The cytotoxicity evaluation of magnetic iron oxide nanoparticles on human aortic endothelial cells

One major obstacle for successful application of nanoparticles in medicine is its potential nanotoxicity on the environment and human health. In this study, we evaluated the cytotoxicity effect of dimercaptosuccinic acid-coated iron oxide (DMSA-Fe2O3) using cultured human aortic endothelial cells (HAECs). Our results showed that DMSA-Fe2O3 in the culture medium could be absorbed into HAECs, and dispersed in the cytoplasm. The cytotoxicity effect of DMSA-Fe2O3 on HAECs was dose-dependent, and the concentrations no more than 0.02 mg/ml had little toxic effect which were revealed by tetrazolium dye assay. Meanwhile, the cell injury biomarker, lactate dehydrogenase, was not significantly higher than that from control cells (without DMSA-Fe2O3). However, the endocrine function for endothelin-1 and prostacyclin I-2, as well as the urea transporter function, was altered even without obvious evidence of cell injury in this context. We also showed by real-time PCR analysis that DMSA-Fe2O3 exposure resulted in differential effects on the expressions of pro- and anti-apoptosis genes of HAECs. Meanwhile, it was noted that DMSA-Fe2O3 exposure could activate the expression of genes related to oxidative stress and adhesion molecules, which suggested that inflammatory response might be evoked. Moreover, we demonstrated by in vitro endothelial tube formation that even a small amount of DMSA-Fe2O3 (0.01 and 0.02 mg/ml) could inhibit angiogenesis by the HAECs. Altogether, these results indicate that DMSA-Fe2O3 have some cytotoxicity that may cause side effects on normal endothelial cells.


Background
The application of magnetic nanoparticles (MNPs) in diagnosis and effective treatment of diseases has become an area of increasing interest in the biomedical sciences [1][2][3][4]. Drug delivery is used to carry drugs region-specifically by attaching them to MNPs and releasing the drug in vivo to the target locale [5][6][7][8][9]. Via AC magnetic fields, the MNPs can mediate hyperthermia for in situ cancer-targeted therapy and be used for in vitro cancer cell-targeted detecting systems [10][11][12][13][14].
Similarly, cells of interest labeling with large amounts of MNPs can be located, tracked, and recovered by imaging techniques such as high-resolution magnetic resonance imaging [15][16][17][18].
MNPs of iron oxide (Fe 3 O 4 , γ-Fe 2 O 3 ) may develop to be the modest and biocompatible one with the rapid progress in biological applications research [19,20]. Many investigations have studied the use of diverse organic coatings as a way of optimizing the delivery of MNPs to or into cell. Several studies have confirmed that a simple dimercaptosuccinic acid (DMSA) coating can enhance the rate of uptake by three orders of magnitude, presumptively by engendering the MNPs with an anionic charge, leading to nonspecific adsorption to the cell surface followed by endocytosis into the cell [21][22][23]. These methods can deliver huge amounts of MNPs into the cells, but a proven concern arises over the impacts that great intracellular concentrations of MNPs might have on normal cell behavior. A quantitative model cell system indicates that intracellular delivery of even restrained levels of iron oxide (Fe 2 O 3 ) nanoparticles may affect cell function. To be more specific, the cytotoxicity investigations show that exposure to mounting concentrations of anionic MNPs, from 0.15 to 15 mM of iron, results in a dose-dependent decreasing viability and capacity of PC12 cells to spread neurites in return for nerve growth factor [24].
In addition to drug delivery, many biomedical applications of MNPs such as magnetic tracking and hyperthermia need a very great deal of MNPs to be injected into blood vessels, which are lined by endothelial cells (ECs), a single squamous epithelial cell layer and an anticoagulant barrier between the vessel wall and blood. EC is involved in the immune and inflammatory response, coagulation, growth regulation, production of extracellular matrix components, and is a modulator of blood flow and blood vessel tone. EC injury, activation, or dysfunction is a hallmark of many pathologic states including atherosclerosis, loss of semi-permeable membrane function, and thrombosis [25]. A wide variety of stimuli can induce programmed cell death (apoptosis) of endothelial cells through extrinsic (death receptor) and/or intrinsic (mitochondria) apoptotic pathway, which is ultimately executed by the intracellular proteases called caspases. There also exist caspase-independent pathways of apoptosis and anti-apoptotic proteins, which can protect cells from apoptosis. These pathways and proteins compose the complicated network of the cell apoptosis [26][27][28][29]. When injecting MNPs into blood vessels, ECs is the first tissue barrier encountered by the MNPs. The focus of this study is thus on the cytotoxicity evaluation of DMSA-coated Fe 2 O 3 nanoparticles (DMSA-Fe 2 O 3 ) on human aortic endothelial cell (HAEC), which is able to proliferate for many generations maintaining its endothelial characteristic and is widely used for in vitro study [30].

