We demonstrated that attenuated total reflectance terahertz time-domain spectroscopy (ATR THz-TDS) is able to monitor oxidative stress response of living human cells, which is proven in this work that it is an efficient non-invasive, label-free, real-time and in situ monitoring of cell death. Furthermore, the dielectric constant and dielectric loss of cultured living human breast epithelial cells, and along with their evolution under oxidative stress response induced by high concentration of H2O2, were quantitatively determined in the work. Our observation and results were finally confirmed using standard fluorescence-labeled flow cytometry measurements and visible fluorescence imaging.
© 2017 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
Oxidative stress is the excessive reactive oxygen species (ROS) mediated damage to cellular macromolecules, which has been implicated in several pathological processes, including cancer, diabetes, myocardial infarction, and neurodegenerative diseases, as well as in cell death and aging [1,2]. ROS are a group of molecules that include highly reactive free oxygen radicals (e.g., superoxide anion [O2-] and the hydroxyl radical [OH-]) and the stable diffusible non-radical oxidants (e.g., hydrogen peroxide [H2O2]) . Although the endogenous ROS is a consequence of normal intracellular metabolism in mitochondria and peroxisomes, as well as from a variety of cytosolic enzyme systems, many environmental stimuli including cytokines, UV radiation, toxins, glucose starvation, heat shock and even growth factors generate high levels of exogenous ROS that can perturb the normal redox balance and shift cells into a state of oxidative stress . Normally, cells do quickly responses to adapt to or resist the stress by a sophisticated enzymatic and non-enzymatic antioxidant defense system, and sustained disturbance of this balance may result in either apoptotic or necrotic cell death [4–6]. A variety of methods including Western Blot, DNA-electrophoresis, colorimetric analysis, flow cytometry or fluorescence microscopy have been proven effective in detecting cell death induced by oxidative stress . However, these tests require substantial manipulation of the cells such as cell fixation and permeabilization, cell components extraction, fluorescent dyes labeling; therefore a real-time and in situ monitoring of oxidative stress responses of living cells is not possible. Some fluorescent dyes developed for live cell imaging suffer from insufficient photostability, photobleaching, or cytotoxic effects [7,8]. Non-destructive tools without complicated cell manipulations, such as spectroscopic techniques, are required for fast and easy to operate.
Terahertz (THz) spectroscopy maybe offers several advantages by quantifying the cellular responses to oxidative stress. It is sensitive enough to directly measure the low frequency vibrations and rotations of biomolecules as well as the translations of the molecular skeleton. It reveals the dynamic process on sub-picosecond to picosecond scales. Moreover, this method is noninvasive and nonionizing for biomolecules so that it is suitable for in vivo and in vitro real-time measurement of tissues and cells [9–11]. However, conventional THz time-domain spectroscopy (THz-TDS), either with transmission or reflection measurement, reaches its limit when applied to characterize samples dissolved in polar liquids due to their exceptionally high absorption in this frequency range. Technical approach to overcome this limitation might be the attenuated total reflectance terahertz time-domain spectroscopy (ATR THz-TDS). In the ATR THz-TDS scheme, the THz pulse is subject to total internal reflection at the upper interface of an ATR prism, generating an evanescent wave that penetrates into monolayer living cells [12–16]. On the other hand, ATR THz-TDS has been proved more sensitive for measuring highly absorptive media than transmission or reflection THz spectroscopy [12,13] or mid-infrared ATR spectroscopy . Thus, it is an ideal tool to analyze the THz properties of living cells cultured in water-rich media. The ATR THz-TDS has been applied to studies of cell permeabilization  as well as studies of characterizing intracellular hydration , and distinguishing cancer cell types .
In this paper, we used home-built silicon-based ATR THz-TDS system to study in situ the cultured living human breast epithelial cells (MCF10A), which are at the same time being monitored by an upright metallographic optical microscope. Living MCF10A cells are successfully cultured on and adherent to the silicon ATR prism. The adherent living cells interacts strongly with THz evanescent wave from silicon prism; hence by detecting the ATR THz spectrum from the silicon-liquid interface, we successfully characterized the THz dielectric responses of MCF10A cells. Further, adherent cells ATR were exposed to high concentration of hydrogen peroxide (10mM, H2O2), and the cell death induced by oxidative stress were monitored in situ by ATR THz-TDS system for more than 2 hours. With fluorescence-based flow cytometry analysis and fluorescence imaging, the validation of THz spectral patterns of cell death process was accomplished. Our findings suggest that ATR THz-TDS is a useful cell analysis platform for non-invasive, label-free, real-time and in situ monitoring cellular processes.
