Potential Biomarker for Triple-Negative Breast Cancer Invasiveness by Optical Redox Imaging

Abstract

Predicting tumor metastatic potential remains a challenge in cancer research and in clinical diagnosis. Cancer invasion to neighboring tissues is a significant event in cancer progression to metastasis. Optical redox imaging (ORI) is based on detecting the endogenous fluorescence signals of reduced nicotinamide adenine dinucleotide (NADH) and oxidized flavin adenine dinucleotide (FAD). Previously, we found that ORI can discriminate between cancer and normal tissue specimens from clinical breast cancer patients and can differentiate the relative invasiveness of melanoma and breast tumors. In this study, we aimed to identify ORI biomarkers to differentiate the invasiveness of four triple-negative breast cancer cell lines (TNBC). Using a fluorescence microscope, we acquired NADH and FAD fluorescent signals from cultured MDA-MB-231, MDA-MB-436, HCC1806, and MDA-MB-468 cells. We found that (1) the redox ratio, FAD/(NADH+FAD), differentiated the four TNBC lines; (2) there was a significant difference of invasive potential between MDA-MB-231 and the other three TNBC lines measured by the transwell invasion assay; and (3) there was a positive logarithmic correlation between the redox ratio and the invasive potential, where the most invasive MDA-MB-231 cells had the highest redox ratio and the least invasive MDA-MB-468 cells had the lowest redox ratio. These results suggest that the redox ratio can potentially be used as a biomarker for TNBC invasiveness and prognosis.

39.1. Introduction

In 2018, there were about 2.1 million newly diagnosed female breast cancer cases worldwide, and breast cancer is the leading cause of cancer death for female in over 100 countries [1]. It was estimated that 20–30% of patients with node-negative breast cancer will ultimately develop metastatic disease with poor prognosis [2]. Furthermore, the current diagnostic methods cannot accurately predict the metastatic potential of individual tumors. Triple-negative breast cancer (TNBC) is a highly aggressive form of breast cancer and has a poorer prognosis compared to other forms of breast cancer [3]. TNBC is also heterogeneous in many aspects, including its invasiveness. Invasion is one of the key steps of cancer metastasis. Thus, to improve TNBC patient survival, it is important to identify novel biomarkers that can more accurately characterize TNBC invasiveness.

Optical redox imaging (ORI), pioneered by Britton Chance et al., is based on detecting the endogenous fluorescence signals of reduced nicotinamide adenine dinucleotide (NADH) and oxidized flavin adenine dinucleotide (FAD) [4–7]. Recently, ORI has been employed in many cancer studies such as discrimination of cancer aggressiveness [8–10], differentiation among receptor status of breast cancers [11], distinction between clinical biopsies of breast cancer and normal tissues [12], and monitoring therapeutic effects [13, 14]. Previously, Sun et al. showed that the redox ratio differentiated among four breast cancer lines of different molecular subtypes in culture, with the most invasive lines showing the largest redox ratio [15]. In this study, we focused on investigating four TNBC cell lines: MDA-MB-231, MDA-MB-436, HCC1806, and MDA-MB-468. These four TNBC lines represent different molecular subgroups within TNBCs [16] and have different invasive potential. Thus, we selected them as our initial models for the discovery of potential redox biomarkers of invasiveness. We performed ORI of these lines and measured their invasive potentials using the Boyden chamber method. We tested the hypothesis that the redox ratio correlates with the invasive potential of these lines.

39.2. Materials and Methods

TNBC cell lines MDA-MB-231, MDA-MB-436, HCC1806, and MDA-MB-468 were cultured in RPMI 1640 medium (phenol red-free, 11.1 mM glucose, 2.1 mM glutamine, Gibco^® catalog no. 11835030) supplemented with 10% FBS and 1% penicillin/strep and maintained at 37 °C and 5% CO. Cells (4 × 10⁴ 2, 200 μl) were seeded on 35 mm glass bottom dishes (MatTek, part no. P35GC-1.5-14-C) and incubated for 24 hours as previously described [17]. Approximately 1 hour before imaging, the medium was changed to live cell imaging solution (LCIS, Life Technology^®) supplemented with 11 mM glucose and 2.1 mM L⁻ glutamine.

NADH and FAD fluorescence images were obtained under 37 °C with a DeltaVision Deconvolution Microscope System that consists of a xenon lamp as the excitation light source, a 12-bit CCD, an objective of 40X/0.95 NA (image matrix size 512 × 512, bin 2 × 2, and pixel size 0.32 μm), proper excitation/emission filters (NADH: 360 ± 20/455 ± 25 nm; FAD: 470 ± 20/520 ± 20 nm), and an exposure time of 3 s. Three to five fields of view (FOV) were taken for each dish. A customized MATLAB^® program was used to extract NADH and FAD signals. Technical details can be found in our previous work [13, 17]. The mean values of FAD, NADH, and redox ratio were averaged across FOVs and then across dishes to obtain the group means and their standard errors (SE).

