HyeRim
Shin
a,
Minjeong
Kwak
b,
Tae Geol
Lee
b and
Ji Youn
Lee
*a
aCenter for Bioanalysis, Division of Chemical and Medical Metrology, Korea Research Institute of Standards and Science, 267 Gajeong-ro, Yuseong-gu, Daejeon 34113, Republic of Korea. E-mail: jylee@kriss.re.kr; Fax: +82-42-868-5801; Tel: +82-42-868-5363
bCenter for Nano-Bio Measurement, Division of Industrial Metrology, Korea Research Institute of Standards and Science, 267 Gajeong-ro, Yuseong-gu, Daejeon 34113, Republic of Korea
First published on 2nd July 2020
Reliable quantification of nanoparticle uptake in mammalian cells is essential to study the effects of nanoparticles in the fields of medicine and environmental science. Most conventional quantification methods, such as electron microscopy or confocal imaging, are laborious and semi-quantitative and therefore not readily applicable to routine analyses. Here, we developed assays to quantify fluorescently labelled nanoparticle uptake in mammalian cells using a flow cytometer. The first approach was to measure the percentage of nanoparticle-containing cells based on a cutoff fluorescence intensity as set from a histogram of control cells, which is a quick and easy way to relatively compare nanoparticle uptake in the same set of experiments. The second approach was to measure the calibrated fluorescence intensity of the nanoparticle-treated cells in molecules of equivalent soluble fluorophore (MESF) values using calibration beads, which allows for comparisons between different sets of experiments. We successfully applied the developed assays to more readily measure fluorescence-labelled silica nanoparticle uptake in A549 lung carcinoma cells in a quantitative rather than semi-quantitative manner. We further tested the assays with nine different types of mammalian cells and investigated the correlation between cell type/size and nanoparticle uptake.
The analytical techniques employed to measure nanoparticle uptake are quite diverse7,27 including transmission electron microscopy (EM),28–30 inductively-coupled plasma mass spectrometry (ICP-MS),31 ICP-atomic emission spectrometry,26 microplate reader,32,33 flow cytometry,5,29,34–38 confocal microscopy,5 transmission X-ray microscopy,39 whole cell tomography,40 confocal Raman microscopy,41 and hyperspectral imaging.42,43 These can be broadly classified into three approaches: EM techniques such as transmission EM (TEM) and scanning EM (SEM), ICP-based techniques, and fluorescence-based techniques such as microplate reader, flow cytometry, and confocal imaging.
Among them, TEM and confocal microscopy are the most widely used for nanoparticle uptake analysis. They have good spatial resolution but poor throughput, making them difficult to apply widely. While the ICP-based techniques have also been chosen for their more quantitative results, they are destructive, not easily accessible, and above all limited to metal particles. At present, there is no primary method to measure nanoparticle uptake, and therefore orthogonal methods often need to be complemented to achieve reliable results.9 For example, Gottstein et al. combined confocal microscopy with flow cytometry to precisely quantify the uptake of polystyrene particles,5 and Valero et al. used flow cytometry and imaging44 to identify molecular targets of drugs.
A flow cytometer is a device that measures the scattering and fluorescence of single particles (e.g. cells) in stream. Unlike other fluorescence-based methods such as microplate readers, it is non-destructive and measures individual cells rather than averaged signals from a population of cells. Forward and side scattering vary upon the size and complexity of the particles, respectively. The introduction of nanoparticles, such as TiO2, ZnO, SiO2, and Ag, to cells significantly increases side scatter, leading to side scattering often being used to quantify nanoparticle uptake in mammalian cells35,45,46 and bacteria.47 In addition, other cellular statuses, such as reactive oxygen species production and genotoxicity, can be measured in parallel using fluorescence channels.38 Since flow cytometry is a non-destructive analysis method, if necessary, the analyzed cells can be recovered using a cell sorter and used in other conventional analyses.