2. Experimental setup and live cell preparation
Experimental setup. We built a balanced ATR-THz element for THz-TDS system, which consist of two HR-Si ATR prisms (FZHR Si, the refractive index in THz band is ~3.42) fixed on a mechanical stage for reference and sample, respectively (Fig. 1(a)). In this way reference and sample measurement can be performed alternately by sliding the prism across the THz beam to avoid long-term instability in conventional TDS measurements (Fig. 1(b)). The sample ATR prism was modified with a reservoir which holds ~3ml of medium above, where ~3 million living cells can attach and be measured (Fig. 1(c)). THz spectroscopy was obtained after growing cells to 100% confluence observed by an upright metallographic microscope above ATR prism. The ATR-THz elements were fitted in a home-built transmission THz-TDS system , which was driven by a femtosecond fiber laser to generate and detect THz pulses with InGaAs photoconductive antennas. The femtosecond fiber laser delivered >100 mW, 63 fs de-chirped pulses (after 3 m polarization-maintaining fiber) at a center wavelength of 1560 nm with a repetition rate of 100MHz.
Live Cell Preparation. Human breast epithelial cells (MCF10A cell line) obtained from ATCC were maintained in plastic culture flasks (polystyrene) using High Glucose medium supplemented with 10% fetal bovine serum (FBS), 100 U/ml penicillin and 100μg/ml streptomycin at 37 °C under 5% CO2. The cells were harvested with 0.25% trypsin when they had reached 80%-90% confluence. The cells were then centrifuged into a pellet and resuspended in culture medium. The cell suspension reached to a cell density of 1 × 106 cells/ml and 3ml of the resulting cell suspension was seeded onto the sample ATR prism and incubated at 37 °C,5% CO2 condition overnight. Before each experiment, the sample ATR prism attached with cells was washed with LCIS (Live Cells Imaging Solution, Invitrogen) and let reach room temperature (25°C) in 3 ml of LCIS, which is an optically clear, physiological solution buffered with HEPES at pH 7.4 that keeps cell healthy for up to 4 hours at ambient atmosphere and temperature.
THz measurements of cells. Data acquisitions are made with sample ATR prism and reference ATR prism, which is measured in the absence of cells. After measured the normal cells in standard LCIS, the cells on the sample ATR prism were exposed to 10 mM H2O2 by adding 3μl of the 10 M stock solution to the cell culture on the sample ATR prism, which has a volume of 3 ml. The cells were incubated for a further ~2 hours with the THz spectrum of the cell monitored at a 6 minutes interval. The standard solution without cells well measured after cell monitoring by scratching cells and washed with LCIS three times. The control (LCIS + H2O2) of H2O2 treated cells was also monitored for ~2 hours after cells monitoring. All the measurements were performed under room temperature at 25°C with ~45% humidity.
Fluorescent labeling and flow cytometry analysis. MCF10A cells were seeded at 2 × 105 per well in 12-well dishes overnight before H2O2-treatment. The cells incubated with 10mM H2O2 for 0, 0.5, 1, 2, hours were harvested and stained using the FITC Annexin V apoptosis detection kit (BD Biosciences) according to the manufacturer’s instructions. In brief, cells were collected and washed with cold phosphatebuffered saline (PBS) and diluted in Annexin-binding buffer with 1 × 106 cells/ml. About 1 × 105 cells were incubated with 5 μl Annexin V and 5 μl propidium iodide (PI) at room temperature for 15 min in the dark. After stained Annexin V and PI, cells were seed on a microscopy slide and mounted by a fluoroshield with DAPI (4’,6-diamidino-2-phenylindole, Sigma-Aldrich), which can label both viable and dead cells, then observed by fluorescence microscope (IX71, Olympus). Annexin V-FITC/PI-stained cells were also analyzed using a flow cytometer (CytoFlex, Beckman coulter). In total 10,000 cells were analyzed per measurement. Data was analyzed using CytExpert software.
