Magnetic Resonance Imaging of Transplanted Porcine Neonatal Pancreatic Cell Clusters Labeled with Exendin-4-Conjugated Manganese Magnetism-Engineered Iron Oxide Nanoparticles

Recently, we have shown that manganese magnetism-engineered iron oxide nanoparticles (MnMEIO NPs) conjugated with exendin-4 (Ex4) act as a contrast agent that directly trace implanted mouse islet β-cells by magnetic resonance imaging (MRI). Here we further advanced this technology to track implanted porcine neonatal pancreatic cell clusters (NPCCs) containing ducts, endocrine, and exocrine cells. NPCCs from one-day-old neonatal pigs were isolated, cultured for three days, and then incubated overnight with MnMEIO-Ex4 NPs. Binding of NPCCs and MnMEIO-Ex4 NPs was confirmed with Prussian blue staining in vitro prior to the transplantation of 2000 MnMEIO-Ex4 NP-labeled NPCCs beneath the left renal capsule of six nondiabetic nude mice. The 7.0 T MRI on recipients revealed persistent hypointense areas at implantation sites for up to 54 days. The MR signal intensity of the graft on left kidney reduced 62–88% compared to the mirror areas on the contralateral kidney. Histological studies showed colocalization of insulin/iron and SOX9/iron staining in NPCC grafts, indicating that MnMEIO-Ex4 NPs were taken up by mature β-cells and pancreatic progenitors. We conclude that MnMEIO-Ex4 NPs are excellent contrast agents for detecting and long-term monitoring implanted NPCCs by MRI.

Manganese ion shortens the T1 and T2 relaxation time of neighboring water protons [43] and is a potential contrast agent for MRI. Magnetism-engineered iron oxide (MEIO) NPs is a novel class of iron oxide NPs which possess high and tunable nanomagnetism [44]. The addition of manganese, MnMEIO NPs, further enhances MR signal. We then fabricated nanoparticles that consist of a copolymer shell of silane, MnMEIO core and amine-functionalized poly(ethylene glycol) (PEG) [45,46]. The flexible PEG arms reduce non-specific binding of MnMEIO-silane-NH 2 -mPEG NPs to cells by shielding positive charges of non-conjugated reactive amine groups. Furthermore, we demonstrated that specific and effective targeting of mouse epidermal growth factor receptor (EGFR)-expressing tumors could be achieved by conjugating reactive amine groups on MnMEIO-silane-NH 2 -mPEG NPs with EGFR antibody [46]. The glucagon-like peptide-1 (GLP-1) receptor is a specific surface marker of pancreatic islet β-cells and is not found in murine and human islet α-, δand PP-cells [47]. Studies have shown that exendin-4 (Ex4), a GLP-1 analog, can be used as β-cell-specific probes for in vivo MR imaging of implanted insulinoma [48] and native pancreatic islets in mice [49,50]. Following this strategy, we conjugated Mn-MEIO NPs with Ex4 (MnMEIO-Ex4 NPs) as a β-cell-specific MRI probe and confirmed that MnMEIO-Ex4 NPs-labeled mouse β-cells could be detected and traced by MRI after transplantation [51]. As we know, NPCCs are different from mature adult cells since they replicate and differentiate post transplant [8][9][10][11][12]15]. Therefore, in the present study, we further investigated whether or not MnMEIO-Ex4 NPs could be used in imaging NPCC grafts by MRI.

Animals
Male and female one-day-old pigs were obtained from a local slaughterhouse. Eight to twelve-week-old male athymic nude Balb/c mice from the National Laboratory Animal Center (Taipei, Taiwan) were used as recipients of NPCCs. All animal experiments were approved by the Institutional Animal Care and Use Committee of Chang Gung Memorial Hospital.

Preparation, Culture and Labeling of NPCCs
Neonatal pig pancreases were cut into fragments of ~1 to 2 mm 3 and then digested by collagenase type V in a shaking water bath at 37 °C. The digest was cultured in RPMI-1640 medium at 37 °C (5% CO2, 95% air) in humidified air [12,15,39] for three days. NPCCs were then incubated overnight with MnMEIO-Ex4 NPs before in vitro studies and transplantation.

