Magnetic resonance imaging (MRI) and clinically approved superparamagnetic iron oxide nanoparticles (SPIOs) are currently used to visualize transplanted pancreatic islets in animals and in clinical studies in humans [1-3]. Although SPIOs provide a satisfactory means of observing transplanted cells, adverse effects of incorporated SPIOs on the biological properties of cells, cellular function and signaling pathways have been suggested [4-10]. To apply nanoparticles, including SPIOs, in the biomedical field, any effects on the biological characteristics of cells must be completely ruled out . Pancreatic islets, as well as insulinoma cell lines (β-TC3, β-TCtet and β-TC-6) that possess the ability to synthesize and secrete insulin, have been labeled with clinically approved and commercially available SPIOs without decreasing cell viability or hindering insulin secretion [1-3, 12-17]. The effect of SPIOs on the functionality and physiology of pancreatic islets, especially insulin-secreting β-cells, should be further determined.
Insulin and glucose activate various signaling pathways for the regulation of proliferation, survival and insulin biosynthesis and secretion in pancreatic β-cells through multiple complex pathways . Glucose metabolism affects most major signaling pathways, including those of insulin production, storage and secretion [19-21]. Insulin signaling is involved in diverse cellular processes, such as antiapoptosis, cell growth and insulin synthesis [20-22]. Mitogen-activated protein kinases (MAPK) in pancreatic β-cells, particularly extracellular signal-regulated kinase 1 and 2 (ERK1/2), are components of the mechanism by which glucose stimulates insulin gene expression, which is mediated by the activation of transcription factors (Pdx-1, NeuroD1 and E47) [23, 24]. The secreted insulin binds to an insulin receptor and triggers activation of the serine-threonine kinase AKT, also known as protein kinase B, and the transcription factors (Hnf4A, Pdx1, MafA, Isl1 and NeuroD1) that control the expression of insulin and many other genes that are important for regulating insulin secretion .
The development of glucose intolerance in type 2 diabetes and gestational diabetes is associated with body iron stores. Increased iron stores have been found to predict the development of type 2 diabetes and induce chronic diabetes complications, while iron depletion is protective . Reducing body iron improves insulin secretion and action in type 2 diabetic patients . In light of the various linked iron and glucose disorders in the pancreas, it is necessary to explore whether the iron load created by SPIO labeling alters the biological functions of pancreatic β-cells. Recently, we reported a significant increase in the insulin and NeroD1 mRNA levels in SPIO-labeled pancreatic islets . Our previous results imply that SPIO labeling may influence the signal transduction pathways related to insulin biosynthesis, including the AKT and ERK1/2 signaling pathways, in pancreatic β-cells.
To investigate the effect of SPIOs on signal transduction by insulin and glucose in pancreatic β-cells, we focused on evaluating the activation of ERK1/2 and AKT stimulated by insulin and glucose in SPIO-labeled pancreatic β-cells (INS-1 cells).
2.1 Cell Viability, Iron Measurement, Prussian Blue Staining and TEM Images of SPIO-labeled Cells
No decrease in viability was observed in labeled cells incubated in medium containing SPIO (200 µg Fe ml−1 iron concentration) compared with unlabeled cells (Fig. 1A). The iron content of the unlabeled cells was 0.02 ± 0.001 pg, whereas the total iron of the labeled cells increased in a dose-dependent manner; the cellular iron level of the cells at 200 µg Fe ml−1 was 1.05 ± 0.11 pg (Fig. 1B). Prussian blue staining showed that the accumulation of SPIO in the cells increased dose-dependently (Fig. 1C). Using transmission electron microscopy (TEM), the internalized SPIO was found within vesicles (endosomes or lysosomes); the SPIO may have been internalized within lysosomes owing to their proximity to the nucleus in the cells labeled with 200 µg Fe ml−1 (Fig. 1D).
2.2 In vitro MRI Analysis of SPIO-labeled Cell Phantoms
As iron oxide concentrations were increased up to 200 µg Fe ml−1, the signal intensity on T2- and T2*-weighted images gradually decreased in the in vitro MRI analysis of cell phantoms (Fig. 2A). The T2 values at 0, 5, 25, 50, 100 and 200 µg Fe ml−1 were 115 ± 6.7, 89.0 ± 11.3, 53.1 ± 27.7, 37.3 ± 2.0, 21.0 ± 2.0 and 21.0 ± 2.0 ms, respectively (Fig. 2B). The T2* values at 0, 5, 25, 50, 100 and 200 µg Fe ml−1 were 84.4 ± 11.1, 26.4 ± 2.6, 11.0 ± 2.8, 10.3 ± 1.9, 10.2 ± 2.3 and 10.0 ± 1.9 ms, respectively (Fig. 2B).