Complex permittivity retrieval. In order to obtain the complex permittivity (dielectric constant and dielectric loss) of living cells in culture, the complex electric field in the frequency domain was calculated from the THz time-domain signal by Fourier transforms with the transfer function H(ω):Eqs. (2) and (3):Eqs. (1)-(3), the complex permittivity of cell culture εs, as well as it real and imaginary parts (εs’(ω) and εs”(ω) are the relative permittivity and dielectric loss, respectively) is determined using Eqs. (4) and (5):
3. Results and discussion
To optimize the THz spectral measured from the living cells and minimize the disturbance from the culturing medium, we measured MCF10A cells by the home-built ATR THz-TDS system (Fig. 1(a)). Here, we proposed a method with two ATR prisms to alternatively measure the reference and sample signal, which enables to perform long term ATR measurement without suffering from phase errors from thermal drift in environment . In order to verify the accuracy of two ATR prisms method, LCIS have been measured several times in different days. Compare to traditional method, which a reference measurement is performed only once at first, the result using our proposed method have a much smaller error bar (Fig. 1(b)). For cell growth observation on the sample ATR prism (Fig. 1(c)), living cells were observed on an upright metallographic microscope for the reflective substrate. The morphology of cells grown on the ATR prism were comparable to the conventional cell culture on polystyrene (Fig. 1(d) and 1(e)), suggesting that Si substrate was suitable for cell growth, which was consistent with the previously report .
The time-domain THz pulses of living cells were collected at three different times and then averaged to confirm the repeatability. The average amplitudes of MCF10A cells cultures and LCIS were smaller than those of the reference pulse due to the absorption (Fig. 2(a)). However, the amplitudes of the average spectra of MCF10A cell culture were generally higher than those of LCIS, indicating bulk water absorbed more THz waves than live cell monolayer. Through a fast Fourier transformation of the time-domain signals, we obtained the frequency-domain spectra (Fig. 2(b)). Consistent with the time-domain result, the amplitudes of the LCIS were lower than those of the MCF10A cells in the frequency domain ranging from 0.2 to 1.0 THz. A vacuum or a non-polar gas such as nitrogen is not suitable for living cell culture and measurement, thus strong water absorption at 0.558 THz were clearly observed by water vapor effects .
After calculation (detail see methods), the determined complex dielectric constant of MCF10A cells in the LCIS is compared with that of the LCIS. We found the dielectric constant (real part) of MCF10A cells was almost unchanged except for a slight decrease below 0.5 THz (Fig. 2(c)). In contrast, dielectric loss (imaginary part) of MCF10A cells showed a significant decrease from 0.3 to 0.9 THz. These disparities both in the dielectric constant (0.3-0.5 THz) and the dielectric loss (0.3-0.9 THz) were significant since they are greater than the analytical error bars. The tendency of complex dielectric constant of MCF10A below 1.0 THz is consistent with the result reported in a colon carcinoma cells (DLD-1) .
In order to obtain the THz dielectric response of MCF10A cells upon oxidative stress, high concentration of H2O2 (10mM) was added to the cell cultured in sample ATR prism after living cells measurement. High concentration of H2O2 of 10mM was added by two following reasons. First, H2O2 is rapidly degraded when added to the cells, only 1.6% remained after 0.1mM H2O2 added to cells 1 hr . Second, 10mM of H2O2 resulted in rapid induction of biomolecules  and provoked a massive cell death , which would suitable for short-term in situ monitoring of cells that cultured in atmospheric environment. The complex dielectric constants from experimental data fitting of control (LCIS + H2O2) and H2O2-treated cells are shown in Fig. 3. The dielectric constants were unchanged after exposure to H2O2 (Fig. 3(a) and close-up between 0.3 and 0.5 THz showed inset). In contrast, a slight increase of dielectric loss during time course of H2O2 exposing was observed (Fig. 3(b)). This increase had statistical significance between 0.3 and 0.9 THz (inset of Fig. 3(b) shows the close-up between 0.3 and 0.5 THz).
