Nanoscopic Quantication of Sub-mitochondrial Morphology, Mitophagy and Mitochondrial Dynamics in Patients With Mitochondrial Disease

SLC25A46 mutations have been found to lead to mitochondrial hyper-fusion and reduced mitochondrial respiratory function, which results in optic atrophy, cerebellar atrophy, and other clinical symptoms of mitochondrial disease. However, it is generally believed that mitochondrial fusion is attributable to increased mitochondrial oxidative phosphorylation (OXPHOS)[1], which is inconsistent with the decreased OXPHOS of highly-fused mitochondria observed in previous studies. In this paper, we have used the live-cell nanoscope to observe and quantify the structure of mitochondrial cristae, and the behavior of mitochondria and lysosomes in patient-derived SLC25A46 mutant broblasts. The results show that the crista have been markedly damaged in the mutant broblasts, but that there is no corresponding increase in mitophagy. This study suggests that severely damaged mitochondrial cristae might be the predominant cause of reduced OXPHOS in SLC25A46 mutant broblasts. This study demonstrates the utility of nanoscope-based imaging for realizing the sub-mitochondrial morphology, mitophagy and mitochondrial dynamics in live cells, which may be particularly helpful for the quick assessment and diagnosis of mitochondrial abnormalities.


Background
The mitochondron is the cellular organelle which is most intimately associated with energy metabolism in mammals and most other eukaryotes. Mitochondrial dysfunction caused by nuclear DNA (nDNA) or mitochondrial DNA (mtDNA) defects lead to cellular respiratory chain and energy metabolism disorders, resulting in a group of multi-system diseases [1,2]. A number of mitochondrial diseases present aberrant mitochondrial morphology, including mitochondrial fragmentation or excessive mitochondrial fusion, which have an effect on mitochondrial function, leading to dysfunction of vital organs and tissues and accordingly threatening patients' health and survival [3][4][5].
The confocal microscope and transmission electron microscope are commonly used to observe mitochondrial morphology. However, a major shortcoming of the confocal microscope is that its spatial resolution is not high enough to visualize and quantitatively calculate the structure of sub-mitochondria, which is vital for examining the sub-mitochondrial damage that may affect mitochondrial function [6].
Transmission electron microscopy, however, is time-consuming, expensive, and unable to observe the variations of live-cell mitochondrial morphology in a dynamic manner. The spatial resolution of 100-120 nm achieved by the structured illumination microscopy (SIM) is su cient to observe the submitochondrial structure in living cells [7].
In this paper, we have set out to take advantage of the live-cell nanoscope − 3D-SIM to dynamically observe the sub-mitochondrial morphology originating from the broblasts of a patient carrying biallelic mutations in SLC25A46. Combined with the sub-mitochondrial structure identi cation/quanti cation and mitochondria-lysosome interaction quanti cation methods developed by our group [8][9][10][11], we have used this approach to identify and quantify the mitochondrial internal structure (cristae) and mitochondrial-lysosomal interaction of live cells. We consequently provide a new method for the identi cation of the mechanisms of mitochondrial oxidative phosphorylation (OXPHOS) dysfunction.

