Abstract
Silica nanomaterials (NMs) are widely used in semiconductor, agriculture, cosmetics, and biomedical applications, in addition to other industries. We investigated the toxic effect of silica NMs on rheological characteristics of human red blood cells (RBCs), including hemolysis, deformability, aggregation, and morphological changes. Red blood cells were exposed to silica nanoparticles (d =∼200 nm) or silica nanowires (d =∼200 nm, l = 1μm or 10μm) at a range of concentrations and incubation times. Rheological characteristics were measured using microfluidic-laser diffractometry and aggregometry. Overall, at a concentration greater than 12.5μg/ml, the hemolytic activity was shown to be in the order of nanoparticles, short nanowires, and long nanowires. Elongation index (EI) values were insignificant in the RBCs exposed to each of the silica NMs at a concentration of 12.5μg/ml. Aggregation index (AI) values decreased in the short silica nanowires at a concentration of 12.5μg/ml compared to other silica NMs. Therefore, the safe concentration of silica NMs for toxicity, in this study, was considered less than 12.5μg/ml. These hemorheological results provided insight into the interaction between RBCs and silica NMs; they will also help assess the risk of NMs’ toxicity in the blood.
Introduction
Industries have been rapidly developing advanced technologies in the previous half-century that are being applied to the products seen all around us. These products are more compact and lightweight. Recently, products consisting of nanomaterials (NMs) have been used in various fields that require compact and lightweight designs. While these nanotechnologies are applied in all industries, there are many applications in which nanoparticles are exposed to the human body during the manufacturing process or internal use. It has been confirmed that nanoparticles accumulate in the human body through experiments with rats. A number of papers have reported that the accumulation of nanoparticles causes inflammation [9, 23]. Silica is one of the most abundant substances on earth and a material that is easily accessible in sand, stone, and clay. Silica has been used as a food additive and catalyst in a number of areas, such as tiny semiconductors, optics, agriculture, cosmetics, drug delivery, target imaging of cancer, and general industries because of its structural stability [3, 19 31]. Since the technology capable of reducing the size of particles is universal, the use of nano-sized silica particles has been increasing in order to achieve a better effect [6]. Silica nanoparticles have been used in the pharmaceutical field as catalysts, additives, and medicine, as well as in biomedical applications [2, 31].
Although many researchers have studied the impact of nanoparticles on the human body, the assessment of NMs on the human body is still lacking [9, 20]. A majority of in vivo and in vitro experiments analyze nanoparticle toxicity [11, 19]. Silica NMs are not significantly different [5, 17–19]. Nelson et al. [21] studied silica NMs’ toxicity using the embryonic zebrafish. Their in vivo results show that silica NMs with an aspect ratio greater than 1 are highly toxic (LD50 = 110 pg/g). A similar result showed that nickel nanowires with a length of 20μm or greater exhibited more inflammation and fibrosis than those with a length of 5μm or less in mouse models [23]. Furthermore, only some silica nanoparticles have been evaluated for toxicity in the blood and hemolysis [18]. Other studies only took into account the deformability and hemolytic phenomenon [16, 30].
In this study, we considered three kinds of silica NMs that are commonly used. We experimentally measured the extent of hemolysis, deformability, and aggregation of red blood cells (RBC) according to the size, shape, and concentration of silica NMs. We also assessed RBCs morphology with a field-emission scanning electron microsopy (FE-SEM).
Materials and methods
Silica nanomaterials (SiO2-NMs)
Silica nanoparticles (SiO2-NPs) were purchased from nanoComposix, Inc. (San Diego, CA, USA). The diameter of the silica nanoparticles was 198.5±10.5 nm. We also considered silica nanowires (SiO2-NWs) with two different average lengths, 1μm and 10μm, made from NMs in the Nanodevice Laboratory at Korea University. The average diameter of the silica nanowires was 200 nm, and the concentration was 1 mg/ml. We utilized transmission electron microscopes (TEM) and scanning electron microscopes (SEM) for the analysis of the silica NMs as shown in Fig. 1.