Binding of MnMEIO-Ex4 NPs by NPCCs
NPCCs were incubated overnight with MnMEIO-Ex4 NPs and then the binding of MnMEIO-Ex4 NPs was examined by Prussian blue staining. After fixation in 4 vol% formaldehyde solution for 30 min, NPCCs were stained for the presence of iron with Prussian blue, freshly prepared potassium ferrocyanate solution (mixture of equal volume of 4 wt% potassium ferrocyanate with 4 vol% hydrochloric acid) for 30 min. Cells with blue particles were considered bound [51].

Transplantation of MnMEIO-Ex4 NPs-Labeled NPCCs
Two thousand NPCCs labeled with MnMEIO-Ex4 NPs were implanted beneath the left renal capsule of each of six nondiabetic nude mouse. NPCCs were carefully transferred into a PE-50 tubing connected to a 200-μL pipette tip prior to centrifugation.

Animals
Male and female one-day-old pigs were obtained from a local slaughterhouse. Eight to twelve-week-old male athymic nude Balb/c mice from the National Laboratory Animal Center (Taipei, Taiwan) were used as recipients of NPCCs. All animal experiments were approved by the Institutional Animal Care and Use Committee of Chang Gung Memorial Hospital.

Preparation, Culture and Labeling of NPCCs
Neonatal pig pancreases were cut into fragments of~1 to 2 mm 3 and then digested by collagenase type V in a shaking water bath at 37 • C. The digest was cultured in RPMI-1640 medium at 37 • C (5% CO 2 , 95% air) in humidified air [12,15,39] for three days. NPCCs were then incubated overnight with MnMEIO-Ex4 NPs before in vitro studies and transplantation.

Binding of MnMEIO-Ex4 NPs by NPCCs
NPCCs were incubated overnight with MnMEIO-Ex4 NPs and then the binding of MnMEIO-Ex4 NPs was examined by Prussian blue staining. After fixation in 4 vol% formaldehyde solution for 30 min, NPCCs were stained for the presence of iron with Prussian blue, freshly prepared potassium ferrocyanate solution (mixture of equal volume of 4 wt% potassium ferrocyanate with 4 vol% hydrochloric acid) for 30 min. Cells with blue particles were considered bound [51].

Transplantation of MnMEIO-Ex4 NPs-Labeled NPCCs
Two thousand NPCCs labeled with MnMEIO-Ex4 NPs were implanted beneath the left renal capsule of each of six nondiabetic nude mouse. NPCCs were carefully transferred into a PE-50 tubing connected to a 200-µL pipette tip prior to centrifugation. Capsulotomy at the lower pole of the left kidney was performed. The tip of the tubing was then inserted and advanced under the capsule towards the injection site [12,15,39].

Histological Study of MnMEIO-Ex4 NPs-Labeled NPCC Grafts
NPCC grafts were retrieved from six recipients, two at day fifteen, fifty-one, and fiftyfive after implantation, respectively. The graft was fixed in a formalin solution, embedded in paraffin and sectioned. Sections were then stained for β-cells with a guinea pig antiswine insulin antibody, pancreatic progenitors with a rabbit polyclonal anti-SOX9 antibody, and for iron with Prussian blue [15,[38][39][40][41][42]51].

Statistical Analysis
The MR signal intensity was computed as mean and standard deviation. If both samples passed the normality test, the independent t-test was performed. The Mann-Whitney U test (Wilcoxon test) was carried out if any one sample failed with the normality test. The p-value less than 0.05 was considered statistically significant.

Binding of MnMEIO-Ex4 NPs to NPCCs
We have developed and characterized MnMEIO-Ex4 NPs, a novel MR contrast agent, with a z-average diameter of 70.2 ± 2.3 nm, a zeta potential of 0.6 ± 0.1 mV, a polydispersity index (PDI) of 0.36 ± 0.01 and an iron concentration of 0.43 mg/mL [51]. To examine cellular binding of MnMEIO-Ex4 NPs, NPCCs were first incubated overnight with MnMEIO-Ex4 NPs and then stained with Prussian blue. We found there was no blue staining on the surface of NPCCs without MnMEIO-Ex4 NPs loading (Figure 2A) while the blue spots were located on all MnMEIO-Ex4 NPs-loaded NPCCs ( Figure 2B), indicating the binding of MnMEIO-Ex4 NPs to NPCCs. = 0.5 mm. MR signal intensity of the graft at the left kidney and the mirror area at contralateral kidney, a within-subject control, was calculated [37][38][39][40][41][42]51].