We further study whether the THz dielectric response of MCF10A cells upon oxidative stress is caused by cell death with fluorescent-labeled optical imaging. Annexin V is a 35-36 kDa Ca2+-dependent phospholipid-binding protein with high affinity for phosphatidylserine (PS), which is translocated from the inner to the outer leaflet of the plasma membrane in early apoptosis cells . Propidium iodide (PI) is a fluorescent membrane-impermeant dye that can stains the nuclei by intercalating between the stacked bases of nuclei acid . Since PI enters the cell only if the cell membranes become permeable and hence only label late apoptotic or necrotic cells. Utilization of fluorescein isothiocyanate conjugated Annexin V (Annexin V-FITC) and PI is a standard procedure to determine if cells are viable, apoptotic, or necrotic. After stained Annexin V and PI, cells were seed on a microscopy slide and mounted by a fluoroshield with DAPI, which can label both viable and dead cells, then observed by fluorescence microscope. Cells that are considered viable are both Annexin V and PI negative (white arrows indicated in Fig. 4(a)) whereas early apoptotic cells (which exposure PS but with intact plasma membrane) are Annexin V-positive and PI-negative (Fig. 4(b)). Both late apoptotic and necrotic cells are Annexin V+/PI + , their morphology can be discriminated by staining of nucleus with PI. The nuclei of late apoptotic cells are fragmented and condensed (Fig. 4(c)). In contrast, necrotic cells have uncondensed nuclei with prominent nucleoli (Fig. 4(d)).
Furthermore, we used flow cytometry to quantitative analysis the cell death induced by high concentrate H2O2. Four independent experiments in different days were taken and the results of all flow cytometric analysis can be found in Appendix. For flow cytometry assay, in the untreated (t = 0) samples, the majority of cells (93.29% ± 4.32) were viable (Annexin V-/PI-). In contrast, when cells were treated with 10 mM of H2O2, the percentage of viable cells at 1 hr and 2 hr was 81.17% ± 19.44 and 31.93% ± 12.01, respectively. Here, we chose one result from multiple measurements showing in Fig. 5. After 1hr of H2O2-treatment, membrane changes leading PS exposure occur rapidly in apoptosis, and the cell population shifts from the lower left quadrant to the lower right quadrant. After PS exposure to the outer leaflet of the cell membrane, the integrity of plasma membrane was lost, and the population shifts to the upper right quadrant. Notably, some cells also immediately move from the lower left quadrant to the upper right quadrant without passing though intermediate Annexin V+/PI- stage, indicated necrosis occurred simultaneously with apoptosis.
To investigate the relation between MCF10A cell death measured by flow cytometer and the THz dielectric responses of cells upon H2O2 treatment, the peak amplitude (PA) of the THz pulse in time-domain was calculated with time. We define a relative ∆rel and normalized ∆ experimental contrasts, which ∆ measures the cell content with previously used methods . We chose peak amplitude of the THz time-domain signal to calculate the contrast. We define a relative ∆rel in Eq. (6),Eq. (7) stands for its normalized value with respect to t = 0 value:
We found 1-∆ and death rate (including apoptosis and necrosis) of MCF10A cells showed a very good agreement (Fig. 6). There is a slow growth of normalized THz contrast during the first 1 hour, coincide with the slow increase of cell death rate. By comparison, a speed increase of THz contrast as well as a dramatic increase of cell death rate was observed during 1 hr to 2 hr. Since THz signal has been proven sensitive to the ions and proteins concentration in the cell [14, 25]. We speculate that cells were more likely undergo apoptosis during the first one hour, thus the permeability of the cell membrane had little influence. Then, with more and more cells dead, cytoplasm leakage occurred in late apoptotic and necrotic cells, which the integrity of the plasma and nuclear membranes decreases [26, 27], and result in a rapid increase of THz contrast.
