Thank you for visiting Nature.com. The version of browser you are using has limited CSS support. For best results, we recommend that you use a newer version of your browser (or disable Compatibility Mode in Internet Explorer). In the meantime, to ensure ongoing support, we are showing the site without styling or JavaScript.
        Propionic acid (PPA) is used to study the role of mitochondrial dysfunction in neurodevelopmental disorders such as autism spectrum disorder. PPA is known to disrupt mitochondrial biogenesis, metabolism, and turnover. However, the effects of PPA on mitochondrial dynamics, fission and fusion remain problematic due to the complex temporal nature of these mechanisms. Here, we use complementary quantitative imaging techniques to investigate how PPA affects mitochondrial ultrastructure, morphology, and dynamics in neuron-like SH-SY5Y cells. PPA (5 mM) caused a significant decrease in mitochondrial area (p < 0.01), Feret diameter and circumference (p < 0.05), and area 2 (p < 0.01). Mitochondrial event locator analysis showed a significant increase (p < 0.05) in fission and fusion events, thereby maintaining the integrity of the mitochondrial network under stress conditions. In addition, the mRNA expression of cMYC (p < 0.0001), NRF1 (p < 0.01), TFAM (p < 0.05), STOML2 (p < 0.0001) and OPA1 (p < 0.05) was significantly reduced. 01). This illustrates the remodeling of mitochondrial morphology, biogenesis and dynamics to maintain function under stress conditions. Our data provide new insight into the effects of PPA on mitochondrial dynamics and highlight the utility of imaging techniques for studying the complex regulatory mechanisms involved in mitochondrial stress responses.
        Mitochondria are integral participants in a variety of cellular functions beyond their typical roles in energy production and biosynthesis. Mitochondrial metabolism is a key regulator of calcium signaling, metabolic and redox homeostasis, inflammatory signaling, epigenetic modifications, cell proliferation, differentiation and programmed cell death1. In particular, mitochondrial metabolism is critical for neuronal development, survival and function and is widely implicated in various manifestations of neuropathology2,3,4.
        Over the past decade, metabolic status has emerged as a central regulator of neurogenesis, differentiation, maturation and plasticity5,6. Recently, mitochondrial morphology and dynamics have become particularly important components of mitosis, a dynamic process that maintains a pool of healthy mitochondria within cells. Mitochondrial dynamics are regulated by complex interdependent pathways ranging from mitochondrial biogenesis and bioenergetics to mitochondrial fission, fusion, transport and clearance7,8. Disruption of any of these integrative mechanisms impairs the maintenance of healthy mitochondrial networks and has profound functional consequences for neurodevelopment9,10. Indeed, dysregulation of mitochondrial dynamics is observed in many psychiatric, neurodegenerative and neurodevelopmental disorders, including autism spectrum disorders (ASD)11,12.
        ASD is a heterogeneous neurodevelopmental disorder with a complex genetic and epigenetic architecture. The heritability of ASD is undisputed, but the underlying molecular etiology remains poorly understood. Accumulating data from preclinical models, clinical studies, and multi-omics molecular datasets provide increasing evidence of mitochondrial dysfunction in ASD13,14. We previously performed a genome-wide DNA methylation screen in a cohort of patients with ASD and identified differentially methylated genes clustered along mitochondrial metabolic pathways15. We subsequently reported differential methylation of central regulators of mitochondrial biogenesis and dynamics, which was associated with increased mtDNA copy number and altered urinary metabolic profile in ASD16. Our data provide increasing evidence that mitochondrial dynamics and homeostasis play a central role in the pathophysiology of ASD. Therefore, improving the mechanistic understanding of the relationship between mitochondrial dynamics, morphology, and function is a key goal of ongoing research into neurological diseases characterized by secondary mitochondrial dysfunction.
        Molecular techniques are often used to study the role of specific genes in mitochondrial stress responses. However, this approach may be limited by the multifaceted and temporal nature of mitotic control mechanisms. Moreover, differential expression of mitochondrial genes is an indirect indicator of functional changes, especially since only a limited number of genes are typically analyzed. Therefore, more direct methods for studying mitochondrial function and bioenergetics have been proposed17. Mitochondrial morphology is closely related to mitochondrial dynamics. Mitochondrial shape, connectivity, and structure are critical for energy production and mitochondrial and cell survival5,18. Moreover, the various components of mitosis focus on changes in mitochondrial morphology, which may serve as useful endpoints of mitochondrial dysfunction and provide a basis for subsequent mechanistic studies.
        Mitochondrial morphology can be directly observed using transmission electron microscopy (TEM), allowing detailed study of cellular ultrastructure. TEM directly visualizes the morphology, shape and structure of mitochondrial cristae at the resolution of individual mitochondria, rather than relying solely on gene transcription, protein expression or mitochondrial functional parameters in cell populations17,19,20. In addition, TEM facilitates the study of interactions between mitochondria and other organelles, such as the endoplasmic reticulum and autophagosomes, which play key roles in mitochondrial function and homeostasis21,22. Thus, this makes TEM a good starting point for studying mitochondrial dysfunction before focusing on specific pathways or genes. As mitochondrial function becomes increasingly relevant to neuropathology, there is a clear need to be able to directly and quantitatively study mitochondrial morphology and dynamics in in vitro neuronal models.