Results
The reduced metabolic ability of patient-derived SLC24A46 mutant broblasts The mitochondrial respiration function was investigated by examining OCR under both basal conditions and drug-induced mitochondrial stress using the Seahorse assay. The OCR was found to be signi cantly decreased in patient-derived SLC25A46 mutant broblasts compared to normal broblasts (Fig. 1A). After a detailed analysis, the basal respiration, oxygen consumption for ATP production, maximum oxygen consumption capacity of mitochondria, proton-leaked oxygen consumption, non-mitochondrial respiration, and the spare respiratory capacity in patient-derived SLC25A46 mutant broblasts were all lower than that of normal broblasts (Fig. 1B).
The MTT assay re ects the metabolic ability of living cells by measuring the proliferation rates of cells.
The results of this assay for the mutant and normal broblasts showed no apparent difference in the number of living cells between these two types of broblasts on Day 2 after cell seeding. However, subsequent to this time point, the normal broblasts showed vigorous metabolism and rapid proliferation rate on Day 4, Day 6, and Day 8 (Fig. 1D). Thus, the metabolic ability and cell proliferation rate of SLC25A46 mutant broblasts were signi cantly lower than that of normal broblasts (Fig. 1D). The imaging results showed that the cell density of normal broblasts was close to 80-90% on Day 8, while it only reached 40-50% in SLC25A46 mutant broblasts (Fig. 1C).
Mitochondrial hyper-fusion in patient-derived SLC24A46 mutant broblasts with the live-cell nanoscope-3D-SIM imaging system The decreased metabolic ability of mutant broblasts suggested that the mitochondrial function in SLC25A46 mutant broblasts has been disturbed. Sanger sequencing results showed a homozygous, missense point mutation (c.1005A > T; p. Glu335Asp) in SLC25A46 mutant broblasts (Fig. 2D). To examine whether this mutation causes any changes in mitochondrial morphology, we used a nanoscope − 3D-SIM imaging approach to observe mitochondrial morphology in those two human cells. The images showed that the normal broblasts had round or medium length mitochondria ( Fig. 2A), while the SLC25A46 mutant broblasts showed slender, hyper-fused mitochondria (Fig. 2B). Imaris software (Nikon, Tokyo, Japan) was used to identify and analyze the mitochondria morphology (Fig. 2C). The results showed that the number of mitochondria in SLC25A46 mutant broblasts was signi cantly lower than what was observed in normal broblasts. In contrast, the average area and volume of mitochondria in mutant cells were signi cantly greater than those in normal broblasts (Fig. 2E). The comparative analysis of mitochondrial morphology showed aberrant hyper fusion of mitochondria in the patientderived SLC24A46 mutant broblasts.
Severe damage of mitochondrial cristae in patient-derived SLC24A46 mutant broblasts Previously, mitochondrial fusion was considered to facilitate OXPHOS, and an increase of mitochondrial fusion will improve the mitochondrial OXPHOS level [12,13]. Mediated mitochondrial fusion was therefore regarded as a new therapeutic target for mitochondrial diseases [14,15]. However, our group found that the highly-fused mitochondria from SLC25A46 mutant broblasts resulted in reduced OXPHOS [16]. What is the underlying cause of this rare condition? One possibility is alterations in the cristae, a most important structures of the inner mitochondrial membrane (IMM), which are deemed as the core of ATP production and mitochondrial respiratory function [17,18]. Therefore, we decided to investigate whether structural defects of mitochondrial cristae lead to decreased OXPHOS.
Using algorithm-based SIM imaging technology previously developed by our team [8], we identi ed and extracted cristae rst, then quantitatively analyzed the mitochondrial cristae for human-derived normal and patient-derived SLC25A46 mutant broblasts. The images showed that the mitochondrial cristae structure was visible and abundant in normal broblasts (Fig. 3A). In contrast, the cristae structure was damaged or even vanished in SLC25A46 mutant broblasts (Fig. 3B). After quanti cation analysis, the mean cristae number ( A similar tendency of mitophagy was observed in normal and SLC25A46 mutant broblasts Mitophagy is the general process by which the cell removes severely damaged mitochondria, consequently achieving the purpose of "quality control" of mitochondria within living cells [19,20]. We observed highly-fused mitochondria with severely damaged cristae structures in SLC25A46 mutant broblasts. This raised the obvious questions of whether or not these abnormal mitochondria induce mitophagy? Using the SIM image-based mitochondria-lysosome co-location analysis method in living cells [9], we can observe and quantify mitophagy in normal and SLC25A46 mutant broblasts (Fig. 4C).
Our results demonstrate that only slight levels of mitophagy are occurring in both of these cell lines (Fig. 4A, Fig. 4B). After quantitative analysis, there was no statistically signi cant difference in the value of mitochondrial -lysosome co-location between the two cell lines. Western Blot also con rmed that the values of the LC3-II/LC3-I ratio were comparable between normal and SLC25A46 mutant broblasts (Fig. 4d), which was consistent with the results of the SIM image-based analysis method. In addition, with this nanoscope, we can straightforwardly monitor the mitochondrial dynamics and the mitochondrialysosome interaction dynamics (Fig. 4e).
This novel nanoscope combined with a quanti cation analysis strategy can not only be used to observe mitochondrial morphology, but also to detect and quantify the damage of structures in sub-mitochondria, assess the extent of mitophagy, and monitor the dynamics of mitochondria and lysosome (Fig. 5).
