EDB Fibronectin-Specific SPECT Probe 99mTc-HYNIC-ZD2 for Breast Cancer Detection.

Extradomain-B fibronectin (EDB-FN), an oncofetal isoform of FN, is a promising diagnostic and therapeutic target of tumors, including breast cancer. Many EDB-FN-targeted drugs have been developed and have shown therapeutic effects in clinical trials. Molecular imaging to visualize EDB-FN-positive cancers may help select the right patients who will be benefit from EDB-FN-targeted therapy. Although a few EDB-FN-targeted imaging probes have been developed, the complicated manufacturing procedure and expensive material and equipment required limit their application for large-scale screening of EDB-FN-positive cancer patients. Thus, more simple and economic EDB-FN-targeted imaging probes are still urgently needed. Previously, we have identified a breast cancer-targeted peptide, CTVRTSADC. Coincidently, it was later identified as an EDB-FN-targeted peptide and named ZD2. In this study, we found a positive correlation between the binding activity of the ZD2 phage and the expression level of EDB-FN in breast cancer cells. Moreover, we observed the colocalization of the ZD2 peptide with EDB-FN in breast cancer cells. Furthermore, in vivo tumor targeting of the ZD2 phage, near-infrared fluorescence imaging, and flow cytometry showed tumor-specific homing of the ZD2 peptide in mice bearing EDB-FN-positive breast cancers. Importantly, on the basis of this EDB-FN-targeted ZD2 peptide, we developed a kit-formulated probe, 99mTc-HYNIC-ZD2, for single-photon-emission computed tomography (SPECT) imaging of breast cancer. The high tumor uptake of 99mTc-HYNIC-ZD2 demonstrated its feasibility for use in visualizing EDB-FN-positive breast cancers in vivo. This kit-formulated EDB-FN-targeted SPECT probe has potential clinical applications for precision screening of EDB-FN-positive cancer patients who may benefit from EDB-FN-targeted therapy.

Introduction

Breast cancer is the most common malignant cancer among women and the leading cause of cancer-related death worldwide.[1] Early diagnosis is the key to successful treatment of breast cancer.[2] X-ray-based mammography, a simple and economic imaging modality, has been shown to improve the early diagnosis rate of breast cancer.[2] Other anatomy-based imaging modalities, including computed tomography (CT), ultrasonography, and magnetic resonance imaging (MRI), are clinical diagnostic techniques that are routinely used for the diagnosis, staging, and therapeutic response assessments of breast cancer.[3] Although these traditional imaging techniques can provide important anatomical imaging information on breast cancer, they are limited in real-time monitoring of tumor functional molecular processes. Molecular imaging modalities based on special functional imaging probes are playing a major role in the era of precision medicine, as they can provide functional molecular imaging information for real-time monitoring of tumor progression at a molecular level.[4,5] Currently, single-photon-emission computed tomography (SPECT) and positron emission tomography (PET) are the major clinically available molecular imaging modalities. Tumor-targeted peptides can readily be radiolabeled with radionuclides such as 99mTc for SPECT and 18F for PET.[6]

Fibronectin (FN) is an abundant glycoprotein in the extracellular matrix of connective tissues.[7] FN regulates cell adhesion, migration, growth, proliferation, and wound healing during normal tissue development and tissue repair.[7] There are three different FN isoforms generated by alternative splicing of premessenger RNA: extradomain-A (EDA), extradomain-B (EDB), and type III homology-connecting segment.[8] The EDA-FN and EDB-FN isoforms sometimes referred to as oncofetal isoforms, as they are typically expressed during embryonic development but re-expressed during tumor development.[9] Functionally, they are involved in tumorigenesis,[10] angiogenesis,[11] and metastasis;[12] thus, they are promising diagnostic and therapeutic tumor targets.[10] EDB-FN-specific antibodies and peptides that mediate delivery of cytokines, cytotoxic agents, chemotherapeutic drugs, and radioisotopes have been developed.[13−19] Some are being examined in clinical trials and exert therapeutic effects in EDB-FN-positive cancer patients. With the help of EDB-FN-targeted molecular imaging, doctors can select suitable cancer patients for these EDB-FN-targeted therapy.[8] Several EDB-FN-targeted imaging probes have been developed;[8] however, most of them are still far from clinical application because they involve complicated manufacturing process and expensive materials and equipment. Thus, more simple and economic EDB-FN-targeted imaging probes are urgently needed.