In this work, we provide a novel quantitative analysis approach to measure the level of nanoparticle uptake in mammalian cells via flow cytometry. Uptake is measured in two different ways: (1) by the percentage of cells containing nanoparticles, and (2) by the calibrated median fluorescence intensity of the cells. First, we demonstrate our method with a human lung carcinoma cell line, A549, and then apply the method to other cell lines from different tissue origins.
The mechanisms for dye-doped silica nanoparticle synthesis have been well-documented,48,49 where silane-conjugated dye molecules are incorporated into silica nanoparticles covalently when a silica matrix is produced from the precursor, tetraethyl-orthosilicate (TEOS). Since FITC molecules are covalently incorporated into the silica matrix during synthesis, rather than being modified after nanoparticle synthesis, we anticipate that dye molecules will present both the inside and outside of nanoparticles. To confirm how stable the fluorescent molecules of the nanoparticles are, the fluorescence intensity of different concentrations of nanoparticle solutions was measured. We suppose that if nanoparticles were leaked out, the self-quenching effect between fluorescent molecules disappears and the fluorescence intensity changes. However, we observed no significant fluorescence decrease or change over 72 h. Considering the negligible deviation of fluorescence intensity among wells (≤5%), we could conclude that FITC molecules did not leach out from the silica nanoparticles (ESI Fig. S1†).
A549 cells cultured under serum-free conditions for 24 h underwent morphological changes: they lost cell–cell contact and rolled up rather than spreading out on the culture dish. Cell proliferation was inhibited and fewer cells were observed via a micrograph even with the lowest tested nanoparticle concentration, 10 ng mL−1 FITC-SiO2NP25. Most cells detached and died when they were treated with FITC-SiO2NP25 at concentrations higher than 20 ng mL−1 possibly due to the adhesion of bare nanoparticles to the cell membrane. As Lesniak et al. explained,23 bare nanoparticles tend to adsorb to cell membranes to lower their surface energy, and this tendency accelerates as particle concentration increases.51 On the other hand, in the case of A549 cells in culture media containing 10% fetal bovine serum (FBS), the presence of nanoparticles, both NP25 and NP50 at concentrations as high as 100 ng mL−1, did not affect cell morphology, adhesion, or proliferation, with viability remaining high throughout the culture period (NP25 in Fig. 2(A), and NP50 in ESI Fig. S2†). Therefore, we decided to use the serum-containing medium in subsequent nanoparticle uptake analyses.
Before proceeding to the uptake analysis, we performed DLS measurements of the nanoparticles in the culture media because it has been well documented that proteins in serum rapidly adhere to nanoparticles to form a corona layer, which, consequently, changes the particle size and biological effects.33 As Lara et al. reported, certain proteins in the biomolecular corona are identified as key players in nanoparticle recognition and uptake.22 In the case of NP25, the size doubled and the z-average diameter increased to 59 nm. The size of NP50 also increased by the addition of 10% FBS, but by less than 15%, with the z-average diameter increasing to 80 nm. The z-average diameter of the nanoparticles was found to change over time-a 30% increase for NP25 and a 30% decrease for NP50 after 3 weeks (ESI Fig. S3†)-when the solution was stored at 4 °C. Therefore, we used freshly prepared nanoparticle solutions.
For the quantification of nanoparticle uptake, we adapted the flow cytometry method reported by Choi et al.,35 where the authors utilized changes in side scattering as induced by the uptake of silica nanoparticles. Their proposed method is simple and straightforward but the change in side scattering was insignificant with negatively charged particles due to the negative charge of the cell membrane. In our study, the side scattering of A549 cells did not change by the addition of silica nanoparticles presumably because of the negative surface charge and small particle size, while the fluorescence intensity of the nanoparticle-treated cells increased dramatically, with both cases ranging from 10 to 100 ng mL−1 (ESI Fig. S4†). Accordingly, we decided to use a fluorescence channel instead of a side scatter channel to quantify nanoparticle uptake.
