Dual Targeting of Integrin αvβ3 and VEGF Receptor Improves PET Imaging of Breast Cancer

Background: Neuropilin-1 (NRP-1) and integrin α v β 3 receptors are both overexpressed in breast cancer. Methods: We developed and synthesized a heterodimeric tracer consisting of Arg−Gly−Asp (RGD) and Ala-Thr-Trp-Leu-Pro-Pro-Arg (ATWLPPR) peptides that simultaneously targets integrin α v β 3 and NRP-1. We then investigated the diagnostic ecacy of RGD-ATWLPPR heterodimeric peptides in MCF-7 xenograft models in mice. DOTA-conjugated RGD-ATWLPPR peptides were then radiolabeled with 68 Ga, and the receptor-binding characteristics and tumor-targeting ecacy of 68 Ga-DOTA-RGD-ATWLPPR were investigated in vitro and in vivo. We also detected integrin α v β 3 and NRP-1 expression in MCF-7 cells and tumor tissues. Results: The peptide showed high stability in vitro. Static PET/CT imaging studies showed that MCF-7 tumors were clearly visible 30 min and 60 min after 68 Ga-DOTA-RGD-ATWLPPR injection. Further, the heterodimer showed higher tumor uptake than 68 Ga-DOTA-RGD or 68 Ga-DOTA-ATWLPPR alone. High specicity was shown in blocking studies using RGD and ATWLPPR peptides. MicroPET/CT and biodistribution studies using 68 Ga-DOTA-RGD-ATWLPPR showed that the tracer specically targets NRP-1 and integrin α ν β 3 receptors. Conclusions: Compared with 68 Ga-DOTA-RGD and 68 Ga-DOTA-ATWLPPR, 68 Ga-DOTA-RGD-ATWLPPR has higher binding anity, better targeting eciency, and longer tumor retention time. It therefore has potential as an imaging probe for breast cancer detection. in 10 min with a high yield. The results from microPET/CT and biodistribution studies showed that the tracer specically targets integrin α v β 3 and NRP-1. Thus, our probe will successfully localize to cancerous lesions and guide precise treatment. We expect that our high-eciency radiotracer will improve high-quality noninvasive imaging and allow targeted cancer therapies in the future.


Background
Breast cancer is the most common neoplastic disease and the leading cause of cancer-related deaths in over 100 countries. In 2018, the estimated number of new breast cancer cases was about 2.1 million worldwide, and the incidence is still increasing [1,2]. Tumor heterogeneity limits early breast cancer diagnosis and targeted therapy. Obtaining data on breast tumor heterogeneity in diverse populations could inform actions to improve precision care in breast cancer research. At present, traditional imaging approaches for breast cancer, such as digital mammography (DM), ultrasound imaging (US), computed tomography (CT), and magnetic resonance imaging (MRI) display tumor volume characteristics. However, clinicians urgently need to understand the biological changes, expression of receptor proteins in tumor tissues, spatial distribution of drugs, and treatment response in breast cancer patients. To this end, functional imaging of molecules is under development.
ATWLPPR, a new peptide target for molecular tumor imaging, effectively antagonizes the binding of vascular endothelial growth factor (VEGF) 165 to the VEGF2 receptor and the co-receptor neuropilin-1 (NRP-1) [3,4]. The VEGF165 signal is preferentially transduced via the tyrosine kinase receptor VEGF2 and is signi cantly enhanced by the association of VEGF2 with NRP-1 on endothelial cells. VEGFR2 and NRP-1 co-receptors stimulate tumor vascular endothelial cell migration and proliferation and increase vascular permeability [5,6]. In molecular docking studies, ATWLPPR localized in the head of the NRP-1 b1 domain, where Arg formed ionic interactions with Asp1054 of VEGFR2 [7,8]. Thus, ATWLPPR is a good candidate for discovering new drugs.
Widespread applications of RGD peptides have stimulated progress in molecular breast cancer imaging. Synthetic peptides containing the RGD sequence can speci cally bind to integrin α v β 3 , which has an important effect on tumor angiogenesis and metastasis in breast cancer. In the past decade, much research effort has been devoted to the molecular imaging of integrin α v β 3 with RGD peptide. Some of these peptides, including 18 F-Galacto-RGD, 18 F-Fluciclatide, 18 F-RGD-K5, 18 F-Alfatide, 18 F-FPPRGD2, 68 Ga-NOTA-RGD, and 68 Ga-NOTA-PRGD2 are currently under clinical investigation for the identi cation of malignant lesions and quantitative assessment of integrin α v β 3 expression [9][10][11].
Though monomeric target-based tumor imaging has been extensively studied, its application still has limitations: rapid elimination, low binding a nity, and suboptimal in vivo pharmacokinetics. Additionally, for early stage or metastatic tumors, the lack of expression of a single biomarker reduces the signal [12][13][14][15]. These limitations may decrease diagnostic accuracy. Dual-receptor targeting tracers can amplify signals because they represent the sum of monomer binding sites [16][17][18][19]. Our previous study indicated that a single peptide bound to other targeted peptides improves the imaging and diagnostic e ciency of tumor angiogenesis in malignant gliomas [20]. Both integrin ανβ3 and NRP-1 are expressed in breast cancer [21,22]. Thus, we hypothesized that a newly synthesized dipeptide probe, 68 Ga-RGD-ATWLPPR, can bind multiple neovascularization-related targets. We evaluated the feasibility and preclinical diagnostic value of a similar probe, 68 Ga-DOTA-RGD-ATWLPPR, in breast cancer xenograft mice, compared with 68 Ga-labeled monomeric RGD and ATWLPPR tracers.