Therefore, this novel approach is a great approach for the observation and etiological diagnosis of mitochondrial damage in patients with mitochondrial disease. Discussion SLC25A46 is responsible for encoding a mitochondrial solute carrier protein [21]. We identi ed SLC25A46 is the human homolog of Ugo1, a protein of Saccharomyces cerevisiae and located in the mitochondrial outer membrane and involved in mitochondrial fusion [16,22,23]. So far, SLC25A46 has been found to be associated with various human diseases. Homozygous or compound heterozygous mutations of SLC25A46 led to a range of clinical syndromes, with the clinical feature of optic atrophy, cerebellar atrophy, progressive myoclonic ataxia, axonal peripheral neuropathy, autosomal recessive cerebellar ataxias (ARCA), lethal congenital pontocerebellar hypoplasia, and even Parkinson's disease [16,21,[24][25][26][27][28][29]. Mice with Slc25a46 dysfunction developed severe motor impairment, optic atrophy, and developmental defects of the nervous system, as well as premature death [30][31][32].
Currently, SLC25A46 is believed to affect mitochondrial dynamics due to the interaction with OPA1 and MFN2 [33,34]. The hyper-fused mitochondria and reduced mitochondrial respiratory function presented in patient-derived SLC25A46 mutant broblasts have also been con rmed by this study, as well as previous studies, which was supposed to be the pathogenic mechanism of a series of neurological diseases [35]. The MTT assay results from this study also strengthened the idea that the metabolic capacity of SLC25A46 mutant broblasts is signi cantly lower than that of control cells. However, there exists a contradiction between the morphology of highly-fused mitochondria and the decline of mitochondrial function. Traditionally, mitochondrial fusion has been veri ed to be vital for maintaining mtDNA stability and improving the tolerance of cells to high mtDNA mutations [36,37]. At the same time, mitochondrial fusion is also a protective factor for maintaining normal mitochondrial respiration function. The absence of mitochondrial fusion in the cerebellum has also been shown to result in a malformed mitochondrial distribution and function [38]. Moreover, mitochondrial fusion is required to support the normal development of embryos [3]. Why then do the SLC25A46 mutant cells examined in our study show mitochondrial hyper-fusion, but a decrease in mitochondrial respiratory function?
The respiratory function of mitochondria is a series of oxidation-reduction reactions mediated by multiple complexes located on the mitochondrial inner cristae, which eventually produce ATP and provide energy for the tissues and cells in living organisms [39,40]. From this viewpoint, we hypothesized that SLC25A46 mutation causes structural abnormalities of cristae in highly fused mitochondria, consequently affecting mitochondrial respiratory function. Based on the identi cation and quanti cation method of mitochondrial cristae invented by our group, we analyzed the mitochondrial internal cristae of patientderived SCL25A46 mutant broblasts and human-derived normal broblasts. Our results showed that, compared with normal mitochondria, the number of mitochondrial cristae decreased, the length of cristae shortened, and the area of cristae was reduced in the SLC25A46 mutant broblasts. We even observed the disappearance of cristae in some mitochondria. Therefore, we posited that the structure of mitochondrial cristae in the mutant cells was damaged, which accordingly affected the mitochondrial respiratory function, as re ected by decreased aerobic respiration, reduced ATP generation and decreased metabolic capacity. Researchers have suggested that SLC25A46 plays a vital role in the interaction between the major structural proteins of the mitochondrial outer membrane and the mitochondrial cristae, and it is crucial for maintaining the structure and stability of the mitochondrial cristae. Immunoblot analysis revealed that MIC60 and MIC19 -two critical proteins of mitochondrial contact site and cristae organizing system (MICOS) complex obviously decreased in patient-derived SLC25A46 mutant broblast. And immunoprecipitation experiments showed SLC25A46 co-immunoprecipitated with MIC60, MIC19, OPA1 (located on IMM), MFN1 and MFN2 (located on OMM) [33]. MICOS complex, especially MIC60 and MIC19, is a crucial factor in cristae biogenesis [41,42]. Therefore, SLC25A46 is believed to be not only involved in maintaining the stability of OMM, but also an essential protein in the interaction and communication between the OMM and IMM, as well as the formation and maintenance of mitochondrial cristae [33]. They have used transmission electron microscopy and observed the signi cantly decreased cristae number and length from the patient-derived mitochondria [33]. This result from transmission electron microscopy is consistent with our nanoscope-based results.
Mitophagy is a crucial step in mitochondrial quality control, which is used to remove damaged mitochondria [43,44]. Severe injury of mitochondrial cristae can induce mitophagy as well [45,46]. Therefore, we also hypothesized that the damaged mitochondrial cristae would increase the rate of mitophagy in SLC25A46 mutant broblasts. We monitored the dynamic changes of mitochondrial and lysosomal behavior in SLC25A46 mutant broblasts in real-time. We observed a contact and colocalization phenomenon between lysosome and mitochondria after mitochondrial fragmentation in SLC25A46 mutant broblasts. However, using the SIM image-based mitophagy quanti cation method we invented before, we determined that the overall tendency of mitophagy in the SLC25A46 mutant broblasts was not statistically different from that in normal broblasts, although mitophagy did occur in some mitochondria in the SLC25A46 mutant broblasts. Consequently, although the mitochondrial cristae were severely damaged in the SLC25A46 mutant broblasts, the damaged cristae alone did not appear to induce the occurrence of mitophagy. Currently, no studies have reported the mitophagy status of SLC25A46 mutant cells.