Preparation of blood samples and silica NMs’ toxicity assessment
Whole blood was drawn from the antecubital vein of healthy donors with a 21-gauge butterfly infusion into a Vacutainer® (BD, Franklin Lakes, NJ, USA) containing K2-EDTA as the anticoagulant. We followed the human subject protocol and biohazard control guidelines [4]. None of the volunteers took any medicine over the previous week. Whole blood was centrifuged at the speed of 3000 rpm for 12 minutes. Plasma, buffy coat, and the upper layer of cells were separated from the rest of the cells. The plasma was kept in 15-ml conical tubes for measurement. We performed a washing process twice on only the separated RBCs. After washing, the separated RBCs were diluted with phosphate buffered saline (PBS) with a mixing ratio of 1:3 to make a suspension of 25% hematocrit. Each of the diluted RBC suspensions, with a volume of 0.3 ml, was mixed with various concentrations of silica NMs in PBS to a volume of 1.2 ml, which had 5% final hematocrit. A control sample of RBCs suspended in PBS, with a volume of 1.2 ml, not containing silica NMs was prepared. Each sample was thoroughly mixed using a roll mixer at room temperature. In this study, we evaluated the hemorheological properties of RBCs exposed to different concentrations of silica NMs at 1 hour, 2 hours, and 3 hours, respectively. All measurements were performed according to the new guidelines for hemorheologicalmeasurements [4].
Hemoglobin concentration measurement
Hemolysis analysis was used to investigate the influence of the NMs exposed to the RBCs. The final RBC suspension, containing a silica dispersion of 5% hematocrit, was incubated over 2 hours at room temperature. The samples were centrifuged with operating conditions of 5000 rpm for five minutes and sufficiently washed with PBS twice. We measured absorbance at 540 nm wavelength using a UV spectrophotometer (DR/4000, Hach®, Germany) and calculated the quantity of hemoglobin using the cyanmethemoglobin method.
Erythrocyte deformability
We measured RBC deformability using Rheoscan-AnD 300 (RheoMeditech, Seoul, Korea). The sample inlet port and sample waste reservoir consisted of a microchannel used as a test chip; the test sample flows in the channel by applying vacuum pressure at the reservoir and is detected by a laser diode [25]. We obtained different diffraction patterns of RBCs from a charge-coupled device (CCD) camera with different shear stresses. The images of the RBCs were analyzed by a fitting program and their deformability expressed as an elongation index (EI) defined as (L−W)/(L+W) where L and W are the major and minor axes of the ellipse, respectively. A more detailed technical description can be found in [25].
Erythrocyte aggregation
We measured RBC aggregation for RBC samples exposed to NMs using the light-transmitted, microchip-stirring ektacytometer (Rheoscan-AnD 300, RheoMeditech, Seoul, Korea). The samples were washed, and their hematocrit was adjusted to 40% by adding plasma. We injected these samples into the sample inlet port of the microchip (C-01, RheoMeditech, Seoul, Korea), a flat-cylindrical test chamber, consisting of an air outlet and a stirrer. While the stirrer rotates at 1000 rpm for 10 seconds, the RBCs undergo a disaggregating process. After the stirrer stops abruptly, the RBCs aggregate; this phenomenon is recorded using light transmission over 120 seconds. RBC aggregation is expressed as an aggregation index (AI). Further details of RBC aggregation measurements are provided elsewhere [26].
Scanning electron microscopy (SEM)
After the RBC samples were exposed to different concentrations of silica NMs for 3 hours, we observed changes in the RBC appearance using SEM (FE-SEM, FEI Quanta 250 FEG, USA). The RBC samples that were exposed to silica NMs were washed with PBS twice, and the fixing process was continued with 2.5% glutaraldehyde (Sigma Aldrich, MO, USA), for 30 minutes. The samples were washed with PBS twice, and the dehydration process was initiated by gradually increasing the ethanol concentrations to 25% , 50% , 70% , 90% , and 100% , respectively. Each process lasted for 10 minutes. Then, the RBC pellet was fixed to the glass coverslip, and the sputter process to Platinum was initiated. The final samples were observed through a field emission SEM.
Statistical analysis
Data is reported as means±standard deviation (SD). Statistical testing was performed using ANOVA; p < 0.05 was considered statistically significant.
Results
Hemolysis assay
Figure 2 shows the elution amount of hemoglobin from RBCs exposed to SiO2-NPs (d =∼200 nm), SiO2-S-NWs (d =∼200 nm, l =∼1μm), and SiO2-L-NWs (d =∼200 nm, l =∼10μm). Each sample was exposed for 2 hours and mixed with the Drabkin’s reagent in a ratio of 1:500. The hemoglobin concentration was measured on a UV-visible spectrophotometer (DR/4000, Hach®, Germany). The hemolysis of RBCs exposed to SiO2-NPs with a concentration of 25μg/ml was 18.1% . For the same concentration, the hemolysis of samples exposed to SiO2-S-NWs and SiO2-L-NWs were 7.4% and 6.2% , respectively. Overall, our results showed that hemolysis of a sample with NPs takes place much faster than that with NWs.
