**3. Results**

In this study, we evaluated the applicability of impedance measurements for the bioelectric profiling of di fferent cancer cell types treated with substance-selected anticancer agents. More specifically, four cancer cell lines were immobilized in calcium alginate and cultured in di fferent cell population densities (50,000, 100,000, and 200,000/100 μL). Then, 5-fluorouracil (5-FU) was applied, as it constitutes

one of the most common cancer therapeutic drugs. In each case, three frequencies were tested: 1 KHz, 10 KHz, and 100 KHz.

## *3.1. Cell Proliferation*

In order to ensure that calcium alginate was a proper immobilization matrix for the cancer cell culture, we assessed cellular viability with the MTT uptake assay. Cells were cultured in the matrix for 24 h (with and without treatment with 5-FU), and the proliferation was determined microscopically and photometrically after MTT application. Figures 2–5 depict the microscopic observations for three different populations of the four cell lines immobilized in calcium alginate after incubation with MTT.

**Figure 2.** Panoramic view of SK-N-SH immobilized cells in 3D matrix after treatment with MTT for 24 h, showing the viability in three different population densities: (**a**) 50,000 cells; (**b**) 100,000 cells; and (**c**) 200,000 cells. Scale bars = 50 μm.

**Figure 3.** Panoramic view of HEK293 immobilized cells in 3D matrix after treatment with MTT for 24 h showing the viability in three different population densities: (**a**) 50,000 cells; (**b**) 100,000 cells; and (**c**) 200,000 cells. Scale bars = 50 μm.

**Figure 4.** Panoramic view of HeLa immobilized cells in 3D matrix after treatment with MTT for 24 h showing the viability in three different population densities: (**a**) 50,000 cells; (**b**) 100,000 cells; and (**c**) 200,000 cells/100 μL. Scale bars = 50 μm.

**Figure 5.** Panoramic view of MCF-7 immobilized cells in 3D matrix after treatment with MTT for 24 h showing the viability in three different population densities: (**a**) 50,000 cells; (**b**) 100,000 cells; and (**c**) 200,000 cells/100 μL. Scale bars = 50 μm.

Viable cells were dyed purple using the yellow formazan (MTT) by intracellular NAD(P)H-oxidoreductases [43]. We can see that cellular proliferation is affected neither by the immobilization matrix, nor by the increase in the cell population density. Contrary to this observation, the results from the photometric MTT determination presented in Figures 6–9 showed an increase in the absorbance as cell number population densities increase, whereas the addition of 5-FU led to a significant reduction in cell viability (see Table 2) in almost all cell lines. Cell population alterations in the neuroblastoma SK-N-SH cell line (see Figure 6) appear to have a limited impact in MTT absorbance for both cell cases, i.e., treated with 5-FU and untreated. On the other hand, in the case of the remaining cell lines (Figures 7–9), we observed an increase in absorbance proportional to the cell number.

**Figure 6.** Cellular viability of SK-N-SH cells immobilized in 3D matrix after treatment with MTT for 24 h showing the viability in three different population densities (50,000, 100,000, and 200,000 cells/100 μL) ± STD: (**a**) untreated cells (control); (**b**) cells treated with 5-FU. ## < 0.01 significantly different from 100,000 cells/100 μL.

**Figure 7.** Cellular viability of HEK293 cells immobilized in 3D matrix after treatment with MTT for 24 h showing the viability in three different population densities (50,000, 100,000, and 200,000 cells/100 μL) ± STD: (**a**) untreated cells (control); (**b**) cells treated with 5-FU. \* < 0.05 significantly different from 50,000 cells, ## < 0.01 significantly different from 100,000 cells/100 μL.

**Figure 8.** Cellular viability of HeLa cells immobilized in 3D matrix after treatment with MTT for 24 h showing the viability in three different population densities (50,000, 100,000, and 200,000 cells/100 μL) ± STD: (**a**) untreated cells (control); (**b**) cells treated with 5-FU. \* < 0.05, \*\* < 0.01 significantly different from 50,000 cells, # < 0.05, ### < 0.001 significantly different from 100,000 cells/100 μL.

**Figure 9.** Cellular viability of MCF-7 cells immobilized in 3D matrix after treatment with MTT for 24 h showing the viability in three different population densities (50,000, 100,000, and 200,000 cells/100 μL) ± STD: (**a**) untreated cells (control); (**b**) cells treated with 5-FU. \* < 0.05, \*\* < 0.01, \*\*\* < 0.001 significantly different from 50,000 cells.