        In this article, we examine mitochondrial dynamics in a neuronal model of mitochondrial dysfunction in autism spectrum disorder. We previously reported differential methylation of propionyl-CoA carboxylase beta (PCCB) in ASD15, a subunit of the mitochondrial propionyl-CoA carboxylase enzyme PCC. Dysregulation of PCC is known to cause toxic accumulation of propionyl derivatives, including propionic acid (PPA)23,24,25. PPA has been shown to disrupt neuronal metabolism and alter behavior in vivo and is an established animal model for studying neurodevelopmental mechanisms involved in ASD26,27,28. Additionally, PPA has been reported to disrupt mitochondrial membrane potential, biogenesis and respiration in vitro and has been widely used to model mitochondrial dysfunction in neurons29,30. However, the impact of PPA-induced mitochondrial dysfunction on mitochondrial morphology and dynamics remains poorly understood.
        This study uses complementary imaging techniques to quantify the effects of PPA on mitochondrial morphology, dynamics, and function in SH-SY5Y cells. First, we developed a TEM method to visualize changes in mitochondrial morphology and ultrastructure17,31,32. Given the dynamic nature of mitochondria33, we also used mitochondrial event localizer (MEL) analysis to quantify changes in the balance between fission and fusion events, mitochondrial number and volume under PPA stress. Finally, we examined whether mitochondrial morphology and dynamics are associated with changes in the expression of genes involved in biogenesis, fission, and fusion. Taken together, our data illustrate the challenge of elucidating the complexity of the mechanisms regulating mitochondrial dynamics. We highlight the utility of TEM in studying mitochondrial morphology as a measurable convergent endpoint of mitosis in SH-SY5Y cells. Additionally, we highlight that TEM data provide the richest information when combined with imaging techniques that also capture dynamic events in response to metabolic stress. Further characterization of the molecular regulatory mechanisms that support neuronal cell mitosis may provide important insight into the mitochondrial component of nervous system and neurodegenerative diseases.
        To induce mitochondrial stress, SH-SY5Y cells were treated with PPA using 3 mM and 5 mM sodium propionate (NaP). Before TEM, samples were subjected to cryogenic sample preparation using high-pressure freezing and freezing (Fig. 1a). We developed an automated mitochondrial image analysis pipeline to measure eight morphological parameters of mitochondrial populations across three biological replicates. We found that PPA treatment significantly changed four parameters: area 2, area, perimeter, and Feret diameter (Fig. 1b–e). Area 2 decreased significantly with both 3 mM and 5 mM PPA treatment (p = 0.0183 and p = 0.002, respectively) (Fig. 1b), while area (p = 0.003), perimeter (p = 0.0106) and Feret diameter have all decreased significantly. There was a significant reduction (p = 0.0172) in the 5 mM treatment group compared to the control group (Fig. 1c–e). Significant reductions in area and circumference showed that cells treated with 5 mM PPA had smaller, more rounded mitochondria, and that these mitochondria were less elongated than those in control cells. This is also consistent with a significant decrease in the Feret diameter, an independent parameter indicating a decrease in the largest distance between particle edges. Changes in the ultrastructure of the cristae were observed: the cristae became less pronounced under the influence of PPA stress (Fig. 1a, panel B). However, not all images clearly reflected the ultrastructure of the cristae, so a quantitative analysis of these changes was not carried out. These TEM data may reflect three possible scenarios: (1) PPA enhances fission or inhibits fusion, causing existing mitochondria to shrink in size; (2) enhanced biogenesis creates new, smaller mitochondria or (3) induces both mechanisms simultaneously. Although these conditions cannot be distinguished by TEM, significant morphological changes indicate changes in mitochondrial homeostasis and dynamics under PPA stress. We subsequently explored additional parameters to further characterize these dynamics and the potential mechanisms underlying them.
        Propionic acid (PPA) remodels mitochondrial morphology. (a) Representative transmission electron microscopy (TEM) images showing that mitochondrial size decreases and mitochondria become smaller and more rounded with increasing PPA treatment; 0 mM (untreated), 3 mM and 5 mM, respectively. Red arrows indicate mitochondria. (b–e) SH-SY5Y cells treated with PPA for 24 h were prepared for TEM and the results were analyzed using Fiji/ImageJ. Four of the eight parameters showed significant differences between control (untreated, 0 mM PPA) and treated (3 mM and 5 mM PPA) cells. (b) Region 2, (c) Area, (d) Perimeter, (e) Feret diameter. One-way analysis of variance (control vs. treatment) and Dunnett’s multiple comparison test were used to determine significant differences (p < 0.05). Data points represent the average mitochondrial value for each individual cell, and error bars represent the mean ± SEM. Data shown represent n = 3, at least 24 cells per replicate; a total of 266 images were analyzed; * indicates p < 0.05, ** indicates p < 0.01.