Conclusions
Overall, this study suggests that severely damaged mitochondrial cristae may be the predominant cause of reduced mitochondrial respiratory dysfunction in SLC25A46 mutant broblasts, but that the damaged mitochondrial cristae do not induce a signi cant increase in mitophagy.
Through the usage of the SIM-based live-cell nanoscope and the quanti cation methods we developed, we can examine the morphology of the outer mitochondrial membrane (OMM), inner mitochondrial membrane (IMM), the rate of mitophagy, and also perform quantitative calculations of all these phenomena. Simultaneously, we can dynamically observe the behavior of mitochondria and lysosomes.
We have thus achieved a comprehensive observation of mitochondrial morphology, internal structure, and mitophagy using a single technique. Therefore, this nanoscope is exceptionally suited for the observation and calculation of the mitochondria and sub-mitochondrial structures in live cells from patients with various mitochondrial diseases. The operation of the device is simple, rapid and accurate, which is helpful for the quick assessment and diagnosis of mitochondrial abnormalities.

Cell culture
The human-derived normal broblasts and patient-derived SLC25A46 mutant (c.1005A > T, p.Glu335Asp) broblasts cell lines were acquired after informed consent was obtained from the patients. The cells were cultured in Dulbecco's modi ed Eagle's medium (DMEM) medium (Gibco, Thermo Fisher Scienti c, USA) with 10% FBS (Gibco, Thermo Fisher Scienti c, USA) and 100 units/ml Anti-Anti (containing streptomycin and penicillin) (Gibco, Thermo Fisher Scienti c, USA) and incubated in a 5% CO2, 37℃ and 100% humidity incubator.

Nanoscope -3D-SIM imaging
The cells were seeded in a glass-bottom culture dish (MatTek Life Sciences, USA) and cultured for 24 hours in 2 ml DMEM containing 10%FBS and 100 units/ml Anti-Anti. Before imaging, cells were rst washed three times with a pre-warmed DMEM medium and then were incubated in a DMEM medium containing 100 nM Mito-Tracker Green (Invitrogen, USA) for half an hour. Cells for mitophagy analysis were co-incubated in DMEM medium containing 100 nM Mito-Tracker Green (Invitrogen, USA) and Lyso-Tracker Red (Invitrogen, USA) for half an hour. Cells were then washed three times with DMEM. The stained cells were photographed using the 3D-structure illumination microscope (Nikon, Tokyo, Japan).

