Microfluidic Single-Cell Monitoring Versus Microplate Bulk-Cell Measurement

1. Introduction

Many methods for measuring cell changes have been developed, and these include bulk-cell measurements conducted with microplates [1], flow cytometry [2,3], and cell imaging [4]. On the other hand, single-cell monitoring has been achieved on different single cells using flow cytometry [2,3], and on the same single cells by the microfluidic chip [5,6,7]. While the microfluidic chips have been used to provide real-time changes monitored on the same single cells, there are concerns about the type of information obtained in measuring only a limited amount of single cells [8,9,10]. So, a knowledge gap exists that the limitations of the microfluidic method may be unclearly founded in the light of the merits of the bulk-cell method. Therefore, in this study, the two methods are directly compared, and the comparison was made using a well-established cell assay: the cell calcium assay. In order to implement the assay for direct comparison, the stimulant or drug for the assay should be selected, and so is the type of cells that is to be chosen.

In cell biology, intracellular calcium plays pivotal role as a universal second messenger that regulates a vast range of cellular functions. For example, the increase of cytosolic Ca²⁺ concentration or [Ca²⁺]_i is associated with the activation of G-protein-coupled receptors (GPCRs) which are considered the drug targets of 50-60% of therapeutic agents [11,12]. Consequently, measuring cytosolic calcium is a crucial assay to be selected for our comparison study. Many microfluidic studies have been conducted using different human cell types, e.g., muscle [13,14], osteoblast [15], glioma [6,16], endothelial [17], intestinal [18,19], kidney [20], tumor microtubes [21], and others such as CHO cells [22].

A549 cells, which belong to a type of lung cancer cell [23], have also been studied for cell calcium. A549 cells express H1 and H2 histamine receptors and their responses to histamine to produce the increase in cytosolic calcium concentration has been previously studied [25,40]. The measurement of the increase in cytosolic calcium concentration can be achieved by using a fluorescent calcium indicator dyes, such as Fura-2 or Fluo-4 [24,25,26,27].

Therefore, we selected to use A549 cells for the cell assay and chose the cell stimulant to be histamine. Since all these previous studies using microplates and flow cytometry were not performed at the single-cell scale, direct comparison is not straightforward. This study is novel in that it directly compares bulk-cell and single-cell calcium measurements performed under identical experimental conditions.

2. Materials and Methods

2.1. Reagents

The reagents used consist of DMSO (99.9%), DMEM/high-glucose medium (supplemented with sodium pyruvate, 10% fetal bovine serum (FBS), and 1% penicillin, Hanks’ balanced salt solution (HBSS), phosphate-buffered saline (PBS), 0.05% trypsin-EDTA and ionomycin.

The increase in [Ca²⁺]_i in mammalian cells has been evaluated using fluorescent calcium indicator dyes, such as Fura 2 [1,28], or Fluo 4 [29]. Here, the probe Fluo 4 AM ester (50 µg, special packaging) was used. It was first dissolved in 50 µL DMSO to prepare a 1 mg/mL stock solution. Since the fluorescence probe is a light-sensitive dye, it must be stored in dark at -20 °C. Before use, the stock solution was freshly diluted in HBSS to make a 5.0 µM working solution. Trypan blue solution (0.4%) was used to evaluate the viability of the cell under the microscope.

Histamine, which is endogenous in humans and released to other tissues upon stimulation by an allergen or other irritants [30,31,32,33,34,35], is also available commercially. Histamine dihydrochloride (98%) was obtained from Thermo Scientific Chemicals; a stock solution of 100 μM was prepared by dissolving the compound in deionized water. This stock solution was subsequently diluted in HBSS to create working solutions of 5 μM and 10 μM.

2.2. Cell Culture

A549 is a line of wild-type human alveolar epithelial cells. A549-hACE2 cell line, which is a commonly used cellular model for the study of respiratory infections, has been generated from the A549 lung carcinoma cell line [36]. A549-hACE2 cells were stably transfected to express the human ACE2 (hACE2) gene. In contrast to their parental cell line, A549-hACE2 cells are permissive to infection by SARS-CoV-2 and/or spike-pseudotyped lentiviral particles. Accordingly, the cells are ideal for studying the SARS-CoV-2 Spike protein, viral entry into host cells, as well as for screening small molecule inhibitors and neutralizing antibodies.

Briefly, the frozen cell vial was thawed by gentle agitation in a 37 °C water bath. To reduce the possibility of contamination, the vial’s cap and O-ring were kept out of water. Thawing should be rapid (approximately 2 minutes). Then, the vial was removed from the water bath as soon as the contents are thawed, and the vial was decontaminated by dipping in, or spraying with, 70% ethanol. After the cap was removed, the cells were transferred to a larger tube containing 15 mL of pre-warmed growth medium. Selection antibiotics was not added until the cells have been passaged twice.