Previously, we have identified several breast cancer-targeted peptides through phage-display peptide library screening against breast cancer MDA-MB-231 cells.[20] The neuroplin-1-targeted peptide CLKADKAKC (CK3) was translated into near-infrared fluorescence (NIRF) imaging and SPECT imaging probes for the diagnosis of breast cancer.[20] Here, we focus on another breast cancer-targeted peptide, CTVRTSADC, which was later identified as an EDB-FN-targeted peptide named ZD2 by Lu et al.[21] They labeled this EDB-FN-targeted ZD2 peptide with the NIRF dye Cy5 to form an NIRF imaging probe, Cy5-ZD2, for NIRF imaging of prostate cancer.[21] However, due to its poor tissue penetration, immature instrument, and high autofluorescence background, the NIRF imaging modality is not routinely used clinically. Here, we developed a more clinically applicable, simple, economic, and kit-formulated SPECT probe, 99mTc-HYNIC-ZD2. Compared with other developed EDB-FN-targeted imaging probes, this SPECT probe is more suitable for large-scale screening of EDB-FN-positive breast cancer patients, who may benefit from EDB-FN-targeted therapy.

Results

Identification of ZD2

Previously, we have identified 12 breast cancer-targeted peptides by screening the Ph.D.-CX7C peptide phage-display library against MDA-MB-231 breast cancer cells. To identify more breast cancer-targeted peptides, here, we randomly picked, purified, and examined 60 individual phages enriched with MDA-MB-231 cancer cells using the enzyme-linked immunosorbent assay (ELISA) (Figure S1). Fifty-four phage clones with a higher binding activity as compared to that of the control phage were identified and selected for DNA sequencing. This approach has identified another dominant peptide, CTVRTSADC (5 of 54 positive clones). Interestingly, this peptide has also been identified by us in our previous study[20] and later by Lu et al.[21] as being an EDB-FN-targeted peptide named ZD2. Here, we synthesized three ZD2-based peptides, including biotin-ZD2, Cy5-ZD2, and HYNIC-ZD2, for subsequent studies (Figure ).

Figure 1

Structures of (A) biotin-ZD2, (B) Cy5-ZD2, and (C) 99mTc-HYNIC-ZD2.

In Vitro ZD2 Specificity

Before the application of this peptide for the imaging of breast cancer in vivo, we used real-time polymerase chain reaction (RT-PCR) to examine the messenger RNA (mRNA) expression levels of EDB-FN in nine different breast cancer cells, including MDA-MB-468, MDA-MB-231, MDA-MB-415, MCF-7, BT-549, HCC1937, SK-BR-3, BT474, and MDA-MB-453 cells (Figure S2). Three representative breast cancer cell lines, including MCF-7 (low EDB-FN expression), MDA-MB-231 (intermediate EDB-FN expression), and MDA-MB-468 (high EDB-FN expression) (Figure A), were selected for further study. To link the expression level of EDB-FN to the binding activity of ZD2 phage, we examined the binding activity of the ZD2 phage in three selected breast cancer cells, as mentioned above. Consistent with the expression levels of EDB-FN in these cells, the binding activity of the ZD2 phage was low in MCF-7 cells, moderate in MDA-MB-231 cells, and high in MDA-MB-468 cells (Figure B). Linear regression analysis showed that there was a positive correlation between the binding activity of the ZD2 phage and the expression level of EDB-FN in these cells (Figure C). Moreover, we observed that the biotin-labeled ZD2 peptide biotin-ZD2 colocalized with EDB-FN in MDA-MB-468 cells (Figure D) but not in MCF-7 cells (data not shown). Thus, the ZD2 peptide binds to EDB-FN-positive breast cancer cells in vitro.