        To further characterize how mitochondrial dynamics respond to PPA, we stained mitochondria with tetramethylrhodamine ethyl ester (TMRE) and used time-lapse microscopy and MEL analysis to localize and quantify mitochondria after 24 hours at 3 and 5 mM PPA. Treatment of fission and fusion events. (Fig. 2a). After MEL analysis, mitochondria were further analyzed to quantify the number of mitochondrial structures and their average volume. We observed a small but significant increase in the number of fission events occurring at 3 mM [4.9 ± 0.3 (p < 0.05)] compared to fission [5.6 ± 0.3 (p < 0.05) )] and fusion [5.4 ± 0.5 (p < 0.05)] and fusion [5.4 ± 0.5 (p < 0.05)] 0.05)] <0.05)] events were significantly increased at 5 mM compared to control (Fig. 3b). The number of mitochondria significantly increased at both 3 [32.6 ± 2.1 (p < 0.05)] and 5 mM [34.1 ± 2.2 (p < 0.05)] (Fig. 3c) , while the average volume of each mitochondrial structure remained unchanged (Fig. 3c). 3d). Taken together, this suggests that remodeling of mitochondrial dynamics serves as a compensatory response that successfully maintains the integrity of the mitochondrial network. The increase in the number of fission events at 3 mM PPA suggests that the increase in mitochondrial number is partly due to mitochondrial fission, but given that the average mitochondrial volume remains essentially unchanged, biogenesis cannot be ruled out as an additional compensatory response. However, these data are consistent with the smaller, round mitochondrial structures observed by TEM and also demonstrate significant changes in mitochondrial dynamics induced by PPA.
        Propionic acid (PPA) induces dynamic mitochondrial remodeling to maintain network integrity. SH-SY5Y cells were cultured, treated with 3 and 5 mM PPA for 24 hours and stained with TMRE and Hoechst 33342 followed by MEL analysis. (a) Representative time-lapse microscopy images depicting color and binarized maximum intensity projections at time 2 (t2) for each condition. Selected regions indicated in each binary image are enhanced and displayed in 3D at three different time frames (t1-t3) to illustrate the dynamics over time; fusion events are highlighted in green; fission events are highlighted in green. Displayed in red. (b) Average number of dynamic events per condition. (c) Average number of mitochondrial structures per cell. (d) Average volume (µm3) of each mitochondrial structure per cell. Data shown are representative of n = 15 cells per treatment group. Error bars shown represent mean ± SEM, scale bar = 10 μm, * p < 0.05.
        Propionic acid (PPA) causes transcriptional suppression of genes associated with mitochondrial dynamics. SH-SY5Y cells were treated with 3 and 5 mM PPA for 24 h. Relative gene quantification was performed using RT-qPCR and normalized to B2M. Mitochondrial biogenesis genes (a) cMYC, (b) TFAM, (c) NRF1 and (d) NFE2L2. Mitochondrial fusion and fission genes (e) STOML2, (f) OPA1, (g) MFN1, (h) MFN2 and (i) DRP1. Significant differences (p < 0.05) were tested using one-way ANOVA (control vs. treatment) and Dunnett’s multiple comparison test: * indicates p < 0.05, ** indicates p < 0.01, and **** indicates p < 0.0001. Bars represent mean expression ± SEM. Data shown represent n = 3 (STOML2, OPA1, TFAM), n = 4 (cMYC, NRF1, NFE2L2), and n = 5 (MFN1, MFN2, DRP1) biological replicates.
        Data from TEM and MEL analyzes together indicate that PPA alters mitochondrial morphology and dynamics. However, these imaging techniques do not provide insight into the underlying mechanisms driving these processes. We therefore examined the mRNA expression of nine key regulators of mitochondrial dynamics, biogenesis, and mitosis in response to PPA treatment. We quantified cell myeloma oncogene (cMYC), nuclear respiratory factor (NRF1), mitochondrial transcription factor 1 (TFAM), NFE2-like transcription factor BZIP (NFE2L2), gastrin-like protein 2 (STOML2), optic nerve atrophy 1 (OPA1 ), Mitofusin 1 (MFN1), Mitofusin 2 (MFN2) and dynamin-related protein 1 (DRP1) after 24 hours of treatment with 3 mM and 5 mM PPA. We observed 3 mM (p = 0.0053, p = 0.0415 and p < 0.0001, respectively) and 5 mM (p = 0.0031, p = 0.0233, p < 0.0001) PPA treatment. (Fig. 3a–c). The decrease in mRNA expression was dose-dependent: the expression of cMYC, NRF1 and TFAM decreased by 5.7, 2.6 and 1.9 times at 3 mM, respectively, and by 11.2, 3 and 2.2 times at 5 mM. In contrast, the central redox biogenesis gene NFE2L2 was not altered at any concentration of PPA, although a similar dose-dependent trend of decreased expression was observed (Fig. 3d).
        We also examined the expression of classical genes involved in the regulation of fission and fusion. STOML2 is thought to be involved in fusion, mitophagy and biogenesis, and its expression was significantly reduced (p < 0.0001) by 3 mM (2.4-fold change) and 5 mM (2.8-fold change) PPA (Fig. 1). 3d). Similarly, OPA1 fusion gene expression was decreased at 3 mM (1.6-fold change) and 5 mM (1.9-fold change) PPA (p = 0.006 and p = 0.0024, respectively) (Fig. 3f). However, we did not find significant differences in the expression of fusion genes MFN1, MFN2 or fission gene DRP1 under 24-h PPA stress (Fig. 3g–i). In addition, we found that the levels of four fusion and fission proteins (OPA1, MFN1, MFN2 and DRP1) did not change under the same conditions (Fig. 4a–d). It is important to note that these data reflect a single point in time and may not reflect changes in protein expression or activity levels during the early stages of PPA stress. However, significant reductions in the expression of cMYC, NRF1, TFAM, STOML2, and OPA1 indicate significant transcriptional dysregulation of mitochondrial metabolism, biogenesis, and dynamics. In addition, these data highlight the utility of imaging techniques to directly study end-state changes in mitochondrial function.