Using a centrifuge (Eppendorf 5804), the tube was spun at 1400 rpm for 5 min using a rotor (F34-6-38, 11.5 cm radius) with a g-force of 252 g. The supernatant containing the cryoprotective agent was removed and cells were resuspended with 1 mL of growth medium. The contents were transferred to a T-25 tissue culture flask containing 5 mL of growth medium without selective antibiotics. The culture was incubated at 37 °C in 5% CO₂. The confluency of the cells was monitored according to Figure 1.

2.3. Single Cell Assay Using the Microfluidic Chip

To conduct single cell assay, a glass chip that consisted of three reservoirs, three channels, and one chamber was employed (Figure 2a), as previously reported [6]. The fabrication of the glass chip was previously reported [29,37]. This chip was comprised of a chamber and a V-shaped cell retention structure for single-cell isolation; whereas Reservoir 2 was used for drug delivery. The channel was 40 μm deep, while the reservoirs (2.5 mm in diameter) were 0.6 mm deep [6].

An optical system with an inverted microscope (TE300, Nikon) and a CCD camera was used for bright-field imaging, see Figure 2b. For easy and clear microscopic observation, a TV monitor was used, with a video capture card installed in a computer for image capture according to Figure 2.

Cell introduction and selection were conducted after the chip was washed, as previously reported [16]. Briefly, the cell medium from Reservoir 1 was removed as much as possible (Figure 3). Then, the cell suspension was resuspended gently and a small aliquot (5 μL) was added into Reservoir 1. When a small amount of medium was removed from Reservoir 3, the cell should move from left to right (Figure 3). When the cell moved closer to the entrance of the cell retention structure, some cell medium was added to the right (Reservoir 3). In this manner, the liquid pressure difference was established, and a desired cancer cell would be moved and stabilized in position within or near the cell retention structure.

Single-cell intracellular calcium measurement was performed on this fluorescence microscope platform following a previously published procedure from our group [1,14]. Using the fluorescence intensities obtained from the microfluidic method, the change in cell calcium concentration, [Ca²⁺]_i can be calculated using Equation (1):

$${{[\mathrm{Ca}}^{2+}]}_i=K_{\mathrm{d}}{\displaystyle \frac{F-F_{\min }}{F_{\max }-F}}$$

where F shows measured fluorescence; K_d represents the dissociation constant of Fluo-4 Ca²⁺ complex and has a value of 0.35 μM [38]; F_min is the background fluorescence intensity (in the calcium-free surrounding solution); F_max indicates maximum fluorescence induced by ionomycin that facilitates calcium entry through the cell membrane to saturate the Fluo 4 dye in the cell.

2.4. Bulk Cell Analysis Using the Microplate Reader

A black clear-bottom cell-culture grade COC-coated 96-well plate (model 3880, Corning NY USA) was used. Wells A1-6, B1-6, C1-6 and D1 were used, see Figure 4. The wells are seeded with A549 cells (A1 to C3) and ACE2-enriched A549 cells (A4-C6) at a density of 10,000 cells per well. Cells were counted using a hemocytometer. Experiments in triplicate (A, B, C) were caried out at 24 h after cell seeding to ensure cell adherence to the well bottoms and desired cell confluency.

To load the calcium dye Fluo-4, cells in wells (A1-6, B1-6, C1-6) were treated with freshly prepared 5 μM Fluo-4 AM working solutions and incubated for 45 min at 37° C. After incubation, the extracellular dye solution was removed and wells were washed and then topped up with HBSS. The total volume per each well was 200 μL. Using a Tecan M200 plate reader (Tecan, Switzerland), first measurements were made to all wells (in the bottom reading mode) to establish the baseline. A cell-free well (D1) was also measured to establish F_min, see equation 1.

Then, the cells in each well were treated with different concentrations of histamine, i.e., 1 μM (1A-C and 4A-C), 5 μM (2A-C and 5A-C), 10 μM (3A-C and 6A-C). Measurements were performed with the M200 plate reader. The temperature inside the plate reader was maintained at 37° C throughout the experiments to mimic physiological conditions. The following settings were used on the Tecan iControl software for measurements: fluorescence bottom reading (which is a more sensitive mode for adherent cells compared to top reading mode), 470 nm excitation wavelength, 530 nm emission wavelength, 25 flashes, 20 µs integration, 0 µs lag time, 0 µs settle time.

Using the fluorescence intensities obtained, i.e., F_min, F and F_max, the change in cell calcium concentration, [Ca²⁺]_i can be calculated using equation 1, where F shows the average of the measured fluorescence (three-replicates or n = 3); F_min is the background fluorescence intensity (in the calcium-free well); F_max indicates maximum fluorescence induced by ionomycin added to the cell in the wells.