Histological Study of MnMEIO-Ex4 NPs-Labeled NPCC Grafts
NPCC grafts were retrieved from six recipients, two at day fifteen, fifty-one, a fifty-five after implantation, respectively. The graft was fixed in a formalin soluti embedded in paraffin and sectioned. Sections were then stained for β-cells with a gui pig anti-swine insulin antibody, pancreatic progenitors with a rabbit polyclonal ti-SOX9 antibody, and for iron with Prussian blue [15,[38][39][40][41][42]51].

Statistical Analysis
The MR signal intensity was computed as mean and standard deviation. All sta tics were analyzed by PASW Statistics 21 (IBM Corp. Released 2012. IBM SPSS Statis for Windows. Armonk, NY: IBM Corp.). For paired comparisons of mean values of graft at the left kidney and the mirror area at the contralateral kidney, we first check the normality of the distribution of the variable by using the Kolmogorov-Smirnov tes both samples passed the normality test, the independent t-test was performed. T Mann-Whitney U test (Wilcoxon test) was carried out if any one sample failed with normality test. The p-value less than 0.05 was considered statistically significant.

Binding of MnMEIO-Ex4 NPs to NPCCs
We have developed and characterized MnMEIO-Ex4 NPs, a novel MR contr agent, with a z-average diameter of 70.2 ± 2.3 nm, a zeta potential of 0.6 ± 0.1 mV, a p ydispersity index (PDI) of 0.36 ± 0.01 and an iron concentration of 0.43 mg/mL [51]. examine cellular binding of MnMEIO-Ex4 NPs, NPCCs were first incubated overni with MnMEIO-Ex4 NPs and then stained with Prussian blue. We found there was blue staining on the surface of NPCCs without MnMEIO-Ex4 NPs loading ( Figure 2 while the blue spots were located on all MnMEIO-Ex4 NPs-loaded NPCCs ( Figure 2 indicating the binding of MnMEIO-Ex4 NPs to NPCCs.

In Vivo MR Images of MnMEIO-Ex4 NPs-Labeled NPCC Grafts
For in vivo MRI, 2000 MnMEIO-Ex4 NPs-labeled NPCCs were transplanted under the left kidney capsule of six nude mice. After transplantation, these mice were scanned by a 7.0 T MRI machine at various time points for 8, 50 ( Figure 3) and 54 (Figure 4) days. The MR images of the MnMEIO-Ex4 NPs-labeled NPCC graft revealed persistent hypointense areas located at the site of implantation (indicated by arrows in Figures 3A,B and 4A,B). The quantitative analysis showed a significant (62-88%) reduction of the MR signal intensity in the graft on left kidney when compared to the mirror area on the contralateral kidney at all time points (p < 0.001) (Figures 3C and 4C). This indicates that MnMEIO-Ex4 NPs can be applied in tracing NPCCs grafts for a long period of time.

Histological Studies of MnMEIO-Ex4 NPs-Labeled NPCC Grafts
MnMEIO-Ex4 NPs-labeled NPCC grafts were removed from six recipients, two at day 15, 51, and 55 post transplantation, respectively. To examine the graft histology, anti-insulin and anti-SOX9 antibodies as well as Prussian stain were used to stain pancreatic β-cells, pancreatic progenitors and iron, respectively. As shown in Figure 5, insulin (upper panel) and iron (lower panel) staining were positive and colocalized in 15-, 51-, and 55-day grafts. We found abundant SOX9-positive cells in ( Figure 6A) and outside ( Figure 6B) pancreatic ducts of the 51-day graft. However, colocalization of SOX9 and Prussian blue staining was only observed in those outside pancreatic ducts ( Figure 6B).