        Fusion and fission factor protein levels did not change after propionic acid (PPA) treatment. SH-SY5Y cells were treated with 3 and 5 mM PPA for 24 h. Protein levels were quantified by Western blot analysis, and expression levels were normalized to total protein. Average protein expression and representative Western blots of target and total protein are shown. a – OPA1, b – MFN1, c – MFN2, d – DRP1. Bars represent mean ± SEM, and data shown are representative of n = 3 biological replicates. Multiple comparisons (p < 0.05) were performed using one-way analysis of variance and Dunnett’s test. The original gel and blot are shown in Figure S1.
        Mitochondrial dysfunction is associated with multisystem diseases ranging from metabolic, cardiovascular and muscular diseases to neurological diseases1,10. Many neurodegenerative and neurodegenerative diseases are associated with mitochondrial dysfunction, highlighting the importance of these organelles throughout the lifespan of the brain. These diseases include Parkinson’s disease, Alzheimer’s disease and ASD3,4,18. However, access to brain tissue to study these diseases is difficult, especially at the mechanistic level, making cellular model systems a necessary alternative. In this study, we use a cellular model system using PPA-treated SH-SY5Y cells to recapitulate the mitochondrial dysfunction observed in neuronal diseases, particularly autism spectrum disorders. Using this PPA model to study mitochondrial dynamics in neurons may provide insight into the etiology of ASD.
        We explored the possibility of using TEM to view changes in mitochondrial morphology. It is important to note that TEM must be used correctly to maximize its effectiveness. Preparation of cryo-specimens allows for better preservation of neuronal structures by simultaneously fixing cellular components and reducing the formation of artifacts34. Consistent with this, we observed that neuron-like SH-SY5Y cells had intact subcellular organelles and elongated mitochondria (Fig. 1a). This highlights the utility of cryogenic preparation techniques for studying mitochondrial morphology in neuronal cell models. Although quantitative measurements are critical for objective analysis of TEM data, there is still no consensus on what specific parameters should be measured to confirm mitochondrial morphological changes. Based on a large number of studies that have quantitatively examined mitochondrial morphology17,31,32, we developed an automated mitochondrial image analysis pipeline that measures eight morphological parameters, namely: area, area2, aspect ratio, perimeter, circularity, degree , Feret diameter. and roundness.
        Among them, PPA significantly reduced area 2, area, perimeter, and Feret diameter (Fig. 1b–e). This showed that mitochondria became smaller and more rounded, which is consistent with previous studies showing a decrease in mitochondrial area after 72 hours of PPA30-induced mitochondrial stress. These morphological features may indicate mitochondrial fission, a necessary process to sequester damaged components from the mitochondrial network to promote their degradation through mitophagy35,36,37. On the other hand, the decrease in average mitochondrial size may be associated with increased biogenesis, which results in the formation of small nascent mitochondria. Increased fission or biogenesis represents a compensatory response to maintain mitosis against mitochondrial stress. However, decreased mitochondrial growth, impaired fusion, or other conditions cannot be excluded.
        Although the high-resolution images created by TEM allow the determination of morphological characteristics at the level of individual mitochondria, this method produces two-dimensional snapshots at a single point in time. To study dynamic responses to metabolic stress, we stained mitochondria with TMRE and used time-lapse microscopy with MEL analysis, which allows high-throughput 3D visualization of changes in the mitochondrial network over time33,38. We observed subtle but significant changes in mitochondrial dynamics under PPA stress (Fig. 2). At 3 mM, the number of fission events increased significantly, while fusion events remained the same as in the control. An increase in the number of both fission and fusion events was observed at 5 mM PPA, but these changes were approximately proportional, suggesting that fission and fusion kinetics reach equilibrium at higher concentrations (Fig. 2b). The average mitochondrial volume remained unchanged at both 3 and 5 mM PPA, indicating that the integrity of the mitochondrial network was preserved (Fig. 2d). This reflects the ability of dynamic mitochondrial networks to respond to mild metabolic stress to effectively maintain homeostasis without causing network fragmentation. At 3 mM PPA, the increase in fission is sufficient to promote transition to a new equilibrium, but more profound kinetic remodeling is required in response to stress induced by higher concentrations of PPA.
        The number of mitochondria increased at both PPA stress concentrations, but the average mitochondrial volume did not change significantly (Fig. 2c). This may be due to increased biogenesis or increased division; however, in the absence of a significant decrease in mean mitochondrial volume, it is more likely that biosynthesis increases. However, the data in Figure 2 do support the existence of two compensatory mechanisms: an increase in the number of fission events, consistent with upregulation of mitochondrial fission, and an increase in the number of events, consistent with mitochondrial biogenesis. Ultimately, dynamic compensation for mild stress may consist of simultaneous processes involving fission, fusion, biogenesis, and mitophagy. Although previous authors have shown that PPA enhances mitosis30,39 and mitophagy29, we provide evidence for remodeling of mitochondrial fission and fusion dynamics in response to PPA. These data confirm the morphological changes observed by TEM and provide further insight into the mechanisms associated with PPA-induced mitochondrial dysfunction.