3. Results and Discussion

Figure 5 illustrates the change in the fluorescent intensity due to histamine as measured in a single-cell of wild type A549. The changes occurred after dye loading (1, 2, 3) but before ionomycin treatment (4). It is because histamine can interact with the cells via histamine receptors [39,40]. All the histamine receptors are subtypes of GPCRs which will produce cytosolic calcium increase upon activation [3,41,42,43].

The variations in intensities in the fluorescence measurements were converted to calcium concentrations using equation 1. The average of three experiments is shown in Figure 6 (blue columns). It is seen that the cytosolic calcium concentration of the wild type A549 cell is low at rest (i.e., before reagent stimulation) but it increases after the addition of histamine. The increase of cytosolic calcium in A549 cells is greater when the histamine concentrations are raised from 5 µM to 10 µM, and finally to 100 µM.

The changes of cytosolic calcium due to histamine are also measured in ACE2-enriched A549 cells, see Figure 6, which shows the concentrations of cytosolic calcium in single wild-type (blue) and ACE2-enriched A549 cells (orange). After adding different concentrations of histamine from 5 µM to 100 µM to the cells, there are increases in [Ca²⁺]_i. The increase in the cytosolic calcium concentration of wild-type A549 cells is greater than those of ACE2-enriched A549 cells. As compared to ACE2-enriched cells, the wild type A549 has a lower resting level of calcium, but its increase after the addition of histamine is more intense, i.e., for 100 µM histamine, an increase of 6.5-fold (wild type) vs. 2.6-fold (ACE2) was observed. The greater increases in the wild-type A549 cells may be caused by physical or biochemical reasons; while the physical reason may be more histamine receptors in the unaltered wild-type cells, the biochemical reason is probably the absence of interactions of histamine and ACE2 receptors, which are both GPCRs [28,43].

Figure 7 illustrates the changes in the concentration of cytosolic calcium measured by bulk cell analysis of wild-type and ACE2-enriched A549 cells. The same pattern of cytosolic calcium increases in A549 cells in response to histamine stimulation compared to that of ACE2-enriched cells is observed. This observation strengthens the notion that ACE2-enrichment on A549 cells results in smaller response as compared to wild-type A549 cells.

A direct comparison of the total cells, reagents (histamine and ionomycin) used is shown in Table 1. This comparison study confirms the results obtained from the bulk-cell measurements to be as good as those from the single-cell monitoring. While the bulk-cell measurements is an established method, it uses more reagents, i.e., 20 μL of 10X reagent in each well vs. 1 μL of 1X reagent in the microchip. For instance, for 10 μM ionomycin used to treat the cells in order to obtain Fmax, 20 μL of 100 μM was used in each microplate well, but only 1 μL of 10 μM ionomycin was needed in each single-cell experiment, see Table 1.

This high-throughput need has been achieved using droplet microfluidics in which thousands of cells are studied in thousands of nano-sized droplets. Furthermore, the issue of cell heterogeneity may be dealt with by obtaining cell information of thousands of cells in one experiment.

For instance, the droplet-based microfluidic device has allowed the investigations of thousands of individual cellular interactions in natural killer cells [44], of cytokine secretions/ immune interactions in single T cells [45,46], of exosome secretion/ drug interaction patterns of single tumor cells [47,48]. Such high throughput microfluidic methods have been used for single-cell transcriptome studies as well [49,50].

Heterogeneity has also been shown using droplet microfluidics for human cells [44,45], and also for bacteria, such as P. aeruginosa and S. aureus [51], or E. coli [52,53], or fungal pathogenicity on humans C. albicans [54].

4. Conclusions

The single-cell approach measures the calcium response in a single cell upon treatment with three concentrations of histamine. This method provides the temporal information for the cell to generate a calcium response upon reagent treatment, and the change of that response with time progression. However, our microchip method lacks the sample throughput to measure more cells, and the ability to detect the cell heterogeneity present in a cell population. In contrast, the bulk-cell calcium measurement approach offers the average cellular responses by analyzing a population of cells collectively. This approach measures more cells and allows researchers to understand the collective cell calcium response. Our two approaches do provide the cellular responses upon the reagent of several concentrations, though not the behaviors of individual cells in a heterogeneous cell population. Recent advances in microfluidic methods do address and overcome these limitations.

In our case, when cells and reagents are sufficient, the bulk-cell measurement approach is a well-established high-throughput method. But when both materials are limited, the single-cell approach provides an alternative way to obtain results, and even a better way to obtain cellular response of a reagent on the same single cell, which is not obscured by the average results generated by the bulk-cell measurements.

Furthermore, both approaches yielded consistent histamine-induced calcium responses, i.e., the wild-type A549 cells exhibited stronger responses than ACE2-enriched cells. The consistent results obtained by the two methods may well suggest a reason for the difference, perhaps due to a physical reason (more histamine receptors in the unaltered wild-type cells) or a biochemical reason (no competitive interactions of histamine with any ACE2 receptors).