Quantitative analysis of interactive behavior of mitochondria and lysosomes using structured illumination microscopy

Abstract

Super-resolution optical microscopy has extended the spatial resolution of cell biology from the cellular level to the nanoscale, enabling the observation of the interactive behavior of single mitochondria and lysosomes. Quantitative parametrization of interactions between mitochondria and lysosomes under super-resolution optical microscopy, however, is currently unavailable, which has severely limited our understanding of the molecular machinery underlying mitochondrial functionality. Here, we introduce an M-value to quantitatively investigate mitochondria and lysosome contact (MLC) and mitophagy under structured illumination microscopy. We found that the M-value for an MLC is typically less than 0.4, whereas in mitophagy it ranges from 0.5 to 1.0. This system permits further investigation of the detailed molecular mechanism governing the interactive behavior of mitochondria and lysosomes.

1. Introduction

The crosstalk between mitochondria and lysosomes is involved in many cellular processes [1]. For instance, mitophagy, a process that selectively removes redundant or damaged mitochondria, plays an important role in regulating the number of intracellular mitochondria and maintaining mitochondrial functions [2]. Dysregulated mitophagy is implicated in many diseases, such as neurodegenerative diseases and cancer [3,4]. To date, mitophagy has often been studied at the cellular level through methods such as flow cytometry [5], enzyme-linked immunosorbent assay [6], western-blot [7], and confocal fluorescence microscopy [8,9]. These methods only report the cumulative level of mitophagy and ignore individual mitophagy events from the fusion between individual mitochondria and lysosome pair [10]. Although confocal fluorescence microscopy can detect mitophagy using mitophagy-specific dye, it is difficult to distinguish subcellular structures at a resolution beyond 200 nm [11,12]. Moreover, confocal fluorescence microscopy does not provide details on the interactive behavior of individual mitochondria and lysosome pairs. Thus, a novel strategy is needed to capture detailed information on the crosstalk between individual mitochondria and lysosome pairs.

Recently developed super-resolution microscopy (SRM) techniques such as stimulated emission depletion (STED) [13], structured illumination microscopy (SIM) [14,15], and stochastic optical reconstruction microscopy (STORM) [16], as well as other single-molecule super-resolution imaging techniques [17,18], have provided new tools for investigating interactions between organelles at the subcellular level [19]. The interaction between mitochondria and lysosomes has been brought into focus using these techniques. For example, SIM has revealed a new type of interaction: mitochondria and lysosome contact (MLC) [1,20]. MLC is defined as contact between mitochondria and lysosome in the membrane formed at close range (~10 nm), allowing them to communicate in the dynamics process from contact to separation in healthy cells [1,20]. Unfortunately, a parametrization system for the assessment of interactive behaviors such as MLC and mitophagy under SRM is currently unavailable.

To fill this gap, in this toolbox paper we propose an analysis system, M-value, to quantitatively analyze the crosstalk between mitochondria and lysosomes at the subcellular level. The M-value in MLC is less than 0.4, whereas the M-value in mitophagy ranges from 0.5 to 1.0. Thus, this M-value system provides a robust platform to quantitatively analyze the interactive behavior of subcellular organelles under super-resolution microscopy.

2. Materials and methods

2.1. Materials

Mito-Tracker Green (#M7514, MTG) and Lyso-Tracker Red (#L12492, LTR) were obtained from Invitrogen (Eugene, Oregon, USA). Autophagosome dye (#NY561, DAPGreen) and mitophagy dye (#MD01-10, Kumamoto, Japan) were obtained from Dojindo Laboratories. Carbonyl cyanide 3-chlorophenylhydrazone (#045200, CCCP) was obtained from Thermo Fisherscientific (Grand Island, NY, USA). Penicillin-streptomycin (#15140163, 10,000 units/ml), fetal bovine serum (#26140079, FBS), Dulbecco’s modified Eagle’s medium (#11965118, DMEM), and other cell culture reagents were obtained from Gibco BRL (Grand Island, NY, USA). Primary and secondary antibodies used in this study were GAPDH (#5174, Cell Signaling Technology, Beverly, MA, USA), FIP200 (#12436, Cell Signaling Technology, Beverly, MA, USA), ATG13 (#13273, Cell Signaling Technology, Beverly, MA, USA), and HRP-linked anti-rabbit IgG (#7074, Cell Signaling Technology, Beverly, MA, USA). HeLa cells were a generous gift from Dr. Carolyn M. Price’s lab (University of Cincinnati). SLC-80 cells were isolated from healthy human fibroblasts by Dr. Taosheng Huang (Cincinnati Children’s Hospital, Cincinnati, OH, USA).