        Because neither TEM nor MEL analysis provided direct evidence of the gene regulatory mechanisms underlying the observed morphological changes, we examined the RNA expression of genes involved in mitochondrial metabolism, biogenesis, and dynamics. The cMYC proto-oncogene is a transcription factor involved in the regulation of mitochondria, glycolysis, amino acid and fatty acid metabolism40. In addition, cMYC is known to regulate the expression of nearly 600 mitochondrial genes involved in mitochondrial transcription, translation, and complex assembly, including NRF1 and TFAM41. NRF1 and TFAM are two central regulators of mitosis, acting downstream of PGC-1α to activate mtDNA replication. This pathway is activated by cAMP and AMPK signaling and is sensitive to energy expenditure and metabolic stress. We also examined NFE2L2, a redox regulator of mitochondrial biogenesis, to determine whether the effects of PPA might be mediated by oxidative stress.
        Although NFE2L2 expression remained unchanged, we found a consistent dose-dependent decrease in the expression of cMYC, NRF1 and TFAM after 24 h of treatment with 3 mM and 5 mM PPA (Fig. 3a–c). Downregulation of cMYC expression has previously been reported as a response to mitochondrial stress42, and conversely, downregulation of cMYC expression can cause mitochondrial dysfunction by remodeling mitochondrial metabolism, network connectivity, and membrane polarization43. Interestingly, cMYC is also involved in the regulation of mitochondrial fission and fusion42,43 and is known to increase DRP1 phosphorylation and mitochondrial localization during cell division44, as well as mediate mitochondrial morphological remodeling in neuronal stem cells45. Indeed, cMYC-deficient fibroblasts exhibit reduced mitochondrial size, consistent with changes induced by PPA43 stress. These data illustrate an interesting but as yet unclear relationship between cMYC and mitochondrial dynamics, providing an interesting target for future studies of PPA stress-induced remodeling.
        The reduction of NRF1 and TFAM is consistent with the role of cMYC as an important transcriptional activator. These data are also consistent with previous studies in human colon cancer cells showing that PPA reduced NRF1 mRNA expression at 22 hours, which was associated with ATP depletion and increased ROS46. These authors also reported that TFAM expression increased at 8.5 hours but returned to baseline levels at 22 hours. In contrast, Kim et al. (2019) showed that TFAM mRNA expression was significantly decreased after 4 h of PPA stress in SH-SY5Y cells; however, after 72 hours, TFAM protein expression was significantly increased and mtDNA copy number was significantly increased. Thus, the decrease in the number of mitochondrial biogenesis genes that we observed after 24 hours does not exclude the possibility that the increase in the number of mitochondria is associated with activation of biogenesis at earlier time points. Previous studies have shown that PPA significantly upregulates PGC-1α mRNA and protein in SH-SY5Y cells at 4 hours 30 minutes, while propionic acid enhances mitochondrial biogenesis in calf hepatocytes via PGC-1α at 12 hours 39 minutes. Interestingly, PGC-1α is not only a direct transcriptional regulator of NRF1 and TFAM, but has also been shown to regulate the activity of MFN2 and DRP1 by regulating fission and fusion47. Taken together, this highlights the close coupling of mechanisms regulating mitochondrial compensatory responses induced by PPA. Moreover, our data reflect significant dysregulation of transcriptional regulation of biogenesis and metabolism under PPA stress.
        The STOML2, OPA1, MFN1, MFN2 and DRP1 genes are among the central regulators of mitochondrial fission, fusion and dynamics37,48,49. There are many other genes involved in mitochondrial dynamics, however, STOML2, OPA1 and MFN2 have previously been found to be differentially methylated in ASD cohorts,16 and several independent studies have reported changes in these transcription factors in response to mitochondrial stress50,51. 52. The expression of both OPA1 and STOML2 was significantly reduced by 3 mM and 5 mM PPA treatment (Fig. 3e, f). OPA1 is one of the classical regulators of mitochondrial fusion through direct interaction with MFN1 and 2 and plays a role in cristae remodeling and mitochondrial morphology53. The precise role of STOML2 in mitochondrial dynamics remains unclear, but evidence suggests that it plays a role in mitochondrial fusion, biogenesis, and mitophagy.
        STOML2 is involved in maintaining mitochondrial respiratory coupling and formation of respiratory chain complexes54,55 and has been shown to profoundly alter the metabolic characteristics of cancer cells56. Studies have shown that STOML2 promotes mitochondrial membrane potential and biogenesis through interaction with BAN and cardiolipin 55, 57, 58. Additionally, independent studies have shown that the interaction between STOML2 and PINK1 regulates mitophagy59,60. Notably, STOML2 has been reported to directly interact with and stabilize MFN2 and also plays an important role in stabilizing long OPA1 isoforms by inhibiting the protease responsible for OPA1 degradation53,61,62. The reduction in STOML2 expression observed in PPA reactions may make these fusion proteins more susceptible to degradation through ubiquitin- and proteasome-dependent pathways48. Although the precise role of STOML2 and OPA1 in the dynamic response to PPA is unclear, decreased expression of these fusion genes (Figure 3) may disrupt the balance between fission and fusion and lead to decreased mitochondrial size (Figure 3). 1).