Reagents and instruments
Peptides were purchased from ChinaPeptides Co. Ltd (Shanghai, China). All other chemicals were purchased from Sigma-Aldrich Chemical (USA), unless otherwise indicated. Chemicals were used directly without further puri cation. Rabbit anti-avβ3 and anti-NRP-1 monoclonal antibodies were purchased from Abcam (Shanghai, China). MCF-7 cell lines were purchased from the cell bank of the Chinese Academy of Sciences in Shanghai and utilized for cell studies. [ 68 Ga] GaCl 3 was produced with a 68 Ge/ [ 68 Ga] generator (China Isotope & Radiation Corporation, China). RGD-ATWLPPR heterodimer was synthesized using our published method [20]. Analytical HPLC had a ow rate of 1 mL/min with a C18 column (10 μm, 250 mm × 4.6 mm, Macherey-Nagel, Nucleosil 100-10). Micro-positron emission tomography/computed tomography (PET/CT) was performed using an Inveon microPET/CT scanner (Siemens).
Subsequently, Peptide DOTA-RGD-ATWLPPR identi cation was performed using mass spectrometry (MS), and we chose the MS procedure for selecting small molecular samples, and the parameters were properly adjusted to obtain the mass spectrum and store the map. Animal experiments were approved by the Xiamen University Institutional Animal Care and Use Committee. Female BALB/c nude mice (6-8 weeks old, 18-20 g) were obtained from the Xiamen University Laboratory Animal Center (Xiamen, China). Each mouse was injected subcutaneously with 5×10 6 cancer cells. Two to three weeks after inoculation, the mice were used for biodistribution and imaging studies.
Tumor-speci c integrin α v β 3 and NRP-1 expression For these experiments, 2 × 10 5 cells were seeded on coverslips in each well of a 6-well plate. MCF-7 cells were washed 3 times with phosphate-buffered saline (PBS) and xed with 4% paraformaldehyde for 20 min. The cells were blocked with PBS containing 1% goat serum for 1 h. Subsequently, the cells were incubated with 5 ug/ml mouse anti-avβ3 monoclonal antibody (Abcam, Cat. No.ab190147) overnight at 4°C and were washed 5 times with PBS for a total of 3 h, incubated with Alexa Fluor™ 594-labeled tyramide for 1 h at 25 °C, or rabbit anti-neuropilin-1 monoclonal antibody (Abcam, Cat. No.ab81321) overnight at 4°C and were then washed 5 times with PBS and incubated with Alexa Fluor™ 488-labeled tyramide for 1 h at room temperature. The samples were observed under laser scanning confocal microscope. MCF-7 tumor tissues were collected, xed in 4% paraformaldehyde, dehydrated, and embedded in para n. The tumor sections (5 μm) were washed three times with PBS. The sections were rinsed with EDTA buffer and blocked with 5% bovine serum albumin. The sections were incubated with primary antibodies (anti-NRP-1, anti-ανβ3; 1:100; Abcam) at 4 °C overnight. After incubation with biotinylated secondary antibody and avidin-biotin-peroxidase reagents for 10 min, 3'3-diaminobenzidine was added for color development. observed using light microscopy.
In vitro cell uptake and blocking studies MCF-7 cells were cultivated to 90% con uence and counted. The cells were seeded into 12-well plates at a density of 5 × 10 5 cells per well and incubated overnight at 37 °C to allow adherence. Then, uptake of 68 Ga-DOTA-RGD-ATWLPPR, 68