Significantly, this work demonstrated that ATR THz-TDS allow noninvasive, label-free, real time and in situ monitoring of living cells in physiological conditions. The THz dielectric constants of living cells and the surrounding medium are distinguished below 1.0 THz. The THz dielectric responses and relative changes of peak amplitudes give access to dynamics of cell death when cell were exposed to H2O2. Measurement of H2O2-induced cell death in living cells has been demonstrated using MCF10A cells, but other cell lines such as PC-3 has also been successfully tested for other ongoing studies, whose results will be reported in separate publications. In summary, this method has been demonstrated to be a viable tool for monitoring cell viability during oxidative stress without the need for additional labeling. This method can be also applied for studying antioxidant activity of drugs.
Results of all flow cytometric analysis
Original data from four independent experiments in different days was shown in Fig. 7. Annexin V FITC vs PI with gates for four populations and the summary of assay results with statistics for H2O2-treated cells were automatically calculated in the report. Figure 5 is the result of experiment 2 (Exp-2). Then we calculated the cell death rate by add up Annexin V+/PI-, Annexin V+/PI+ and Annexin V-/PI+ cells together in each experiment and statistically analyzed with OrignPro 9.1 software (Table 1). Data represent mean ± SEM.
National Basic Research Program of China (973 Program, No. 2015CB755405); National Key R&D Program (No. 2016YFC0101003); National Science and Technology Ministry of Science and Technology Support Program (No. 2015BAI01B01); Postdoctoral Science Foundation of China (No. 2016M592706); National Natural Science Foundation of China (NSFC) (No. 61427814); Foundation of President of China Academy of Engineering Physics (No. 201501033).
The authors declare that there are no conflicts of interest related to this article.
References and links
1. U. Krengel and S. Törnroth-Horsefield, “Coping with oxidative stress,” Science 347(6218), 125–126 (2015). [PubMed]
2. T. R. Figueira, M. H. Barros, A. A. Camargo, R. F. Castilho, J. C. B. Ferreira, A. J. Kowaltowski, F. E. Sluse, N. C. Souza-Pinto, and A. E. Vercesi, “Mitochondria as a source of reactive oxygen and nitrogen species: from molecular mechanisms to human health,” Antioxid. Redox Signal. 18(16), 2029–2074 (2013). [PubMed]
3. D. R. Gough and T. G. Cotter, “Hydrogen peroxide: a Jekyll and Hyde signaling molecule,” Cell Death Disease 2, e213 (2011).
4. T. Finkel and N. J. Holbrook, “Oxidants, oxidative stress and the biology of ageing,” Nature 408(6809), 239–247 (2000). [PubMed]
5. S. Fulda, A. M. Gorman, O. Hori, and A. Samali, “Cellular stress responses: cell survival and cell death,” Int. J. Cell Biol. 2010, 214074 (2010). [PubMed]
6. D. Trachootham, W. Lu, M. A. Ogasawara, R. D. Nilsa, and P. Huang, “Redox regulation of cell survival,” Antioxid. Redox Signal. 10(8), 1343–1374 (2008). [PubMed]
7. O. Kepp, L. Galluzzi, M. Lipinski, J. Yuan, and G. Kroemer, “Cell death assays for drug discovery,” Nat. Rev. Drug Discov. 10(3), 221–237 (2011). [PubMed]
8. A. Gomes, E. Fernandes, and J. L. Lima, “Fluorescence probes used for detection of reactive oxygen species,” J. Biochem. Biophys. Methods 65(2-3), 45–80 (2005). [PubMed]
9. X. Yang, X. Zhao, K. Yang, Y. Liu, Y. Liu, W. Fu, and Y. Luo, “Biomedical applications of terahertz spectroscopy and imaging,” Trends Biotechnol. 34(10), 810–824 (2016). [PubMed]
10. G. J. Wilmink and J. E. Grundt, “Invited Review Article: Current state of research on biological effects of terahertz radiation,” J. Infrared Milli. TeraHz. Waves 32(10), 1074–1122 (2011).
11. J. Oin, Y. Ying, and L. Xie, “The detection of agricultural products and food using terahertz spectroscopy: a review,” Appl. Spectrosc. Rev. 48(6), 439–457 (2013).
12. P. U. Jepsen, D. G. Cooke, and M. Koch, “Terahertz spectroscopy and imaging – modern techniques and applications,” Laser Photonics Rev. 5(1), 124–166 (2011).