Figure 2

ZD2 peptide binds to EDB-FN-positive breast cancer cells. (A) Relative mRNA levels of EDB-FN in MCF-7, MDA-MB-231, and MDA-MB-468 breast cancer cells, examined by RT-PCR. The expression level of EDB-FN in MCF-7 cells was used as the control. (B) Binding of the ZD2 phage to MCF-7, MDA-MB-231, and MDA-MB-468 cells. Wild-type M13 phage without the coding peptide was used as the control phage. (C) Positive correlation between the EDB-FN mRNA expression levels and the binding affinities of the ZD2 phage in these cells. (D) Confocal immunofluorescence images of MDA-MB-468 cells incubated with the biotin-labeled ZD2 peptide. The Pearson correlation coefficient was about 0.65.

In Vivo Tumor Homing of ZD2

To determine whether the ZD2 peptide can specifically home to breast cancer tissues in vivo, we injected MDA-MB-468 cells under the mammary fat pad to form EDB-FN-positive breast cancer mice models. First, we examined whether the ZD2 phage would home to tumors in vivo. ZD2 phages were intravenously injected into mice bearing MDA-MB-468 tumors. The uptake of the ZD2 phage in different organs and tumor tissues was examined. There was more accumulation of the ZD2 phage in tumor tissues as compared to that of the control phage (Figure S3). We also observed that predominant phage uptake occurred in organs of the mononuclear phagocyte system, including the liver and spleen, but there were no significant differences between the accumulation of the ZD2 phage and control phage in these organs (data no shown).

We then labeled the ZD2 peptide with the NIRF dye Cy5 to form the NIRF imaging probe Cy5-ZD2, which was intravenously injected into mice bearing MDA-MB-468 tumors under the mammary fat pad. The Cy5-CG7C peptide, without any organ and tissue accumulation except in the kidney for excretion, was used as a negative control. Whole-body NIRF imaging showed that the intravenously injected Cy5-ZD2 peptide but not the control Cy5-CG7C peptide specifically homed to tumors (Figure A). NIRF imaging of the collected organs and tumors further demonstrated specific accumulation of the Cy5-ZD2 peptide into tumors but not into other organs (Figure B). Confocal imaging of the tumor tissues showed that the Cy5-ZD2 peptide but not the control Cy5-CG7C peptide specifically accumulated into tumors (Figure C). Additionally, we found that the Cy5-ZD2 peptide colocalized with EDB-FN in tumor tissues (Figure S4). Quantification of the collected organs and tumors indicated that the radiant efficiency produced by Cy5-ZD2 was about 15-fold over that by control Cy5-CG7C (Figure D).

Figure 3

Tumor-specific homing of Cy5-ZD2 in mice bearing MDA-MB-468 tumors. (A) NIRF imaging showed in vivo tumor homing of Cy5-labeled ZD2 (Cy5-ZD2). Approximately 200 μg of Cy5-ZD2 or the Cy5-labeled control peptide CG7C (Cy5-CG7C) was intravenously injected into mice bearing MDA-MB-468 tumors and allowed to circulate for 48 h before imaging. (B) Tumors and organs were collected and imaged 48 h post injection of the Cy5-labeled peptides. (C) Imaging of the tumor accumulation of Cy5-ZD2. (D) Quantification of Cy5-ZD2 and Cy5-CG7C peptides in tumors and organs using Living Image software. Statistical analyses were performed using Student’s t-test. n = 3; error bars, SD; ***p < 0.001.