Statistical analysis
All data are presented as the mean ± SD from at least three independent experiments. statistical analyses were performed using RStudio (R 3.6.1). Means were compared using one-way ANOVA. P < 0.05 was considered statistically signi cant.

Chemistry and radiochemistry
In this study, DOTA was conjugated to RGD-ATWLPPR via glutamate to obtain DOTA-RGD-ATWLPPR.
Coupling was performed by reacting the DOTA carboxyl group and the free amino group on the polypeptide, and the expected products were con rmed by both analytical HPLC and MS. The

Integrin α v β 3 and NRP-1 expression in breast cancer tumor cells and tissues
Some of the differences observed between cancer tissues and cell lines indicate that the expression levels may be different between in vitro culture and in vivo tumor tissue. Immuno uorescent staining for integrin α v β 3 and NRP-1 was performed in MCF-7 cells (Fig. 3a); cellular immuno uorescence revealed that the cytoplasmic location of NRP-1 was strong in MCF-7 cells, whereas a much weaker uorescence signal was found for α v β 3 . To con rm the weak expression of α v β 3 in MCF-7 cells, western blotting (WB) was performed. We found that integrin αvβ3 expression was weakly positive in MCF-7 cells. Positive α v β 3 and NRP-1 are displayed with DAB stain; immunohistochemistry showed that tumor cells strongly expressed α v β 3 and NRP-1 (Fig. 3c). As shown in Fig. 3d, Expression of α v β 3 and NRP-1 was determined by immunohistochemical semi-quantitative analysis; αvβ3 and NRP-1 showed different expression levels in the same MCF-7 xenograft tissue, proving that MCF-7 xenograft tissue was heterogeneous.

Micropet/ct Imaging Of Tumor-bearing Mice
Mouse MCF-7 xenograft model were imaged by microPET using 68 Ga-DOTA-RGD-ATWLPPR, which can accurately display the tumor location. Representative coronal and transverse images were obtained at 30 and 60 min after injection of 68 Ga-DOTA-RGD-ATWLPPR. In contrast, MCF-7 tumor xenograft mice were injected with only normal saline or 3.7 MBq 68 Ga iron to serve as the control groups at 30 min. As shown in Fig. 5a, no gamma signal was detected in the normal saline group at 30 min, and the background signal was clean. We also added a control group that was injected only 68 Ga iron; the transverse and coronal microPET/CT images showed non-speci c diffuse signals throughout the mice body, but the tumor (yellow arrow) showed no obvious higher absorption. However, microPET/CT Images were acquired at 30 and 60 min after injection of 68 Ga-DOTA-RGD-ATWLPPR (3.7-7.4 MBq; 100-200 µCi; n = 3/group), the tumor (yellow arrows) has an obvious high absorption and contrast. Quantitative analysis showed that the average T/M ratios at 30-and 60-min post-injection were 5.21 ± 0.82 and 4.06 ± 0.12, respectively. In addition to tumor uptake, prominent 68 Ga-DOTA-RGD-ATWLPPR uptake was observed in the kidneys and bladder. We performed 60-min dynamic 68 Ga-DOTA-RGD-ATWLPPR PET scans of the heart, tumor, muscle, and kidneys. The unbound tracer was rapidly cleared from the blood and tumor/non-tumor contrast remained relatively stable for 1 h, which could be consistent with static PET images. The results showed that excluding simple blood-pool effects of the tracer and drug delivery would ensure be achieved by active targeting (Fig. 5b).
The 68 Ga-DOTA-RGD-ATWLPPR biodistribution results are shown in Fig. 8a. The tumor uptake of 68 Ga-DOTA-RGD-ATWLPPR was 1.24 ± 0.73 %ID/g, 2.05 ± 0.83 %ID/g, 3.90 ± 0.27 %ID/g, and 2.72 ± 0.37 %ID/g at 5, 15, 30, and 60 min, respectively. 68 Ga-DOTA-RGD-ATWLPPR uptake decreased from 15 to 60 min in most examined organs, and the tumor uptake was higher than the background uptake at all time points. These results indicate that renal clearance is the major metabolic pathway, which is consistent with the imaging results. These results indicate that 68 Ga-DOTA-RGD-ATWLPPR has high speci city for tumor targeting. Comparing binding e ciency after 68 Ga-DOTA-RGD-ATWLPPR, 68 Ga-DOTA-RGD and 68 Ga-DOTA-ATWLPPR injection at 30 min indicated signi cantly higher uptake and T/M of 68 Ga-DOTA-RGD-ATWLPPR in MCF-7 tumor-bearing nude mice.