We used two different approaches, as illustrated in Fig. 2(B). First, we quantified the percentage of cells containing FITC-SiO2NPs, as assigned by the cells with higher FITC intensities than the threshold intensity. The threshold value was the sum of the median intensity and robust standard deviation (SD) of control (i.e. untreated) cells. We used median and robust SD to obtain statistically meaningful measurements by excluding outliers in the broad histograms. Second, we calculated molecules of equivalent soluble fluorophore (MESF)52 values of the nanoparticle-treated cells. While the fluorescence intensity as measured using flow cytometers is an arbitrary value changed by instrument settings, MESF is a calibrated value that can be easily compared directly with the results of other experiments.
Following nanoparticle treatment, the resulting histogram from the FITC channel shifted to larger intensities in a dose-dependent manner, while other scattering parameters remained unchanged (ESI Fig. S2†). The percentage of cells containing FITC-SiO2NPs also increased in a dose-dependent manner, with most cells appearing to contain nanoparticles following treatment at concentrations higher than 20 ng mL−1 (Fig. 2(C)). In the case of NP50, the histogram shift was large, even at the lowest concentration of 10 ng mL−1. Many studies have reported that 50 nm is the optimum size for cellular uptake.16,53 However, since we do not have information on the FITC content inside the nanoparticles or the size-dependent difference in cellular uptake between the tested nanoparticles, the exact reason for the histogram shift cannot be specified in the present work.
To obtain MESF values for the nanoparticle-treated cells, we employed commercial hard-dyed beads (Rainbow Calibration Particles). We used six-peak particles dyed to six different fluorescence intensities and adjusted the voltage of the FITC channel to include all six peaks in the histogram. Reproducibility was reasonably sufficient to use the same calibration curve for different experiments performed over a period of about 3 months throughout the entire study (ESI Fig. S5†). Even with different voltage settings, correlations between MESF and FITC intensities did not change as long as all six peaks of the calibration particles were fit within the detection range. We subtracted the basal fluorescence level of the untreated cells from the MESF of the nanoparticle-containing cells. Similar to the percentage of cells containing FITC-SiO2NPs, the MESF increase following nanoparticle uptake was dose-dependent, with the increase by NP50 treatment greater than that by NP25 treatment when the cells were treated with the same concentrations (Fig. 1(D)).
To demonstrate the versatility of the developed method, we quantified nanoparticle uptake in various cell lines from different organs, namely lungs, skin, intestine, and blood, which can be primarily exposed to nanoparticles. The list includes Beas-2B, A549, HCC827, and H1975 as lung cell lines, HaCaT, MNT1, and Hs294 T as skin cell lines, Caco2 as an intestinal cell line, and THP-1 cells as an immune cell line. Their representative micrographs are shown in ESI Fig. S6.† For THP-1 cells, we used both naïve and phorbol 12-myristate 13-acetate (PMA)-activated cells, referred to as THP-1 (M0).
First, we analyzed the percentage of cells containing FITC-SiO2NPs in the nine tested cell lines (Fig. 4A). Lung cell lines and THP-1 cells, both naïve and M0, showed a similar pattern as A549 cells: a high uptake percentage for NP50 and a dose-dependent uptake increase for NP25. In general, the level of nanoparticle uptake was lower in skin and intestinal cells than in lung and immune cells. However, Hs294 and Caco2 cells treated with NP50 displayed uptake percentages comparable to the lung cell lines. Therefore, we could not conclude that the level of nanoparticle uptake can be classified according to the organ from which the cell line was derived.
