in Z08698 were substituted by N11 and Y25 in Z08699. Previous findings on HER2-targeting affibody molecules (20) have shown that substitution of two amino acids in the binding site by two more lipophilic ones can cause higher hepatic uptake. Overall, ¹¹¹In-HEHEHE-Z08698-NOTA provided higher tumor-to-blood ratio, and should be considered as superior.

In a previous study, we evaluated targeting of BT-474 xenografts using HEHEHE-Z08699 labeled with ^99mTc(CO)₃ (4 h p.i.) (13). A comparison with ¹¹¹In-HEHEHE-Z08699-NOTA suggests that the combination of ¹¹¹In as a label and NOTA as a chelator provided a marked increase in tumor uptake (5±1 vs. 1.7±0.7 %IA/g, for ¹¹¹In-HEHEHE-Z08699-NOTA and ^99mTc(CO)₃-HEHEHE-Z08699, respectively). ¹¹¹In-HEHEHE-Z08699-NOTA provided ~2-fold higher tumor-to-stomach and tumor-to-lung ratios, the latter of which is important in the context of breast cancer as lung metastases are extremely common in this disease. This finding suggests that the HEHEHE-containing NOTA-conjugated affibody molecule is a preferable platform for further development of HER3-imaging agents.

Few studies are available on the development of imaging agents visualizing HER3 in vivo. One potential agent, ¹¹¹In-DTPA-HRG provided a tumor uptake of 2.1±0.4 %IA/g, and a tumor-to-blood ratio of 7 at 2 days p.i. in mice bearing HER3-positive xenografts (25). An anti-HER3 antibody (⁸⁹Zr-RG7116) demonstrated the highest tumor uptake and imaging contrast at 6 days p.i. (10). Another PET imaging probe based on a F(ab′)2 antibody fragment was presented in an abstract (11). This probe demonstrated comparable contrast to the anti-HER3 affibody molecules shown in the present study, although the time point was not specified (11).

¹¹¹In-HEHEHE-Z08698-NOTA has reasonable imaging contrast (tumor-to-blood ratio of 12±3, tumor-to-muscle ratio of 16±4, tumor-to-lung ratio of 2.3±0.3, and tumor-to-bone ratio of 8.9±0.5) already at 4 h p.i. This indicates that this variant labeled with short-lived ⁶⁸Ga (T½ =67.6 min) and ¹⁸F (T½ =110 min) may be useful for imaging using PET. The labeling chemistry should be re-optimized, and imaging properties have to be re-evaluated for these conjugates. However, the current study provides a good rationale for future studies.

The suggested imaging agent may enable a non-invasive and repeatable visualization of HER3 expression in metastatic cancer. Detection of HER3 overexpression onset suggests including anti-HER3 therapy to the treatment of disseminated cancer. Use of HER3-imaging agents in combination with ¹⁸F-FDG PET may initially detect metastases and provide their anatomical location prior to imaging of HER3 using affibody molecules, as has been described for HER2-imaging affibody molecules.

In conclusion, ¹¹¹In-HEHEHE-Z08698-NOTA is a promising imaging agent for visualization of HER3-expression in cancer metastasis using SPECT. A good imaging contrast at 4 h after injection indicates that this conjugate may be used for patient stratification for anti-HER3 therapy.