ENO1‐targeted superparamagnetic iron oxide nanoparticles for detecting pancreatic cancer by magnetic resonance imaging

Abstract The aim of this study was to investigate in vitro magnetic resonance imaging (MRI) of PDAC using ENO1‐targeted superparamagnetic iron oxide nanoparticles and xenograft models. Expression level and location of ENO1 protein in pancreatic cancer cell lines of CFPAC‐1 and MiaPaCa‐2 were detected by Western blotting, flow cytometry and confocal microscopy. Dex‐g‐PCL/SPIO nanoparticles targeting ENO1 were constructed with ENO1 antibody and characterized by MRI. In addition, ENO1‐Dex‐g‐PCL/SPIO nanoparticles were tested to assess their efficacy on the detection of PDAC using in vitro and in vivo MRI. The results showed that ENO1 was expressed in both human PDAC cell lines of CFPAC‐1 and MiaPaCa‐2, demonstrating that the localization of cytoplasm and membrane was dominant. It was confirmed that ENO1 antibody was connected to the SPIO surface in ENO1‐Dex‐g‐PCL/SPIO nanoparticles. The nanoparticles had satisfactory superparamagnetism and significantly enhance the detection of PDAC by in vivo and in vitro MRI. In conclusion, ENO1 can serve as a membrane protein expressed on human PDAC cell lines. ENO1‐targeted SPIO nanoparticles using ENO1 antibody can increase the efficiency of detection of PDAC by in vitro and in vivo MRI.

Molecular imaging, defined as the characterization and measurement of biological processes at the cellular and molecular levels, has significantly attracted scholars' attention in the detection of malignant diseases. Combining magnetic resonance imaging with advanced contrast agents, it is possible to perform molecular analysis of the targeted histiocytosis and make early detection of cancer possible. Superparamagnetic iron oxide (SPIO) has been widely applied in MRI as a contrast agent, and conjugation of SPIO with specific ligands, such as antibodies, peptides and nucleotides, can result in the appearance of a variety of targeted SPIO nanoparticles. [3][4][5][6][7] Targeted SPIO nanoparticles are ideal carriers for the fabrication of imaging probes with their own characteristics and the realization of PDAC molecular imaging.
Enolase 1 (ENO1), also named as pyruvate dehydrogenase 1, is a glycolytic enzyme, acting as a multifunctional moonlight protein. 8,9 It is up-regulated at the mRNA/protein level in PDAC cell lines and tissues and is involved in development, invasion, metastasis and chemoresistance of PDAC. 10,11 ENO1 can be a cytoplasmic protein with enzymatic activity and be translated into C-Myc promoter-binding protein (MBP-1) in the nucleus 12 ; furthermore, ENO1 can serve as a membrane protein expressed on the cell membrane, acting as a plasminogen receptor. 13,14 Additionally, ENO1 located on cell membrane of PDAC might be beneficial for magnetic resonance molecular imaging with SPIO nanoparticles.
In the present study, we constructed ENO1-targeted Dex-g-PCL/SPIO nanoparticles with ENO1 antibody. The ability of ENO1targeted SPIO nanoparticles in detecting pancreatic cancer by MRI was assessed in both in vitro and in vivo experiments. The successful construction of ENO1-targeted SPIO nanoparticles would facilitate early and accurate detection PDACs, paving the way for pancreatic cancer treatment. This original method of constructing nanoparticles provided new approaches for targeting cancer cells, which might also be used in the early diagnosis of other diseases.

| Cell lines
Human PDAC cell lines of CFPAC-1 and MiaPaCa-2 were purchased from American Type Culture Collection (Manassas, VA, ATCC).

| Western blot analysis
To determine the expression level of ENO1 protein in CFPAC-1 and MiaPaCa-2 cells, the cells (1 × 10 7 ) were harvested for Western blot analysis. In addition, 20 μg of total protein was separated by sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE), followed by transferring into nitrocellulose filter (NC) membranes (EMD Millipore) using a semi-dry electrophoretic transfer cell system (Bio-Rad Laboratories, Inc). The membranes were blocked with 5% non-fat milk for 1 hour and then incubated with anti-ENO1 (Abcam) or an antibody specific to glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (Sigma-Aldrich Corp.) overnight at 4°C.
The membranes were then incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies (Santa Cruz Biotechnology, Inc) for 1 hour at room temperature. Blots were detected with enhanced chemiluminescence (ECL) reagents (EMD Millipore) and exposed to a chemiluminescent imaging system for 5 minutes.

| Flow cytometry analysis
Here, PDAC cells of CFPAC-1 and MiaPaCa-2 were spun down and re-suspended in 500-µL phosphate-buffered saline (PBS), in which cells were incubated with a primary anti-ENO1 (Abcam) or an isotype-matched negative control antibody for 30 minutes at 4°C, and the analysis was conducted using the FACScan flow cytometer (BD Biosciences).