13. K. Shiraga, T. Suzuki, N. Kondo, K. Tanaka, and Y. Ogawa, “Hydration state inside HeLa cell monolayer investigated with terahertz spectroscopy,” Appl. Phys. Lett. 106, 253701 (2015).
14. M. Grognot and G. Gallot, “Quantitative measurement of permeabilizattion of living cells by terahertz attenuated total reflection,” Appl. Phys. Lett. 107, 103702 (2015).
15. K. Shiraga, Y. Ogawa, T. Suzuki, N. Kondo, A. Irisawa, and M. Imamura, “Characterization of dielectric responses of human cancer cells in the terahertz region,” J. Infrared Milli. Terahz. Waves 35, 493–502 (2014).
16. K. Shiraga, Y. Ogawa, T. Suzuki, N. Kondo, A. Irisawa, and M. Imamura, “Determination of the complex dielectric constant of an epithelial cell monolayer in the terahertz region,” Appl. Phys. Lett. 102, 053702 (2013).
17. Y. Zou, J. Li, Y. Cui, P. Tang, L. Du, T. Chen, K. Meng, Q. Liu, H. Feng, J. Zhao, M. Chen, and L. G. Zhu, “Terahertz spectroscopic diagnosis of myelin deficit brain in mice and rhesus monkey with chemometric techniques,” Sci. Rep. 7(1), 5176 (2017). [PubMed]
18. A. Soltani, L. Duschek, S. F. Busch, T. Probst, E. Castro-Camus, and M. Koch, “A novel accurate method for attenuated total reflection spectroscopy,” International Conference on Infrared, Millimeter, and Terahertz Waves, (2014)
19. K. Wehbe, J. Filik, M. D. Frogley, and G. Cinque, “The effect of optical substrates on micro-FTIR analysis of single mammalian cells,” Anal. Bioanal. Chem. 405(4), 1311–1324 (2013). [PubMed]
20. M. Exter, C. Fattinger, and D. Grischkowsky, “Terahertz time-domain spectroscopy of water vapor,” Opt. Lett. 14(20), 1128–1130 (1989). [PubMed]
21. N. Driessens, S. Versteyhe, C. Ghaddhab, A. Burniat, X. De Deken, J. Van Sande, J. E. Dumont, F. Miot, and B. Corvilain, “Hydrogen peroxide induces DNA single- and double-strand breaks in thyroid cells and is therefore a potential mutagen for this organ,” Endocr. Relat. Cancer 16(3), 845–856 (2009). [PubMed]
22. S. Y. Sung, H. Kubo, K. Shigemura, R. S. Arnold, S. Logani, R. Wang, H. Konaka, M. Nakagawa, S. Mousses, M. Amin, C. Anderson, P. Johnstone, J. A. Petros, F. F. Marshall, H. E. Zhau, and L. W. Chung, “Oxidative Stress Induces ADAM9 Protein Expression in Human Prostate Cancer Cells,” Cancer Res. 66(19), 9519–9526 (2006). [PubMed]
23. D. Wlodkowic, W. Telford, J. Skommer, and Z. Darzynkiewicz, “Apoptosis and beyond: cytometry in studies of programmed cell death,” Methods Cell Biol. 103, 55–98 (2011). [PubMed]
24. D. V. Krysko, T. Vanden Berghe, K. D’Herde, and P. Vandenabeele, “Apoptosis and necrosis: Detection, discrimination and phagocytosis,” Methods 44(3), 205–221 (2008). [PubMed]
25. J. B. Masson, M. P. Sauviat, J. L. Martin, and G. Gallot, “Ionic contrast terahertz near-field imaging of axonal water fluxes,” Proc. Natl. Acad. Sci. U.S.A. 103(13), 4808–4812 (2006). [PubMed]
26. G. Kroemer, B. Dallaporta, and M. Resche-Rigon, “The mitochondrial death/life regulator in apoptosis and necrosis,” Annu. Rev. Physiol. 60, 619–642 (1998). [PubMed]
27. G. Denecker, D. Vercammen, W. Declercq, and P. Vandenabeele, “Apoptotic and necrotic cell death induced by death domain receptors,” Cell. Mol. Life Sci. 58(3), 356–370 (2001). [PubMed]