2.2. Cell culture

Cells were cultured in DMEM supplemented with 10% FBS, penicillin-streptomycin (100 units/ml) in a 5% CO₂ humidified incubator at 37 °C.

2.3. Live cell labeling

Cells were incubated with 100 nM MTG for 30 min, further co-incubated with 200 nM LTR at 37 °C for another 30 min in fresh DMEM, washed three times with fresh DMEM, and observed using a fluorescence microscopy confocal laser scanning microscopy or OMX 3D-SIM super-resolution microscope.

2.4. Confocal laser scanning microscopy

The images were obtained using an LSM-710 confocal laser scanning microscope (Carl Zeiss, Inc.) equipped with a 63 × /1.49 numerical aperture oil-immersion objective, and they were analyzed with ZEN 2012 (Carl Zeiss, Inc.) and ImageJ software (National Institutes of Health). All fluorescence images were analyzed with ImageJ software (https://imagej.nih.gov/ij/).

2.5. OMX 3D-SIM imaging and analysis

A total of 2 × 10⁵ cells were seeded on a glass-bottom microwell dish and incubated with 2 ml of DMEM medium supplemented with 10% FBS for 24 h. After treatment, the cells were washed three times with pre-warmed fresh DMEM medium, stained with 100 nM MTG for 30 min, co-incubated with 200 nM LTR at 37 °C for another 30 min, and washed with fresh DMEM three times. Finally, cells were cultured in a phenol-free medium (#1894117, Gibco, Grand Island, NY, USA) and observed under an OMX 3D-SIM super-resolution microscope (Bioptechs, Inc) equipped with an Olympus 100 × /1.49 numerical aperture oil-immersion objective lens and solid-state lasers.

SIM images were analyzed with ImageJ software equipped with a fluorescence intensity measurement tool. The intensity profile along the perpendicular (white solid lines, Fig. S1) of mitochondria-lysosome contact (white dash lines, Fig. S1) was used for the M-value calculation. For co-localization, Pearson correlation coefficient (PCC, the degree of overlap between two fluorescent channels, pixel-based) was analyzed using ImageJ software equipped a colocalization analysis plugin. For more information, please refer to https://imagejdocu.tudor.lu/plugin/analysis/colocalizationfinder/start#colocalization_finder.

2.6. CRISPR/Cas9-mediated knockout of FIP200 and ATG13 in HeLa cells

The pX458 plasmid (pSpCas9(BB)-2A-GFP; Addgene) was used as the cloning backbone for expressing sgFIP200 and sgATG13. Two complementary oligos for each sgRNA were denatured, annealed, and ligated into linearized pX458 vector digested by Bbsl (New England Biolabs). Empty constructs and pooled pX458-sgRNA were transiently transfected into HeLa cells, respectively, using Lipofectamine 3000 (Invitrogen). After 48 h the transfected cells were sorted based on the fluorescence of GFP (reporter) using a FACSAria cytometer (BD Biosciences). Individual sorted cells were cultured in a 96-well plate and subjected to Western blot analyses. At least three different clones were pooled for functional experiments. The sgRNA sequences targeting FIP200 and ATG13 were based on the published literature using CRISPR/Cas9 library [21,22]. The sgRNA sequences of FIP200 and ATG13 are listed below:

• sgFIP200: CAGGTGCTGGTGGTCAATGG

• sgATG13-1: TCACCCTAGTTATAGCAAGA

• sgATG13-2: CAGTCTGTTGTACACCGTGT

• sgATG13-3: GACTGTCCAAGTGATTGTCC

2.7. Western-blot assay

Cells were cultured in 3.5 cm diameter plates (80–90% confluence), washed by PBS buffer, and lysed for 15 min on ice using a RIPA buffer (#C2978, Sigma, St. Louis, MO, USA) containing an anti-protease mix (#PI78415, Thermo Scientific, Waltham, MA, USA). Protein concentration was measured by BCA assay (#23225, Thermo Scientific, Waltham, MA, USA). Equal amounts of proteins were subjected to SDS-PAGE and immunoblotting as described previously [23].

2.8. Electron microscopy

Cells were removed from the Petri dish with a flat scraper and collected by centrifugation at 1,000g. The sample was fixed in Karnovsky, then fixed in 2% osmium tetroxide, dehydrated in a series of graded ethanol, and embedded in Araldite. The sections were cut using the LKB for ultra-thin sections and collected on a Formvar-coated grid. Sections of 70 nm thick were stained with uranyl acetate and lead citrate and evaluated in TEM (JEM 100CX II, Tokyo, Japan) with an acceleration voltage of 80 kV.

3. Results and discussion

3.1. MLC in living HeLa cells under SIM

To investigate MLC events in living HeLa cells using super-resolution imaging, we first incubated cells with a commercially available mitochondrial probe (Mito-Tracker Green, MTG, 100 nM) and a lysosomal probe (Lyso-Tracker Red, LTR, 200 nM) for 30 min, and we then observed mitochondria (green) and lysosomes (red) under a SIM (Fig. 1A). The intracellular mitochondria showed a spherical, rod-shaped, or filamentous arrangement in an irregular manner at a resolution of approximately 150 ± 24.67 nm (n = 10) (Fig. S2). Mitochondria are approximately 0.3 pm in width and 0.6–10.0 pm in length (Fig. S3), which is consistent with our previous results [12]. In particular, there were more filamentous mitochondria than spherical and rod-shaped ones, a benchmark of healthy cells. In contrast, the diameter of spherical lysosomes (red color) was 0.6 ± 0.16 pm (n = 200) (Fig. S4). Both MTG and LTR allow for high-resolution staining under SIM, which enabled us to capture and understand the dynamic process of MLC (Fig. 1A and B). MLC events occur frequently in living cells (Fig. 1B) with a dynamic process from contact to separation (Fig. 1B, white arrows). In addition, we observed that mitochondria were surrounded by multiple lysosomes (Fig. 1C). Various types of MLC events [1,12], such as point contact (Figs. 1E–7), extended contact (Figs. 1E–8), and surrounding contact (Figs. 1E–9) were also observed.

Fig. 1.

Whole cell quantitative analysis of mitochondrial-lysosome contact (MLC) in living cells. (A) SIM image of mitochondria (green) and lysosomes (red). 1–11 represent MLC events. (B) A representative dynamic process of MLC in a living HeLa cell. White solid lines indicate the fluorescence intensity shown in (D). (C) Partial enlargement of (A). White solid lines indicate the fluorescence intensity shown in (G). (E) and (H) represent MLC events. White solid lines indicate the fluorescence intensity shown in (F). (I) Schematic diagram of calculation formula for the M-value of MLC. (J) M-value range of MLC dynamic events. Scale bars: A 5.0 μm; B, C, E, and H 0.5 μm; insets of A 0.5 μm. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