| Immunofluorescence analysis
To determine the cellular location of ENO1 protein in CFPAC-1 and

| Construction and characterization of ENO1targeted Dex-g-PCL/SPIO nanoparticles
The first is to prepare SPIO nanoparticles, in which 1 mmol Fe(acac)3 with 5 mmol 1,2-hexadecanol, 3 mmol oleic acid and 3 mmol oleylamine were mixed in 10 mL benzyl ether, heated to 300°C for 1 hour under argon protection. The solution was precipitated with methanol at room temperature, and the obtained black solid product was kept in n-hexane. Next, poly(epsilon-caprolactone)-grafted dextran (Dex-g-PCL) was prepared as previously reported. 15 Dex-g-PCL/ SPIO nanoparticles in the aqueous phase were prepared through the emulsion evaporation method. Next, 50 µL of Enolase-1 monoclonal antibody (0.5 mg/mL, Abcam) was mixed to the Dex-g-PCL/SPIO nanoparticles solution at room temperature for 15 minutes, and 2 mg of carbodiimide was added at room temperature for 30 minutes. The sample was transferred to a dialysis bag and dialysed for 24 hours in PBS solution at 4°C. Afterwards, suspension of ENO1-Dex-g-PCL/ SPIO nanoparticles was taken out from the dialysis bag and stored at 4°C until use. The size of the nanoparticles was examined using a transmission electron microscope (Joel, JEM-1011). The diameters of nanoparticles were measured by Malvern Zetasizer (Malvern Panalytical Ltd.). The nanoparticles were also detected with X-ray diffractometer (Jordan Valley Semiconductors, Ltd.) and Fourier infrared spectroscopy (Fairborn, China). The binding specificity of ENO1-Dex-g-PCL/SPIO nanoparticles to their targets in cells was examined through Prussian blue staining after cultivating for 6 hours.

| Magnetic property and T2 relativity analysis
Magnetic properties of ENO1-Dex-g-PCL/SPIO nanoparticles were detected by superconducting quantum interference magnetometer (Quantum Design Inc) at room temperature after lyophilization.
Next, the efficiency of T2 relaxation for nanoparticles was examined by 1.5T MRI (Siemens Magnetom Trio Tim, Siemens AG) as previously reported. 16 We prepared SPIO and ENO1-SPIO granule solutions with iron concentrations of 30, 15, 7.5, 3.75, 1.875 and 0.9375 ug/mL and then added 1 mL 2% agarose to prepare a final concentration of 1% agarose. Next, 1% agarose was used as a blank control to detect the T2 value and T2* value of different concentrations of iron particle solution. The corresponding relaxation rates R2 and R2* were calculated as 1/T2 and 1/T2*, respectively.

| In vitro MRI
Here, CFPAC-1 and MiaPaCa-2 cells were plated at a density of 3 × 10 5 cells/well in a 24-well plate and then treated with 10-ul SPIO (Fe with concentration of 0.1 mg Fe/mL), ENO1-SPIO, and PBS control for 2 hours, and scanned for MRI as well.

| Pancreatic cancer xenograft model and in vitro MRI
In this study, 12 male BALB/c nude mice (age, 7-week-old; weight, 20-25 g) were kept in specific pathogen-free (SPF) facilities. After that, 2 × 10 7 CFPAC-1 cells were subcutaneously injected for about 3 weeks as previously reported. 17 When the tumours grew to about 10 mm diameter in size, the tumour would be appropriate Tumour-bearing nude mice scanning was performed on a 1.5T MR scanner (Siemens Magnetom Trio Tim; Semens AG) using an animal coil. The mice were anaesthetized using Ketamine (4.0 mg/100 g).
Mice were scanned in advance, and 2, 4, 12 and 24 hours after injection with the SPIO or ENO1-SPIO (body weight, 0.1 mL/20 g) through the tail vein.
The imaging sequences are as follows: TR of 4200 ms and TE of 36 ms were used for T2-weighted fast-spin-echo imaging, a field of view (FOV) of 40 mm × 40 mm and slice thickness of 1 mm. Region of interest (ROI) was utilized for signal measurement, and 20-mm 2 field at the largest level of tumour imaging was selected as ROI.

| Histologic analysis
Mice were killed in a CO 2 chamber after MRI. Tissues were collected and kept in formaldehyde solution for 24 hours at 4°C, and then embedded in paraffin. Tissue sections were stained with haematoxylin and eosin (H&E) for histology or Prussian blue to detect blue iron nanoparticles. Immunohistochemistry (IHC) of the tissues was carried out according to the previously reported protocols. 11

| Statistical analysis
Data were expressed as mean ± standard deviation (SD). Differences between two groups were analysed by the Student's t test. To analyse differences among three or more groups, one-way analysis of variance (ANOVA) was utilized. The statistical analysis was undertaken by GraphPad Prism 6 software; P-values less than 0.05 were statistically considered significant.