**Table 2.** Significant differences (Student's T-test) between cell populations before and after treatment with 5-FU. \*\*< 0.01, \*\*\* < 0.001.


For further analysis, all cell line combinations were compared using a Student's T-test in each population density, with or without the addition of 5-FU. As shown in Table 3, in the 50,000 cell/100 μL population density, treatment with 5-FU did not significantly affect MTT uptake. However, it seems that the other two population densities contributed to differential viability results, i.e., with or without treatment with 5-FU.


**Table 3.** Significant di fferences (Student's T-test) in cell viability between di fferent cell lines X cell population densities before and after treatment with 5-FU. \* < 0.05, \*\*< 0.01, \*\*\* < 0.001.

#### *3.2. Comparative Bioelectrical Profiling Results among Di*ff*erent Immobilized Cell Lines*

This experimental approach refers to the analysis of bioelectrical impedance-based measurements on various cancer cell types in di fferent population densities. Calcium alginate was once again chosen as the 3D immobilization matrix used for each cancer cell culture. Figures 10–13 depict the absolute values of the di fferences between the mean blank values and the mean impedance values for three population densities tested for each cell line chosen in three di fferent frequencies (1 KHz, 10 KHz, and 100 KHz).

**Figure 10.** Normalized values of the mean impedance magnitude for untreated (control) immobilized SK-N-SH cancer cell lines tested at three frequencies (1 KHz, 10 KHz, 100 KHz) for three di fferent population densities ± STD: (**a**) 50,000 cells; (**b**) 100,000 cells and (**c**) 200,000 cells/100 μL.

**Figure 11.** Normalized values of the mean impedance magnitude for untreated (control) immobilized HEK293 cancer cell lines tested at three frequencies (1 KHz, 10 KHz, 100 KHz) for three di fferent population densities ± STD: (**a**) 50,000 cells; (**b**) 100,000 cells and (**c**) 200,000 cells/100 μL.

**Figure 12.** Normalized values of the mean impedance magnitude for untreated (control) immobilized HeLa cancer cell lines tested at three frequencies (1 KHz, 10 KHz, 100 KHz) for three different population densities ± STD: (**a**) 50,000 cells; (**b**) 100,000 cells and (**c**) 200,000 cells/100 μL.

**Figure 13.** Normalized values of the mean impedance magnitude for untreated (control) immobilized MCF-7 cancer cell lines tested at three frequencies (1 KHz, 10 KHz, 100 KHz) for three different population densities: (**a**) 50,000 cells; (**b**) 100,000 cells and (**c**) 200,000 cells/100 μL.

As indicated by many studies, and also in our case, a frequency-dependent impedance response is observed. Each cell line x population density combination corresponds to a different pattern in the impedance magnitude measured in each frequency. Thus, in neuroblastoma SK-N-SH cells (see Figure 10), we can observe that the normalized impedance value drops when we move from 50,000 to 100,000 cells, and then increases at 200,000 cells/100 μL at each frequency. A similar pattern is observed in the case of HEK293 cells, but as shown in Figure 11, at 200,000 cells/100 μL population density, the impedance value is higher than the respective impedance in 50,000 cells. However, HeLa and MCF-7 cell cultures (Figures 12 and 13) appear to have a totally different behavior. More specifically, at a lower frequency (1 KHz), HeLa cells initially demonstrated a high impedance value (54, 23 Ohm), and as population density increased, a dramatic decrease was observed (1, 60 Ohm). In contrast, MCF-7 cells demonstrated the opposite trend, starting with 66, 72 Ohm at 50,000 cells and ending up at 79, 23 Ohm at 200,000 cells/100 μL. Furthermore, at the other two frequencies, we observed differential fluctuations for both cell lines. In other words, each cell line x population density combination was characterized by its own unique impedance behavior fingerprint.

#### *3.3. Comparative Bioelectrical Profiling Results among Di*ff*erent Immobilized Cell Lines Treated with 5-FU*

At this experimental stage, the evaluation of 5-FU applied to the previous immobilized cancer cell cultures was implemented to investigate the effects of this widely-used anticancer medicine through impedance measurements at specific frequencies. Figures 14–17 depict the normalized impedance values after the subtraction of the mean blank values from the mean impedance values in three population densities tested for each cell line chosen in three different frequencies, with (yellow bars) or without (blue bars) 5-FU. The statistical significances after pair comparisons for all combinations of the cell populations are represented in Tables 4–7.