2.3 Effect of SPIOs on the ERK1/2 and AKT Phosphorylation and the Expression of Insulin Protein
We next investigated whether the SPIOs activate ERK1/2 and AKT in the absence of any stimuli and found that SPIO labeling (25–200 µg Fe ml−1) for 24 h induced both ERK1/2 and AKT phosphorylation (Fig. 3A). As shown in Fig. 3(B), there was an approximately 2-fold increase in ERK1/2 and AKT phosphorylation in the cells labeled with 200 µg Fe ml−1 (p-ERK1/2 of labeled vs unlabeled, 2.26 ± 1.6 vs 1 ± 0.0, p = 0.015; p-AKT of labeled vs unlabeled, 2.7 ± 0.4 vs 1 ± 0.0, p = 0.03). We further evaluated the effect of SPIO labeling at the highest iron concentration (200 µg Fe ml−1) on signal transduction by glucose and insulin (Figs 4A and B, 5A and BF). As shown in Fig. 4(A and B), glucose (25 mM) enhanced ERK1/2 and AKT phosphorylation, but the SPIO labeling (200 µg Fe ml−1) significantly reduced the glucose-stimulated phosphorylation of ERK1/2 and AKT in the labeled cells compared with the unlabeled cells (p-ERK1/2 of labeled vs unlabeled, 1.46 ± 0.51 vs 4.2 ± 0.63, p = 0.003; p-AKT of labeled vs unlabeled, 2.92 ± 0.19 vs 4.7 ± 0.48, p = 0.01). To evaluate the intracellular insulin protein levels, a Western blot was performed using total protein lysate of unlabeled or labeled cells cultured in 2.5 or 25 mM glucose for 12 h. The intracellular insulin protein levels stimulated by glucose were significantly decreased in the labeled cells in comparison with the unlabeled cells (Fig. 4A and B, labeled vs unlabeled, 0.34 ± 0.1 vs 1.54 ± 0.66, p = 0.003). As shown in Fig. 5A and B, the insulin-stimulated phosphorylation of ERK1/2 increased dose-dependentlyand slightly decreased in the labeled cells at 200 µg Fe ml−1 (labeled vs unlabeled at 1 µM insulin, 4.8 ± 1.0 vs 5.4 ± 0.8, p = 0.85). Furthermore, a significant decrease in the 1 µM insulin-stimulated phosphorylation of AKT was observed in the labeled cells compared with the unlabeled cells (labeled vs unlabeled, 0.6 ± 0.5 vs 1.4 ± 0.5, p = 0.018), and the intracellular insulin protein level was also significantly reduced in the labeled cells treated with 1 µM insulin (labeled vs unlabeled, 0.58 ± 0.2 vs 0.99 ± 0.1, p = 0.005).
SPIOs have been widely applied to non-invasively monitor the delivery, biodistribution and fate of transplanted cells and to predict the cellular therapeutic efficacy using MRI. Clinical and experimental approaches using clinically approved SPIOs, including Feridex and Resovist, have been performed to label diverse cell types, such as stem cells, lymphocytes and pancreatic islets, for cell transplantation therapy [1, 28-30]. The SPIOs incorporated into cells have no serious deleterious effects on cell viability or growth [6, 31, 32], but they show diverse changes in biological properties, for example, alterations in gene expression [6, 10], changes in the differentiation capacity of stem cells [4, 7, 33], the immunomodulation of macrophages [5, 34] and the modulation of signaling molecules such as AKT and ERK1/2 [8, 9]. Even in the absence of significant iron overload, dysfunctions of pancreatic β-cells are closely associated with iron-storage disease . Iron overload is linked to insulin resistance and glucose intolerance in patients with type 2 diabetes, and these conditions are improved by iron depletion or treatment with iron chelators . SPIO labeling was found not to change in vitro viability [2, 30], but it did alter insulin and BETA2 gene expression in the pancreatic islets of rats . Altogether, these observations indicate that the iron released from SPIOs may modulate insulin and glucose action in pancreatic β-cells.
Glucose is the most physiologically relevant stimulus for insulin biosynthesis and secretion; insulin is stored in the pancreatic β-cells until an appropriate stimulus triggers its release . The secreted insulin plays an autocrine role in stimulating insulin biosynthesis . Consequently, a parallel stimulation of insulin biosynthesis is necessary to replenish intracellular insulin stores. Many studies have reported that the insulin levels released from pancreatic islets stimulated by high glucose are similar between unlabeled islets and labeled islets [1, 2, 6, 15, 17]. However, it has not been reported whether SPIO may cause a change in intracellular insulin protein in glucose- or insulin-stimulated β-cells. In this study, we evaluated the phosphorylation of ERK1/2 and AKT as two major signaling molecules to mediate glucose- or insulin-stimulated insulin biosynthesis [18, 36] and intracellular insulin in SPIO-labeled INS-1 cells.