        On the other hand, OPA1 protein expression remained unchanged after 24 h, while the mRNA and protein levels of MFN1, MFN2 or DRP1 did not change significantly after PPA treatment (Fig. 3g-i, Fig. 4). This may indicate that there are no changes in the regulation of these factors involved in mitochondrial fusion and fission. However, it is worth noting that each of these four genes is also regulated by posttranscriptional modifications (PTMs) that control protein activity. OPA1 has eight alternative splice variants that are proteolytically cleaved in mitochondria to produce two distinct isoforms 63 . The balance between long and short isoforms ultimately determines the role of OPA1 in mitochondrial fusion and maintenance of the mitochondrial network64. DRP1 activity is regulated by calcium/calmodulin-dependent protein kinase II (CaMKII) phosphorylation, while DRP1 degradation is regulated by ubiquitination and SUMOylation65. Finally, both DRP1 and MFN1/2 are GTPases, so activity may be influenced by the rate of GTP production in mitochondria 66 . Therefore, although the expression of these proteins remains constant, this may not reflect unchanged protein activity or localization67,68. Indeed, existing PTM protein repertoires often serve as the first line of defense responsible for mediating acute stress responses. In the presence of moderate metabolic stress in our model, it is likely that PTM promotes increased activity of fusion and fission proteins to sufficiently restore mitochondrial integrity without requiring additional activation of these genes at the mRNA or protein level.
        Taken together, the above data highlight the complex and time-dependent regulation of mitochondrial morphology and the challenges of elucidating these mechanisms. To study gene expression, it is first necessary to identify specific target genes in the pathway. However, our data show that genes in the same pathway do not respond in the same way to the same stress. In fact, previous studies have shown that different genes in the same pathway may exhibit different temporal response profiles30,46. In addition, there are complex post-transcriptional mechanisms that disrupt the relationship between transcription and gene function. Proteomic studies can provide insight into the impact of PTMs and protein function, but they also pose challenges including low-throughput methods, high signal-to-noise ratios, and poor resolution.
        In this context, studying mitochondrial morphology using TEM and MEL has great potential to address fundamental questions about the relationship between mitochondrial dynamics and function and how this influences disease. Most importantly, TEM provides a direct method for measuring mitochondrial morphology as a convergent endpoint of mitochondrial dysfunction and dynamics51. MEL also provides a direct method for visualizing fission and fusion events in a three-dimensional cellular environment, allowing quantification of dynamic mitochondrial remodeling even in the absence of changes in gene expression33. Here we highlight the utility of mitochondrial imaging techniques in secondary mitochondrial diseases. These diseases are typically characterized by chronic mild metabolic stress characterized by subtle remodeling of mitochondrial networks rather than acute mitochondrial damage. However, the mitochondrial compensation required to maintain mitosis under chronic stress has profound functional consequences. In the context of neuroscience, a better understanding of these compensatory mechanisms may provide important information about the pleiotropic neuropathology associated with mitochondrial dysfunction.
        Ultimately, our data highlight the utility of imaging techniques for understanding the functional consequences of the complex interactions between gene expression, protein modifications, and protein activity that control neuronal mitochondrial dynamics. We used PPA to model mitochondrial dysfunction in a neuronal cell model to gain insight into the mitochondrial component of ASD. SH-SY5Y cells treated with PPA showed changes in mitochondrial morphology: mitochondria became small and round, and cristae were poorly defined when observed by TEM. MEL analysis shows that these changes occur concomitantly with an increase in fission and fusion events to maintain the mitochondrial network in response to mild metabolic stress. Moreover, PPA significantly disrupts the transcriptional regulation of mitochondrial metabolism and homeostasis. We identified cMYC, NRF1, TFAM, STOML2, and OPA1 as key mitochondrial regulators disrupted by PPA stress and may play a role in mediating PPA-induced changes in mitochondrial morphology and function. Future studies are needed to better characterize PPA-induced temporal changes in gene expression and protein activity, localization, and post-translational modifications. Our data highlight the complexity and interdependence of the regulatory mechanisms mediating the mitochondrial stress response and demonstrate the utility of TEM and other imaging techniques for more targeted mechanistic studies.
        The SH-SY5Y cell line (ECACC, 94030304-1VL) was purchased from Sigma-Aldrich. SH-SY5Y cells were grown in Dulbecco’s modified Eagle’s medium/F-12 nutrient mixture (DMEM/F-12) and L-glutamine (SC09411, ScienCell) in 25 cm2 flasks supplemented with 20% fetal bovine serum (FBS) (10493106, ThermoFisher Scientific) and 1% penicillin-streptomycin (P4333-20ML, Sigma-Aldrich) at 37 °C, 5% CO2. Cells were subcultured to 80% confluence using 0.05% trypsin-EDTA (15400054, ThermoFisher Scientific), centrifuged at 300 g and plated at a density of approximately 7 × 105 cells/ml. All experiments were performed on undifferentiated SH-SY5Y cells between passages 19–22. PPA is administered as NaP. Dissolve NaP powder (CAS No. 137-40-6, chemical formula C3H5NaO2, P5436-100G, Sigma-Aldrich) in warm MilliQ water to a concentration of 1 M and store at 4 °C. On the day of treatment, dilute this solution with 1 M PPA to 3 mM and 5 mM PPA in serum-free medium (DMEM/F-12 with L-glutamine). Treatment concentrations for all experiments were no PPA (0 mM, control), 3 mM, and 5 mM PPA. Experiments were carried out in at least three biological replicates.