Histological Studies of MnMEIO-Ex4 NPs-Labeled NPCC Grafts
MnMEIO-Ex4 NPs-labeled NPCC grafts were removed from six recipients, two day 15, 51, and 55 post transplantation, respectively. To examine the graft histology, ti-insulin and anti-SOX9 antibodies as well as Prussian stain were used to stain panc atic β-cells, pancreatic progenitors and iron, respectively. As shown in Figure 5, insu (upper panel) and iron (lower panel) staining were positive and colocalized in 15-, 5 and 55-day grafts. We found abundant SOX9-positive cells in ( Figure 6A) and outs ( Figure 6B) pancreatic ducts of the 51-day graft. However, colocalization of SOX9 a Prussian blue staining was only observed in those outside pancreatic ducts (Figure 6B

Discussion
Previously, we coated SPIO NPs with chitosan [36,37] and successfully image CSPIO NPs-labeled NPCC grafts by MRI [39]. Since SPIO NPs are taken up through en

Discussion
Previously, we coated SPIO NPs with chitosan [36,37] and successfully imaged CSPIO NPs-labeled NPCC grafts by MRI [39]. Since SPIO NPs are taken up through endocytosis by cells [19,25,41], those MR images are not necessarily representative of β-cells. To specifically image transplanted β-cells, we conjugated an MR contrast agent MnMEIO NPs with GLP-1 analog Ex4 which can bind GLP-1 receptors on the surface of β-cells. Our results showed that MnMEIO NPs were taken up by β-cells through receptor-mediated endocytosis and MnMEIO-Ex4 NPs were safe and effective for the detection and long-term tracing of transplanted mouse islet β-cells by MRI [51]. In this study, we further demonstrated that MnMEIO-Ex4 NPs could bind NPCCs and MnMEIO-Ex4 NPs-labeled NPCC grafts could be visualized and monitored by MRI for a long period of time.
For in vivo MR imaging, we transplanted 2000 MnMEIO-Ex4 NPs-labeled NPCCs beneath the left renal capsule in each nude mouse. During the 50-and 54-day follow-up, there was a reduction in the MR signal intensity of the graft on the left kidney by 62-88% compared to the mirror areas on the contralateral kidney. These findings are consistent with our previous observation with CSPIO NPs-labeled NPCC grafts which showed that 60-80% reduction of the MR signal intensity [39]. This indicates that MnMEIO-Ex4 NPs are as effective as CSPIO NPs in detecting NPCC grafts.
Through GLP-1 receptors, MnMEIO-Ex4 NPs can be taken up by β-cells (i.e., receptormediated endocytosis) [51]. Therefore, the labeled cells in NPCC grafts showed hypointense areas on in vivo MR images. This notion is confirmed by our histological studies of colocalization of insulin/iron and SOX9/iron staining in NPCC grafts. In fact, NPCCs are clusters of pancreatic cells containing ducts, endocrine and exocrine cells [8,9,15]. In mouse, rat, and human pancreases, GLP-1 receptors express not only on islet β-cells but also in duct tissues [47,52,53] where progenitor cells are located. In NPCC graft, we found abundant cells stained positive for insulin and SOX9, indicating the existence of mature β-cells and pancreatic progenitors. We only observed the colocalization of the SOX9 and Prussian blue staining in cells outside but not in pancreatic ducts. Presumably, SOX9-positive cells in pancreatic ducts are newly formed progenitors [54] which were not present during the labeling of MnMEIO-Ex4 NPs before transplantation. That's why we did not find colocalization of SOX9 and Prussian blue staining in those cells. Taken together, MnMEIO-Ex4 NPs were taken up by NPCCs with GLP-1 receptors (i.e., mature β-cells and pancreatic progenitors), and these cells showed positive MR images. In this regard, MnMEIO-Ex4 NPs are superior to CSPIO NPs. To the best of our knowledge, we are the first to apply GLP-1 receptor probes in imaging NPCC grafts.

Conclusions
In addition to MnMEIO-Ex4 NPs-labeled mouse islet isografts [49], in this study, we further extended the application of MnMEIO-Ex4 NPs in tracing NPCCs by MRI. Our results showed that MnMEIO-Ex4 NPs bound NPCCs in vitro and NPCCs grafts revealed persistent positive MR images for up to 54 days after transplantation. Histological studies also confirmed colocalization of the insulin/iron and SOX9/iron staining in NPCC grafts. We conclude that MnMEIO-Ex4 NPs are excellent contrast agents for detecting and longterm monitoring transplanted NPCCs.