Finally, the collected organs and tumors were further homogenized and analyzed by flow cytometry. Consistent with previous results, the accumulation of Cy5-ZD2 into tumor tissues was more than that of Cy5-CG7C (Figure ). These findings indicate that the ZD2 peptide specifically homes to EDB-FN-positive breast cancer in vivo.

Figure 4

Biodistribution of Cy5-ZD2 and control Cy5-CG7C in organs and tumor analyzed by flow cytometry. Note that the accumulation of Cy5-ZD2 into tumors was more than that of Cy5-CG7C.

SPECT Imaging of Breast Cancer with 99mTc-HYNIC-ZD2

According to previous studies, an HYNIC-modified peptide can be easily radiolabeled with 99mTc to form a SPECT probe 99mTc-HYNIC-peptide.[6] Referring to previous studies by Guggenberg et al., all radiolabeling processes would form a read-to-use kit, and the final radiolabeled products can be purified on a mini column.[22] Importantly, all materials and equipment required are simple and affordable. Thus, we developed a kit-formulated EDB-FN-targeted SPECT probe. 99mTc-HYNIC-ZD2 (Figure A), for the large -scale of screening EDB-FN-positive breast cancer patients who may benefit from EDB-FN-targeted therapy. 99mTc-HYNIC-ZD2 was first examined by radio-high-performance liquid chromatography (radio-HPLC) for radiolabeling yield measurement (Figure S5) and then was intravenously injected into mice that had EDB-FN-positive MDA-MB-468 breast cancer under the mammary fat pad (Figure B). SPECT imaging showed that 99mTc-HYNIC-ZD2 accumulated in the tumor and kidney 2 h post intravenous injection. (Figure C,D). The biodistribution of 99mTc-HYNIC-ZD2 showed that 99mTc-labeled ZD2 accumulated in the tumors and kidneys for excretion 1, 2, and 3 h after intravenous injection (Table ). Note that there was still relatively high accumulation of 99mTc-HYNIC-ZD2 in the lung except kidney (Table ). The reason behind 99mTc-HYNIC-ZD2 having a relatively high lung accumulation is deserved to be further researched in the future. Our present findings demonstrate that 99mTc-HYNIC-ZD2 is applicable for SPECT imaging of EDB-FN-positive breast cancer.

Figure 5

Tumor-specific homing of 99mTc-HYNIC-ZD2 in mice bearing MDA-MB-468 tumors. (A) CT imaging of the same mouse showed the tumor under the mammary fat pad. (B) White-light imaging of a mouse bearing an MDA-MB-468 tumor under the mammary fat pad. (C) SPECT imaging of the same mouse with 99mTc-HYNIC-ZD2. Note that 99mTc-HYNIC-ZD2 accumulated in the tumor and kidney 2 h post intravenous injection.

Table 1

Biodistribution of 99mTc-HYNIC-ZD2 in Mice Bearing MDA-MB-468 Tumorsa

	1 h		2 h		3 h
organs	%ID/g	T/B ratio	%ID/g	T/B ratio	%ID/g	T/B ratio
brain	0.0076 ± 0.0005	56.2	0.0045 ± 0.0045	42.0	0.0030 ± 0.0002	15.4
bone	0.0216 ± 0.0051	19.6	0.0211 ± 0.0211	9.1	0.0069 ± 0.0016	6.7
muscle	0.0616 ± 0.0148	6.9	0.0154 ± 0.0154	12.4	0.0109 ± 0.0026	4.2
intestine	0.0567 ± 0.0016	7.5	0.0331 ± 0.0331	5.8	0.0174 ± 0.0005	2.6
liver	0.0617 ± 0.0018	6.9	0.0317 ± 0.0317	6.0	0.0239 ± 0.0007	1.9
spleen	0.0815 ± 0.0221	5.2	0.0609 ± 0.0609	3.1	0.0344 ± 0.0093	1.3
skin	0.0821 ± 0.0055	5.2	0.0382 ± 0.0382	5.0	0.0200 ± 0.0013	2.3
heart	0.0894 ± 0.0179	4.8	0.0511 ± 0.0511	3.7	0.0254 ± 0.0051	1.8
lung	0.2706 ± 0.0335	1.6	0.0989 ± 0.0989	1.9	0.0790 ± 0.0098	0.6
kidney	4.6816 ± 0.4054	0.1	0.5984 ± 0.5984	0.3	0.3813 ± 0.0330	0.1
tumor	0.4249 ± 0.0305	1.0	0.1907 ± 0.1907	1.0	0.0458 ± 0.0033	1.0