        SH-SY5Y cells were seeded into 25 cm5 flasks at a rate of 5.5 × 105 cells/ml and grown for 24 hours. The PPA treatment was added to the flask before 24 h of incubation. Collect cell pellets following normal mammalian tissue subculture protocols (described above). Resuspend the cell pellet in 100 µl 2.5% glutaraldehyde, 1× PBS and store at 4 °C until processing. SH-SY5Y cells were briefly centrifuged to pellet the cells and remove 2.5% glutaraldehyde, 1× PBS solution. Resuspend the sediment in a 4% agarose gel prepared in distilled water (the ratio of agarose to sediment volume is 1:1). Agarose pieces were placed on grids on flat plates and coated with 1-hexadecene before high-pressure freezing. Samples were frozen in 100% dry acetone at -90°C for 24 hours. The temperature was then raised to -80°C and a solution of 1% osmium tetroxide and 0.1% glutaraldehyde was added. Samples were stored at -80°C for 24 hours. After this, the temperature was gradually increased to room temperature over several days: from – 80 °C to – 50 °C for 24 hours, to – 30 °C for 24 hours, to – 10 °C for 24 hours and finally to room temperature. temperature.
        After cryogenic preparation, the samples were impregnated with resin and ultrathin sections (∼100 nm) were made using a Leica Reichert UltracutS ultramicrotome (Leica Microsystems). Sections were stained with 2% uranyl acetate and lead citrate. Samples were observed using a FEI Tecnai 20 transmission electron microscope (ThermoFisher (formerly FEI), Eindhoven, The Netherlands) operating at 200 kV (Lab6 transmitter) and a Gatan CCD camera (Gatan, UK) equipped with a Tridiem energy filter.
        In each technical replicate, at least 24 single cell images were acquired, for a total of 266 images. All images were analyzed using the Region of Interest (ROI) macro and the Mitochondria macro. The mitochondrial macro is based on published methods17,31,32 and allows semi-automated batch processing of TEM images in Fiji/ImageJ69. In a nutshell: the image is inverted and inverted using rolling ball background subtraction (60 pixel radius) and an FFT bandpass filter (using 60 and 8 pixel upper and lower bounds respectively) and vertical line suppression with an orientation tolerance of 5%. The processed image is automatically thresholded using a maximum entropy algorithm and a binary mask is generated. Image regions associated with manually selected ROIs in raw TEM images were extracted, characterizing mitochondria and excluding the plasma membrane and other high-contrast regions. For each extracted ROI, binary particles larger than 600 pixels were analyzed, and particle area, perimeter, major and minor axes, Feret diameter, roundness, and circularity were measured using the built-in measurement functions of Fiji/ImageJ. Following Merrill, Flippo, and Strack (2017), area 2, particle aspect ratio (major to minor axis ratio), and shape factor (FF) were calculated from these data, where FF = perimeter 2/4pi x area. The definition of the parametric formula can be found in Merrill, Flippo, and Strack (2017). The macros mentioned are available on GitHub (see Data Availability Statement). On average, approximately 5,600 particles were analyzed per PPA treatment, for a total of approximately 17,000 particles (data not shown).
        SH-SH5Y cells were placed in 8-chamber culture dishes (ThermoFisher, #155411) to allow adhesion overnight and then incubated with TMRE 1:1000 (ThermoFisher, #T669) and Hoechst 33342 1:200 (Sigma-Aldrich, H6024 ). dyeing. Images were acquired using 405 nm and 561 nm lasers over a 10 min environment, and raw images were acquired as z-stacks containing 10 image micrographs with a z step of 0.2 μm between image frames at 12 subsequent time points . Images were collected using a Carl Zeiss LSM780 ELYRA PS.1 super-resolution platform (Carl Zeiss, Oberkochen, Germany) using an LCI Plan Apochromate 100x/1.4 Oil DIC M27 lens. Images were analyzed in ImageJ using a previously described pipeline and the ImageJ plugin to measure fusion and fission events, average number of mitochondrial structures, and average mitochondrial volume per cell33. MEL macros are available on GitHub (see Data Availability Statement).