The Boyden chamber method was employed to measure the invasive potential (IVP) of the TNBC cells as previously described [17, 18]. Each upper chamber of a transwell insert (Corning, catalog No. 3422) was coated with 20 μl Matrigel (Corning, catalog no. CB356238) in a 1:1.5 dilution of Matrigel by FBS-free RPMI 1640 medium. Cells were cultured in serum-free RPMI media for 24 hours and seeded at a density of 5 × 10⁴ cells with 300 μL FBS-free RPMI 1640 medium in the upper chamber. The lower chamber was filled with 500 μL complete RPMI 1640 cell culture medium (containing 10% FBS as chemotactic stimulus). The chambers were incubated at 37 °C for 20 hours. The invaded cells were fixed in methanol and stained with DAPI (300 nM) and tile imaged with a fluorescence microscope to cover the whole area of the chamber membrane. The total numbers of invading cells were quantified using ImageJ (NIH, United States). Triplicate wells were used in each experiment, and the experiment was repeated three times.

One-way ANOVA with post-hoc Tukey multiple comparisons was performed with Prism 8 (GraphPad^®) for statistical analysis of the group differences, and p < 0.05 was considered statistically significant.

39.3. Results

Figure 39.1 shows the typical redox images for each cell line. The quantitative analysis showed that MDA-MB-231 had the highest FAD intensity while HCC1806 had the lowest FAD intensity among the four TNBC cell lines (Fig. 39.2). On the other hand, MDA-MB-468 had the highest NADH intensity and was significantly different from the other three TNBC cell lines (p < 0.001). Furthermore, statistical analysis indicated significant differences in redox ratios among four lines except between HCC1806 and MDA-MB-468. The rank order of the redox ratio from the highest to the lowest was found as follows: MDA-MB-231 > MDA-MB-436 > HCC18 06 > MDA-MB-468.

Fig. 39.1.

Typical redox images of FAD, NADH, and redox ratio FAD/(NADH+FAD) for (a) MDA-MB-231, (b) MDA-MB-436, (c) HCC1806, and (d) MDA-MB-468. The color bars of the FAD and NADH images indicate the signal intensity in arbitrary units. Note that FAD color bars were scaled differently for better displaying of all cells. The color bars of the redox ratio images indicate a range from 0 to 1

Fig. 39.2.

Optical redox imaging quantifications. The left y-axis indicates NADH and FAD intensity in arbitrary units, and the right y-axis indicates FAD/(NADH+FAD) redox ratio. Results are reported as mean ± SE. *p < 0.05, **p < 0.01, ***p < 0.001, n = 4–9 dishes

Based on the total number of cells that invaded through Matrigel-coated membrane, we found that MDA-MB-231 had a significantly higher invasive potential than the other three TNBC cell lines (p < 0.001) (Fig. 39.3a). The IVP of the other three lines were not significantly different. The rank order of the four cell lines from the most invasive to the least invasive was same as the rank order of the redox ratios. A positive linear correlation was found between logarithm of IVP, ln(IVP), and the redox ratio with a borderline significance (Fig. 39.3b; R² = 0.86; p = 0.07).

Fig. 39.3.

IVP quantification and its correlation with ORI of 4 TNBC lines. (a) IVP normalized to that of MDA-MB-231 cells, n = 6 – 15 wells. (b) Correlation of the logarithm of IVP versus the redox ratio FAD/(NADH+FAD). A trend line, y = 15x − 7.2, was generated with a correlation coefficient of R² = 0.86. Results are reported as mean ± SE. ***p < 0.001

39.4. Discussion and Conclusions

In this study, we showed that the redox ratios obtained by ORI can differentiate among four TNBC cell lines with different invasiveness. The invasive potential results were consistent with previous studies [15, 19, 20]. Both the redox ratios and the invasive potentials of the four lines exhibited the same rank order, despite the insignificant differences among the invasion of three less invasive TNBC lines. We also found a logarithmic correlation between the invasive potentials and the redox ratios with a borderline significance.

However, this study had several limitations. First, this study has a limited number of TNBC cell lines. Second, redox status may change with metabolic microenvironments such as nutrition levels [15]. Third, the O₂ level in cell culture is quite different from that of tumors in vivo. More studies need to be done in vitro under various microenvironments (including hypoxia) or in tumors in vivo to confirm and generalize the results of this study.