Then, we analyzed MESF–FITC values from the median fluorescence intensity of an ensemble of cells treated with FITC-SiO2NP25 (Fig. 4B). Since the background level of each cell line in the FITC channel was not negligible and different for each cell line (ESI, Fig. S7†), we subtracted the background values for an objective comparison. Here, we present only the NP25 results because, for some cell lines, the FITC histogram for the highest nanoparticle concentration was truncated due to a high level of fluorescence. Cell lines were roughly divided into three groups based on the MESF level of the 50 ng mL−1 NP25-treated samples. The highest MESF-level group includes a skin cell line (Hs294T) and two lung cell lines (Beas-2B and HCC629); the middle includes two lung cell lines (H1975 and A549), immune cell lines (THP-1 and THP-1 (M0)), and a skin cell line (MNT1); and the lowest includes a skin cell line (HaCaT) and the intestinal cell line (Caco2). In general, lung cells showed higher MESF values while skin and intestinal cells showed lower MESF values; however, exceptions were observed, similar to the uptake percentage results. Notably, the three skin cell lines, Hs294T, MNT1, and HaCaT, all presented different levels of MESF values, with Hs294T showing very high results.
Since one of the main parameters affecting the cellular uptake of nanoparticles is cell size, and we already have that information in the form of forward scatter (FSC) from the flow cytometer data, we investigated the correlation between cell size and MESF value. Because FSC reflects the size of cells in the form of spheres and may not directly reflect the area of the cells in culture, we further estimated the cell areas of the eight adherent types of cells (except THP-1, which is a suspension type) using a cytoplasm staining dye, carboxyfluorescein diacetate succinimidyl ester (CFSE). The FSC (i.e. FSC-A) and CFSE-area results showed good linear correlation (R2 = 0.7851, ESI, Fig. S8A†), although the measured area depends somewhat on the flatness of the cells. For example, Caco2 cells maintained a round shape with a convex edge, and thus displayed a smaller attached area than other cell lines, resulting in a relatively lower CFSE-area value compared to FSC. In contrast, Beas-2B cells tended to spread out over the surface and remained flat, resulting in relatively a higher CFSE-area value compared to FSC (Fig. S8B†).
Interestingly, when we plot the cell size, both FSC and CFSE-area, versus MESF–FITC, the two parameters displayed a meaningful correlation, with the tested cell lines again roughly dividing into three groups (Fig. 5). The Hs294 T, Beas-2B, and HCC827 cells showed the largest area and the highest MESF–FITC; the A549, H1975, THP-1 (M0), THP-1, and MNT1 cells showed moderate area and MESF–FITC; and the Caco2 and HaCaT cells showed the smallest area and the lowest MESF–FITC. It is well known that nanoparticle uptake depends on both the physicochemical properties of the nanoparticles, such as size, surface charge, and shape, and the cells, such as tissue origin15,33 and cellular mechanical state.54 Our test results from numerous cell types also suggest that the surface area of the cells plays an important role in determining the level of nanoparticle uptake, regardless of the cell type, presumably due to the increased frequency of interaction between nanoparticles and cells.
Fig. 5 Correlation between cell size and nanoparticle uptake. Plots between median MESF values with (A) FSC-A obtained by flow cytometry and (B) CFSE-area obtained by image analysis. |
Lastly, we investigated the uptake dynamics for a few cell lines, namely A549, HCC827, HaCaT, and THP-1. Sample preparation for flow cytometry is relatively simple and fast, making it suitable for monitoring nanoparticle uptake dynamics. For A549 cells, a steady increase of nanoparticle uptake over time until 5 h (ref. 19) and until 24 h (ref. 53) has previously been reported. We observed a similar steady linear increase of nanoparticle uptake until 24 h (ESI, Fig. S9A–C†). Interestingly, while all tested adherent cells showed a linear increase in MESF–FITC values over time until 24 h, THP-1 cells showed an immediate uptake of nanoparticles upon nanoparticle exposure (Fig. S9D†). Even after 5 min, the minimal time to transfer and load a sample for flow cytometer measurement, after treatment, the level of MESF–FITC was similar to that of the 1 h sample. Macrophages are known to engulf foreign substances by their nature, and are known to act as scavengers for the delivery of nanoparticles using the blood stream.18 While a detailed investigation into the nanoparticle uptake mechanism in THP-1 cells is not the scope of this paper, our proposed method can be useful in analyzing nanoparticle uptake dynamics because the sample preparation time is significantly shorter than other assays.