To apply nanoparticles, including SPIOs, in the biomedical field, the effects of nanoparticles on the biological characteristics of cells must be completely ruled out [8, 11, 33, 37, 38]. Accumulating evidence strongly suggests that nanoparticles, including the iron component of SPIOs, have the ability to induce the production of reactive oxygen species (ROS), which function as second messenger molecules in normal physiological processes and can directly activate proteins [8, 39]. At different concentrations of SPIO, INS-1 exhibited varying profiles of ERK1/2 and AKT activation. All of the iron concentrations produced a two-fold increase in ERK1/2 phosphorylation, and AKT was highly phosphorylated only when cells were incubated with 200 µg Fe ml−1. These results imply that ERK1/2 and AKT can be activated through ROS triggered by iron component overload of SPIO or other mechanisms. The signaling mechanism by which SPIO induces ERK1/2 and AKT phosphorylation activation should be defined.
The phosphorylation of ERK1/2 and AKT in response to glucose or insulin in INS-1 cells and pancreatic β-cells is induced via different mechanisms. Insulin induces tyrosine-phosphorylation of the insulin receptor subunit-2 (IRS-2) and increases PI3K, which in turn mediates AKT phosphorylation . The effect of glucose on AKT phosphorylation occurs secondary to autocrine activation of insulin signaling proteins by secreted insulin , but the activation of ERK1/2 by glucose is not required for insulin secretion [41, 42]. This study demonstrated that AKT phosphorylation decreases when labeled INS-1 cells are stimulated with glucose or high insulin concentrations and ERK1/2 phosphorylation decreases in labeled INS-1 cells stimulated with high glucose. These results suggest that SPIOs modulate the insulin-mediated IRS-2/PI3K/AKT pathway, leading to decreased AKT phosphorylation only when INS-1 cells are stimulated with glucose or a high insulin concentration and may regulate ERK1/2 through a different mechanism in glucose- or insulin-stimulated INS-1 cells.
Furthermore, labeled INS-1 cells exhibited a decrease in the intracellular insulin content below basal values when stimulated by glucose or insulin. The possible mechanism to explain these results would be that SPIO labeling suppresses the glucose- or insulin-stimulated insulin biosynthesis by decreasing AKT and ERK1/2 phosphorylation, which causes labeled INS-1 cells not to replenish intracellular insulin stores. The effects of SPIOs may be more complex, and there may be intricate molecular mechanisms by which AKT and ERK1/2 phosphorylation is modulated by SPIO.
According to our MRI study, INS-1 cells incubated with SPIO at iron concentrations higher than 25 µg Fe ml−1 did not show further reductions in T2* values. On the other hand, T2 values of the cells continuously decreased far beyond 25 µg Fe ml−1. Therefore, compared with spin-echo-based sequences, gradient-echo-based sequences are much more sensitive to the susceptibility effect arising from iron oxides. Thus, the gradient-echo-based sequences with higher sensitivity allow the detection of low iron concentrations. The toxic effects are limited because of the low SPIO doses. From the perspective of using SPIO-labeled cells for clinical or preclinical application therapies, it is mandatory to optimize protocols to yield high-efficiency cell labeling while preserving the biological features of the labeled cells before transplantation.
SPIO labeling caused significant decreases in glucose- and insulin-stimulated AKT phosphorylation and in glucose-stimulated ERK1/2 phosphorylation, which resulted in a reduction in the insulin produced upon stimulation with glucose or insulin in INS-1 cells. We suggest that the signal transduction responses to chronic iron overload resulting from SPIO labeling in preparation for transplanted islets imaging should be further studied. Our findings suggest that an evaluation of the effects of SPIO labeling on islet functionality and homeostasis should be required before it is used for labeling transplanted islets in clinical practice.
5.1 SPIO Labeling and Cell Viability
INS-1 cells were maintained in RPMI-1640 medium supplemented with 10% fetal calf serum, 1 mM sodium pyruvate, 50 µM β-mercaptoethanol, 10 m m sodium HEPES buffer, pH 7.3, 100 U ml−1 penicillin and 100 µg ml−1 streptomycin. The clinically approved SPIO, Resovist, was obtained from Schering (Berlin, Germany). For the SPIO labeling, the INS-1 cells were incubated in a culture medium containing Resovist (5-200 µg Fe ml−1) without transfection agents at 37 °C for 24 h. After labeling with the SPIO, the cell viability was evaluated using a standard 3-[4,5-dimethylthialzol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) assay (Sigma, St. Louis, MO, USA).