Western blot
Protease inhibitor cocktail (Sigma, USA) and 2 × RIPA lysis and extraction buffer (Thermo sher Scienti c, USA) were added to the centrifuged cell pellets, and then were sonicated for 5 minutes each time, three times in total. The protein concentration was measured by using the Pierce BCA Protein Assay Kit (Thermo sher Scienti c, USA). 30ug protein for each sample and 4X NuPAGE LDS Sample Buffer (Thermo sher Scienti c, USA) were mixed at 4:1 ratio and denatured at 95℃ for 5 minutes, and then separated in 4-12% Bis-Tris gel (Invitrogen, USA). The gel was transferred onto a PVDF membrane (Invitrogen, USA) through the iBlot 2 gel transfer device (Life Technologies, USA). The transferred PVDF membrane was placed in the Intercept Blocking Buffer (LI-COR Biosciences, USA) for 45 minutes, and then incubated overnight in the primary antibody, rabbit anti-LC3B (cell signaling technology, USA) diluted in the blocking buffer at a ratio of 1:200 with Tween 20 diluted in the blocking buffer at a ratio of 1:1000. Rabbit anti-GAPDH (cell signaling technology, USA) was also diluted in the blocking buffer at a ratio of 1:2000 and set as the loading control. The next day, the PVDF membrane was washed for 10 minutes each time, three times in total. Then, the membrane was incubated in the secondary antibody, IRDye 800CW Goat anti-Rabbit IgG (LI-COR Biosciences, USA), for 120 minutes. The bands were detected by the LI-COR Odyssey Clx Imaging System (LI-COR Biosciences, Lincoln, NE).

Sanger sequencing for mutation detection
To detect the point mutation of SLC25A46 in human-derived normal and patient-derived broblasts, genomic DNA was extracted using DNeasy Blood & Tissue Kit (Qiagen, USA). PCR products of 186 bp in length were ampli ed using GoTag mater mixes (Promega, USA). The following primer set was used for the ampli cation: Forward: TGCCAGTCTTTGTTCTGACG and Reverse: CCAAACACTCCTTCCTCCTG. The reactions were performed following the thermal cycling program: 95 °C for 2 minutes, followed by 30 cycles of 95 °C for 30 seconds, 56 °C for 30 seconds, and 72 °C for 30 seconds. A nal extension step was then performed at 72 °C for 4 minutes.

Oxygen consumption rate (OCR) measurement
Human-derived normal and patient-derived SLC25A46 mutant cells were seeded at a density of 1.0 × 10 4 cells/well with DMEM supplemented with 10% FBS in XFe96 cell culture plates (Agilent Technologies, USA). After incubation for 24 hours, the DMEM medium was removed and changed with the warmed XF DMEM Medium supplemented with 1 mM sodium pyruvate, 10 mM glucose and 2 mM L-glutamine at pH 7.4. All cells were treated with 1 µM oligomycin A, 1 µM FCCP, and 500 nM rotenone/antimycin A. The OCRs of the cells was assessed by using the XF Cell Mito Stress Test Kit (Agilent Technologies, USA). The Seahorse XF96 analyzer (Agilent Technologies, USA) was used for OCR measurement.
Cell proliferation rate measurement (MTT assay) Human-derived normal and patient-derived SLC25A46 mutant broblasts were seeded in 96 well plates (Corning, USA) at a density of 3.0 × 10 3 cells/well with DMEM supplemented with 10% FBS and incubated at 37˚C, 5% CO2. 10 µl of MTT solution (Roche, USA) was added to 100 ul culture medium in each well at a nal concentration of 0.5 mg/ml. The following process was implemented according to the manual provided by the kit. The absorbance was detected at 570 nm by the microplate reader (BioTek, USA).

Statistical analysis
Graphpad Prism 7 software was used to display data. Independent-Samples T-test was used for statistical analysis. * was de ned as P < 0.05, ** as P < 0.01, *** as P < 0.001, and **** as P < 0.0001.

Availability of data and materials
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.