Discussion
Radionuclide imaging is an attractive, repeatable, and e cient approach to noninvasively detect biological markers in diseased tissue [23]. However, the target receptors must be highly expressed on the cancer cell surface or cancer microenvironment compared with normal tissues. To improve the relatively low binding a nity and imperfect pharmacokinetics of monomeric peptides, many heterodimer peptides have been developed for radionuclide imaging. To improve tumor imaging results, we used a probe that targets two receptors associated with neovascularization in breast cancer. Previous research by our group has demonstrated that the 18 F-labeled heterodimeric peptide RGD-ATWLPPR coupled with NOTA shows good tumor targeting in dual-receptor positive U87MG tumors compared with targeting by a single receptor [20]. However, 18 F-NOTA-RGD-ATWLPPR showed unfavorable hepatobiliary and renal excretion in mice, leading to signi cant liver uptake. Further, the labeling e ciency was between 30-40% with tracer de ciencies. This study is the rst to combine 68 Ga with RGD-ATWLPPR and to use DOTA-coupled dipeptides to synthesize a heterodimeric probe for breast cancer imaging.
We found that 68 Ga-DOTA-RGD-ATWLPPR exhibited good stability, pharmacokinetics and a high T/M ratio in MCF-7 mouse xenograft models, which makes it very suitable for use as breast cancer tracers.
Furthermore, both integrin α v β 3 and NRP-1 receptors are related to neovascularization. 68 Ga-DOTA-RGD-ATWLPPR can improve the sensitivity of invasive tumor imaging and predict prognosis in breast cancer patients. Our results showed that 68 Ga-DOTA-RGD-ATWLPPR is superior to the corresponding monomeric probes due to the additional binding sites. As we expected, 68 Ga-DOTA-RGD-ATWLPPR binds to NRP-1 using the ATWLPPR moiety. The RGD moiety binds with nearby integrin α v β 3 . However, the length of the linker between the two peptide ligands that will improve the exibility of the fusion peptide and achieve optimal binding remains unclear.
In terms of the 68 Ga labeling strategy, the radiolabeling process is simple, and the obtained 68 Ga-DOTA-RGD-ATWLPPR tracer is very stable in PBS. Compared with the uoride labeling method, the 68 Ga labeling reaction time is 10 min, which is shortened by about 25 min [25]. The radiochemical yield increased from 40-60% to 86.1 ± 4.8%, while maintaining very high speci c activity. 68  The dynamic quantitative biodistribution data are consistent with the PET/CT images. The 68 Ga-DOTA-RGD-ATWLPPR dimer had a higher tumor uptake rate and longer tumor retention time than those of the 68 Ga-DOTA-RGD and 68 Ga-DOTA-ATWLPPR monomers. The high signal observed in the kidney had little effect on breast cancer detection because lesions were mainly concentrated in the upper abdomen.
Taken together, our ndings demonstrate that our dual-targeting peptide binding tracer shows improved imaging potential for breast tumors expressing integrin α v β 3 or NRP-1. These two receptors are highly expressed in cancer neovascularization, such as that observed in non-small cell lung cancer, pancreatic cancer, thyroid cancer, colorectal cancer, and ovarian cancer [25,26].Thus, our dual-target probe can be used for imaging of other cancer types. Further, our probe could be used to treat breast cancer. By increasing the target points on tumor cells, the probe has a stronger binding ability on tumors, which could prevent the growth of malignant tumors. However, for treatment, it is necessary to further optimize the probe residence time in the tumor. Our ultimate goal is to develop dual-receptor targeted drugs for radiotherapy.
The subcutaneous tumor cell xenograft model used in our study may exhibit different biological behavior than human breast cancer tumors. Therefore, future studies using patient-derived tumor xenografts should be performed. Third, even though single-dose administration and very early observation were performed, the integrity of evaluation may have been affected.