5.2 Prussian Blue Staining and TEM
For the Prussian blue staining to detect the SPIO, the cells were fixed in 2% paraformaldehyde, incubated for 30 min with 5% potassium ferrocyanide in 5% hydrochloric acid, and counterstained with Nuclear Fast Red for 5 min. Representative labeled cells were examined under a light microscope to determine the intracellular SPIO distribution.
For the TEM observation of the intracellular distribution of SPIO, the cells were fixed with 2.5% glutaraldehyde, treated with 2% osmium tetroxide in 0.1 mM cacodylate buffer for 2 h, dehydrated with graded ethanol from 50 to 100% and propylene oxide and embedded in pure Epon resin at 60 ºC for 3 days. Ultrathin sections were cut with glass knives and a Diatome diamond knife (Reichert-Jung, Vienna, Austria) using an ultramicrotome (RMC MTXL; Tucson, AZ, USA), stained with lead citrate and uranyl acetate and observed with a JEM-100 CX transmission electron microscope (Jeol, Tokyo, Japan).
5.3 Intracellular Iron Concentration and in vitro MRI
For the iron uptake assays, the cells were incubated with the SPIO (5-200 µg Fe ml−1) for 24 h. The collected cells were then washed with phosphate-buffered saline, resuspended in 6 N hydrochloric acid and incubated at 70 ºC for 60 min. The total iron content was determined using a total iron reagent set (Pointe Scientific, Canton, MI, USA), and the average iron amount per cell was calculated.
For the in vitro MRI phantom, labeled and unlabeled cells (1 × 106) were suspended in 1% agarose and transferred to 1.5 ml microcentrifuge tubes. The in vitro MRI was performed with a wrist coil using a 3 T MR scanner (Trio, A Tim System, Piscataway, NJ, USA). For the measurement of T2 values, we used a spin-echo pulse sequence with the following imaging parameters: repetition time (TR)/echo time (TE) = 5000/16, 20, 32, 40, 48, 50, 60, 64 80, 100, 150 or 200 ms; flip angle = 90°/180°; field of view = 60 × 120 mm; matrix = 256 × 256; slice thickness = 2 mm; and the number of excitations = 1. For the measurement of T2* values, we used a gradient-echo pulse sequence with the following imaging parameters: repetition time (TR)/echo time (TE) = 800/4.3, 11.8, 19.3, 26.8 or 34.3 ms; flip angle = 20°; field of view = 60 × 120 mm; matrix = 256 × 96; slice thickness = 2 mm; and the number of excitations = 2. The T2 and T2* values were estimated by fitting the decreased signal intensities with the increasing TEs into a mono-exponential function.
5.4 Western Blotting
The signal transduction by glucose and insulin in labeled and unlabeled cells was assessed using Western blotting. After labeling with SPIO (5–200 µg Fe ml−1) for 24 h, the unlabeled and labeled cells were incubated with medium containing 2.5 m m glucose and 2% fetal bovine serum for 5 h and were subsequently treated with glucose (25 mM) or insulin (0.1 and 1 µM) for 12 h. The cells were lysed in RIPA buffer containing a protease inhibitor cocktail (Sigma), and the proteins were separated using SDS–PAGE and transferred to nitrocellulose membranes. The membranes were blocked with 5% skim milk in Tris-buffered saline and incubated with primary antibodies against phosphorylated-ERK1/2, phosphorylated-AKT (ser437), ERK1/2, AKT (Cell Signaling Technology, Danvers, MA, USA), insulin and β-actin (Sigma) overnight at 4 °C, followed by incubation with horseradish peroxidase-conjugated secondary antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA) at room temperature for 30 min. The blots were developed using Enhanced Chemiluminescence Reagents (Amersham Biosciences, Piscataway, NJ, USA). The relative intensity of the bands observed by Western blotting was analyzed using the Image J program.
5.5 Statistical Analysis
The data are presented as the means ± standard error of more than five independent experiments. Comparisons were performed using the ANOVA test. Differences were considered significant at p values < 0.05.
This work was supported by National Research Foundation of Korea grants funded by the Korea government (MEST; 2009-0073532, 2011-0003657, 2012-0001949, 2011-0000174), by a grant (A062260) from the Innovative Research Institute for Cell Therapy, Republic of Korea and by a grant (03-2008-0120) from the SNUH Research Fund.