Quantification of the organ imaging confirmed that 99mTc-HYNIC-ZD2 accumulated in the tumors and kidneys for excretion 1, 2, and 3 h after intravenous injection. The biodistribution data and T/B ratios were reported as an average plus the standard deviation at each time point.

Discussion

EDB-FN is overexpressed in many tumors, including but not limited to breast cancer,[23−25] liver cancer,[26] lung cancer,[27] colorectal cancer,[28] head and neck cancer,[29] lymphoma,[30,31] glioma,[32] and melanoma.[33] Kaczmarek et al. used the EDN-FN antibody BC-1 to perform immunohistochemical analysis in normal, hyperplastic, and neoplastic human breast tissues.[23] They found that 90 of 97 (93%) patients with invasive ductal carcinomas and 10 of 14 (71.4%) patients with invasive lobular carcinoma reacted positively to the EDN-FN antibody BC-1.[23] Matsumoto et al. analyzed human breast cancer tissues by in situ hybridization using probes specific to the EDN-FN alternative splicing site.[24] They observed that 33% of the invasive ductal carcinomas demonstrated EDB-FN-positive mRNA signals.[24] D’Ovidio et al. found that EDB-FN staining was significantly more in the stroma of poorly differentiated and metastatic breast cancer tissues, although it was not correlated with the progression of tumor vascularization.[25] All of these researches suggest diagnostic and therapeutic significance of EDB-FN in breast cancer.

EDB-FN-specific antibodies have been explored for the molecular imaging of EDB-FN in tumors. Monica et al. explored a radiolabeled EDB-FN antibody, 123I-L19(scFv)2, for immunoscintigraphic imaging of 20 patients diagnosed with colorectal, lung, and brain cancers.[34] They observed that 16 of 20 patients had different levels of antibody accumulation either in the primary tumors or metastases.[34] Poli et al. used 124I-labeled L19SIP in immunoPET to predict the doses delivered to tumor lesions and healthy organs by subsequent 131I-labeled L19SIP radioimmunotherapy in patients with brain metastases from solid tumors.[35] They showed that ImmunoPET with 124I-labeled L19SIP offered an advantage over conventional 131I imaging, that is, accuracy of dosimetric results from the treatment with radretumab.[35] Dietmar et al. radiolabeled 99mTc with three L19 derivatives for imaging of tumor angiogenesis. They found that 99mTc-AP39 showed the most favorable biodistribution and tumor-imaging properties.[36] Compared with these EDB-FN antibody L19-based imaging approaches, those involving the use of the ZD2 peptide as an EDB-FN-targeted imaging vehicle may offer several advantages, including low immunogenicity, easy synthesis, enhanced tissue penetration, ready diffusion, and simple radiolabeling. Moreover, compared with those of 123I and 124I, the optimal nuclear properties and ease of availability at a low cost of 99mTc make this radionuclide more suitable for the formation of SPECT imaging probes.