| ENO1 expression of pancreatic cancer cell lines
To clarify the expression of ENO1 and the location of pancreatic cancer lines, CFPAC-1 and MiaPaCa-2 cells were analysed by Western blotting, flow cytometry and immunofluorescence staining. Results of Western blotting showed that ENO1 was expressed in both MiaPaCa-2 and CFPAC-1 cells ( Figure 1A). Flow cytometry, using ENO1 antibody, further confirmed the cell-surface expression of pancreatic cell lines ( Figure 1B). Immunofluorescence analysis confirmed that ENO1 displayed a predominant cytoplasm and membrane localization ( Figure 1C). These results suggested that ENO1 is a protein target, which might be used in magnetic resonance molecular imaging with SPIO nanoparticles.  (Figure 2A,B).

| Characterization of ENO1-Dex-g-PCL/SPIO nanoparticles
The nanoparticles were detected by Fourier transform infrared spectroscopy. The characteristic of C = O absorption peak appeared in the absorption spectrum, indicating that the amide bond was formed by the carbodiimide method between the carboxyl group located at the surface of the particle and the amino group of ENO1 antibody, which confirmed that ENO1 antibody was connected to the SPIO surface ( Figure 2C).
In addition, X-ray diffractometry revealed that the diffraction line of the complex had seven characteristic peaks at 2θ angles, which were located at 30.05, 35.59, 43.21, 53.62, 57.23, 62.77 and 74.15.
The position and relative intensity of the peak indicated that the sample was mainly consisted of Fe 3 O 4 with the complete crystal structure ( Figure 2D).

CFPAC-1 cells incubated with ENO1-Dex-g-PCL/SPIO nanoparticles,
demonstrating that the absorption of antibody-modified polymer nanoparticles depends on the binding of antibody ligands ( Figure 2E).
The hysteresis curve demonstrated that ENO1-Dex-g-PCL/SPIO nanoparticles had a satisfactory property of superparamagnetism ( Figure 2F). Significant decreases of T2 and T2* relaxation time were detected at different iron concentrations with SPIO and ENO1-SPIO.

| In vivo MRI of ENO1-Dex-g-PCL/SPIO nanoparticles
The results showed that the tissues had equal and higher signal on the T2 sequence during MRI, the signal was not uniform, and necrosis was observed inside the tumour ( Figure 4A). After non-targeted contrast agent SPIO was injected into the control group, the signal intensity gradually increased 1 hour after the early non-specific enhancement, which reached the peak after 2 hours, and was fully recovered to the pre-injection level after 24 hours. After the injection of ENO1-SPIO targeting contrast agent, the T2 signal intensity of tumour tissue significantly decreased, and the tumour gradually darkened over time, and the peak of enhancement was observed at 24 hours ( Figure 4B). IHC staining demonstrated ENO1 expression in pancreatic tumour tissues of xenograft model ( Figure 4C). Prussian blue staining revealed that more positive iron particles were found in ENO1-SPIO group compared with SPIO group ( Figure 4D).

| D ISCUSS I ON
Poor prognosis of PDAC indicates the necessity of discovering new approaches for early detection, and molecular imaging is an can enhance glycolytic process in tumour cells because of the Warburg effect. 25,26 Studies revealed that ENO1 belongs to a group of moonlight protein, which plays a key role in many biological and pathophysiological processes. 27 In the majority of cases, ENO1 is a cytoplasmic protein with enzymatic activity, and ENO1 can also serve as a membrane protein expressed on cell membrane, acting as a plasminogen receptor promoting cell invasion; besides, ENO1 can be translated into MBP-1 in the nucleus, leading to tumour suppression. 12 Previous studies showed that ENO1 located on the surface in tumour cells is subjected to post-translational modifications, including acetylation, methylation and  Results were achieved from representative experiments in triplicate and were shown as mean ± standard deviation (SD). *P < .05, **P < .01 F I G U R E 4 Detection of pancreatic tumour by in vivo MRI of ENO1-Dex-g-PCL/SPIO nanoparticles. A, MRI of ENO1-Dex-g-PCL/SPIO nanoparticles in a pancreatic cancer xenograft model. B, Compared with PBS control group and SPIO group, the T2 signal intensity of tumour tissue significantly decreased, and the tumour gradually darkened over time, in which the peak of enhancement was at 24 h in ENO1-SPIO group. C, IHC staining (40×) of the pancreatic tumour tissues 24 h after injection with ENO1-SPIO nanoparticles. D, Prussian blue staining (40×) of the pancreatic tumour tissues 24 h after injection with ENO1-SPIO or SPIO. More positive iron particles were found in ENO1-SPIO group. Results were achieved from representative experiments in triplicate and were shown as mean ± standard deviation (SD). *P < .05, **P < .01

| CON CLUS IONS
ENO1 can serve as a membrane protein expressed on human PDAC cell lines. ENO1-targeted SPIO nanoparticles using ENO1 antibody can increase the efficiency of detection of PDAC by in vitro and in vivo MRI.

CO N FLI C T O F I NTE R E S T
All authors declare that they have no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data were available on request from the corresponding author.