Preparation of DMSA-Fe 2 O 3 nanoparticles
The DMSA-Fe 2 O 3 was prepared by co-authors Dr. Fei Xiong, Dr. Yu Zhang, and Dr. Ning Gu. The characterization data, such as transmission electronic microscopy (TEM) images, crystal structure, surface charge, and magnetic measurements and Fourier transform infrared spectroscopy measurements were determined as the previous report in Dr. Gu's Lab [31]. In the present study, quasi-spherical DMSA-Fe 2 O 3 with an average diameter of 10 nm, was diluted in deionized water to 1 mg/ml, and then further diluted in tested concentrations with cell culture medium before using.

Location of DMSA-Fe 2 O 3 in the HAEC
For TEM analysis, the HAECs incubated with 0.02 mg/ml DMSA-Fe 2 O 3 for 24 h were washed with PBS and routinely fixed, dehydrated, and embedded [32]. Ultrathin sections (80 nm) were transferred to the 200 mesh copper grid, stained with 5% lead tetraacetate, air-dried, and then examined with a TEM (JEM-1010, JEOL, Akishima-shi, Japan) at 80 kV.

Cell viability/cytotoxicity assay
The cytotoxicity of DMSA-Fe 2 O 3 against HAECs was investigated by the tetrazolium dye (MTT) assay [33]. For the dose-dependent effect, the DMSA-Fe 2 O 3 , diluted with culture medium at graded concentrations from 0.001 to 0.2 mg/ml, was applied to the HAECs for 24 h. For the time-dependent effect, 0.05 mg/ml of DMSA-Fe 2 O 3 was applied to the cells for 4, 24, 48, and 72 h, respectively. After washing with PBS, the cells were incubated with MTT solution at 37°C for 2 h, and the dyes were dissolved by dimethyl sulfoxide (DMSO) for 15 min. Absorbance was examined at 595 nm with the Ultra Microplate Reader ELX808IU, and cell viability was calculated as a percentage of control cells treated without DMSA-Fe 2 O 3 . Each experiment was repeated at least three times independently.

Assessments of HAEC injury markers and endocrine factors
In this study, HAECs were co-cultured with 0.02 mg/ml of DMSA-Fe 2 O 3 for 24 h. Then, the cell culture supernatant was centrifuged at 8000 × g, 4°C for 30 min to remove the rest of the nanoparticles and cell debris. ET-1, PGI-2, and NO concentrations in the supernatant were measured using ELISA kits according to the manufacturer's instructions, respectively. Lactate dehydrogenase (LDH) and urea were determined using an automatic biochemistry analyzer (Olympus AU5400, Olympus Corporation, Shinjuku-ku, Japan).

Real-time PCR analysis of HAEC gene expression
Thirty-eight genes related to apoptosis cascade, endoplasmic reticulum (ER) stress, oxidative stress, adhesion molecules, and calcium-handling proteins were detected by real-time PCR. In this study, HAECs were incubated with 0.02 mg/ml of DMSA-Fe 2 O 3 for 24 h. The total RNA (300 ng) extracted from HAECs was reversetranscribed using the PrimeScript ™ RT reagent Kit, and then the cDNA was amplified using the SYBR Premix Ex Taq ™ according to the following cycle conditions: 30 s at 95°C for 1 cycle, 5 s at 95°C, and 30 s at 60°C for 40 cycles (AB 7900HT Fast Real-Time PCR system). All real-time PCR reactions were performed in triplicate. The housekeeping gene GAPDH was used as an internal control. The fold changes of target gene expression relative to those of the control group were analyzed by the 2 −ΔΔCT method [34], divided into different ranges and depicted as different colors.