Conclusion
In this study, we showed that 68 Ga-DOTA-RGD-ATWLPPR e ciently targets tumors with a high tumor/non-tumor ratio and demonstrated the potential of 68 Ga-DOTA-RGD-ATWLPPR as an imaging probe for breast cancer detection. 68 Ga-DOTA-RGD-ATWLPPR can be synthesized in 10 min with a high yield. The results from microPET/CT and biodistribution studies showed that the tracer speci cally targets integrin α v β 3 and NRP-1. Thus, our probe will successfully localize to cancerous lesions and guide precise treatment. We expect that our high-e ciency radiotracer will improve high-quality noninvasive imaging and allow targeted cancer therapies in the future. Abbreviations PET/CT; positron emission tomography/computed tomography; RGD:Arg-Arg-Gly-Asp; ATWLPPR:Thr-Trp-Leu-Pro-Pro-Arg; ANOVA:analysis of variance; DOTA:dodecane tetraacetic acid; MBq:megabecquerel, 106 Bq, equivalent to 1 Rutherford; ID:inject dose Availability of data and materials: The data that support the nding of this study are available from the corresponding author upon reasonable request.
Competing interest: The authors declare that they have no competing interests.  Figure 1 DOTA-RGD-ATWLPPR peptide identi cation a High-performance liquid chromatographic analysis of DOTA-RGD-ATWLPPR heterodimer peptides. HPLC analysis showed that the product eluted with a retention time of 9.13 minutes, and the peak area was dose-dependent. b Mass spectrogram of heterodimer DOTA-RGD-ATWLPPR peptides, DOTA-RGD-ATWLPPR with calculated molecular weight of 1998.24 obtained by mass spectrometry. c Chemical structure of DOTA-RGD-ATWLPPR.  Integrin αvβ3 and NRP-1 expression in breast cancer tumor cells and tissues a Immuno uorescent staining for integrin αvβ3 and NRP-1 in MCF-7 cells. Cellular immuno uorescence revealed that the cytoplasmic location of NRP-1 was strong in MCF-7 cells, while a much weaker uorescence signal for αvβ3 was found in MCF-7 cells. b To con rm the weak expression of αvβ3 in MCF-7 cells, we performed western blotting (WB) and found that integrin αvβ3 was weakly expressed in MCF-7 cells. c

Figures
Representative image of integrin αvβ3 and NRP-1 staining in MCF-7 xenograft tissue. Magni cation is ×200 and ×400. Positive αvβ3 and NRP-1 are displayed with DAB stain. d Expression of the αvβ3 and NRP-1 were determined by immunohistochemical semi-quantitative analysis; αvβ3 and NRP-1 showed different expression levels in the same MCF-7 xenograft tissue, proving that MCF-7 xenograft tissue was heterogeneous.

Figure 5
Representative whole-body coronal and transverse 68Ga-DOTA-RGD-ATWPPR PET/CT images of MCF-7 xenograft mice a Images were acquired at 30 min after injection of normal saline, only 68Ga iron or 68Ga-DOTA-RGD-ATWLPPR (3.7-7.4 MBq; 100-200 μCi; n=3/group), respectively. The tumor is indicated by the yellow arrow. In the normal saline group, the nude mice showed no signal. After injection of only 68Ga iron as a control group, the tumor showed no obvious high absorption, but the background noise signal of the whole body of the nude mice was very strong. MicroPET/CT images were acquired at 30 and 60 min after injection of 68Ga-DOTA-RGD-ATWLPPR, the average tumor-to-muscle (T/M) ratios at 30 and 60 min post-injection were 5.21±0.82 and 4.06±0.12, respectively. Highly speci c uptake of 68Ga-DOTA-RGD-ATWLPPR in breast tumor tissue. b Semi-quantitative 68Ga-DOTA-RGD-ATWLPPR tissue uptake in tumor, muscle, kidney, and heart from dynamic PET scans and 60 min MIP image. As time passed, 68Ga-DOTA-RGD-ATWLPPR was excreted from the blood. Figure 6