Schubert and Müller-Goymann [24] suggested that a hemolysis test should be within 5% for toxicological safety. The hemolytic concentrations of blood samples exposed to SiO2-NPs, SiO2-S-NWs, and SiO2-L-NWs, with concentrations of 12.5μg/ml, were 5.2% , 5% , and 4.7% , respectively. The hemolytic concentrations of blood samples exposed to SiO2-NPs, SiO2-S-NWs, and SiO2-L-NWs, with a concentration of 50μg/ml, were 21.1% , 16.3% , and 6% , respectively. The hemolytic concentrations of blood samples exposed to SiO2-NPs, SiO2-S-NWs, and SiO2-L-NWs, with a concentration of 100μg/ml, were 47.5% , 14.1% , and 12.7% , respectively. Overall, the hemolytic activity was shown to be in the order of NPs, SiO2-S-NWs, and SiO2-L-NWs. Therefore, the safe range of toxicity for silica NMs in this study was considered 12.5μg/ml. Based on the above results, concentrations of silica NMs were chosen in the range of 0 to 12.5μg/ml, in order to evaluate the deformability and aggregation of RBCs shown in the following two sections.
RBC deformability
RBC deformability is represented by EI over a range of wall shear stresses (0–20 Pa). For simple comparison, EI was measured at 3 Pa. Fig. 3 illustrates the comparison of EI values of RBCs, according to the concentration and incubation time of exposure to silica NMs. The graph in Fig. 3A shows the EI values of RBCs exposed to different concentrations of silica NMs for 3 hours. The EI values of RBCs exposed to each of the different types of NMs, with concentrations of 12.5μg/ml, were shown to be insignificant, but the RBCs exposed to SiO2-S-NWs had a lower value of deformability than the RBCs exposed to SiO2-NPs or SiO2-L-NWs. On the other hand, RBCs exposed to SiO2-NP and SiO2-L-NWs, with different concentrations, showed a similar pattern of EI values. The graph in Fig. 3B represents EI values of RBCs according to the exposure times of silica NMs with a concentration of 12.5μg/ml. The EI value for each of the different types of NMs was found to be insignificant, compared to the value of the control RBCs.
Erythrocyte aggregation
Figure 4 is an illustration of a comparison of AI values of the RBCs in accordance with the concentration and incubation time of exposure to the silica NMs. Unlike RBC deformability, RBC aggregation showed a notable difference. The graph in Fig. 4A shows the AI values of RBCs exposed to silica NMs with different concentrations for 3 hours. The AI values of RBCs exposed to SiO2-NPs, with a concentration of 6.25μg/ml or greater decreased. The AI values for the SiO2-S-NWs, with concentrations of 12.5μg/ml or greater also decreased. The RBCs exposed to SiO2-L-NWs, at any concentration, were found to have almost the same EI values as the control RBCs. The graph in Fig. 4B indicates the AI values of RBCs according to the exposure time of silica NMs with a concentration of 12.5μg/ml. The AI values of SiO2-NPs, SiO2-S-NWs, and SiO2-L-NWs rapidly decreased in 1 hour, and all three materials had no significant change in AI value in 2 hours. Finally, the AI values of SiO2-S-NWs dropped more than that of the other two materials in 3 hours.
SEM
Samples exposed to the three different silica NMs, with concentrations of 100μg/ml, were incubated for 2 hours, and their shape changes were observed using SEM. A control sample without silica NMs had no change of appearance, as shown in Fig. 5A. However, the samples exposed to silica NMs, with concentrations of 100μg/ml, for 2 hours were observed to have the shape of echinocytes. In addition, some of the SiO2-NPs were attached to the surface of the RBCs (Fig. 5B). SiO2-S-NWs, with the above concentration, were tangled with the surface of the RBCs as shown in Fig. 5C, but none of the RBC shape changes were different from the SiO2-NPs. Unlike SiO2-S-NWs, SiO2-L-NWs at the same concentration were attached to the bottom of the RBCs (Fig. 5D), not tabled with the surface of the RBCs. These materials did not affect the RBC shape changes.