Effects of DMSA-Fe 2 O 3 on HAEC tube formation
The tube formation assay is one of the most widely used assays to model the reorganization stage of angiogenesis in vitro. The assay measures the ability of endothelial cells, plated at subconfluent densities with the appropriate extracellular matrix support, to form capillary-like structures. In this study, the Matrigel basement membrane matrix was used as extracellular matrix support to observe whether angiogenesis of HAEC can be intervened by DMSA-Fe 2 O 3 or not. For HAEC tube formation, 50 μl/well of the Matrigel basement membrane matrix was added to a 96-well plate and allowed to gel for 60 min at 37°C. Then, HAECs were seeded at a density of 1.5 × 10 4 cells/well on the surface of the gel in the presence or absence of conditioned DMSA-Fe 2 O 3 and incubated for 14 h at 37°C in a CO 2 incubator. Meanwhile, the high urea solution (6M urea) was used as a positive control for inhibition of tube formation. The cultures on the gel were fixed for 10 min in 25% glutaraldehyde, washed, and stained with Mayer's hematoxylin. Each well was inspected under a light microscope at ×100 magnification and captured more than three pictures from different fields. Image-Pro plus (IPP) 6.0 for Windows software (Media Cybernetics, Inc., Rockville, MD, USA) was used to measure the length of tube formation on each picture. The average data from the same well was calculated as its quantitative value.

Statistical analysis
The data were represented as mean ± SD of no less than three independent experiments. Statistical analysis was performed using a student's t test. A value of p < 0.05 was considered statistically significant.

Results and discussion
Endocytosis of DMSA-Fe 2 O 3 by HAECs We were able to recognize the DMSA-Fe 2 O 3 inside the HAECs and distinguish them from the cellular structures by their high electron density on TEM.

HAECs viability studies
The tetrazolium dye (MTT) assay has been used for detecting the number of viable cells (proliferation) and loss of viable cells (cytotoxicity) resulting from toxic materials since only living cells can reduce the MTT to its insoluble form, formazan, which can be quantitatively measured after dissolved in DMSO by a spectrophotometer, and the resultant value is linked to the number of living cells [35].
In the present study, the viability of HAECs was apparently decreased with increased DMSA-Fe 2 O 3 concentrations compared with that of control cells (Figure 2a

Effects of DMSA-Fe 2 O 3 on HAEC injury markers and endocrine factors
LDH is a cytoplasmic enzyme which can be released to the extracellular space because of the disturbances of the cellular integrity induced by pathological conditions. Therefore, supernatant LDH of cultured HAECs is detected as a marker for cell injury [36]. We found that there was no difference in LDH released from the HAECs incubated with 0.02 mg/ml DMSA-Fe 2 O 3 for 24 h and the control cells ( Figure 3). This finding was consistent with the results of little cytotoxicity effect in MTT assay (Figure 2a) and cell membrane integrity changes shown by TEM (Figure 1c,d).
We then examined whether the endocrine function of HAECs was changed when exposed to this low dose of DMSA-Fe 2 O 3 that did not cause measurable cell injury. ECs can regulate blood pressure and blood flow by releasing vasodilators such as NO and PGI-2, as well as vasoconstrictors, including ET-1. So, the endocrine function of cultured HAECs can be assessed by detecting the above-mentioned factors in the supernatant. We found that the release of NO was not changed in the HAECs treated with 0.02 mg/ml DMSA-Fe 2 O 3 for 24 h (Figure 3). NO released toward the vascular lumen is the most important stimulator for vascular dilator and a potent inhibitor of platelet aggregation and adhesion. NO protects against the onset and later steps in atherogenesis, and thus is one of the most important protective molecules in the vasculature. Endothelial NO synthase (eNOS) is the predominant NOS isoform in the vasculature responsible for most of the vascular NO production. A functional eNOS oxidizes its substrate L-arginine to L-citrulline and NO. Our results indicate that the eNOS function in the HAECs is not affected by treatment with 0.02 mg/ml DMSA-Fe 2 O 3 for 24 h.
In contrast to the release of NO, the release of another vasodilator PGI-2 and the vasoconstrictor ET-1 was significantly decreased in the HAECs treated with 0.02 mg/ml DMSA-Fe 2 O 3 for 24 h (Figure 3, p < 0.01 vs. control group). Besides its function as an effective vasodilator, PGI-2 can prevent platelet plug formation by inhibiting platelet activation. PGI-2 is produced in endothelial cells from prostaglandin H 2 by the action of the enzyme PGI-2 synthase. ET-1 is secreted constitutively by endothelial cells from its inactive intermediate, big ET-1, through the action of endothelin-converting enzyme, which is present at the EC surface and on intracellular vesicles. Expression and release of PGI-2 and ET-1 in the ECs are regulated by complex signals; we did not study the mechanism for their reducing expressions and/or release in this study. However, our results demonstrate that the endocrine functions of HAECs are sensitive to DMSA-Fe 2 O 3 treatment, and these functions may be interfered before severe cell injuries occur.
In addition to the cellular-releasing function of these vessel tone regulators, we also studied the cellular uptake function by examining the urea transporter function. The transporter for urea is expressed in the vascular endothelium that transports urea into the cell. Urea plays a significant role in the endothelial cell, and previous studies have revealed that uremic levels of urea (25 mM) inhibit L-arginine transport in cultured endothelial cells [37]. In this study, we found that the urea concentration in the HAECs treated with 0.02 mg/ml of DMSA-Fe 2 O 3 for 24 h was significantly higher than that in control cells (Figure 3, p < 0.05). This observation suggests that the function of urea transporter in the HAECs is also inhibited by the DMSA-Fe 2 O 3 exposure.