Kim et al. have identified several EDB-FN-specific aptides using their designed aptide libraries.[37] They conjugated these EDB-FN specific aptides with superparamagnetic iron oxide nanoparticles (SPIONs) to form APTEDB–SPIONs for MRI of Lewis lung carcinoma.[38] In the following study, they continued to developed other nanoparticles APTEDB–TCL–SPION for MRI of breast tumor-initiating cells.[39] These studies demonstrate that these contrast-enhanced MRI agents have the potential to enable the visualization of EDB-FN-positive cancers. Unfortunately, MRI in molecular imaging is limited by its relatively low sensitivity. Moreover, the complicated formation of SPIONs, the cost of the materials, and the time-consuming examining process may limit the use of MRI for the large-scale screening of EDB-FN-positive cancer patients.

EDB-FN-targeted peptides have been identified and explored for MRI of EDB-FN-positive tumors. Using phage-display technology,[40] Pilch et al. identified two FN–fibrin complex-specific cyclic peptides, CGLIIQKNEC (CLT1) and CNAGESSKNC (CLT2). Following study, Lu et al. translated the CLT1 peptide into MRI contrast agents.[41] They also identified a novel EDB-FN-targeted peptide, CTVRTSADC (ZD2), which was translated into the NIRF imaging agent Cy5-ZD2 for prostate cancer.[21] They labeled this EDB-FN-targeted ZD2 peptide with the NIRF dye Cy5 to form a NIRF imaging probe, Cy5-ZD2, for NIRF imaging of prostate cancer.[21] Coincidently, in a previous study, we have reported this ZD2 peptide as being a breast cancer-targeted peptide.[20]

Although several EDB-FN-targeted imaging probes have been developed, their complicated manufacture, expensive materials and equipment required, time-consuming detection procedure, and low sensitivity may limit their clinical application for large-scale of screening of potential cancer patients who may benefit from EDB-FN-targeted therapy. Thus, to develop more clinically applicable, simple, and economic imaging probes, here, we developed a kit-formulated SPECT probe, 99mTc-HYNIC-ZD2. This novel probe has excellent targeting properties, fast clearance, and a high tumor-to-organ ratio, suggesting its potential future application for the screening of breast cancer and other cancers with EDB-FN overexpression. Compared with other developed EDB-FN-targeted imaging probes, this probe has several advantages, including ease of mass production, simple radiolabeling procedure, ease of availability for routine clinical use, and low cost. However, it is noteworthy that 99mTc-HYNIC-ZD2 has a relative high accumulation in the lungs, which may limit its future application for the detection of tumors with lung metastasis.

Conclusions

In this study, we showed that the ZD2 peptide specifically bound to EDB-FN-positive breast cancer cell lines in vitro and homed to EDB-FN-positive breast cancer in vivo. On the basis of this EDB-FN-targeted peptide, we developed a more clinically applicable, simple, economic, and kit-formulated probe for SPECT imaging of breast cancer. The high tumor uptake of this probe in mice bearing EDB-FN-positive breast cancers suggests its feasibility of application in the SPECT imaging of EDB-FN-positive breast cancer and other cancers with EDB-FN overexpression.

Experimental Section

Materials and Methods

Reagents

Antifibronectin monoclonal antibody (BC-1; ab154210); anti-mouse Alexa Fluor 488 (Molecular Probes; Thermo Fisher Scientific), and chemical agents, including 4′,6-diamidine-2′-phenylindole dihydrochloride (DAPI), streptavidin-Cy3, ethylenediammonium diacetate (EDDA), tricine, mannitol, and SnCl2·2H2O, were purchased from Sigma-Aldrich (St. Louis, MO). 99mTcO4– was obtained from HTA Co., Ltd., China.

Cell Lines

MDA-MB-468, MDA-MB-231, MDA-MB-415, MCF-7, BT-549, HCC1937, SK-BR-3, BT474, and MDA-MB-453 breast cancer cells were purchased from American Type Culture Collection (Manassas, VA) and maintained in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum, penicillin, and streptomycin.

Mice

We conducted animal experiments according to the institutional guidelines of the Guangdong Province. All experiments were approved by the Use Committee for Animal Care. To create breast cancer mice models, 2 × 106 MDA-MB-468 cells were injected into the mammary fat pad of each NOD/SCID mouse.