        SH-SY5Y cells were grown in six-well plates at a density of 0.3 × 106 cells/mL for 24 hours before treatment. RNA was extracted using the Quick-RNA™ Miniprep protocol (ZR R1055, Zymo Research) with slight modifications: add 300 μl of RNA lysis buffer to each well before removal and lyse each sample as a final step with 30 μl of DNase/RNase elution. -free water. All samples were checked for quantity and quality using a NanoDrop ND-1000 UV-Vis Spectrophotometer. Total protein from cell lysates was obtained using 200 μl RIPA lysis buffer, and protein concentration was quantified using the Bradford protein assay70.
        cDNA synthesis was performed using the Tetro™ cDNA Synthesis Kit (BIO-65043, Meridian Bioscience) according to the manufacturer’s instructions with some modifications. cDNA was synthesized in 20-μl reactions using 0.7 to 1 μg of total RNA. Primers were selected from previously published papers 42, 71, 72, 73, 74, 75, 76, 77, 78 (Table S1) and accompanying probes were designed using the PrimerQuest tool from Integrated DNA Technologies. All genes of interest were normalized to the nuclear B2M gene. The gene expression of STOML2, NRF1, NFE2L2, TFAM, cMYC and OPA1 was measured by RT-qPCR. The master mix included LUNA Taq polymerase (M3003L, New England Biolabs), 10 μM forward and reverse primers, cDNA, and PCR-grade water to yield a final volume of 10 μL for each reaction. Expression of division and fission genes (DRP1, MFN1/2) was measured using TaqMan multiplex assays. Luna Universal Probe qPCR Master Mix (M3004S, New England Biolabs) was used according to the manufacturer’s instructions with minor modifications. The multiplex RT-qPCR master mix includes 1X LUNA Taq polymerase, 10 μM forward and reverse primers, 10 μM probe, cDNA, and PCR-grade water, resulting in a final volume of 20 μL for each reaction. RT-qPCR was performed using Rotor-Gene Q 6-plex (QIAGEN RG—serial number: R0618110). Cycling conditions are shown in Table S1. All cDNA samples were amplified in triplicate and a standard curve was generated using a series of tenfold dilutions. Outliers in triplicate samples with cycle threshold standard deviation (Ct) >0.5 were removed from the analysis to ensure data reproducibility30,72. Relative gene expression was calculated using the 2-ΔΔCt79 method.
        Protein samples (60 μg) were mixed with Laemmli loading buffer at a 2:1 ratio and run on a 12% colorless protein gel (Bio-Rad #1610184). Proteins were transferred to a PVDF (polyvinylidene fluoride) membrane (#170-84156, Bio-Rad) using the Trans-Blot Turbo system (#170-4155, Bio-Rad). The membrane was blocked and incubated with the appropriate primary antibodies (OPA1, MFN1, MFN2, and DRP1) (diluted 1:1000) for 48 hours, followed by incubation with secondary antibodies (1:10,000) for 1 hour. Membranes were then imaged using Clarity Western ECL Substrate (#170-5061, Bio-Rad) and recorded using a Bio-Rad ChemiDoc MP system. ImageLab version 6.1 was used for Western blot analysis. The original gel and blot are shown in Figure S1. Antibody information is provided in Table S2.
        Data sets are presented as the mean and standard error of the mean (SEM) of at least three independent samples. Data sets were tested for normality using the Shapiro-Wilks test (unless otherwise stated) before assuming a Gaussian distribution and equal standard deviations and proceeding with the analyses. In addition to analyzing the data set using Fisher’s MEL LSD (p < 0.05), one-way ANOVA (treatment vs. control mean), and Dunnett’s multiple comparison test to determine significance (p < 0.05). Significant p values ​​are shown in the graph as *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. All statistical analyzes and graphs were performed and generated using GraphPad Prism 9.4.0.
        Fiji/ImageJ macros for TEM image analysis are publicly available on GitHub: https://github.com/caaja/TEMMitoMacro. The Mitochondrial Event Locator (MEL) macro is publicly available on GitHub: https://github.com/rensutheart/MEL-Fiji-Plugin.
        Meiliana A., Devi N.M. and Vijaya A. Mitochondria: master regulators of metabolism, homeostasis, stress, aging and epigenetics. Indonesian. Biomedical Science. J. 13, 221–241 (2021).
        Ben-Shachar, D. Multifaceted mitochondrial dysfunction in schizophrenia, complex I as a possible pathological target. Schizophrenia. resource. 187, 3–10 (2017).
        Bose, A. and Beal, M.F. Mitochondrial dysfunction in Parkinson’s disease. J. Neurochemistry. 139, 216–231 (2016).
        Sharma V.K., Singh T.G. and Mehta V. Stressed mitochondria: targets of invasion in Alzheimer’s disease. Mitochondria 59, 48–57 (2021).
        Belenguer P., Duarte J.M.N., Shook P.F. and Ferreira G.K. Mitochondria and the brain: bioenergetics and more. Neurotoxins. resource. 36, 219–238 (2019).
        Rangaraju, V. et al. Pleiotropic mitochondria: the impact of mitochondria on neuronal development and disease. J. Neuroscience. 39, 8200–8208 (2019).
        Cardaño-Ramos, C. and Morais, V.A. Mitochondrial biogenesis in neurons: how and where. internationality. J. Mohr. the science. 22, 13059 (2021).
        Yu, R., Lendahl, U., Nister, M. and Zhao, J. Regulation of mammalian mitochondrial dynamics: opportunities and challenges. front. endocrine. (Lausanne) 11, 374 (2020).
        Khacho, M. and Slack, R.S. Mitochondrial dynamics in the regulation of neurogenesis: from the developing to adult brain. development. dynamic. 247, 47–53 (2018).