Gene expression on HAECs
Endothelial cell death, which can be caused by environmental stresses such as oxidative stress, endoplasmic reticulum stress, and adhesion molecules, is mostly apoptotic [26]. We thereby examined gene expression related to the apoptosis cascade, endoplasmic reticulum stress, oxidative stress, adhesion molecules, and calciumhandling proteins (Figure 4). After the HAECs were incubated with 0.02 mg/ml of DMSA-Fe 2 O 3 for 24 h, the expressions of most of genes involved in the apoptosis cascade and calcium-handling proteins were changed from 0.5-to 1.5-fold compared to those of HAECs without DMSA-Fe 2 O 3 treatment, except MAPK14 (mitogenactivated protein kinase 14, MAPK14, also called p38-α), CASP3 (caspase 3), and BCL2 (Bcl-2). Caspase 3 [38] and Bcl-2 [27], which promote cell death and inhibit cell death, respectively, were increased by over 1.5-fold in mRNA expression in the experiment group. In contrast, the expression of proapoptotic MAPK14 [39] in DMSA-Fe 2 O 3treated HAECs was decreased to less than 0.5-fold to that of the control cells. Therefore, the DMSA-Fe 2 O 3 caused differential effects on the expression of pro-and antiapoptosis genes of HAECs; this may explain why the viability of HAECs was not changed at this low concentration of DMSA-Fe 2 O 3 , which might not be sufficient to activate the cell apoptosis pathway.
In this study, the expressions of all four tested genes involved in ER stress, were down-regulated in DMSA-Fe 2 O 3 -treated HAECs (Figure 4), especially the AFT4 gene (activating transcription factor 4), whose expression was decreased by over 50%. In contrast, most of the examined genes related to oxidative stress showed an increased change in DMSA-Fe 2 O 3 -treated HAECs with the expression of SOD2 (superoxide dismutase 2) and PTGS2 (cyclooxygenase-2, COX-2) elevated to 1.96and 2.44-fold, respectively. COX-2 is unexpressed under the normal conditions but elevated during an inflammation. The data suggest that oxidative stress, not ER stress, is sensitive to DMSA-Fe 2 O 3 . In addition, the expression of NOS3 (eNOS) was mildly decreased in DMSA-Fe 2 O 3treated HAECs, which was consistent to the result of NO concentration (Figure 3).
We found up-regulation of gene expression for cellcell contact and adhesion including ICAM1 (intercellular adhesion molecule 1, ICAM-1), VCAM1 (vascular cell adhesion protein 1, VCAM-1), and SELE (endothelialleukocyte adhesion molecule 1, E-selectin) (3.3-, 4.9-, and 8.1-fold, respectively, Figure 4). ICAM-1 is a type of intercellular adhesion molecule which continuously presents in low concentrations in the membranes of leukocytes and endothelial cells, and greatly increases upon cytokine stimulation. VCAM-1 and E-selectin are cell adhesion molecules expressed only after the endothelial cells being stimulated by cytokines and thus play an important role in inflammation. Thus, together with the data from genes associated with oxidative stress, the results of adhesion molecular genes indicate that inflammation response is likely evoked in HAECs following 0.02 mg/ml DMSA-Fe 2 O 3 treatment before the onset of cell death.

Conclusions
In summary, the present study shows that DMSA-Fe 2 O 3 nanoparticles absorbed by the HAECs can cause a dosedependent cytotoxic event. HAECs exposed to even a small amount of DMSA-Fe 2 O 3 may have impaired endocrine function and angiogenic functions without obvious cell toxicity. Furthermore, the genes related to oxidative stress and inflammation response were activated. Therefore, cautious evaluation of DMSA-Fe 2 O 3 nanoparticles in vivo is needed before applying them in medicine.  Length of tube networks formed by HAEC cultured on Matrigel. Image-Pro plus 6.0 for Windows software was used to measure the length of tube networks (pixels). The stained cells were inspected under a light microscope at ×100 magnification and captured more than three pictures from different fields. The average data from the same well was calculated as its quantitative value. Data are expressed as mean ± SD. **p < 0.01 vs. control.