3.2. Quantitative analysis of whole-cell MLC

Recently, it has been reported that dysregulated contact sites formed by mitochondria and lysosomes are linked to Parkinson’s disease [1,24]. We reason that proteins involved in MLC will provide a new perspective and drug targets for the treatment of this disease. A quantitative analysis platform to analyze MLC at the subcellular level would therefore allow us to understand the biological functions and evaluate the therapeutic effects of different drugs. However, the traditional co-localization quantitative parameters (such as PCC and Mander’s coefficients) report quantification values based on pixels. These pixels contain the fluorescence scattering caused by the probe (Fig. 1J), which does not reflect the real contact between two organelles. Here, we introduce a quantitative analysis system, M-value, which is derived from the full-width at half-maximum (FWHM) of organelle images, for understanding MLC events at the subcellular level. FWHM refers to the full width of the image at half-maximum value and can directly reflect the resolution of the image. MTG-stained mitochondria and LTR-stained lysosomes result in yellow spots when they merge (Fig. 1H). Herein, we use the following formula to define the M-value of the merge:

$$M=\frac{Y}{(R_{\mathrm{FHWM}},G_{\mathrm{FHWM}})\min }$$

R_FWHM indicates the FWHM of red (lysosome); G_FWHM indicates the FWHM of green (mitochondria); Y indicates the merging distance between R_FWHM and G_FWHM, and (R_FWHM, G_FWHM) _min indicates the minimum value in R_FWHM or G_FWHM.

We will demonstrate that the M-value is a simple and well-defined number that can be used to quantitatively assess MLC events in live cells. We first observed that a typical MLC event (Fig. 1B, D) generates an ensemble of M-values below 0.4. We hypothesize that an M-value of 0.4 can be the upper bound to characterize dynamic MLC events. To support this idea, we calculated the M-values of a large number of MLC events (n = 100) (Fig. 1C, E, F, G, and Figs. S5–9) and found that the M-values of all MLC events were indeed below 0.4 (Fig. 1J).

3.3. Quantitative analysis of whole-cell mitophagy

Mitophagy has been commonly studied by co-localizing fluorescently stained mitochondria and autophagosomes with confocal microscopy in cells [25,26]. However, this strategy is not suitable for the quantitative analysis of mitochondria-lysosome interactions at the subcellular level. To address this issue, we set out to use the M-value to quantitatively analyze mitophagy.

Mitophagy was triggered by the incubation of HeLa cells with 10.0 μM carbonyl cyanide m-chlorophenyl hydrazone (CCCP), a common mitochondria damage inducer [12], for 12 h (Fig. 2A). Treatment with CCCP broke mitochondria into spherical shapes of varying sizes compared to untreated cells, which primarily showed filamentous mitochondria (Fig. 1A). In addition, CCCP treatment induced more frequent mitophagy (Frame 1–3 in Fig. 2A), as evidenced by the appearance of more yellow spots resulting from fusion between mitochondria (green) and lysosomes (red) (Fig. 2A inset 1–3 and Fig. 2B). This result indicates that more mitophagy occurs under pathological conditions.

Fig. 2.

Quantitative analysis of fusion between mitochondria and lysosome in CCCP-treated cells. (A) Frame 1–3 of mitochondrial and lysosome fusion events. White solid lines indicate fluorescence intensity shown in (C). (B) 3D-SIM surface plot of representative fusion between mitochondria and lysosome. (D) Representative fusion events in (A). White solid lines indicate fluorescence intensity shown in (E). Scale bars: A 5.0 μm; B and D 0.5 μm; inset of A 0.5 μm.

We next calculated the M-value of the mitophagy after the cell was exposed to CCCP for 12 h (Fig. 2A blue enlargement and Fig. 2D). The M-value of the mitophagy event is greater than 0.5, and some even reach close to 1.0 (Fig. 2C, E). To test whether the M-value can be applied to all mitophagy events, we performed quantitative analyses of more mitophagy events (n = 100) (Figs. S10–13). The M-values in all mitophagy events consistently range from 0.5 to 1.0 (Fig. 3A). Thus, the M-value platform provides a robust system to differentiate MLC (0–0.4) and mitophagy (0.5–1.0) in physiological and pathological conditions. This platform is also suitable for other cell lines such as SLC-80 cells (Fig. S14).

Fig. 3.