In summary, we developed flow cytometry-based assays to quantify the level of nanoparticle uptake in mammalian cells and validated the assays with different types of cell lines. In-house synthesized FITC-labeled silica nanoparticles of two different sizes, average diameter 25 nm and 50 nm, were employed. The first approach to measure nanoparticle uptake simply quantified the percentage of nanoparticle-containing cells based on the median fluorescence intensity of the control cell population. Although this method allows for a rapid assessment of nanoparticle uptake, it does not reflect the amount of nanoparticles present in the cell. The second approach, therefore, measured the fluorescence intensity of a population of nanoparticle-treated cells with values normalized to calibration particles, MESF. This quantitative measure also provides information about the amount of nanoparticles in the cells, and enables the comparison of experiments from different setups. Assays with a flow cytometer are high-throughput measurements providing extensive information on thousands of individual cells. By using multiple fluorescence channels, numerous cellular states such as cell cycle, oxidative stress, and genotoxicity can be simultaneously monitored to investigate the correlation between cellular status and nanoparticle uptake. In addition, the assays are non-destructive and the cells can be recovered, allowing for further downstream analysis of the nanoparticle-treated cells. We examined our assays with A549 cells and other cell lines from different tissue origins, and found that the level of nanoparticle uptake differed among the tested cell lines and largely depended on cell shape and adhered surface area rather than tissue origin. One of the important factors involved in nanoparticle uptake, but not covered in detail in the current study, is corona proteins. Corona proteins are frontline molecules interacting with the cellular membrane during uptake and their crucial role has been highlighted in multiple articles. Recent studies focus on not only analyzing the composition and amount of the corona layer using mass spectrometry,19 but also understanding the underlying mechanism of nanoparticle uptake by cells through specific proteins in the corona layer. We believe that our quantification assays could accelerate the understanding of the interactions of nanoparticles with organisms along with those pioneering studies.
All procedures for FITC-SiO2NP synthesis were conducted while minimizing exposure to light. For 25 nm nanoparticle synthesis, we modified the method from Quan et al.55 In brief, an APTES–FITC mixture was prepared by mixing 2 mg of FITC, 400 μl of butanol, and 2 drops of APTES for 24 h. The APTES–FITC mixture was then mixed with 70 ml of 0.82 μM L-arginine solution, 12 ml of butanol, and 18 ml of TEOS, and allowed to react in a round-bottom flask under vigorous stirring at 65 °C for 18 hours. The aqueous phase, which contains highly monodisperse FITC-SiO2NPs, was obtained after removing the organic phase using a glass funnel. For 50 nm nanoparticle synthesis, we modified the method from Imhof et al.56 In brief, an APTES–FITC mixture was prepared by mixing 12 μg of FITC, 10 ml of ethanol, and 14 μg of APTES for 24 hours. The APTES–FITC mixture was mixed with 44.25 ml of ethanol, 2.1 ml of ammonia, and 0.75 ml of TEOS, and then kept in a round-bottom flask under moderate stirring at room temperature for 20 hours. The solvent was exchanged with deionized water using microcentrifugation (15000 rpm, 10 min) and washed twice.
The morphology and size information of the FITC-SiO2NPs were obtained by SEM and DLS. SEM images of the 25 nm and 50 nm FITC-SiO2NPs were obtained using a HITACHI S-4800 and ZEISS GeminiSEM 500, respectively. The average particle size was obtained by measuring at least 100 single particles using ImageJ software. The hydrodynamic size and zeta potential were determined using a Zetasizer Nano ZSP (Malvern). DLS results are reported as the average of three runs.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/d0nr01627f |
This journal is © The Royal Society of Chemistry 2020 |