Peptide Synthesis and Labeling

Biotin-CTVRTSADC, biotin-CG7C, Cy5-CTVRTSADC, Cy5-CG7C, and HYNIC-CTVRTSADC were synthesized at the Chinese Peptide Company (Hangzhou, China) using standard solid-phase Fmoc chemistry. The peptides were purified to a minimum purity of 95% by HPLC. The peptides were isolated by lyophilization.

ELISA

ELISA was performed as previously described.[20] Briefly, MDA-MB231 cells were plated in 96-well plates 24 h before ELISA. The wells were washed with phosphate-buffered saline (PBS) three times and fixed with 0.25% glutaral. Then, 3% hydrogen peroxide and 2% bovine serum albumin (BSA) were added to the wells to inhibit the endogenous peroxidase activity and block nonspecific binding, respectively. About 109 transducing units of phages were added and incubated for 2 h. After washing with 0.05% Tris-buffered saline with Tween 20 (TBST) three times, horseradish peroxidase-conjugated anti-M13 monoclonal antibodies (1:5000) were added, and the cells were incubated for 1 h at 37 °C. Tetramethylbenzidine was added for 30 min after finally washing with 0.05% TBST three times. The reaction was stopped by added 2 M H2SO4, and the absorbance was recorded against that of the blank at 450 nm with an ELISA plate reader.

Immunofluorescence

MDA-MB-468 and MCF-7 cells were plated on coverslips in 24-well culture plates. The wells were incubated with 50 mM biotin-CTVRTSADC (ZD2) or biotin-CG7C at 37 °C for 2 h. After washing three times with PBS, the cells were fixed with 4% paraformaldehyde for 10 min, blocked with 2% BSA for 30 min, and incubated with antifibronectin monoclonal antibody (BC-1; ab154210) (1:500) for 2 h. After washing with PBS three times, anti-mouse Alexa Fluor 488 (1:1000) and streptavidin-Cy3 (1:3000) were added and the cells were incubated for 1 h. The nuclei were stained with DAPI after three more washes. For the staining of the tumor tissues, the sections were fixed with 4% paraformaldehyde for 10 min, blocked with 2% BSA for 30 min, and then incubated with antifibronectin monoclonal antibodies (BC-1; ab154210) (1:500) for 12 h. After washing with PBS three times, anti-mouse Alexa Fluor 488 (1:1000) was added and the sections were incubated for 1 h. The nuclei were stained with DAPI after three more washes. Confocal images were acquired using a confocal laser-scanning system and FluoView application software FV10-ASW 3.0 (Olympus, Tokyo, Japan).

In Vitro Cellular Binding and in Vivo Tumor Targeting of the ZD2 Phage

In vitro cellular binding of the ZD2 and control phages was performed as previously described.[20] For in vivo tumor targeting of the ZD2 and control phages, about 109 transducing units of the ZD2 or control phage in 200 μL of DMEM were injected intravenously into mice bearing MDA-MB-468 tumors. The phages were allowed to circulate for 10 min, and the mice were perfused through the heart with 20 mL of DMEM. The organs and tumor were collected, weighed, and homogenized. The homogenates were washed three times with DMEM containing a protein inhibitor cocktail and PMSF. The organ- and tumor-binding phages were recovered and infected with ER2738 bacteria. The mixture was diluted and plated on agar plates.

NIRF Imaging

NIRF optical imaging was performed using the IVIS Spectrum (Xenogen Corporation, Caliper Life Sciences, Hopkinton, MA). The data were analyzed using Living Image version 3.0 (Caliper Life Sciences, Hopkinton, MA). The NIRF activity in different organs was quantified on the basis of the following procedures: a region of interest was defined, NIRF measurements were carried out, and the measures were expressed as efficiency using the Living Image software.