Discussion
The hemolysis test was used as one of the toxicity measurements; 5% or less of hemolysis is considered safe [24]. The evaluation of toxicity using the hemolysis test was described in the following studies. For amorphous silica nanoparticles with a diameter of 100 nm to 300 nm and a concentration of 20μg/ml, approximately 9% of the hemolysis took place. For mesoporous silica nanoparticles, the hemolysis did not take place at the same concentration [28]. Kim et al. [11] reported that the hemolysis levels were not significantly changed even at high concentrations of lead for less than 3 hours, but when the RBCs were exposed to lead with a concentration of 5μM and incubated for 4 hours, the hemoglobin levels significantly decreased. Our study similarly confirmed a reduction in hemoglobin concentration of RBCs exposed to nanoparticle materials. It was observed that the degree of hemolysis increased as the concentration of silica NMs exposed to the RBC increased, a trend found in the study of the hemolysis phenomenon of porous silica particles [29]. The RBCs exposed to silica NPs had more hemolytic phenomenon than the silica NWs, as shown in Fig. 2.
Our collaborators performed an acute toxicity test on daphnia magna and silica nanoparticles with diameters of approximately 15 nm and silica nanowires with diameters of 50–500 nm and lengths of 10–30μm. They found that silica NPs were classified as acute category 3 (10 mg/l≤EC50≤100 mg/l), and silica NWs did not belong to this category [27]. Therefore, silica NPs were considered more toxic on the daphnia magna than silica NWs. We also observed that silica NPs had higher hemolytic activity than silica NWs. This trend was also found in the toxicity evaluation of silver nanoparticles and nanowires [13]. The cytotoxicity study of Adili et al. [1] reported that silica nanowires were nontoxic at concentrations below 190μg/ml but showed a more severe toxicity than silica nanoparticles, which produced results contradictory to ours.
The in vivo study of Poland et al. [23] showed that 20-μm nickel nanowires had more inflammation than 5-μm nickel nanowires, which is contradictory to our finding that the hemolysis is in the order of SiO2-NPs, SiO2-S-NWs, and SiO2-L-NWs. This difference can be explained by the inflammation field of toxicity for carbon nanotube (CNT), whereby a longer length of CNT leads to an inflammatory response because macrophages do not engulf it completely [12, 15]. However, in in-vitro studies we should focus on the size of the nanowires that can enter the cell. Since a short length of nanowire is relatively easier to penetrate into a cell than a longer nanowire, a short nanowire has greater impact on the hemorheological parameters of the RBC.
We considered the 12.5μg/ml concentration of silica NMs where hemolysis occurred within 5% . In the silver (Ag) NMs study of Kim and Shin [13], the concentration was determined to be 150μg/ml for the occurrence of 5% hemolysis. In case of the CNT, 5% hemolysis occurs at a concentration of 0.1μg/ml [8]. Therefore, it is suggested that CNT has a greater hemolysis than silica or Ag NMs at the same concentration.
Similarly, Fig. 3 shows that EI is decreased as the concentration of silica NMs is increased over time. The EI values of RBCs exposed to each of the different types of NMs, with concentrations of 12.5μg/ml, were shown to be insignificant. However, silica nanoparticles with a diameter of 50 nm had lower EI values than those with a diameter of 100 nm at the same concentration of 20μg/ml in Park’s thesis study [22]. We can infer that the smaller size of the NMs can facilitate the reduction of the deformability.
We observed reduced aggregation of RBCs on exposure to silica NMs in 2 hours as shown in Fig. 4. Further, the higher concentration of NMs shows a tendency to decrease the degree of aggregation proportionally. Similar to the above experiments with hemoglobin concentration and deformability, a comparable trend was found; silica nanoparticles have a lower aggregation than silica nanowires, which implied that nanoparticles had a greater reduction than nanowires.
Conclusions
The purpose of our study was to evaluate silica NMs’ toxicity in RBCs with hemorheological parameters, such as hemolysis, deformability, and aggregation. We found that silica nanoparticles resulted in larger rheological changes than nanowires through the above experiments. In addition, this study suggested that silica NMs at a concentration of 12.5μg/ml or less are not toxic to RBCs Finally, our experiments showed the possibility of rheological changes in RBCs exposed to silica NMs, but it seemed to need additional literature reviews and subsequent research. If those studies are clearly identified, we can interpret the toxicity level of the NMs.
Footnotes
Acknowledgments
This research was supported by the Nano Material Technology Development Program (Green Nano Technology Development Program) through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science, and Technology (No. 2011-0020090). The authors also wish to acknowledge Prof. Jung-Hee Park for providing SEM images in 
.