**Figure 14.** Normalized values of the mean impedance magnitude for control immobilized SK-N-SH cells (blue bars) and immobilized SK-N-SH cells treated with 5-FU (yellow bars), tested at three different cell population densities ± STD: (**a**) 50,000; (**b**) 100,000 and (**c**) 200,000/100 μL for three different frequencies (1 KHz, 10 KHz, and 100 KHz).

**Figure 15.** Normalized values of the mean impedance magnitude for control immobilized HEK293 cells (blue bars) and immobilized HEK293 cells treated with 5-FU (yellow bars), tested at three different cell population densities ± STD: (**a**) 50,000; (**b**) 100,000 and (**c**) 200,000/100 μL for three different frequencies (1 KHz, 10 KHz, and 100 KHz).

**Figure 16.** Normalized values of the mean impedance magnitude for control immobilized HeLa cells (blue bars) and immobilized HeLa cells treated with 5-FU (yellow bars), tested at three different cell population densities ± STD: (**a**) 50,000; (**b**) 100,000 and (**c**) 200,000/100 μL for three different frequencies (1 KHz, 10 KHz, and 100 KHz).

**Figure 17.** Normalized values of the mean impedance magnitude for control immobilized MCF-7 cells (blue bars) and immobilized MCF-7 cells treated with 5-FU (yellow bars), tested at three different cell population densities ± STD: (**a**) 50,000; (**b**) 100,000 and (**c**) 200,000/100 μL for three different frequencies (1 KHz, 10 KHz, and 100 KHz).

**Table 4.** Significant differences (Student's T-test) between population densities for the SK-N-SH cell line before and after treatment with 5-FU. \*\*< 0.01, \*\*\* < 0.001.


**Table 5.** Significant di fferences (Student's T-test) between population densities for HEK293 cell line before and after treatment with 5-FU. \* < 0.05, \*\*< 0.01, \*\*\* < 0.001.


**Table 6.** Significant differences (Student's T-test) between population densities for HeLa cell line before and after treatment with 5-FU. \* < 0.05, \*\*< 0.01, \*\*\* < 0.001.


**Table 7.** Significant di fferences (Student's T-test) between population densities for MCF-7 cell line before and after treatment with 5-FU. \* < 0.05, \*\*< 0.01, \*\*\* < 0.001.


When the anticancer agen<sup>t</sup> 5-fluorouracil was applied for 24 h, we observed that in most cases (18 out of 27), 5-FU treatment gave higher impedance normalized values in comparison to cells with no treatment. The SK-N-SH cell line (see Figure 14) followed mostly the opposite pattern when 5-FU was applied compared to the rest of the cell lines, especially at a frequency of 1 KHz. In the case of 200,000 cells/100 μL population density, the cell impedance magnitude followed a downward trend, as opposed to cells under 5-FU exposure, that gave higher values in the frequency range of 1–100 KHz.

Figure 15 summarizes the results for the HEK293 cell line. Similar to the aforementioned observations, impedance significantly dropped with an increase in frequency in all population densities, and treatment with 5-FU led to higher normalized values (see Table 5). An exception can be observed at the frequencies of 10 KHz and 100 KHz in 50,000 cell population density, and also at the frequency of 10 KHz in 200,000 cells/100 μL population density, where the values obtained from control cells were higher than the respective values with 5-FU.

As depicted in Figure 16, HeLa cells treated with 5-FU follow a frequency-dependent, downward motif for every population density. On the other hand, untreated cells do not show the same pattern, since a downward trend is observed at 50,000 cells/100 μL, followed by an upward tendency at 200,000 cells/100 μL, and general non-linear behavior at 100,000 cells/100 μL. Once again, in almost all cases, the response of the cells treated with 5-FU is significantly higher in comparison with control cells (Table 6). The only exception is observed in 50,000 and 200,000 cells/100 μL population densities at 100 KHz. The latter frequency gave low impedance values for both the treated and untreated cell populations.

In the MCF-7 cell line (Figure 17), we observed particularly high impedance values when 5-FU was applied for every population density tested when compared with untreated cells, especially at a frequency of 1 KHz. These values significantly increased with population density at the same frequency (Table 7). The impedance values of cells not treated with 5-FU depicted low variations

between di fferent cell population densities. A general observation is that as frequency magnitude increases, the normalized impedance values decrease.