Flow Cytometry

To compare the uptake efficiencies of the Cy5-labeled peptides by the cells in the tumor tissues and normal organs, the tumor tissues and normal organs (e.g., heart, liver, spleen, lung, kidney, brain, tumor tissues) were collected and dispersed into single-cell suspensions with a homogenizer. The cell suspensions were analyzed by flow cytometry. The total number of gated cells (1 × 106) was measured in each tissue.

Kit Formulation

Kit formulation was performed as previously described.[22] Briefly, we dissolved EDDA in water with gently heating (20 mg/mL) and tricine/mannitol in water (40 mg/mL tricine, 100 mg/mL mannitol). A 24 mL stock solution (10 mg/mL EDDA, 20 mg/mL tricine, 50 mg/mL mannitol) was prepared by mixing 12 mL of EDDA and 12 mL of tricine/mannitol solution and purged with nitrogen in a rubber-sealed vial for 15 min. Before use, we dissolved HYNIC-ZD2 in water (500 μg/mL) and SnCl2·2H2O in nitrogen-purged 0.1 N HCl (1 mg/mL). The final solution contains 1 mL of EDDA/tricine/mannitol (10 mg of EDDA, 20 mg of tricine, 50 mg of mannitol), 0.04 mL of HYNIC-ZD2 (20 μg), and 0.04 mL of SnCl2·2H2O (40 μg) in sterile glass vials. The vials were lyophilized with a freeze dryer (AdVantage 2.0; SP Scientific), capped under vacuum, and stored at 4 °C.

Preparation of 99mTc-HYNIC-ZD2 and SPECT Imaging

Before radiolabeling, a Sep-Pak Light C18 Cartridge (Waters, Milford, MA) was conditioned with 5 mL of 95% ethanol and equilibrated with 5 mL of saline. First, we added 1 mL of 0.2 M Na2HPO4 to the kit formulation vial; then, Na99mTcO4 (20–25 mCi) generator eluate, saline to a final volume of 2 mL. The vial was incubated in a boiling water bath for 10 min and cooled at room temperature for another 10 min. The solution in the vial was loaded onto the prepared C18 Cartridge, which was washed with 5 mL of saline to remove hydrophilic impurities. Finally, 99mTc-HYNIC-ZD2 was eluted with 0.5 mL of 95% ethanol. The sample was examined by radio-HPLC for radiolabeling yield measurement and diluted with saline and intravenously injected into NOD/SCID mice bearing MDA-MB-468 tumors. Each tumor-bearing mouse was imaged at 2 h post injection.

Biodistribution of 99mTc-HYNIC-ZD2

The biodistribution of 99mTc-HYNIC-ZD2 was performed as previously described.[42] Mice were first anesthetized with sodium pentobarbital at a dose of 50.0 mg/kg and then sacrificed by cervical dislocation at 1, 2, and 3 h post injection. The organs of interest were collected, weighed, and measured for radioactivity in a γ-counter. The percentage of injected dose per gram of tissue mass (%ID/g) was calculated to indicate organ uptake. The organ uptake data and tumor-to-organ ratio were calculated on the basis of results from three mice at each time point.

SPECT Imaging

SPECT imaging of 99mTc-HYNIC-ZD2 was performed as previously described.[42] Briefly, the imaging study was performed using female NOD/SCID mice bearing MDA-MB-468 xenografts. Sodium pentobarbital at a dose of 50.0 mg/kg was first intraperitoneally injected to anesthetize mice, and then, 99mTc-HYNIC-ZD2 (∼400 μCi) in 0.2 mL of saline was intravenously injected and allowed to circulate 2 h post injection. SPECT images were acquired using a two-head γ-camera with a parallel-hole, low-energy, and high-resolution.

Data Analysis and Statistics

Statistical analyses were performed using SPSS Statistics 13.0 (SPSS Inc.). The data were analyzed using Student’s t-test and linear regression.