M-value of WT cells with and without CCCP treatment. (A) M-value distribution of 100 MLC events in normal cells and 100 mitophagy fusions in CCCP-treated cells. (B) A MLC event at different angles. The solid white line indicates the fluorescence intensity shown in (C). (D) A mitophagy at different angles. The white solid line indicates the fluorescence intensity shown in (E). (F) Cells incubated with mitophagy dye with or without CCCP treatment. Scale bars: B 0.5 μm, F 5.0 μm.

To verify the applicability of using the M-value for MLC and mitophagy, we next applied the M-value to quantitative analysis of an MLC (Fig. 3B, C) and a mitophagy event (Fig. 3D and E) at different rotation angles (0°-180°). We found that the M-value for MLC and mitophagy is consistent at different observation angles. Finally, CCCP-induced mitophagy was confirmed with a commercially available mitophagy detection dye [27] that generates higher red fluorescence compared with untreated cells (Fig. 3F), as well as electron microscopy showing a representative structure of mitochondrial engulfment in autolysosomes (Fig. S15).

3.4. Comparison of interactive behavior of mitochondrial and lysosome using epi-illumination fluorescence microscopy; confocal microscopy; and SIM

To demonstrate the advantages of SIM, we compared the performance of epi-illumination fluorescent microscopy, confocal microscopy, and SIM in resolving mitochondria-lysosome interaction using the same staining process with MTG and LTR (Fig. 4). Mitochondria and lysosomes appeared as green and red plaques in untreated cells under an epi-illumination fluorescence microscope, and a large area of yellow plaque was generated in the overlaid image (Fig. 4A). One cannot capture any information about MLC events at this spatial resolution, which may even cause a misinterpretation of mitophagy in untreated cells. Fig. 4B showed the morphology of mitochondria and lysosomes under a confocal microscope. Although confocal microscopy successfully located mitochondria and lysosomes at the subcellular level, its spatial resolution (~500 nm, Fig. S16) is not enough for the quantitative investigation of the detailed mitochondria-lysosome interaction below 100 nm compared with SIM (Fig. 4C).

Fig. 4.

Comparison of interactive behavior of mitochondrial and lysosome using fluorescence microscopy, confocal microscopy, and SIM. Mitochondrial and lysosome were stained with MTG and LTR. Images of untreated cells under fluorescence microscopy (A), confocal microscopy (B), and SIM (C). Images of CCCP-treated cells under fluorescence microscopy (D), confocal microscopy (E), and SIM (F). Scale bars: 1.0 μm.

In order to further investigate the mitochondria-lysosome interaction under pathological conditions using different optical microscopy techniques, we incubated 10.0 μM CCCP with HeLa cells for 12 h. As expected, neither epi-illumination fluorescence microscopy (Fig. 4D) nor confocal microscopy (Fig. 4E) was able to clearly capture mitophagy at the subcellular level. In contrast, SIM provides excellent imaging quality that can resolve the mitochondria-lysosome interaction (Fig. 4F). We also compared M-value with PCC value in quantifying MLC in SIM images (Fig. S17). Value 1 indicates that there must be lysosomes where there are mitochondria; value −1 indicates that there is no lysosome overlap with mitochondria; and value 0 indicates the random distribution of mitochondria and lysosomes [28]. Results showed that the pixel-based PCC value was higher than the M-value, indicating that the M-value is more accurate for investigating MLC.

3.5. Application of M-value to ATG13 and FIP200 knockout cells

Since the M-value system is feasible to quantitatively separate MLC (< 0.4) and mitophagy (0.5–1.0), we then applied M-value to analyze the crosstalk between mitochondria and lysosomes in ATG13 knockout (ATG13 KO) and FIP200 knockout (FIP200 KO) HeLa cells. Both ATG13 and FIP200 play important roles in the process of autophagy and mitophagy (Fig. 5A) [29]. ATG13 and FIP200 are important components of ULK1 complex that play an important role in maintaining the stability and kinase activity of ULK1. When nutrients are deficient, the activity of mTORCl is inhibited, and ULK1 and ATG13 are rapidly dephosphorylated, leading to the activation of ULK1 kinase to induce the occurrence and development of autophagy [30].

Fig. 5.

ATG13/FIP200 knockout cells under SIM. (A) Illustration of the role of ATG13 and FIP200 in mitophagy. (B) Western blot for detecting the protein expression of ATG13 and FIP200. Images of untreated and CCCP-treated ATG13 KO cells (C) and FIP200 KO cells (D). (E) Representative M-value calculations indicate MLC events in ATG13KO cells (1,2) and FIP200 KO cells (3,4) with or without CCCP treatment. Scale bars: C 5.0 μm, E 0.5 μm.

Next, the endogenous ATG13 and FIP200 in wild-type (WT) cells were knocked out with a CRISPR/Cas9 gene editing assay (Fig. 5B). The ATG13 KO and FIP200 KO cells were then exposed to media with or without 10.0 μM CCCP for 12 h. In untreated ATG13 KO cells (Fig. 5C), mitochondria showed a continuous rod shape, and they interacted with lysosomes as MLC (M-value, 0.283) (Figs. 5E–1). After CCCP treatment, damaged mitochondria appeared as granules of various sizes (Fig. 5C) similar to those in CCCP-treated WT cells (Fig. 2A). MLC (M-value, 0.241) also occurred under CCCP treatment (Figs. 5E–2). In contrast, compared to CCCP-treated WT cells (Fig. 2A), mitophagy events in CCCP-treated ATG13 KO cells significantly decreased. In FIP200 KO cells, no mitophagy was observed before or after CCCP-treatment (Fig 5D, E-3, E-4) (Fig. S18), whereas MLC still occurred. These results demonstrated that ATG13 and FIP200 are required for mitophagy. The M-value system is applicable to distinguish the two events in autophagy-defective cells.

4. Conclusion

Mitochondria are important organelles for energy conversion in eukaryotic cells and are involved in a variety of biological processes such as intracellular homeostasis, proliferation, senescence, and death [31,32]. At the same time, functions of mitochondria are regulated by their interactions with lysosomes through processes such as MLC and mitophagy [1,33]. MLC is widely found in healthy cells, and it was recently discovered that RAB7 GTP hydrolysis plays a key role in the MLC process [1,34]. Mitophagy is a selective type of autophagy [35] that regulates the number of mitochondria in cells and maintains normal mitochondrial function [3,36]. Herein, the interactive behaviors of individual mitochondria and lysosome pairs were observed under SIM. Compared to conventional tools such as epi-illumination and confocal fluorescence microscopy, SIM offers improved spatial resolution that captures detailed information regarding the dynamic mitochondria-lysosome interactions in live cells.

Another bottleneck that limits our understanding of mitochondria-lysosome interaction is the lack of a quantitative analysis platform for data generated from SIM. To address this challenge, we developed an M-value platform based on the FWHM of organelle images. This platform allows us to successfully differentiate MLC from mitophagy because their M-values are bounded within different ranges: an M-value below 0.4 implies MLC, and an M-value ranging from 0.5 to 1.0 implies mitophagy. In addition, this platform can be used to analyze mitochondria and autophagosomes interactions (Fig. S19). Thus, the M-value quantitative analysis system based on the super-resolution microscopy technique opens a new avenue to adopt automatic image analyzing software for high-throughput study of subcellular organelle interactions.

We envision that the M-value platform will specifically benefit high-throughput drug screening as an evaluation index of pharmacodynamics. Currently, high-throughput drug screening mainly relies on reporter gene systems, fluorescent labeling detection, micro-chemical technology, and fluorescence imaging technology [37]. Among them, the combination of fluorescence imaging and automatic analysis is an excellent screening strategy for micro-, sub-cellular image analysis in living cells [38]. However, the confocal microscopy systems commonly used in these high-throughput screening systems cannot provide sufficient spatial resolution to capture organelle interactions in live cells [39,40]. Combining artificial intelligence and SIM imaging with the M-value system will significantly retrench the process of early drug discovery and shorten the drug screening cycle to promote the success rate of novel drug discovery.